Angiotensinogen is primarily, although by no means exclusively, synthesized in the liver, particularly the pericentral zone of the hepatic lobules. In humans, it is coded by a single gene, composed of five exons and four introns, that spans about 13 kb of genomic sequence on chromosome 1 (1q42-q43). It is translated to a 453 amino acid globular glycoprotein with a molecular weight between 45 and 65 kDa, depending on the extent of its glycosylation, that then undergoes posttranslational cleavage of a 24– or 33–amino acid signal peptide, giving rise to the mature circulating form of angiotensinogen. New approaches to target the RAS are either direct antisense oligonucleotides that inhibit angiotensinogen RNA translation or small interfering RNA (siRNA) that function via the RNA interference pathway in the liver. Since all angiotensins stem from angiotensinogen, lowering angiotensinogen has the potential to circumvent the RAS escape phenomenon.

Structurally, angiotensinogen bears substantial homology to the serpin superfamily of protease inhibitors and, like many members of its family, behaves as an acute-phase reactant in the inflammatory setting, reflecting the presence of an acute-phase response element that binds the transcription factor, NF-κB (nuclear factor kappa-light-chain-enhancer of activated B cells).

Like angiotensinogen, the gene-encoding renin is also located on the long arm of chromosome 1 (1q32) and contains 10 exons and 9 introns, similar to other aspartyl proteases. Unlike humans and rats that have only a single renin gene, the mouse has two genes, Ren1 and Ren2, expressed primarily in the submandibular gland and kidney, respectively.

Following its synthesis as a 406–amino acid preprohormone, the 23–amino acid leader sequence of preprorenin is cleaved in the rough endoplasmic reticulum, giving rise to prorenin (also called inactive renin and “big” renin), which may be then rapidly secreted directly from the Golgi apparatus or from protogranules. Alternatively, and virtually exclusively in the JGA, prorenin may be packaged into mature, dense granules that instead of being immediately secreted, undergo further processing to the active enzyme, renin (active renin). Contrasting with the more constitutive secretion of prorenin, the release of renin-containing granules is tightly regulated.

Mature, active renin is a variably glycosylated 340–amino acid 37–40-kDa aspartyl protease that is active at neutral pH, and in contrast to the more promiscuous activities of most other proteases in this class, it has only a single known substrate, cleaving the decapeptide angiotensin I from the amino terminal of angiotensinogen. Although the kidney produces both renin and prorenin, a range of extrarenal tissues including the adrenals, gonads, and placenta produce prorenin and contribute to its presence in plasma. However, as evidenced by the near total absence of active renin in anephric patients, the kidney, and the JGA in particular, appears to be the only source of circulating renin in humans.

Factors that chronically stimulate renin secretion, such as a low-sodium diet and ACE inhibition, lead to an increase in the number of renin-secreting cells rather than an increase in cell size or the number of granules that each JGA cell contains. This expansion of the renin-secreting mass occurs proximally by metaplastic transformation of smooth muscle cells within the walls of the afferent arteriole. Although sometimes mentioned, ectopic renin expression within the extraglomerular mesangium appears to be an uncommon event. The mechanisms that inhibit renin expression in nonexpressing cell types remain poorly understood.

Prorenin activation

Prorenin is maintained as an inactive zymogen through the occupation of its catalytic cleft by its prosegment. Removing this prosegment by either proteolytic or nonproteolytic means yields active renin, a term that denotes its enzymatic activity rather than its amino acid sequence ( Fig. 11.2 ).

The conformational changes and the expression of immunoreactive epitopes associated with the activation of prorenin are depicted.

The main body of the molecule (blue) , the substrate-binding cleft, and the prosegment (black line) are shown. The closed triangle represents the epitope of the main body expressed by PR _c (prorenin in the inactive closed conformation), PR _oi (prorenin in the inactive intermediary open conformation), PR _o (prorenin in the active open conformation), and renin. The closed circle (yellow) represents the epitope of the main body, expressed by PR _o and renin, but not by PR _c and PR _oi . The open circles represent epitopes of the prosegment expressed by PR _o but not by PR _c and PR _oi .

Modified from Schalekamp MA, Derkx FH, Deinum J, et al. Newly developed renin and prorenin assays and the clinical evaluation of renin inhibitors. J Hypertension. 2008;26:928–937.

Within the dense core secretory granules of the JGA, acidification by vacuolar adenosine triphosphatases (ATPases) provides the optimal pH for the prosegment-cleaving enzymes (proconvertase 1 and cathepsin B) and may also assist the pH-dependent, nonenzymatic activation of prorenin as well. Although various peptidases such as trypsin, plasmin, and kallikrein can also cleave the prosegment of prorenin in vitro, these do not appear to contribute to the generation of renin in the in vivo setting. Although traditionally viewed as occurring only in the JGA, cell culture–based studies suggest that proteolytic activation of renin can also occur in cardiac and vascular smooth muscle cells by as yet unidentified serine proteases. The significance of these findings in the intact organism, however, remains to be established.

In addition to proteolytic cleavage of its prosegment, prorenin can also be reversibly activated nonenzymatically by a conformational change such that the prosegment no longer occupies the enzymatic cleft. Under usual circumstances, <2% of prorenin is in this open active conformation. This process can, however, be induced by acid (pH 4.0) ^17,18 and to a lesser extent by cold. More recently, the putative (pro)renin receptor (PRR; discussed later) has also been shown to nonproteolytically activate prorenin.

Regulation of renin secretion

Mechanical, neurologic, and chemical factors regulate the activity of the RAS by modulating renin secretion.

Renal baroreceptor

Renin-expressing juxtaglomerular (JG) cells possess an intrinsic pressure-sensing mechanism(s) that regulates renin synthesis and release in response to changes in perfusion pressure. The existence of a renal baroreceptor mechanism was first conceptualized by Skinner and colleagues to explain how renin secretion increases when afferent arteriolar perfusion pressure falls. Studies in conscious dogs show that changes in renal perfusion pressure have only a small effect on renin secretion until a threshold of about 90 mm Hg is reached, below which renin secretion abruptly increases, doubling with every 2 to 3 mm Hg fall in pressure. Accordingly, reduction in pressure below this level profoundly stimulates renin secretion, thereby acutely activating the RAS and resulting in a range of angiotensin II–dependent phenomena that collectively serve to restore systemic pressure. Despite the importance of the baroreceptor function, several decades of research have not identified precisely how the pressure signal is transduced into renin release, though postulated mediators include Piezo1, a mechanosensitive ion channel, endothelins (ETs), and prostaglandins.

Neural control

The JGA is endowed with a rich network of noradrenergic nerve endings and their β1 receptors. Stimulation of the renal sympathetic nerve activity leads to renin secretion that is independent of changes in RBF, glomerular filtration rate (GFR), or Na ⁺ reabsorption. Moreover, this effect can be blocked surgically (denervation) and pharmacologically, by the administration of β-adrenoreceptor blockers. The role of cholinergic, dopaminergic, and adrenergic activation is controversial, though these agents have also been shown to modulate renin release under certain circumstances.

Tubule control

Chronic diminution in luminal NaCl delivery to the macula densa is a potent stimulus for renin secretion, reflecting a coordinate interaction between a range of mediators including adenosine, nitric oxide (NO), and prostaglandins that affect renin release and its transcription. This mechanism is thought to account for the chronically high plasma renin activity (PRA) in persons ingesting a low-salt diet.

Metabolic control

The tricarboxylic acid (TCA) cycle provides a final common pathway by which carbohydrates, fatty acids, and amino acids converge in the process of adenosine triphosphate (ATP) generation by aerobic electron transfer. Although the TCA cycle operates within mitochondria, its intermediates can be detected within the extracellular space, increasing in abundance when local energy supply and demand are mismatched or when cells are exposed to hypoxia, toxins, or injury. Succinate, for instance, has been shown to stimulate renin release and its intravenous administration leads to hypertension, though the mechanisms underlying this effect have only recently been unraveled. In 2004, He and colleagues reported that α keto-glutarate and succinate are ligands for the previously orphaned G protein–coupled receptors (GPCRs), GPR99 and GPR91, respectively, and that succinate-induced hypertension is abolished in GPR91-deficient mice. Indeed, in follow-up studies from this group, GPR91 was localized to the apical plasma membrane of macula densa cells where succinate stimulation was shown to activate p38 and Erk 1/2 mitogen-activated protein (MAP) kinases (MAPKs), inducing cyclooxygenase-2 (COX-2)-dependent synthesis of prostaglandin E2, a well-established paracrine mediator of renin release. Moreover, the ability of tubule succinate to induce JG renin secretion suggests that this phenomenon is likely an important determinant of JGA function in both physiologic and pathophysiologic settings. In diabetic rats, for instance, elevated succinate has been detected in both plasma and urine.

Vitamin D receptor

The vitamin D receptor (VDR) is a negative regulator of the RAS such that VDR-null mice display a marked increase in renin expression and angiotensin II production in conjunction with hypertension and cardiac hypertrophy. Importantly, these effects occurred independently of calcium and parathyroid hormone, both of which have been reported to also modulate renin expression. The molecular basis for the interaction between vitamin D and renin expression has also been, at least partly, unraveled in a series of studies by Yuan and colleagues. Under usual circumstances, activation of the cyclic adenosine monophosphate (cAMP)–protein kinase A pathway by the sympathetic nervous system or macula densa leads to phosphorylation of a cAMP response element (CRE) binding protein (CREB) and recruitment of CREB-binding protein/p300 to the CRE in the promoter region of the renin gene. VDR-bound 1,25 (OH) ₂ D ₃ , however, blocks binding of CREB to the CRE DNA cis -element, leading to a reduction in renin gene transcription ( Fig. 11.3 ).

Model of 1,25(OH) ₂ D ₃ -induced transrepression of renin gene expression.

The cAMP–PKA pathway activates CREB by phosphorylation, leading to recruitment of CBP/p300. In the presence of 1,25(OH) ₂ D ₃ , liganded VDR interacts with CREB and blocks its binding to CRE, leading to reduction of renin gene transcription.

cAMP, Cyclic AMP; CBP, CREB-binding protein; CRE, cAMP response element; CREB, CRE-binding protein; D, 1,25(OH) ₂ D ₃ ; P, phosphorylation; PKA, protein kinase A; Pol II, RNA polymerase II; VDR, vitamin D receptor.

Modified from Yuan W, Pan W, Kong J, et al. 1,25-Dihydroxyvitamin D ₃ suppresses renin gene transcription by blocking the activity of the cyclic AMP response element in the renin gene promoter. J Biol Chem. 2007;282:29821–29830.

From a clinical perspective, this interaction between vitamin D and renin may explain the well-documented inverse relationship among plasma vitamin D ₃ , blood pressure, and PRA. While several studies examining the effects of vitamin D supplementation as a kidney-protective or antihypertensive measure have been undertaken, their findings have been mixed and long-term, randomized controlled trials with clinically meaningful endpoints have not been performed.

Other local factors

In addition to the factors discussed earlier, a large range of locally produced biologically active molecules have also been shown to alter renin secretion. These include peptides (ANP, kinins, vasoactive intestinal polypeptide, ET, calcitonin gene–related peptide [CGRP]); amines (dopamine and histamine); and arachidonic acid derivatives.

ACE is a zinc-containing dipeptidyl carboxypeptidase that cleaves the terminal histidyl-leucine from angiotensin I to form the octapeptide angiotensin II. In contrast to the single-substrate specificity of renin, ACE is not specific, cleaving the two terminal acids from peptides with the C′-terminal sequence R ₁ -R ₂ -R ₃ -OH, where R ₁ is the protected (noncleaved) amino acid, R ₂ is any nonproline l -amino acid, and R ₃ is any nondicarboxylic (cysteine, ornithine, lysine, arginine) l -amino acid with a free carboxyl terminal. Importantly, therefore, ACE also catalyzes the inactivation of bradykinin. Although encoded by a single ACE gene, two distinct tissue-specific messenger RNAs (mRNAs) are transcribed, each with different initiation and alternative splice sites. The somatic form, present in almost all tissues, is a 1306–amino acid, 140–160-kDa glycoprotein with two active sites, whereas the 90- to 100-kDa testicular or germinal form is found exclusively in postmeiotic male germ cells and contains a single active site and appears to be involved in spermatogenesis. ^{, ,} The somatic form of ACE is widely distributed with activity present in not only tissues but also most biological fluids. In the human kidney, ACE is present to the greatest extent within proximal and distal tubules; however, both its magnitude of expression and its site-specific distribution may be altered by disease.

The AT ₁ R mediates most of the known physiologic effects of angiotensin II. The gene for this widely distributed 359–amino acid, 40-kDa, seven-transmembrane GPCR is located on chromosome 3 in humans.

Within the kidney, AT ₁ Rs are widely expressed. In the glomerulus, they are found in both afferent and efferent arterioles, as well as in the mesangium, in the endothelium, and in podocytes. Consistent with angiotensin II’s role in Na ⁺ reabsorption, AT ₁ Rs are highly abundant on the brush borders of proximal tubule epithelial cells. Prominent expression has also been found in renal medullary interstitial cells, located between the renal tubules and vasa recta, where angiotensin II is purported to have a potential role in the regulation of medullary blood flow.

Angiotensin II binding to AT ₁ R initiates cell signaling by several different pathways that have been mostly studied in vascular smooth muscle cells. These include G protein–mediated pathways and the activation of tyrosine kinases, NADH/NADPH oxidases, and serine/threonine kinases.

Tyrosine kinases

Angiotensin II binding to the AT ₁ R “transactivates” a number of nonreceptor tyrosine kinases (Src, Pyk2, focal adhesion kinase [FAK], and Janus kinase [JAK]), as well as the growth factor receptor tyrosine kinases for epidermal growth factor (EGF) ^, and platelet-derived growth factor (PDGF). ^, By binding to the AT ₁ R, angiotensin II initiates the translocation of tumor necrosis factor–α (TNF-α)–converting enzyme (TACE, ADAM17) to the cell surface. TACE then cleaves TNF-α from its membrane-associated precursor (pro–TNF-α), allowing it to bind to the EGF receptor (EGFR) on the cell surface. This ligand-receptor interaction then induces EGFR autophosphorylation and activates its downstream signaling pathways that include Akt, Erk-1/2, and mammalian target of rapamycin ( Fig. 11.4 ). The in vivo relevance of this transactivation pathway has been confirmed. Using mice that express a dominant negative form of EGFR, Lautrette and colleagues showed that despite similar blood pressures, mutant mice infused with angiotensin II had less proteinuria and renal fibrosis than their wild-type counterparts. Consistent with these findings and the pivotal role of the RAS in diabetic nephropathy, studies using a specific EGFR tyrosine kinase inhibitor (i.e., PKI 166) have also shown a reduction in early structural injury in a rat model of diabetic nephropathy.

Angiotensin II (ANG II) binds to its angiotensin II type 1 (AT ₁ ) receptor, a G protein–coupled receptor lacking intrinsic tyrosine kinase activity.

Through as-yet-undescribed mechanisms, this interaction leads to the translocation of the metalloprotease tumor necrosis factor-α (TNF-α) –converting enzyme (TACE) from the cytosol to the cell surface, where it cleaves TNF-α from its membrane-associated promolecule, allowing it to bind and activate the epidermal growth factor (EGF) receptor. Erk 1/2, Extracellular signal–regulated kinases 1 and 2; mTOR, mammalian target of rapamycin; P13K, phosphatidylinositol-3-kinase.

Modified from Wolf G. “As time goes by”: angiotensin II–mediated transactivation of the EGF receptor comes of age. Nephrol Dial Transplant. 2005;20:2050–2053.

Like EGFR, the transactivation of the PDGF receptor (PDGFR) by AT ₁ R is also complex, involving the adaptor protein Shc. ^, In addition to studies that have explored the angiotensin-PDGFR interaction in cell culture or organ baths, a further report has shown that despite continued hypertension, inhibition of the PDGFR kinase in vivo can also dramatically attenuate angiotensin II–induced vascular remodeling.

Angiotensin type 1 receptor internalization

In addition to the conventional ligand-receptor–mediated pathways, a range of other signaling mechanisms that involve the AT ₁ R have also been described. These include the discovery of receptor–interacting proteins, heterologous receptor dimerization, and ligand-independent activation ( Fig. 11.5 ). These additional insights, although adding greater complexity to our understanding of the RAS, also provide the potential for new therapeutic targets in disease prevention and management.

Regulation of angiotensin receptors.

ARAP1, AT ₁ receptor–associated protein 1; ATBP50, AT ₂ receptor–binding protein of 50 kDa; ATIP, angiotensin type 2 receptor–interacting protein; AT ₁ receptor, angiotensin type 1 receptor; AT ₂ receptor, angiotensin type 2 receptor; ATRAP, AT ₁ receptor–associated protein; PLZF, promyelocytic leukemia zinc finger.

Modified from Mogi M, Iwai M, Horiuchi M. New insights into the regulation of angiotensin receptors. Curr Opin Nephrol Hypertens . 2009;18:138–143.

Following ligand binding and the initiation of signal transduction, AT ₁ Rs are rapidly internalized, followed by either lysosomal degradation or recycling back to the plasma membrane. Several mechanisms account for AT ₁ R internalization including interaction with caveolae, phosphorylation of its carboxyl terminal by G-protein receptor kinases, and association with the newly described AT ₁ R interacting proteins. To date, two such interacting proteins, AT ₁ r eceptor– a ssociated p rotein (ATRAP) and A T ₁ r eceptor– a ssociated p rotein-1 (ARAP1), have been described. ATRAP interacts with the C terminal of AT ₁ R, downregulating cell surface AT ₁ R expression and attenuating angiotensin II–mediated effects. ARAP1, though somewhat similar to ATRAP, promotes AT ₁ R recycling to the plasma membrane such that its kidney-specific overexpression induces hypertension and renal hypertrophy.

The traditional actions of angiotensin II relate primarily to its effects on vascular tone and fluid balance that are mediated by its actions on the vasculature, heart, kidney, brain, and adrenal glands by the AT ₁ R. In vascular smooth muscle, stimulation of AT ₁ Rs by angiotensin II induces cell contraction and consequent vasoconstriction. In the adrenal cortex, this ligand-receptor interaction stimulates aldosterone release, thereby promoting sodium reabsorption in the distal nephron. Moreover, angiotensin II will directly enhance sodium retention by the proximal tubule and in the brain, it will stimulate thirst and salt craving. Additional effects include sympathoadrenal stimulation and the augmentation of cardiac contractility. Together, these effects serve to maintain extracellular fluid volume and systemic blood pressure. Given the central role that the kidney has in the regulation of these key aspects of mammalian homeostasis, it is not surprising that angiotensin II should have profound effects on renal physiology.

In humans, the AT ₂ R is a 363–amino acid protein that maps to the X chromosome and is highly homologous to its rat and mouse counterparts. Like AT ₁ R, AT ₂ R is also a seven-transmembrane GPCR, though it shares only 34% homology.

Despite substantial research, the actions of AT ₂ R are still not well understood and remain somewhat controversial. In general, however, the actions of AT ₂ R stimulation oppose those of AT ₁ R. For instance, whereas AT ₁ R vasoconstricts and promotes Na ⁺ retention, AT ₂ R stimulation leads to vasodilation and natriuresis, consistent with its abundance on the epithelium of the proximal tubule. The vasodilatory effects of AT ₂ R stimulation are mediated by increasing NO synthesis and cyclic guanosine monophosphate (cGMP) by bradykinin-dependent and bradykinin-independent mechanisms. Its natriuretic effects, however, seem to be dependent on angiotensin II’s conversion to angiotensin III by aminopeptidase N.

Like the AT ₁ R, the activity of AT ₂ R may also be modulated by oligomerization, in association with various interacting proteins and ligand-independent effects.

In 2002 an apparently novel, 350-amino acid, single-transmembrane protein that binds both renin and prorenin with high affinity was identified. Ligand binding to this protein was shown to induce a fourfold increase in the catalytic cleavage of angiotensinogen, as well as stimulating intracellular signaling with activation of MAPKs extracellular signal–regulated kinases 1 and 2 (ERK1/2), leading to it being named the (pro)renin receptor ([P]RR). The designation (pro)renin refers to its ability to interact with both renin and prorenin.

Given its localization to the mesangium in initial studies, its actions in augmenting local angiotensin II production, and its ability to increase mesangial transforming growth factor-β (TGF-β) production, the (P)RR was understandably implicated in the pathogenesis of kidney disease. However, despite the appeal, it has been difficult to reconcile this view of the (P)RR with a number of other experimental findings, regarding both its potentially pathogenetic role and pattern of distribution within the kidney and its homology to other proteins. For instance, given the purported pathogenetic role of the (P)RR, the increased abundance of renin that follows the use of ACE inhibitors and ARBs would be expected to be adverse, yet these classes of drugs have been repeatedly shown to be kidney protective. Second, although the (P)RR was initially localized to the glomerular mesangium, more recent detailed studies have shown that (P)RR is primarily expressed in the collecting duct. Third, although initially reported as having no homology with any known membrane protein, database interrogation shows that the (P)RR is identical to two other proteins: CAPER (endoplasmic reticulum–localized type 1 transmembrane adaptor precursor) and ATP6ap2 (ATPase, H ⁺ transporting, lysosomal accessory protein 2), a protein that associates with the vacuolar H ⁺ -ATPase. Indeed, the predominant expression of the (P)RR at the apex of acid-secreting cells in the collecting duct, in conjunction with its colocalization and homology with an accessory subunit of the vacuolar H ⁺ -ATPase, suggests that the (P)RR may function primarily in urinary acidification. However, the vacuolar H ⁺ -ATPase is not restricted to the kidney but is widely distributed in the plasma membrane and the membranes of organelles in several tissues where it functions, not only in urinary acidification but also in endocytosis, conversion of proinsulin to insulin, and osteoclast bone resorption. Whereas the prevailing data indicate that (P)RR is an accessory subunit of the vacuolar H ⁺ -ATPase that also binds renin and prorenin, the precise functions of the prorenin-binding and renin-binding subunit remain to be unraveled in the kidney and elsewhere.

In 2000, two groups independently reported the existence of the first ACE homolog, ACE2, an apparently novel zinc metalloprotease but with considerable homology (40% identity and 61% similarity) to ACE. ^, The gene-encoding ACE2, located on the X chromosome (Xp22), contains 18 exons, several of which bear considerable similarity to the first 17 exons of human ACE . Its transcript is 3.4 kb, generating an 805–amino acid peptide that is most highly expressed in kidney, heart, and testis but is also present in plasma and urine. ^, In contrast to ACE, ACE2 functions as a carboxypeptidase, removing the terminal phenylalanine from angiotensin II to yield the vasodilatory heptapeptide, angiotensin 1–7 (Ang [1–7]). ACE2 may also indirectly lead to the formation of Ang (1–7) by cleaving the C-terminal leucine from angiotensin I, thereby generating angiotensin 1–9 (Ang [1–9]), which may then give rise to Ang (1-7) under the influence of ACE or neutral endopeptidase (NEP). Thus ACE2 contributes to both angiotensin II degradation and Ang (1–7) synthesis. Accordingly, ACE and ACE2 were initially viewed as having opposing actions with regard to vascular tone and tissue injury. However, emerging data suggest that the situation is far from clear. For instance, while lentivirus-induced overexpression of ACE2 in the heart exerted a protective influence following experimental myocardial infarction, ACE2 overexpression led to cardiac dysfunction and fibrosis, despite lowering systemic blood pressure.

In the kidney, ACE2 colocalizes with ACE and angiotensin receptors in the proximal tubule, while in the glomerulus it is predominantly expressed within podocytes and to a lesser extent in mesangial cells, contrasting the endothelial predilection of ACE at that site. Numerous studies have explored changes in ACE2 expression in human kidney disease, as well as in a range of animal models, reporting both increased and decreased levels. As such, it is uncertain whether increased ACE2 might be detrimental or a beneficial response to injury. With this in mind, the findings of intervention studies are of particular importance.

In experimental diabetic nephropathy, for instance, pharmacologic ACE2 inhibition with MLN-4760 led to worsening albuminuria and glomerular injury; similar findings were reported in ACE2 knockout mice that were crossed with the Akita model of type 1 diabetes. As might be expected from these findings, augmenting ACE2 activity by the infusion of human recombinant protein (hrACE2) was shown to attenuate diabetic kidney injury in the Akita mouse. In this study, hrACE2 improved kidney structure and function and also showed that the protective effects were likely due to reduction in angiotensin II and an increase in Ang (1–7) signaling. In addition to its role in the RAS, ACE2 has been shown to be the receptor for the severe adult respiratory syndrome (SARS) coronavirus. The relevance of some of these findings to the human setting will hopefully be clarified when results of the recently initiated clinical trials of a recombinant human ACE2 peptide (GSK2586881) become available.

In the traditional view of the RAS, angiotensin II functions as a hormone that, in classical endocrine fashion, circulates systemically to act at sites distant from those where it was formed. However, since the cloning of its components, it has become increasingly clear that there is an additional local, tissue–based RAS that functions quasiindependently from its systemic counterpart, acting in paracrine, autocrine, and possibly even intracrine modes. This is most clearly seen in the kidney where pioneering work of Navar and colleagues has shown that the kidney possesses all the necessary molecular machinery to synthesize angiotensin II and other bioactive angiotensin peptides. Moreover, their concentrations in glomerular filtrate, tubule fluid, and interstitium are between 10- and 1000-fold higher than in plasma. ^,

Within the kidney, renin-expressing cells have traditionally been considered to be terminally differentiated and confined to the JGA. However, in a series of elegant studies using a fate-mapping cre-loxP system, Sequeira López and colleagues showed that renin-expressing cells are precursors to a range of other cell types in the kidney, including those of the arteriolar media, mesangium, Bowman capsule, and proximal tubule. While normally quiescent, these cells may undergo metaplastic transformation to synthesize renin when homeostasis is challenged. Such threats include those related to volume depletion and also tissue injury. For instance, in the setting of single nephron hyperfiltration and consequent progressive dysfunction that follows renal mass reduction, Gilbert and colleagues noted the de novo expression of renin mRNA and angiotensin II peptide in tubule epithelial cells.

In addition to resident kidney cells, infiltrating mast cells may also contribute to activation of the local RAS in disease. Traditionally associated with allergic reactions and host responses to parasite infestation, mast cells have been increasingly recognized for their role in inflammation, immunomodulation, and chronic disease. In the kidney, interstitial mast cell infiltration accompanies most forms of chronic kidney disease (CKD), where their abundance correlates with the extent of tubulointerstitial fibrosis and declining GFR, though not proteinuria. Notably, mast cells have been shown to synthesize renin, such that their degranulation will release large quantities of both renin and chymase, accelerating angiotensin II formation in the local environment.

Peptide hormones traditionally bind to their cognate receptors on the plasma membrane and produce their effects through the generation of secondary intermediates. However, emerging evidence suggests that certain peptides may also act directly within the cell’s interior, having arrived there by either internalization or intracellular synthesis. For instance, angiotensin II has been localized within the cytoplasm and nucleus, but its introduction into the cytoplasm was shown more than a decade ago to have major effects on intracellular calcium currents. Uptake of angiotensin II from the extracellular space likely contributes to its intracellular activity; however, newer studies have focused predominantly on its endogenous synthesis. Consistent with the potential role for intracellular angiotensin II, transgenic mice that express an enhanced cyan fluorescent protein–angiotensin II fusion protein that lacks a secretory signal so that it is retained intracellularly develop hypertension with microthrombi in glomerular capillaries and small vessels. To date, numerous canonical and noncanonical pathways in the cytoplasm, nucleus, and mitochondria have been implicated in the intracrine RAS, ^, providing a new forefront for the role of the RAS in physiology, pathophysiology, and therapeutics.

In a critical series of experiments in the 1980s, Brenner and colleagues studied the hemodynamic effects of renal mass ablation in 5/6 nephrectomized rats, a now well-established model of progressive kidney disease. In the setting of nephron loss, those glomeruli that remain undergo compensatory enlargement with increased single-nephron GFR (SNGFR) and elevations in intraglomerular pressure (P _GC ), argued to initiate a vicious cycle of glomerulosclerosis and progressive loss of function. That this phenomenon might be related to angiotensin II was suggested by previous work in which angiotensin II infusion was demonstrated to also result in elevated P _GC . Together, these studies suggested that intraglomerular hypertension, as a consequence of angiotensin II’s action, was a pivotal factor underlying the inexorable progression of kidney disease and that strategies to reduce P _GC would lead to its amelioration. Indeed, in proof-of-concept studies, blockade of angiotensin II formation with the ACE inhibitor, enalapril, was shown to dilate the glomerular efferent arteriole, reduce P _GC , and slow disease progression in 5/6 nephrectomized rats. By contrast, combination therapy with hydralazine, reserpine, and hydrochlorothiazide, though equally effective in lowering systemic blood pressure, failed to ameliorate intraglomerular hypertension and disease progression. These studies were soon followed by similar ones in other disease models, particularly diabetes, which like the 5/6 nephrectomized rats is also characterized by increased SNGFR and elevated P _GC . ^, For a more detailed discussion of glomerular hemodynamic changes after 5/6 nephrectomy, see Chapter 50 .

Considerable research has also focused on many of the nonhemodynamic effects of angiotensin II. For instance, in addition to their effects on P _GC , ACE inhibitors and ARBs are also highly effective in reducing interstitial fibrosis and tubule atrophy, each close correlates of progressive kidney dysfunction. Underlying these effects is the ability of angiotensin II to potently induce expression of the profibrotic and proapoptotic growth factor and TGF-β in a range of kidney cell types. ^, Consistent with these in vitro studies, TGF-β overexpression is seen in both the glomerular and tubulointerstitial compartments in 5/6 nephrectomized rats and diabetic rats, where studies also showed that both ACE inhibitors and ARBs were effective at reducing TGF-β and disease progression. ^, Similarly, in human diabetic nephropathy, the ACE inhibitor perindopril was found to reduce TGF-β mRNA in a sequential renal biopsy study and losartan was shown to lower urinary TGF-β excretion. For further discussion of the nonhemodynamic effects of angiotensin II, see Chapter 50 ^.

The development of proteinuria is both a cardinal manifestation of glomerular injury and a pathogenetic factor in the progression of renal dysfunction. While P _GC remains an important factor in determining the transglomerular passage of albumin, more recent work has focused on the potential contribution of the podocyte. Indeed, podocyte injury is a cardinal manifestation of proteinuric renal disease, where foot process effacement has been shown to be prevented by both ACE inhibition and angiotensin receptor blockade. In consideration of its crucial role in the development and function of the glomerular filtration barrier, other studies have focused on the podocyte slit pore membrane protein nephrin. Of note, podocytes express the AT ₁ R and respond to the addition of angiotensin II to the cell culture medium by dramatically decreasing their expression of nephrin. Consistent with these findings, the reduction in nephrin expression in patients with diabetic nephropathy was shown to be ameliorated by ACE inhibitor treatment for 2 years. For further discussion of the pathophysiology of proteinuria in CKD, see Chapter 29 , Chapter 49 .

Inflammatory cell infiltration is a long-recognized feature of CKD that is attenuated in rodent models by agents that block the RAS. In the in vitro setting, angiotensin II activates NF-κB by both AT ₁ R- and AT ₂ R-dependent pathways, stimulating the expression of a number of potent chemokines such as MCP-1 and regulated on activation, normal T cell expressed and secreted (RANTES), as well as cytokines, like IL-6. In addition to angiotensin II, angiotensin (1–7), acting via the Mas receptor, activates NF-κB, inducing proinflammatory effects in the kidney under both basal and disease settings.

In addition to macrophages, mast cells, and other components of the innate immune system, the adaptive immune system also appears to be involved in the pathogenesis of angiotensin II–mediated organ injury. Of note, suppression of the adaptive immune system prevents the development of angiotensin II–dependent hypertension in experimental models and adoptive transfer of CD4 ⁺ CD25 ⁺ regulatory T cells is able to ameliorate angiotensin II–dependent injury.

Despite the fact that patients with long-standing diabetes characteristically have low plasma renin, suggesting that the RAS is not activated by the disease, agents that block the RAS are the mainstay of therapy in diabetic nephropathy. Compounding this apparent paradox, although PRA is normal or low in diabetes, plasma prorenin is characteristically elevated. This dichotomy suggests differences in cell-specific responses to diabetes since the JGA is the primary source of renin secretion while prorenin is secreted by a much wider range of cell types. Peti-Peterdi and colleagues have proposed a model for the (pro)renin paradox of diabetes. Although early diabetes would lead to augmented succinate and enhanced JGA renin release, elevated angiotensin II levels would thereafter suppress JGA renin secretion. Contrasting this negative feedback at the JGA, angiotensin II has been shown to have the opposite effect in the tubule, with diabetes causing a 3.5-fold increase in collecting duct renin that could be reduced by AT ₁ R blockade. These observations also support a role for the intrarenal component of the RAS in CKD progression, discussed earlier.

ET receptor antagonists have been employed to study the role of ETs in renal pathophysiology in a range of experimental models including the rat remnant kidney, lupus nephritis, and diabetes. In the remnant kidney model of progressive kidney disease, although beneficial effects have been reported with nonselective ET receptor antagonists, selective ET-A receptor inhibition appears to yield superior outcomes, with concomitant inhibition of ET-B receptors potentially abrogating any beneficial effects.

Data with regard to an effect of high glucose concentrations on ET synthesis and secretion are conflicting. Mesangial cell p38 MAPK activation in response to ET-1, angiotensin II, and PDGF is enhanced in the presence of high glucose levels. By contrast, mesangial contraction in response to ET-1 is diminished under high glucose conditions. ^, Circulating ET-1 concentrations are elevated in animal models of both type 1 and type 2 diabetes. Increased expression of ET-1 and its receptors has been found in glomeruli and tubule epithelial cells, ^, although increased expression of ET receptors has not been a universal finding. Diabetes also causes an increase in renal ECE1 expression, the effect being synergistic with that of radiocontrast media.

A number of researchers have investigated the effect of both nonselective and selective ET-A receptor antagonists in experimental diabetic nephropathy. In streptozotocin-diabetic rats, the nonselective ET receptor antagonist bosentan has yielded conflicting results, ^, whereas another nonselective ET receptor antagonist, PD142893, improved kidney function when administered to streptozotocin-diabetic rats that were already proteinuric. In addition, acute ET-A receptor antagonism has been shown to improve oxygen availability in the kidneys of streptozotocin-diabetic rats. In Otsuka Long Evans Tokushima Fatty (OLETF) rats with type 2 diabetes, selective ET-A receptor blockade attenuated albuminuria, without affecting blood pressure, whereas ET-B receptor blockade had no effect. In a study of streptozotocin-diabetic apolipoprotein E knockout mice, the kidney-protective effects of the predominant ET-A receptor antagonist avosentan were comparable with or superior to the ACE inhibitor quinapril. Supporting a protective role for the ET-B receptor, diabetic ET-B receptor–deficient rats developed severe hypertension and progressive kidney failure.

Accumulation of reactive oxygen species plays a major role in the pathogenesis of diabetic complications, particularly diabetic nephropathy, ^, and several observations suggest that the ET system may contribute to oxidative stress. In low-renin hypertension, ET-1 increases superoxide in carotid arteries and ET-A receptor blockade decreases vascular superoxide generation. ^, Similarly, ET-1 infusion increased urinary excretion of 8-isoprostane prostaglandin F ₂ α in rats, which is indicative of increased generation of reactive oxygen species. By contrast, however, other preclinical studies have suggested a predominantly proinflammatory role for ET-1 in diabetic nephropathy. For instance, the selective ET-A receptor antagonist ABT-627 prevented the development of albuminuria in streptozotocin-diabetic rats without an improvement in markers of oxidative stress but with a reduction in macrophage infiltration and urinary excretion of TGF-β and prostaglandin E ₂ metabolites.

Both plasma and urinary ET-1 levels are increased in patients with CKD ^{, ,} with plasma ET-1 levels inversely correlating with estimated GFR. In a study of hypertensive patients with CKD, both selective ET-A receptor blockade and nonselective ET receptor blockade lowered blood pressure. However, ET-A receptor blockade increased both RBF and effective filtration fraction and decreased RVR, whereas dual blockade had no effect.

In a study of 22 nondiabetic individuals with CKD, intravenous infusion of the ET-A receptor antagonist BQ-123 reduced pulse-wave velocity and proteinuria to a greater extent than the calcium channel antagonist nifedipine, which comparably lowered blood pressure. These findings suggest a potential blood pressure–independent mechanism of action for the antiproteinuric effect observed. In a subsequent study by the same investigators, 27 subjects with proteinuric CKD were treated with the ET-A receptor antagonist sitaxsentan for 6 weeks in a three-way crossover study design. Sitaxsentan treatment was associated with a reduction in blood pressure, proteinuria, and pulse-wave velocity, whereas nifedipine reduced pulse-wave velocity and blood pressure but had no effect on urine protein excretion. Subsequently, sitaxsentan was also shown to restore the nocturnal dip in blood pressure in people with CKD. A fall in GFR with sitaxsentan therapy observed in this study is analogous to that seen with RAS blockade. Although no clinically significant adverse effects were seen, sitaxsentan development has subsequently been halted due to hepatotoxicity.

The antiproteinuric effect of the ET-A receptor antagonist, zibotentan, has been evaluated in combination with the SGLT2 inhibitor, dapagliflozin, in persons with eGFR ≥20 mL/min/1.73 m ² and urinary albumin-creatinine ratio (UACR) 150 to 1500 mg/g. The rationale for this trial design was to assess if the effects of zibotentan were additive to dapagliflozin and to explore whether the natriuretic effects of an SGLT2 inhibitor may ameliorate the fluid retention associated with ET-A receptor antagonists. The original study design included groups randomized to placebo, dapagliflozin, and dapagliflozin combined with three doses of zibotentan: 5 mg/day, 1.5 mg/day and 0.25 mg/day (in addition to RAAS inhibitor treatment if tolerated). Randomization to the placebo and high-dose (5 mg) zibotentan groups was discontinued on the advice of the data safety monitoring committee due to evidence of a high incidence of fluid retention and publication of trials establishing the kidney-protective efficacy of SGLT2 inhibitors. Among 447 participants randomized to the lower doses of zibotentan (mean GFR 46.7 mL/min/1.73 m ² ; median UACR 565.5 mg/g), zibotentan 1.5 mg/day and 0.25 mg/day plus dapagliflozin were associated with a 33.7% (90% CI, 42.5–23.5%; P < 0·0001) and 27.0% (90% CI, 38.4–13.6%; P = 0·0022) reductions in UACR versus dapagliflozin alone after 12 weeks. Fluid retention (increase in body weight ≥3%) was observed in 18%, 9%, and 8% in the three groups, respectively. A post hoc analysis of all the original groups that included 508 participants found that the risk of fluid retention increased with higher zibotentan dose and lower GFR but was decreased when zibotentan was combined with dapagliflozin.

The effect of the ET-A receptor antagonist avosentan was examined, in addition to standard treatment with an ACE inhibitor or ARB, in a placebo-controlled trial of 286 patients with diabetic nephropathy and macroalbuminuria. At 12 weeks, avosentan was found to decrease urine albumin excretion rate without affecting blood pressure. These results led to the initiation of the ASCEND trial (a randomized, double-blind, placebo-controlled, parallel group study to assess the effect of the endothelin receptor antagonist avosentan on time to doubling of serum creatinine, end-stage renal disease, or death in patients with type 2 diabetes mellitus and diabetic nephropathy). ASCEND set out to examine the effects of avosentan, on top of RAS blockade, in 1392 individuals with type 2 diabetes and nephropathy but was terminated after a median duration of 4 months due to adverse events. In that study, despite a more than 40% reduction in urine albumin-to-creatinine ratio with avosentan, adverse events, predominantly fluid overload and congestive heart failure, occurred more frequently in those receiving active therapy than placebo.

Since the publication of ASCEND, it has become apparent that the adverse outcomes were likely related to the relatively high dose of avosentan that was selected and that ET-A receptor antagonists likely have a relatively narrow therapeutic window. However, if this therapeutic window is appropriately targeted and patients carefully selected, then ET-A receptor antagonists may still find a clinical niche for the treatment of diabetic nephropathy. The Reducing Residual Albuminuria in Subjects with Diabetes and Nephropathy with Atrasentan trial and an identical study conducted in Japan (RADAR/JAPAN) explored the effects of two different doses of the ET-A receptor antagonist atrasentan (0.75 and 1.25 mg/day, respectively) in 211 patients with type 2 diabetes and albuminuria (estimated GFR [eGFR] 30–75 mL/min/1.73 m ² ). In comparison with placebo, 0.75 and 1.25 mg/day atrasentan, on top of maximum tolerated doses of ACE inhibitor or ARB, reduced urine albumin-to-creatinine ratio by an average of 35% and 38%, respectively, without major side effects. Atrasentan treatment was also associated with decreases in 24-hour blood pressure, low-density lipoprotein cholesterol, and triglycerides. Importantly, despite a comparable lowering of albuminuria to the 0.75 mg/day dose of atrasentan, the 1.25 mg/day atrasentan dose was accompanied by more fluid retention. In a post hoc analysis, fluid retention was more likely in participants with lower eGFR and a higher dose of atrasentan, whereas the degree of urinary albumin lowering was not linked to the degree of fluid retention. Thus it is likely that the antiproteinuric- and fluid-retaining effects of ET receptor antagonists are mediated by different mechanisms; plausibly, the former by vascular or glomerular actions, and the latter by direct effects of ET receptor blockade on sodium transport in the renal tubule.

To reduce the risk of adverse effects, investigators included an “enrichment period” in the Study of Diabetic Nephropathy with Atrasentan (SONAR) trial, during which all participants received atrasentan. Only those who evidenced at least a 30% reduction in albuminuria and who did not develop edema progressed to the next phase of the trial to be randomized to continue atrasentan or placebo ( n = 2648 of an initial 5117). The primary endpoint of doubling of serum creatinine, kidney failure, or death from kidney failure was reduced by 35% (HR 0.65; 95% CI, 0.49–0.88; P = 0·0047). Fluid retention and anemia occurred more frequently in the group who received atrasentan, but there was no statistically significant increase in hospital admissions with heart failure (HR 1.33; 95% CI, 0.85–2.07; P = 0·208). The relative and absolute effects of the 0.75 mg/day dose of atrasentan on kidney and heart failure events according to baseline eGFR and UACR were evaluated in a post hoc analysis of the SONAR trial. Atrasentan reduced the relative risk of the primary kidney outcome (composite of doubling of serum creatinine, kidney failure, or kidney death) consistently across baseline urine ACR and eGFR subgroups. The absolute risk reduction was greater among patients in the lowest eGFR and highest albuminuria category who were at highest baseline risk. Conversely, the relative and absolute risks of heart failure hospitalization were similar across baseline UACR and eGFR subgroups. (For further discussion of the treatment of diabetic kidney disease, see Chapter 41 )

Distinct from ET receptor antagonism, blockade of ET-1-induced signaling has been explored in the clinical setting with the use of the combined ECE and neprilysin inhibitor daglutril. In an 8-week, endothelin converting enzyme (ECE) crossover design study of participants with type 2 diabetes, blood pressure <140/90 mm Hg, and urinary albumin excretion 20–999 μg/min, daglutril (300 mg/day) did not significantly reduce albuminuria compared with placebo, although blood pressure was reduced. The failure to reach the primary endpoint of albuminuria reduction may relate to concurrent neprilysin inhibition, which may diminish ET-1 degradation. Alternatively, it may reflect an overall diminution of ET-1 with consequent decreased activation of ET-B, as well as ET-A.

Sparsentan is a dual endothelin and angiotensin receptor antagonist. In the phase 3 PROTECT trial, treatment with sparsentan was compared with irbesartan, an angiotensin II receptor blocker, in patients with IgA nephropathy. This double-blind, randomized trial was conducted across 134 clinical sites in 18 countries spanning the Americas, Asia, and Europe. Treatment with sparsentan led to greater proteinuria reduction and better preservation of kidney function than maximally titrated irbesartan. Patients receiving sparsentan experienced a slower decline in eGFR, with a chronic 2-year slope (weeks 6–110) of −2.7 mL/min per 1.73 m ² per year compared with −3.8 mL/min per 1.73 m ² per year in the irbesartan group (difference: 1.1 mL/min per 1.73 m ² per year; 95% CI 0.1–2.1; P = 0.037); the difference in total 2-year slope (day 1–week 110) was–2.9 mL/min per 1.73 m ² per year versus–3.9 mL/min per 1.73 m ² per year (difference: 1.0 mL/min per 1.73 m ² per year, 95% CI–0·03 to 1.94; P = 0·058)

The ALIGN study evaluated the impact of atrasentan, a selective endothelin type A receptor antagonist, on albuminuria in IgA nephropathy using the 24-hour urinary protein-to-creatinine ratio as a surrogate endpoint. In this trial, patients with at least 1 g of daily urinary protein excretion and an eGFR of at least 30 mL/min per 1.73 m ² were randomly assigned to receive either atrasentan (0.75 mg/day) or placebo for 132 weeks. While the long-term effect on kidney disease progression remains unclear, interim results showed a significant reduction in proteinuria with atrasentan compared with placebo at 36 weeks. Ongoing analysis is examining atrasentan’s effect on eGFR.

Administration of the selective ET-A receptor antagonist ambrisentan preserved GFR and prevented the development of albuminuria in a humanized mouse model of sickle cell disease (SCD). However, the kidney-protective effects of ambrisentan were only partially recapitulated by treatment with the combined ET-A/ET-B receptor antagonist A-182086, highlighting the importance of selectively targeting the ET-A receptor in SCD.

A series of studies in pigs provide support for ET-A receptor blockade in the treatment of renovascular disease. In an initial study, investigators treated pigs with unilateral renal artery stenosis with an ET-A receptor antagonist beginning at the onset of renovascular disease and continuing for 6 weeks. In these experiments, researchers observed that ET-A receptor blockade preserved renal hemodynamics, renal function, and microvascular architecture in the stenotic kidney. Similarly, ET-A receptor blockade (but not ET-B receptor blockade) reversed microvascular rarefaction and diminished renal inflammation and fibrosis when it was initiated 6 weeks after the induction of renal artery stenosis. Finally, ET-A receptor blockade also led to an improvement in microvascular density and renal function recovery compared with placebo when it was administered following percutaneous transluminal renal angioplasty/stenting.

ET-1 may play a role in sepsis-mediated acute kidney injury (AKI), although experimental findings have been conflicting, dependent to some extent on the ET receptor antagonist employed. For example, in a rat model of early normotensive endotoxemia, neither an ET-A receptor antagonist nor a combined ET-A/ET-B receptor blockade improved GFR, whereas ET-B receptor blockade alone resulted in a marked reduction in RBF. By contrast, in a porcine model of endotoxemic shock, the dual ET receptor antagonist tezosentan attenuated the decrease in RBF and increase in plasma creatinine. Pointing to a role of ET-1/ET-A signaling in the progression from AKI to CKD, transient unilateral renal ischemia-induced upregulation of ET-1 and ET-A receptor in mice and ET-A receptor antagonism (but not ET-B receptor antagonism) prevented progressive kidney injury.

Urinary ET-1 excretion is correlated with disease activity in persons with systemic lupus erythematosus (SLE), and serum from such patients has been shown to stimulate ET-1 release from endothelial cells in culture. In accordance with a pathogenetic role for the ET system in SLE, the ET-A receptor antagonist FR139317 attenuated kidney injury in a murine model of lupus nephritis. ^Moreover, endothelin receptor antagonism decreased renal immune cell infiltration in mice with SLE.

Sparsentan, a dual ET-A receptor/ARB antagonist, has also been evaluated for the treatment of primary focal and segmental glomerulosclerosis (FSGS). A phase 2 study examined the effect of three different doses of sparsentan (200, 400, and 800 mg/day) when compared with the ARB, irbesartan (300 mg/day), in 96 participants over an 8-week period and reported a mean reduction in proteinuria in sparsentan-treated patients of 45% in comparison with a 19% reduction in those receiving irbesartan. In a subsequent phase 3 trial, 371 participants with FSGS (without known secondary cause) were randomized to treatment with sparsentan or irbesartan for 108 weeks. At a prespecified interim analysis after 36 weeks, a greater proportion of those receiving sparsentan had achieved partial remission of proteinuria than those receiving irbesartan (42% vs. 26%; P = 0.009), but there was no difference in the primary endpoint of GFR slope at 108 weeks.

The Zibotentan Better Renal Scleroderma Outcome Study (ZEBRA) was a three-part phase 2 study (ZEBRA 1, ZEBRA 2A, and ZEBRA 2B) exploring the safety and therapeutic potential of the ET-A receptor antagonist zibotentan in acute and chronic renal complications of scleroderma (NCT02047708). The primary outcome measure was the plasma level of soluble vascular cell adhesion molecule-1 as a biomarker of scleroderma renal involvement. Zibotentan was generally well tolerated. ZEBRA 1 did not show any effect of zibotentan on serum sVCAM-1 but was associated with numerical improvement in eGFR at 26 weeks that was more marked at 52 weeks.

Plasma ET-1 concentrations are increased in individuals with cirrhosis and ascites and in patients with type 2 hepatorenal syndrome (diuretic-resistant or refractory ascites with slowly progressive renal decline) in whom systemic vasodilation accompanies paradoxical renal vasoconstriction. To investigate the therapeutic potential of ET receptor antagonism in this setting, the combined ET-A/ET-B receptor blocker tezosentan was administered to six patients in an early-phase clinical trial. In this study, treatment was discontinued early in five patients, in one case because of systemic hypotension and in four because of concerns about worsening renal function. These adverse effects are consistent with a dose-dependent decline in kidney function in patients with acute heart failure treated with tezosentan, and they highlight the need for caution with the use of ET receptor antagonists in certain patient populations.

A role for ET-1 in the development of preeclampsia is suggested by the observations that infusion of fms-like tyrosine kinase-1 and TNF-α into pregnant rats induced ET-A–dependent hypertension, whereas ET-A receptor antagonism attenuated placental ischemia-induced hypertension in a rat model. Despite the mechanistic role of ET-1 in the pathogenesis of preeclampsia, however, ET receptor antagonists are unlikely to be used in this condition given their known teratogenicity.

ANP is a 28–amino acid peptide comprising a 17–amino acid ring linked by a disulfide bond between two cysteine residues and a COOH-terminal extension that confers its biologic activity ( Fig. 11.7 ). The gene for ANP, NPPA, is found on chromosome 1p36 and encodes the precursor preproANP, which is between 149 and 153 amino acids in length according to the species of origin. Human preproANP consists of 151 amino acids and is rapidly processed to the 126–amino acid proANP. ANP is identical in mammalian species except for a single amino acid substitution at residue 110, which is isoleucine in rat, rabbit, and mouse and methionine in human, pig, dog, sheep, and cow.

Molecular structure of the natriuretic peptides.

ANP, Atrial natriuretic peptide; BNP, brain natriuretic peptide; CNP, C-type natriuretic peptide; DNP, Dendroaspis natriuretic peptide.

Modified from Cea LB. Natriuretic peptide family: new aspects. Curr Med Chem Cardiovasc Hematol Agents. 2005;3:87–98.

ANP synthesis occurs primarily within atrial cardiomyocytes, in which it is stored as proANP, the main constituent of the atrial secretory granules. The major stimulus to ANP release is mechanical stretch of the atria that is secondary to increased wall tension. In addition to atrial stretch, ANP synthesis and release may be stimulated by neurohumoral factors such as glucocorticoids, ET, vasopressin, and angiotensin II, partly through changes in atrial pressure and partly through direct cellular effects. Although ANP mRNA levels are approximately 30- to 50-fold higher in the cardiac atria than in the ventricles, ventricular expression is dramatically increased in the developing heart and in conditions of hemodynamic overload such as heart failure and hypertension. Beyond the heart, ANP has also been demonstrated in the kidneys, brain, lungs, adrenal glands, and liver. In the kidneys, alternate processing of proANP adds four amino acids to the NH ₂ terminus of the ANP peptide to generate a 32–amino acid peptide: proANP 95-126 or urodilatin.

ANP is stored, primarily as proANP, in the secretory granules of the atrial cardiomyocytes and is released by fusion of the granules with the cell surface. During this process, proANP is cleaved to an NH ₂ -terminal 98–amino acid peptide (ANP 1-98) and the COOH-terminal 28–amino acid biologically active fragment (ANP 99-126). Both fragments circulate in the plasma; further processing of the NH ₂ -terminal fragment leads to the generation of peptides ANP 1-30 (long-acting NP), ANP 31-67 (vessel dilator), and ANP 79-98 (kaliuretic peptide), all of whose biologic actions may be similar to those of ANP.

The BNP gene, NPPB, is located only about 8 kb upstream of the ANP gene on the short arm of chromosome 1 in humans, which suggests that the two genes may share both evolutionary origin and transcriptional regulation. By contrast, NPPC, the gene-encoding CNP, is found separately, on chromosome 2. CNP is highly conserved across species; thus it may represent the evolutionary ancestor of ANP and BNP. BNP, like ANP, is synthesized as a preprohormone, between 121 and 134 amino acids in length, according to species of origin. Human preproBNP (134 amino acids) is cleaved to produce the 108–amino acid precursor proBNP. Further processing leads to the production of the 32–amino acid, biologically active BNP (which corresponds to the C terminal of the precursor), as well as a 76–amino acid N-terminal fragment (NT-proBNP). Active BNP, NT-proBNP, and proBNP all circulate in the plasma. Circulating BNP contains the characteristic 17–amino acid ring structure closed by a disulfide bond between two cysteine residues, along with a nine–amino acid N-terminal tail and a six–amino acid C-terminal tail (see Fig. 11.7 ).

The term brain natriuretic peptide is somewhat misleading, given that the primary sites of synthesis of BNP are the cardiac ventricles, and expression also occurs, to a lesser extent, in atrial cardiomyocytes. Like ANP, expression of BNP is regulated by changes in intracardiac pressure and stretch. However, unlike ANP, which is stored and released from secretory granules, BNP is regulated at the gene expression level and is synthesized and secreted in bursts. BNP expression is increased in heart failure, hypertension, and kidney failure. Its plasma half-life is approximately 22 minutes; by contrast, the half-life of circulating ANP is 3 to 5 minutes, and the half-life of the biologically inactive NT-proBNP is 120 minutes. This difference is relevant to the utility of NP measurement as a biologic marker of cardiorenal disease. Changes in pulmonary capillary wedge pressure may be reflected by plasma BNP concentrations every 2 hours and by NT-proBNP levels every 12 hours. ^, The physiologic actions of BNP are similar to those of ANP including effects on the kidneys (natriuresis and diuresis), vasculature (hypotension), endocrine system (inhibition of plasma renin and aldosterone secretion), and the brain (central vasodepressor activity).

As is the case for ANP and BNP, CNP is derived from a prepropeptide that undergoes posttranslational proteolytic cleavage. The initial translation product preproCNP is 126 amino acids in length and is cleaved to produce the 103–amino acid prohormone. Cleavage of proCNP yields two mature peptides made up of 22 and 53 amino acids: CNP and NH ₂ -terminally extended forms of CNP, respectively. Of the 17 amino acids within the CNP ring structure, 11 are identical to those in the other NPs, although, uniquely, CNP lacks an amino tail at the carboxy terminus (see Fig. 11.7 ). Whereas ANP and BNP are ligands for a guanylyl cyclase–coupled receptor, the NPR-A receptor, CNP is a specific ligand for the NPR-B receptor. CNP primarily functions in an autocrine/paracrine manner with effects on vascular tone and muscle cell growth. Expression of the CNP gene by the endothelial cells, the presence of CNP receptors on vascular smooth muscle cells, and the antiproliferative effect of CNP on vascular smooth muscle cells suggest that CNP is produced by the endothelium and acts on adjacent cells, serving as an autocrine/paracrine endothelium-derived locally active vasoregulatory system. Accordingly, plasma concentrations of CNP are low, although they are increased in the conditions of heart failure and kidney failure. CNP is present in the heart, kidneys, and endothelium, and its receptor is also expressed in abundance in the hypothalamus and pituitary gland, which suggests that the peptide may also play a role as a neuromodulator or neurotransmitter. Regulation of CNP expression is distinct from that of ANP and BNP and is controlled by a number of vasoactive mediators including insulin, vascular endothelial growth factor, TGF-β, TNF-α, and IL-1β.

The principal enzymes responsible for the conversion of proANP, proBNP, and proCNP to their active forms are the serine proteases, corin and furin. Corin converts proANP to ANP, furin converts proCNP to CNP, and both corin and furin cleave proBNP. Corin is highly expressed in the heart and to a lesser extent in the kidney and it is the rate-limiting enzyme in ANP activation. In response to pressure overload, corin-deficient mice develop hypertension together with cardiac hypertrophy and dysfunction. ^, In the kidneys, corin colocalizes with ANP and decreased urinary corin excretion has been observed in patients with CKD. Interestingly, studies combining observations made in organ-specific corin-deficient mice together with human correlative experiments have identified a role for impaired uterine corin/ANP function in the pathogenesis of preeclampsia.

The physiologic role of DNP has been controversial since its original identification in the venom of the Green Mamba snake, D. angusticeps, in 1992. ^, DNP is a 38–amino acid peptide that shares the 17–amino acid ring structure common to all NPs, except that it has unique N- and C-terminal regions (see Fig. 11.7 ). Immunoreactivity for DNP has been reported in human plasma and atrial myocardium, and DNP has also been described in rat and rabbit kidneys, rat colon, rat aortic vascular smooth muscle cells, and pig ovarian granulosa cells. DNP binds to NPR-A ³⁰⁵ and the clearance receptor NPR-C, which may be of particular relevance in view of the peptide’s apparent resistance to enzymatic degradation. In dogs, either under normal conditions or in a pacing-induced heart failure model, administration of synthetic DNP decreased cardiac filling pressures; increased GFR, natriuresis, and diuresis; and lowered blood pressure, suppressing renin release and increasing plasma and urine cGMP levels. ^, Despite these propitious findings, several aspects of the biologic role of DNP remain contentious. In particular, the gene for the peptide has not been identified in mammals and the fractionation of DNP from human samples has not been reported. These uncertainties have led some authors to question whether DNP is, in fact, expressed at all in humans.

Urodilatin is a structural homolog of ANP that shares the same 17–amino acid ring structure and COOH-terminal tail. It is synthesized in renal distal tubule cells and differentially processed to a 32–amino acid NH ₂ -terminally extended form of ANP. Urodilatin is not found in plasma; instead, it acts in a paracrine manner within the kidneys on receptors in the glomeruli and IMCDs to promote natriuresis and diuresis. Urodilatin is upregulated in diabetic animals and in the remnant kidney and is relatively resistant to enzymatic degradation, which may explain its more potent renal effects.

The natriuretic and diuretic actions of the NPs are consequences of both vasomotor effects and direct effects on the renal tubule. Both ANP and BNP cause an increase in glomerular capillary hydrostatic pressure and a rise in GFR by inducing afferent arteriolar vasodilation and efferent arteriolar vasoconstriction. These contrasting effects of the NPs on the afferent and efferent arterioles differ from the actions of classical vasodilators such as bradykinin. In addition to direct effects on vascular tone, ANP can increase GFR through cGMP-mediated mesangial cell relaxation and consequent changes in the ultrafiltration coefficient. Plasma levels of ANP that do not increase GFR can induce natriuresis, indicating the potential for direct tubule effects, which may involve either locally produced NPs acting in a paracrine manner, such as urodilatin, or circulating NPs. A number of mechanisms may be responsible for the natriuresis including direct effects on sodium transport in tubule epithelial cells and indirect effects through inhibition of renin secretion after increased sodium delivery to the macula densa.

NPs also antagonize vasopressin in the cortical collecting ducts. Similar mechanisms probably underlie the response to ANP, BNP, and urodilatin. By contrast, CNP has little natriuretic or diuretic effect, which may indicate a requirement for the presence of the C-terminal extension of the peptide for renal effects. The NPs may have antifibrotic effects within the kidneys, as evidenced by an increase in renal fibrosis in NPR-A–knockout mice after unilateral ureteric obstruction. In cultured proximal tubule cells, ANP attenuates high glucose–induced activation of TGF-β ₁ , Smad, and collagen synthesis, which illustrates the potentially antifibrotic properties of the peptide in the context of diabetic nephropathy.

All NPs have vasodilatory and hypotensive properties. Heterozygous mutant mice with a disrupted proANP gene display evidence of salt-sensitive hypertension, whereas hypotension is a feature of transgenic mice overexpressing ANP. In a human patient population, a variant in the ANP promoter was associated with both lower levels of plasma ANP and increased susceptibility to early development of hypertension. However, infusion of high concentrations of ANP can actually induce a rise in blood pressure, which suggests that counterregulatory baroreceptors may be activated.

ANP lowers blood pressure through two major direct mechanisms. First, it increases vascular permeability with a shift of fluid from the intravascular to extravascular compartments by capillary hydraulic pressure. Second, ANP increases venous capacitance and lowers preload. In addition, ANP and BNP antagonize the vasoconstrictive effects of the RAS, ET, and the sympathetic nervous system by decreasing sympathetic peripheral vascular tone, thereby suppressing the release of catecholamines and reducing central sympathetic outflow. By lowering the activation threshold of vagal afferents, ANP prevents vasoconstriction and tachycardia that normally follow a reduction in preload and thereby produces a sustained drop in blood pressure. CNP is a more potent vasodilator than either ANP or BNP. In fact, CNP relaxes human subcutaneous resistance arteries, whereas ANP and BNP have no effect.

NPs have a number of other effects on the cardiovascular system distinct from their action on vasomotor tone. For example, NPs play a major role in cardiac remodeling. Mice with genetic deficiencies of ANP exhibit an increase in cardiac mass, whereas heart size is diminished in mice transgenically overexpressing ANP. The antimitogenic and antitrophic effects of NPs, which appear to be mediated by cGMP, have also been demonstrated in a range of cultured cell types, including cultured vascular cells, fibroblasts, and myocytes, and in vivo in response to balloon angioplasty. Further evidence for the role of ANP in mediating cardiac hypertrophy was obtained from population studies, in which variants in either the NPPA promoter (associated with reduced circulating ANP) or the NPR-A gene, NPR1, have been associated with left ventricular hypertrophy. ^, BNP has been shown to have antifibrotic properties within the heart. In vitro, BNP antagonizes TGF-β–induced fibrosis in cardiac fibroblasts, and in vivo, targeted genetic disruption of BNP in mice is associated with an increase in cardiac fibrosis, in the absence of either hypertension or ventricular hypertrophy.

Cardiac CNP is increased in heart failure, where it may play a role in ventricular remodeling. Comparison of plasma CNP levels in samples taken from the aorta and renal vein, at the time of diagnostic heart catheterization, has demonstrated that CNP is indeed synthesized and secreted by the kidney. Moreover, this effect was found to be blunted in patients with heart failure, potentially contributing to renal sodium retention. In rats subjected to unilateral ureteric obstruction, recombinant CNP decreased blood urea nitrogen and creatinine levels and attenuated renal fibrosis.

Even though they do not cross the blood-brain barrier, NPs exert important CNS effects that may augment their peripheral actions. ANP, BNP, and particularly CNP are all expressed within the brain. Circulating NPs may also exert central effects through actions at sites that are outside the blood-brain barrier. The NPR-B receptor is expressed throughout the CNS, which reflects the wide distribution of CNP, whereas the NPR-A receptor is expressed in areas adjacent to the third ventricle, which is indicative of a role of peripherally circulating ANP and BNP, as well as centrally expressed peptides. Complementing their natriuretic and diuretic effects, NPs inhibit both salt appetite and water drinking. ANP also prevents release of vasopressin and possibly adrenocorticotropic hormone from the pituitary gland, whereas sympathetic tone is increased by the actions of the NPs on the brainstem.

Clinical and experimental evidence suggests that NPs play a role in mediating metabolism. Circulating levels of NPs are decreased in obese individuals and among patients with the metabolic syndrome, ^, correlating inversely with both plasma glucose and fasting insulin levels. In accordance with these epidemiologic observations, infusion of ANP activates hormone-sensitive lipase from fat cells, which is indicative of lipolysis. In vitro, ANP inhibits preadipocyte proliferation, the lipolytic properties of the peptide being mediated by cGMP phosphorylation. ^,

Knockout mouse studies have revealed that CNP plays a predominant role in the regulation of skeletal growth, specifically cartilage homeostasis and endochondral bone formation. Mice with genetic deficiencies of either CNP or its receptor NPR-B lack growth of longitudinal bones and vertebrae and have a shortened life span as a consequence of respiratory insufficiency secondary to abnormal ossification of the skull and vertebrae. ^, Transgenic mice that overexpress CNP are relatively protected from glucocorticoid-induced growth retardation. Mutations in the NPR-B gene have also been reported in patients with autosomal recessive skeletal dysplasia and acromesomelic dysplasia–type Maroteaux and obligate carriers of the mutations have heights that are below predicted levels. Accordingly, CNP analog therapy is being investigated as a possible treatment for achondroplasia.

Measurement of circulating levels of either BNP or NT-proBNP has effectively helped guide clinical practice in several aspects of the management of heart failure including diagnosis, screening, prognosis, and monitoring of therapy. The primary role of BNP measurement in the assessment of dyspnea is as a “ruling out” test: A plasma BNP level lower than 100 pg/mL has a negative predictive value for heart failure of 90%. In the ProBNP Investigation of Dyspnea in the Emergency Department (PRIDE) study, an NT-proBNP level lower than 300 pg/mL was optimal in ruling out heart failure, with a negative predictive value of 99%. Screening for BNP and NT-BNP levels has also proven useful in identifying individuals at risk for heart failure for whom aggressive medical therapy should be targeted. The utility of serial BNP or NT-BNP measurements in guiding the treatment of patients with heart failure was the subject of a 2016 Cochrane Review. This review concluded that low-quality evidence shows that NP-guided treatment is associated with a reduction in hospital admissions for heart failure and that low-quality evidence shows uncertainty with respect to the effect of NP-guided treatment on mortality or all-cause hospital admissions. In the interpretation of plasma levels of BNP and NT-proBNP, a number of other biologic variables should be taken into account. NP levels rise with age and are higher in women, the latter effect possibly secondary to estrogen regulation, inasmuch as hormone replacement therapy increases BNP levels. Conversely, NP levels fall with increasing obesity. Although BNP levels of heart failure patients are higher in Asian or African American patients than in Caucasian or Hispanic patients, they provide prognostic value regardless of race or ethnicity.

The interpretation of NP concentrations in patients with kidney disease merits special consideration. NP levels are increased in individuals with impaired kidney function. This increase is probably multifactorial in origin and not solely the consequence of increased intravascular volume. Other factors that contribute to increased NP levels include decreased NP responsiveness, subclinical ventricular dysfunction, hypertension, left ventricular hypertrophy, subclinical ischemia, myocardial fibrosis, and RAS activation, as well as decreased filtration and reduced clearance by NPR-C and NEP. Although, on the basis of observational studies, it has been widely considered that renal clearance plays a greater role in the removal of NT-proBNP from the circulation than removal of BNP, one study has challenged this view. By measuring both NT-proBNP and BNP in the renal arteries and veins of 165 subjects undergoing renal arteriography, investigators found that both NT-proBNP and BNP are equally dependent on renal clearance. However, the NT-proBNP-to-BNP ratio did increase with declining GFR, which suggests that the two peptides may be differentially cleared at GFRs lower than 30 mL/min/1.73 m ² .

Even though both BNP and NT-proBNP are affected by kidney impairment, their clinical utility for the prediction of heart failure persists in CKD patients in the context of appropriately adjusted reference ranges. For example, in the Breathing Not Properly study, BNP cut point values were approximately threefold higher to diagnose heart failure in patients with an estimated GFR lower than 60 mL/min relative to the conventional cut point value of 100 pg/mL. In a cohort of 831 patients with dyspnea and a GFR <60 mL/min, both BNP and NT-proBNP were effective predictors of heart failure, although NT-proBNP was superior in predicting mortality. In asymptomatic patients with CKD, both BNP and NT-proBNP were equivalent and effective in indicating the presence of left ventricular hypertrophy or coronary artery disease. In patients with CKD, BNP and NT-proBNP may be predictive of the progression of kidney function decline ^, and cardiovascular disease and mortality. In a nondialysis CKD population, NT-proBNP, but not BNP, was an independent predictor of death; in 994 black patients with hypertensive renal disease (GFR = 20 to 65 mL/min/1.73 m ² ), NT-proBNP was predictive of cardiovascular disease and mortality, particularly among individuals with proteinuria. In pediatric CKD patients, both BNP and proBNP (but not troponins I and T) were indicative of left ventricular hypertrophy or dysfunction.

BNP and NT-proBNP have been studied extensively in dialysis recipients as both prognostic indicators and markers of volume status. The molecular weights of BNP (3.5 kDa) and NT-proBNP (8.35 kDa) are low enough that both peptides may be cleared by high-flux dialysis. ^, Nevertheless, in contrast to ANP, which falls sharply after either hemodialysis or peritoneal dialysis, levels of BNP and NT-proBNP are less affected. ^, The role of NP levels as indicators of volume status in either hemodialysis or peritoneal dialysis recipients is confounded by the common coexistence of left ventricular abnormalities. ^{, ,} Both BNP and NT-proBNP levels are predictive of mortality, heart failure, and coronary artery disease in the population undergoing dialysis. However, no definite cut point values for diagnosing heart failure in dialysis patients have been defined.

Although CNP usually functions in a paracrine manner, its presence in the plasma may provide utility as a biomarker of cardiovascular risk. In a study of 1841 individuals from the general population, individuals with plasma CNP levels in the highest quartile were at increased risk of myocardial infarction and, unlike BNP levels, plasma CNP levels were unaffected by sex and only weakly associated with age.

ANP has a short half-life and a high total body clearance. Its intravenous administration causes a reduction in blood pressure, diuresis, and natriuresis in healthy individuals; this response is reduced in the setting of acute heart failure. In a 6-year open-label study of 3777 patients with acute heart failure treated with carperitide, clinical improvement was reported in 82%. Whereas early experimental studies were suggestive of a potential benefit of exogenous ANP in AKI, results in patients have generally been disappointing. Nevertheless, the peptide may have a limited role in selected patient populations. For example, low-dose carperitide preserved kidney function in patients undergoing repair of abdominal aortic aneurysm and reduced the incidence of contrast-induced nephropathy in patients after coronary angiography. However, a meta-analysis suggested that recombinant ANP has no effect on mortality in patients with AKI, although a trend toward a reduction in the need for kidney replacement therapy was shown. In a separate meta-analysis of studies conducted in cardiovascular surgery patients, ANP infusion decreased peak serum creatinine concentration, incidence of arrhythmia, and need for kidney replacement therapy, whereas both ANP and BNP decreased the length of intensive care unit and hospital stay. Among 367 high-risk individuals undergoing coronary artery bypass grafting (CABG), recombinant ANP decreased the incidence of major adverse cardiovascular and cerebrovascular events and the need for dialysis, immediately and up to 2 years postoperatively, although survival was unaffected. Similarly, among CKD patients undergoing CABG, those receiving recombinant ANP experienced a smaller rise in serum creatinine, fewer cardiac events, and lower requirement for dialysis, although mortality did not differ from those that did not receive ANP. However, when employed in an effort to treat rather than prevent AKI following cardiac surgery, recombinant ANP had no significant effect on kidney function, the need for kidney replacement, length of stay, or medical costs.

Nesiritide is recombinant human BNP, manufactured from Escherichia coli and identical in structure to native human BNP, with a mean terminal half-life of 18 minutes in patients with heart failure. Intravenous administration of nesiritide lowers pulmonary and systemic vascular resistance, decreases right atrial pressure, and increases cardiac output (presumably through effects on ventricular afterload) in a concentration-dependent fashion. In the kidneys, nesiritide increases RBF and GFR through both direct vasodilatory effects and indirect effects on cardiac output and norepinephrine inhibition. Diuresis and natriuresis may also occur, although these effects are modest and may not be seen at the approved doses. Additional effects of nesiritide may also include inhibition of renin secretion in the kidneys and aldosterone production in the heart and adrenal glands.

In response to meta-analysis data suggesting that nesiritide treatment may be associated with a worsening of renal function and an increase in the rate of early death, ^, the Acute Study of Clinical Effectiveness of Nesiritide in Decompensated Heart Failure (ASCEND-HF) trial was initiated. In this study of 7141 patients hospitalized with acute heart failure, nesiritide neither increased nor decreased the rate of death or rehospitalization, rates of worsening kidney function were unaffected, and there was a small but nonsignificant improvement in self-reported rates of dyspnea. On the basis of these results, the investigators concluded that nesiritide cannot be recommended for routine use in the broad population of patients with acute heart failure.

The effects of urodilatin (ularitide) have been assessed in both heart failure and AKI. However, the diuretic effect of urodilatin appears to be attenuated in heart failure patients, which reflects a blunted response, as observed for ANP and BNP. ^, Similarly, as with ANP and BNP, hypotension appears to be a dose-limiting side effect of ularitide therapy. ^, In the Safety and Efficacy of an Intravenous Placebo-Controlled Randomized Infusion of Ularitide in a Prospective Double-blind Study in Patients with Symptomatic, Decompensated Chronic Heart Failure (SIRIUS II) study, a phase II trial of 221 patients hospitalized for decompensated heart failure, a single 24-hour infusion of ularitide preserved short-term renal function. The NP vessel dilator may offer theoretical advantages for the treatment of acute decompensated heart failure in comparison with current NP-based therapies. ^, In particular, vessel dilator may produce a greater and more sustained natriuresis than does ANP or BNP, without a blunted response in patients with heart failure, and may also improve renal function in the setting of experimental acute renal injury. An alternative therapeutic approach is the development of novel chimeric peptides. For example, researchers have synthesized a peptide (cenderitide) that represents fusion of the 22–amino acid peptide CNP together with the 15–amino acid linear C terminus of DNP. In vitro, this peptide activates cGMP and attenuates cardiac fibroblast proliferation. In vivo, cenderitide is both natriuretic and diuretic and increases GFR with less hypotension than does BNP. ^, Cenderitide is more resistant to degradation by neprilysin than the naturally occurring NPs and is eightfold more potent in inducing glomerular cGMP production than CNP.

Notwithstanding concerns regarding the efficacy and cost-effectiveness of recombinant NP therapy, a major limitation is the requirement for systemic administration, which is unsuitable for chronic treatment. Alternative methods to increase the biologic activity of NPs may offer a more feasible approach for chronic therapy. In particular, inhibition of the enzymatic degradation of NPs by neprilysin has been the focus of drug discovery efforts for a number of years. Neprilysin is a zinc metallopeptidase with catalytic similarity to ACE and with a wide tissue distribution, although abundant at the proximal tubule brush border. Several pharmacologic neprilysin inhibitors have been investigated (e.g., candoxatril, thiorphan, and phosphoramidon). Although these agents, in general, lead to an increase in plasma levels of the NPs and, under some experimental conditions, induce natriuresis and diuresis with peripheral vasodilation, results of clinical trials in hypertension and heart failure have generally been disappointing. Specifically, sustained antihypertensive effects have not been demonstrated, and some researchers have reported a paradoxical rise in blood pressure. This may be a consequence of the induction of both neprilysin and ACE expression with neprilysin inhibition and a consequent increase in angiotensin II levels. The biologic actions of the NPs are, however, restored in the presence of an inhibited RAS and this has led to the development of two classes of agents: 1. vasopeptidase inhibitors that inhibit both neprilysin and ACE and 2. combined angiotensin receptor blockade/neprilysin inhibition, the latter having gained regulatory authority approval for the treatment of heart failure.

The rational design of vasopeptidase inhibitors—such as mixanpril (S21402), CGS30440, aladotril, MDL 100173, sampatrilat, and omapatrilat—was made possible because of the similar structural characteristics of the catalytic sites of both neprilysin and ACE. Despite the theoretical advantages of vasopeptidase inhibitors, phase III clinical studies have not been able to demonstrate superiority of vasopeptidase inhibition over ACE inhibition, and an increase in the incidence of angioedema has raised safety concerns. For instance, in the Omapatrilat Cardiovascular Treatment vs. Enalapril (OCTAVE) trial of 25,302 hypertensive patients, angioedema occurred in 2.17% of omapatrilat-treated patients, in comparison with 0.68% of patients treated with the ACE inhibitor enalapril. In the Omapatrilat Versus Enalapril Randomized Trial of Utility in Reducing Events (OVERTURE) study, the incidence of angioedema was, again, increased among subjects receiving omapatrilat in comparison with those receiving enalapril (0.8% vs. 0.5%). The increased angioedema with vasopeptidase inhibition is likely a consequence of decreased degradation of bradykinin and substance P with combined inhibition of the two metallopeptidases.

The rationale that concurrent RAS blockade may potentiate the therapeutic effects of neprilysin inhibition, yet concurrent ACE inhibition increases the risk of angioedema, encouraged the development of a new class of drug that combines an ARB and neprilysin inhibitor. This class has been termed angiotensin receptor–neprilysin inhibitor (ARNi). In July 2015, the first in the class, valsartan/sacubitril gained approval from the U.S. Food and Drug Administration for the treatment of heart failure with reduced ejection fraction. Valsartan/sacubitril (LCZ696) is a single molecule composed of molecular moieties of the ARB, valsartan, and the neprilysin inhibitor prodrug sacubitril (formerly AHU-377) in a 1:1 ratio. After ingestion, valsartan/sacubitril dissociates into valsartan and sacubitril and sacubitril is subsequently converted to its active form sacubitrilat (LBQ657) by esterases. In a study of 1328 patients, valsartan/sacubitril conferred greater blood pressure lowering than valsartan alone with no cases of angioedema reported. In the Prospective comparison of ARNi with ARB on Management Of heart failUre with preserved ejectioN fracTion (PARAMOUNT) study of 301 individuals with heart failure with preserved ejection fraction, valsartan/sacubitril lowered NT-proBNP levels to a greater extent than valsartan after 12 weeks of treatment and was well tolerated. Although the reduction in NT-proBNP was sustained at 36 weeks, the difference in NT-proBNP levels between participants randomized to valsartan/sacubitril and valsartan was no longer significant. However, left atrial remodeling and heart failure symptoms were improved.

The case for regulatory authority approval for valsartan/sacubitril was based on the findings of the phase III prospective comparison of ARNi with ACEi (Determine Impact on Global Mortality and Morbidity in Heart Failure [PARADIGM-HF] trial). PARADIGM-HF compared the effects of valsartan/sacubitril (200 mg twice daily) and enalapril (10 mg twice daily) in 8442 patients with heart failure (New York Heart Association class II–IV) and a reduced ejection fraction (≤40%). The primary outcome, a composite of cardiovascular death and hospitalization for heart failure, occurred in 21.8% of participants treated with valsartan/sacubitril and 26.5% of participants treated with enalapril (hazard ratio 0.80, confidence interval 0.73–0.87, P <.001). Hypotension was more common in participants receiving valsartan/sacubitril, whereas cough, hyperkalemia, and kidney impairment were more common in those receiving enalapril. Importantly, there was no significant difference in the number of cases of angioedema between participants receiving valsartan/sacubitril than those receiving enalapril, although the number of cases of angioedema was numerically higher in the valsartan/sacubitril group (19 vs. 10, P = 0.13). The effect of valsartan/sacubitril in patients with heart failure and preserved ejection fraction was evaluated in the Prospective Comparison of ARNI with ARB Global Outcomes in Heart Failure with preserved Ejection Fraction (PARAGON-HF) trial (NCT01920711). There was no statistically significant reduction in the primary endpoint of hospital admission with heart failure or cardiovascular death (rate ratio 0.87; 95% CI, 0.75–1.01; P = 0.06). Other molecules combining ARB and NEP inhibition are also currently under development. ^,

Despite the promising findings of PARADIGM-HF, there are some difficulties with the design of the study and theoretical considerations around the use of neprilysin inhibitors in a broad population. In terms of the active comparator in PARADIGM-HF, it is noteworthy that valsartan/sacubitril was not compared with valsartan alone and that the dose of enalapril (10 mg twice daily) may have been insufficient. For instance, in the Cooperative North Scandinavian Enalapril Survival Study (CONSENSUS) of patients with heart failure, the target dose of enalapril was up to 20 mg twice daily. Separately, whether the apparently low risk of angioedema observed in PARADIGM-HF translates to the real-world setting remains to be determined. Only 5% of participants in PARADIGM-HF were Black, a group at increased risk of angioedema associated with ACE inhibition or neprilysin inhibition. Furthermore, 78% of participants in PARADIGM-HF had been previously treated with an ACE inhibitor and inclusion of a run-in period where all participants were exposed to enalapril may have resulted in an underrepresentation in the number of cases of angioedema. Other theoretical risk concerns surrounding chronic neprilysin inhibition largely relate to the other substrates that are normally degraded by neprilysin (see Table 11.2 ). On the basis of the breadth of activity of these substrates, the possibility has been raised that long-term neprilysin inhibition could have deleterious effects on bronchial reactivity, pain, inflammation, tumorigenesis, and neuronal function. Of particular note has been the recognition that neprilysin is important in the metabolism of amyloid-β peptides and that its inhibition could predispose to the development of Alzheimer disease, age-related macular degeneration, and cerebral amyloid angiopathy, which may take many years to manifest. Reassuringly, the PERSPECTIVE trial (NCT02884206) compared the efficacy and safety of valsartan/sacubitril with valsartan on cognitive function in patients with heart failure and preserved ejection fraction and showed that 3-year neprilysin inhibition treatment was not associated with increased Aβ accumulation, determined by PET, or with cognitive deterioration, which was reassuring. Despite these theoretical concerns, valsartan/sacubitril treatment is now recommended in clinical guidelines for the treatment of heart failure with reduced ejection fraction.

The renal effects of valsartan/sacubitril were assessed in a post hoc analysis of participants in the PARAMOUNT trial of individuals with heart failure and preserved ejection fraction. Participants received treatment with valsartan/sacubitril titrated to 200 mg twice daily or valsartan titrated to 160 mg twice daily, which each gave similar levels of systemic exposure to valsartan. ^, In the PARAMOUNT trial, eGFR declined less with valsartan/sacubitril treatment than it did with valsartan treatment over a 36-week period (–1.5 vs.–5.2 mL/min/1.73 m ² , P =.002), whereas the geometric mean of the urinary albumin-to-creatinine ratio increased from baseline in the valsartan/sacubitril group (2.4–2.9 mg/mmol) and was unchanged in the valsartan group (2.1–2.0 mg/mmol; P value for difference between groups =.016). The finding of relative preservation in eGFR with valsartan/sacubitril is consistent with the observation from the PARADIGM-HF trial that valsartan/sacubitril-treated patients experienced less kidney impairment that necessitated cessation of therapy than enalapril-treated patients. Both the effects on eGFR and albuminuria are reminiscent of the effect of systemic ANP administration, raising the possibility that they are a consequence of increased levels of biologically active ANP with neprilysin inhibition. Participants in the PARAMOUNT trial were required to have an eGFR of at least 30 mL/min/1.73 m ² at enrollment, and thus the effects of combination ARB and neprilysin inhibition in more advanced renal disease are currently unknown, as are the effects on hard renal endpoints in at-risk populations.

The renal effects of valsartan/sacubitril have also been evaluated in persons with CKD in the absence of heart failure. In the U.K.’s Heart and Renal Protection (HARP) III trial, 414 participants with eGFR 60 to 45 mL/min/1.73 m ² and UACR >177 mg/g (>20 mg/mmol) or eGFR 44 to 20 mL/min/1.73 m ² regardless of albuminuria were randomized to treatment with valsartan/sacubitril (mean baseline measured 34 GFR mL/min/1.73 m ² or irbesartan (meaan baseline measured GFR 34.7 mL/min/1.73 m ² ). After 12 months, there was no difference in the primary outcome of measured GFR or secondary outcome of UACR. There was also no difference in adverse events or incidence of severe hyperkalemia.

While angiotensin receptor–neprilysin inhibitors (ARNIs) have become increasingly important in managing heart failure (HF) in patients with preserved kidney function, evidence regarding their benefits in HF patients with end-stage kidney disease (ESKD) on dialysis remains limited to two studies: a retrospective study showing improvements in left ventricular ejection fraction and a clinical trial demonstrating enhanced left ventricular echocardiographic parameters after 1 year of therapy. As of 2024, no randomized studies have compared ARNI with ACE inhibitors or ARBs in patients with CKD and an eGFR <30 mL/min per 1.73 m ² , regardless of heart failure status.

The existence of intestinal NPs has been suggested by initial observations that sodium excretion is greater after an oral salt load than after an intravenous salt load. ^, These intestinal peptides include guanylin and uroguanylin. However, a different study of 15 healthy volunteers found that sodium excretion was similar in response to either oral or intravenous sodium load during either a low- or high-sodium–containing diet. Moreover, serum concentrations of either prouroguanylin or proguanylin were unchanged following either oral or intravenous sodium load and showed no correlation with sodium excretion. Collectively, these observations challenge the notion of a gastrointestinal–renal natriuretic axis mediated by the guanylin peptide family. ^, It thus appears likely that the natriuretic, kaliuretic, and diuretic effects of guanylin and uroguanylin, which occur without change in GFR or RBF, are mediated by local production of the peptides within the kidney.

Adrenomedullin is a 52–amino acid peptide originally isolated from human pheochromocytoma cells, although it is synthesized mainly by vascular smooth muscle cells, endothelial cells, and macrophages and is present in the plasma, vasculature, lungs, heart, and adipose tissue. The peptide is upregulated in patients with cardiovascular disease and has positive inotropic and vasodilatory properties. Systemic administration of adrenomedullin induces an NO-dependent natriuresis and an increase in GFR both under normal conditions and in patients with congestive heart failure; it also decreases plasma aldosterone levels without affecting renin activity. Individuals with type 2 diabetes and plasma levels of midregional proadrenomedullin (MR-proADM) peptide in the highest tertile are at an increased risk of severe nephropathy (doubling of plasma creatinine and/or end-stage renal disease), which may reflect a reactive rise in MR-proADM.

Humans possess a single kininogen gene, KNG1, which is localized to chromosome 3q26 and encodes both high-molecular-weight (HMW) kininogens (626 amino acids, 88–120 kDa) and low-molecular-weight (LMW) kininogens (409 amino acids, 50–68 kDa) through alternate splicing from 11 exons spread over a 27-kb genomic region. A second kininogen gene has been identified in mice. In humans, kininogen deficiency may be relatively asymptomatic; the kininogen-deficient Brown Norway Katholiek rat strain, however, shows increased sensitivity to the pressor effects of salt, angiotensin II, and mineralocorticoid. ^,

HMW and LMW kininogen are cleaved by the serine protease kallikrein. The name “kallikrein” is derived from the Greek term kallikreas, meaning “pancreas,” after the work of Frey and colleagues, in the 1930s, who extracted a kinin-producing enzyme from the pancreas of dogs. Since then, 15 tissue kallikreins have been identified, although, in humans, only one (KLK1) is involved in local kinin production. The human kallikrein genes are clustered on chromosome 19 at loci q13.3-13.4. Plasma kallikrein is found in the circulation and is involved largely with the coagulation cascade and activation of neutrophils. The tissue kallikreins are acid glycoproteins that are variably and extensively glycosylated. Human renal kallikrein is synthesized as a zymogen (prekallikrein) with a 17–amino acid signal peptide and a 7–amino acid activation sequence, which must be cleaved to activate the enzyme. In most mammals including humans, tissue kallikrein cleaves kallidin (lys-bradykinin) from kininogens, whereas plasma kallikrein releases bradykinin.

Although the physiologic effects of kallikrein have been attributed to increased kinin generation, the enzyme may also have direct effects on the B2R, as well as actions independent of the kinin receptors. ^, For example, in kininogen-deficient Brown Norway Katholiek rats, local injection of kallikrein into the myocardium after coronary artery ligation had a cardioprotective effect that was abolished by the NO synthase inhibitor Nω-nitro- l -arginine methyl ester and the selective B2R inhibitor icatibant (Hoe 140). As a serine protease, kallikrein may also elicit kinin receptor–independent effects on endothelial cell migration and survival through cleavage of growth factors and matrix metalloproteinases. Transgenic mice overexpressing human kallikrein exhibit a sustained reduction in systemic blood pressure throughout their life span, which is indicative of the lack of sufficient compensatory mechanisms to reverse the hypotensive effect of kallikrein. In humans, polymorphisms of the kallikrein gene KLK1 or its promoter can impair enzymatic activity, potentially influencing both kinin-dependent and kinin-independent effects. Among normotensive men with a common loss-of-function KLK1 polymorphism (R53H), an increase in wall shear stress and a paradoxical reduction in artery diameter and lumen were noted, although flow-mediated and endothelium-independent vasodilation were unaffected.

The kinins are bradykinin and kallidin in humans and bradykinin and kallidin-like peptide in rodents. Plasma aminopeptidase can convert kallidin (10 amino acids: Lys-Arg-Pro-Pro-Gly-Phe-Ser-Pro-Phe-Arg) to bradykinin (9 amino acids: Arg-Pro-Pro-Gly-Phe-Ser-Pro-Phe-Arg) by cleavage of the first N-terminal lysine residue. Cleavage of the carboxy-terminal arginine residue by kininase I (carboxypeptidase-N) and carboxypeptidase-M generates their des-Arg derivatives, which are agonists of the B1R. Removal of two C-terminal amino acids (Phe and Arg) by ACE (kininase II), neprilysin, or ECE is responsible for inactivation of the peptides.

B1R and B2R share 36% homology, and both are GPCRs with seven-transmembrane domains. The genes for the two receptors are in tandem on a compact locus (14q23) separated by only 12 kb. The B2R is the principal receptor mediating the actions of both kinins, is expressed in abundance by vascular endothelial cells and is present in most tissues including those of the kidneys, heart, skeletal muscle, CNS, vas deferens, trachea, intestines, uterus, and bladder. In general, the distribution and action of B1Rs are similar to those of the B2Rs. The B1R, by contrast, is expressed at low levels under normal conditions but is upregulated in response to inflammatory stimuli (e.g., lipopolysaccharide, endotoxins, cytokines such as IL-1β and TNF-α) and in the setting of diabetes and ischemia-reperfusion injury. B2R binds both bradykinin and kallidin, whereas bradykinin has almost no effect at the B1R. The carboxypeptidase required to generate the des-Arg B1R-active kinin fragments is closely associated with the B1R on the cell surface. This association would enable B2R agonists to rapidly activate B1Rs, particularly in response to inflammation.

Ligand binding of both receptor subtypes induces activation of phospholipase C, which results in intracellular calcium mobilization through production of inositol 1,4,5-triphosphate and DAG via activation of G proteins including Ga _q and Ga _i . The physiologic effects of bradykinin receptor activation are mediated through generation of both endothelial NO synthase–derived NO and prostaglandins. B2R activation leads to a rise in intracellular calcium concentrations in vascular endothelial cells. However, bradykinin-induced vasodilation is not abolished by coadministration of NO synthase and COX inhibitors, which indicates that additional effectors are also likely to be involved, possibly an endothelium-derived hyperpolarizing factor. In addition, through binding to both B1R and B2R, bradykinin also increases the expression of inducible NO synthase (iNOS), at least in rodents. It is difficult to induce the iNOS gene in human tissues, especially the vascular endothelium. Mice that have genetic deficiencies of B2R, B1R, or both receptors have been generated; the reported phenotypes of the different knockout strains have been varied, which may be a result of different genetic backgrounds, or, in the case of the single knockouts, differing compensatory effects of the remaining receptor. For example, some studies of B2R-deficient mice revealed an increase in resting systemic blood pressure, an exaggerated pressor response to angiotensin II and salt sensitivity, whereas others revealed no difference in resting blood pressure between B2R- or B1R-deficient mice and wild-type animals. ^, Double B2R-/B1R-knockout mice were also reported to have resting blood pressure identical to that in wild-type mice and were resistant to lipopolysaccharide-induced hypotension. ^, By contrast, transgenic mice expressing the human B2R had a lower resting blood pressure than did wild-type controls. Transgenic mice expressing the rat B1R (as well as their native murine B2R) were normotensive but showed an exaggerated hypotensive response to lipopolysaccharide and, unexpectedly, a hypertensive response to des-Arg bradykinin.

Kallistatin is an endogenous serpin inhibitor of kallikrein that acts by forming a heat-stable complex with the enzyme. Surprisingly, administration of human kallistatin to rodents induced vasodilation and a decline in systemic blood pressure, which was unaltered by either an NO synthase inhibitor or the B2R antagonist icatibant; this suggests that the vasodilatory properties of kallistatin may be mediated through a smooth muscle mechanism independent of bradykinin receptor activation.

With the exception of the metabolites des-Arg-bradykinin and des-Arg-kallidin, kinin-cleavage products are biologically inactive. Kinins are cleaved by a number of enzymes including carboxypeptidases, ACE, and neprilysin. ACE also truncates its own reaction product, bradykinin-(1–7), further to form bradykinin-(1–5). Neprilysin, like ACE, cleaves bradykinin at the 7 to 8 position and has a broad substrate specificity (see Table 11.2 ). The amino terminal of bradykinin possesses two proline residues and is susceptible to cleavage by the proline-specific exopeptidase aminopeptidase P. The resultant peptide, bradykinin-(2–9), may be further cleaved by proteases that include the endothelial enzyme dipeptidyl peptidase-4, which reduces this metabolite to bradykinin-(4–9).

Independent of its ability to generate kinin, tissue kallikrein also exerts separate effects on tubule solute transport by regulating the activity of the epithelial Na ⁺ channel (ENaC), the colonic H ⁺ , K ⁺ ATPase, and the epithelial calcium channel TRPV5 (transient receptor potential channel vanilloid subtype 5). The connecting tubules secrete a large amount of tissue kallikrein, which, through its enzymatic activity, can alter the function of ion transporters expressed on the luminal surface of cells downstream of its site of secretion. For instance, tissue kallikrein may participate in the proteolytic processing of ENaC increasing its activity, whereas tissue kallikrein–deficient mice have decreased ENaC activity. Despite decreased ENaC activity, however, coincident upregulation of ENaC-independent electroneutral NaCl absorption ensures that tissue kallikrein is not essential for sodium homeostasis. ^, The cortical collecting ducts from tissue kallikrein–deficient mice also demonstrate enhanced activity of the colonic H ⁺ , K ⁺ ATPase in intercalated cells, resulting in net K ⁺ absorption. ^, Finally, tissue kallikrein functions to stabilize the TRPV5 channel at the plasma membrane, promoting Ca ²⁺ reabsorption, whereas tissue kallikrein knockout mice exhibit robust hypercalciuria. Distal tubule defects in potassium and calcium handling have also been reported in humans with the loss-of-function R53H polymorphism in the tissue kallikrein gene.