Volume 93, Issue 5 p. 557-563
Free Access

Prorenin and (pro)renin receptor: a review of available data from in vitro studies and experimental models in rodents

Geneviève Nguyen

Geneviève Nguyen

Institut de la Santé et de la Recherche Médicale (INSERM) Unit 833 and Collège de France, 11 place Marcelin Berthelot, 75005, Paris, France

Search for more papers by this author
A. H. Jan Danser

A. H. Jan Danser

Pharmacology, Vascular and Metabolic Diseases Sector, Department of Internal Medicine, Erasmus MC, Dr Molewaterplein 50, 3015 GE Rotterdam, The Netherlands

Search for more papers by this author
First published: 14 April 2008
Citations: 74
Corresponding author G. Nguyen: Institut de la Santé et de la Recherche Médicale, (INSERM) Unit 833 and Collège de France, Experimental Medecine Unit, 11 place Marcelin Berthelot, 75005, Paris, France. Email: [email protected]

Abstract

The discovery of a (pro)renin receptor [(P)RR] and the introduction of renin inhibitors in the clinic have brought renin and prorenin back into the spotlight. The (P)RR binds both renin and its inactive precursor prorenin, and such binding triggers intracellular signalling that upregulates the expression of profibrotic genes, potentially leading to cardiac and renal fibrosis, growth and remodelling. Simultaneously, binding of renin to the (P)RR increases its angiotensin I-generating activity, whereas binding of prorenin allows the ‘inactive’ renin precursor to become fully enzymatically active. Therefore, the (pro)renin receptor system could be considered as having two functions, an angiotensin-independent function related to (P)RR-induced intracellular signalling and its downstream effects and an angiotensin-dependent function related to the increased catalytic activity of receptor-bound (pro)renin. A (P)RR blocker has already been described which blocks both functions, thus preventing diabetic nephropathy, cardiac fibrosis and ocular neovascularization. On-going experimental studies should now determine which of the two functions plays the more important role in pathological situations. The results of these studies are extremely important in view of the clinical use of renin inhibitors, since it is well known that their administration results in increased levels of both renin and prorenin. Although this rise can be interpreted as evidence of effective renin–angiotensin system blockade, it could also result in increased (P)RR activation.

The discovery of a specific receptor for renin and its precursor, prorenin, named the (pro)renin receptor [(P)RR], has modified our conception of renin as an enzyme only responsible for the cleavage of angiotensin (Ang) I from angiotensinogen. Renin and prorenin binding to the receptor allows these enzymes to display enzymatic activity on the cell surface and to trigger intracellular signalling that in turn modifies gene expression. With regard to prorenin, this would imply that finally a function has been found for this ‘inactive’ renin precursor. Information on the role of the (P)RR in organ damage has only been obtained in the past few years. Importantly, blocking prorenin–(P)RR interaction with a putative (P)RR blocker called ‘handle region peptide’ (HRP) was claimed not only to prevent diabetic nephropathy but also to reverse the glomerulosclerosis of diabetic nephropathy (Ichihara et al. 2004, 2006b; Takahashi et al. 2007). If this is true, then it would make the (P)RR a major therapeutic target. Unfortunately, however, not all studies applying HRP support its (P)RR-blocking activity (Batenburg et al. 2007; Feldt et al. 2008a,b; Müller et al. 2008), and thus more work is needed before definite conclusions can be drawn. The aim of this review is to make a critical appraisal of all available evidence on the pathophysiological role of prorenin and the (P)RR and to address the controversies with regard to the putative (P)RR blocker.

Prorenin

Being the precursor of renin, prorenin has always been assumed to have no function of its own. Yet, it circulates in human plasma in excess of renin, sometimes at concentrations that are 100 times higher. A 43-amino acid N−terminal propeptide explains the absence of enzymatic activity of prorenin. This propeptide covers the enzymatic cleft and obstructs access of angiotensinogen to the active site of renin. Prorenin can be activated in two ways: proteolytic or non-proteolytic (Danser & Deinum, 2005). Proteolytic activation is irreversible because it involves removal of the propeptide. Non-proteolytic activation of prorenin is reversible. It can best be imagined as an unfolding of the propeptide from the enzymatic cleft. Non-proteolytic activation can be induced by exposure to low pH (pH 3.3) or cold (4°C; Danser & Deinum, 2005). Non-proteolytically activated prorenin is enzymatically active and can be recognized by monoclonal antibodies that are specific for the active site. Kinetic studies of the non-proteolytic activation process have indicated that an equilibrium exists between the closed (inactive) and open (active) forms of prorenin. The inactivation step is highly temperature dependent and occurs very rapidly at neutral pH and 37°C. Consequently, under physiological conditions, <2% of prorenin is in the open, active form, i.e. displays enzymatic activity, and >98% is closed and inactive.

Prorenin and renin levels are highly correlated, but do not alter in parallel under all circumstances (Danser et al. 1998). Acute stimuli of renin will not affect prorenin levels, whereas chronic stimuli increase both renin and prorenin. This is due to the fact that renin is stored as active enzyme and thus can be released immediately upon stimulation of the juxtaglomerular apparatus. Prorenin, in contrast, is released constitutively and not stored. Chronic stimulation causes more prorenin to be converted to renin, leading to an increased renin/prorenin ratio in plasma (Schalekamp et al. 2008). However, some exceptions to this rule exist. A well-known example is diabetes mellitus complicated by retinopathy and nephropathy (Luetscher et al. 1985). In microalbuminuric diabetic subjects, prorenin is increased out of proportion to renin. This increase starts before the occurrence of microalbuminuria, and the prorenin level, in conjunction with the glycated haemoglobin level, might even be used to predict the occurrence of later microalbuminuria (Deinum et al. 1999). The reason for the elevated prorenin levels in diabetic subjects is unknown. One possibility is that prorenin originates outside the kidney. Indeed, it is prorenin, and not renin, that remains detectable in blood following a bilateral nephrectomy, although its levels are lower than in normal subjects (Danser et al. 1998; Krop et al. 2008). This suggests that, although the kidney is the main, if not the only, source of renin in the body, there are other tissues releasing prorenin into the circulation. For instance, pregnant women have high plasma prorenin levels, derived from the ovaries (Derkx et al. 1987). The function of this prorenin is unknown, as is the function of prorenin in amniotic fluid, in which prorenin was discovered. The reproductive organs, together with the adrenal, eye and submandibular gland are now well-accepted sites of extrarenal renin gene expression (Krop & Danser, 2008). For reasons that are not understood, these tissues exclusively release prorenin. Whether one or more of these organs contributes to the prorenin rise in diabetic subjects with microvascular complications still needs to be investigated.

Interestingly, the renal vasodilator response to captopril in diabetic subjects correlated better with plasma prorenin than with plasma renin (Stankovic et al. 2006). Possibly, therefore, prorenin (and not renin) is responsible for tissue angiotensin generation. Obviously, this would require local conversion prorenin to renin, for which no evidence exists (Lenz et al. 1991). In support of this concept, however, transgenic rodents with (inducible) prorenin expression in the liver display increased cardiac Ang I levels, cardiac hypertrophy, hypertension and/or vascular damage without evidence for increased renin or angiotensin levels in blood (Véniant et al. 1996; Prescott et al. 2002; Peters et al. 2008). Interestingly, increased tissue Ang I formation occurred even when expressing a non-cleavable prorenin variant, i.e. a prorenin variant that cannot be enzymatically cleaved to renin (Methot et al. 1999). Based on such data, it seems logical to assume that tissues are capable of sequestering prorenin, e.g. via a receptor-dependent mechanism, and that this procedure results in prorenin activation, possibly in a non-proteolytic manner. Several candidates for such a binding/uptake mechanism have been put forward, including an intracellular renin-binding protein (RnBP; Maru et al. 1996) and the mannose 6 phosphate/insulin-like growth factor II receptor (M6P/IGF2R; van Kesteren et al. 1997; Saris et al. 2001a; van den Eijnden et al. 2001). The intracellular RnBP was eventually found to inhibit renin, and its deletion affected neither blood pressure nor plasma renin (Schmitz et al. 2000). Furthermore, the M6P/IGF2R, which binds phosphomannosylated (M6P-containing) proteins (such as renin and prorenin), indeed bound and internalized renin and prorenin. It also resulted in proteolytic cleavage of prorenin to renin (Saris et al. 2001a,b). However, such binding and activation did not result in angiotensin generation, either intracellularly or extracellularly, and it is now believed that the M6P/IGF2R serves as a clearance receptor for renin/prorenin (Saris et al. 2002). This leaves the (P)RR as the most promising candidate for the tissue uptake of circulating renin/prorenin.

The (pro)renin receptor

Biochemistry of the (P)RR The (pro)renin receptor is a 350-amino-acid receptor with a single transmembrane domain, like receptors for growth factors (Nguyen et al. 2002). There is no homology with any known protein based on the nucleotide and the amino-acid sequence of the (P)RR. Homologies in the tertiary structure have not yet been determined owing to lack of knowledge of the crystal structure of (P)RR. The receptor binds both renin and prorenin, with affinities in the nanomolar range, and the encoding gene, called ATP6ap2 (see below), is located on the X chromosome in locus p11.4.

The initial characteristics of the (P)RR were (Fig. 1; Nguyen et al. 2002) as follows:

Details are in the caption following the image

The two faces of the (pro)renin receptor
Activation of the angiotensin (Ang) II-dependent pathway is widely accepted to lead to organ damage and fibrosis. The in vivo consequences of activation of the Ang II-independent pathway remain to be demonstrated. Abbreviations: HSP27, heat shock protein 27; TGF-β1, transforming growth factor β1; and PAI-1, plasminogen-activator inhibitor-1.

  • 1

    Renin and prorenin bound to the receptor are not internalized or degraded, i.e. they remain on the cell surface.

  • 2

    Renin bound to the receptor displays increased catalytic activity compared with renin in solution.

  • 3

    Receptor-bound prorenin displays Ang I-generating activity in the absence of cleavage of the prosegment, i.e. receptor binding activates prorenin non-proteolytically, most likely via a conformational change induced by binding per se.

  • 4

    (Pro)renin binding triggers activation of the mitogen-activated protein (MAP) kinase–extracellular signal-regulated kinase (ERK)1/2 signalling pathway.

Further studies on the signalling pathways involved in (P)RR activation confirmed ERK1/2 phosphorylation and showed that it was due to ERK kinase and provoked Ets-like gene (Elk) phosphorylation (Huang et al. 2006, 2007b; Sakoda et al. 2007; Feldt et al. 2008a). Moreover, ERK 1/2 activation resulted in the upregulation of transforming growth factor β1 gene expression, the subsequent upregulation of genes coding for profibrotic molecules, such as plasminogen-activator inhibitor-1, fibronectin and collagens, and the induction of mesangial cell proliferation (Huang et al. 2006, 2007a,b). The ERK1/2 pathway is not the only signalling pathway linked to the (P)RR, since the receptor also appears to activate the MAP kinase p38–heat shock protein 27 cascade (Ichihara et al. 2006b; Saris et al. 2006) and the phosphatidylinositol-3 kinase-p85 (PI3K-p85) pathway (Schefe et al. 2006). Importantly, the latter pathway results in the nuclear translocation of the promyelocytic zinc finger transcription factor, which downregulates the expression of the (P)RR itself (Schefe et al. 2006). In other words, high (pro)renin levels will suppress (P)RR expression, thereby preventing excessive receptor activation.

Prorenin binding and its subsequent non-proteolytic activation was confirmed in both primary cells (Batenburg et al. 2007) and cells with transient overexpression of (P)RR (Nabi et al. 2006). Data in rat aortic vascular smooth muscle cells overexpressing the human (P)RR suggested that prorenin binds with higher affinity to the receptor than renin, so that in vivo, prorenin might be the endogenous agonist of the receptor (Batenburg et al. 2007). The fact that both prorenin and renin are capable of binding to the (P)RR implies that the domains involved in the interaction between (P)RR and the (pro)renin molecule are different from the active site and are not restricted to the prosegment of prorenin. Unfortunately, owing to the difficulties in generating purified recombinant (P)RR, no structure–function studies are currently available that would allow the identification of the domains of the (P)RR and (pro)renin involved in binding. In the absence of such structure–function studies or of an X-ray crystallographic structure of the (P)RR, it is difficult to design antagonists for the (P)RR.

Nevertheless, Suzuki et al. (2003) made the interesting observation that an antibody against a sequence of the prosegment of human prorenin (I11PFLKR15P) was able to open the profragment to yield a ‘non-proteolytically’ activated prorenin, in a manner similar to the putative mechanism of (P)RR binding-induced prorenin activation. They named this region of the prosegment the ‘handle’ region. Based on this observation, Ichihara et al. (2004) tested a 10-amino-acid peptide which encompassed the handle region (HRP), as a blocker of prorenin–(P)RR binding. In a clever set-up of studies in diabetic rodents, they reasoned that diabetes would increase prorenin synthesis, thus creating optimal conditions to test the efficacy of HRP in vivo. Indeed, HRP prevented or even reversed diabetic nephropathy (Ichihara et al. 2004, 2006b; Takahashi et al. 2007), and blocked ischaemia-induced retinal neovascularization and ocular inflammation in endotoxin-induced uveitis (Satofuka et al. 2006, 2007). Moreover, it diminished cardiac fibrosis in stroke-prone spontaneously hypertensive rats (Ichihara et al. 2006a). Taken together, these data strongly suggest that the prorenin–(P)RR axis plays an essential role in end-organ damage in diabetic and inflammatory pathologies. Handle region peptide was subsequently renamed a (P)RR ‘blocker’.

Nevertheless, many questions remain. In vitro studies by others did not support blockade of prorenin binding to its receptor by HRP (Batenburg et al. 2007; Feldt et al. 2008a; Müller et al. 2008), and fluorescein isothiocyanate-labelled HRP also bound to cells devoid of the (P)RR on the plasma membrane (Feldt et al. 2008a). Furthermore, HRP is unlikely to block renin binding to the (P)RR, and thus one may wonder why it is so successful if it does not block renin–(P)RR interaction. Indeed, HRP was ineffective in a high-renin, low-prorenin model, the Goldblatt rat (Müller et al. 2008). At present, it cannot be ruled out that HRP also exerts other, non-(P)RR-related effects, particularly in diabetic animals. Nonetheless, if confirmed, the in vivo results indicate that HRP has a great potential in diabetic nephropathy. Clearly, more work is needed to unravel its mechanism of action, before it can truly be called a (P)RR blocker.

(Pro)renin in experimental models of cardiovascular and renal diseases The high blood pressure occurring in a transgenic rat model targeting human (P)RR expression to vascular smooth muscle cells suggests a pathological role of the (P)RR in raising blood pressure (Burckléet al. 2006). Ubiquitous overexpression of the human (P)RR resulted in HRP-inhibitable proteinuria, glomerulosclerosis (Kaneshiro et al. 2007) and cyclo-oxygenase−2 upregulation (Kaneshiro et al. 2006). Both targeted and ubiquitous (P)RR expression left the plasma levels of renin and angiotensin unaltered, but did cause a rise in plasma aldosterone.

In a Goldblatt model of hypertension, the parallel increases in (P)RR and renin have been suggested to be profibrotic in the clipped kidney (Krebs et al. 2007). The above-described beneficial effects of HRP in diabetic rodents and stroke-prone spontaneously hypertensive rats are also suggestive for a role of the (P)RR in fibrosis and glomerulosclerosis, although no increased (P)RR expression was described in these models (Ichihara et al. 2004, 2006a,b). Moreover, glomerulosclerosis did not occur in transgenic ren-2 rats with inducible prorenin expression (Peters et al. 2008), despite the fact that such rats, following induction, displayed 200-fold higher prorenin levels, with no change in renin. This argues against the concept that prorenin, through a direct interaction with its receptor, induces glomerulosclerosis.

There are two means to establish the role of a receptor in pathology: the use of an antagonist of the receptor; and studies in knock-out mice not expressing the gene encoding for the receptor. The antagonist, as discussed, is not yet ideal, and the total knock-out of the (P)RR is, surprisingly for a component of the renin–angiotensin system (RAS), not possible (Burcklé & Bader, 2006). The generation of (P)RR conditional knock-out mice is thus mandatory. Such animals will allow further establishment of the role of (P)RR in disease.

Unexpected properties and ontogeny of the (P)RR Before the (P)RR was cloned, a truncated form of the (P)RR, composed of the transmembrane and cytoplasmic domains of the (P)RR, had been co-purified with a vacuolar H+-ATPase (V−ATPase; Fig. 2; Ludwig et al. 1998). This V-ATPase is a complex, 13−subunit protein, essential to maintain an acidic pH in intracellular vesicles and to regulate cellular pH homeostasis (Nishi & Forgac, 2002). The link with V−ATPase explains why the gene of the (P)RR is called ATP6ap2 (ATPase-associated protein). Unexpectedly, total ablation of the (P)RR gene in mouse embryonic stem cells is impossible and incompatible with their incorporation into blastocysts (M. Bader, personal communication). Since this contrasts with the ablation of other components of the RAS, it must be concluded that the (P)RR exerts essential, non-RAS-related effects. The necessity of an intact (P)RR/ATP6ap2 gene in early development is stressed by the observations that in zebra fish, the mutation of (P)RR/ATP6ap2 gene provoked the death of the fish before the end of embryogenesis (Amsterdam et al. 2004) and that in rodents (P)RR/ATP6ap2 gene expression is ubiquitous and early in development (Contrepas et al. 2007). While in mice renin expression can be detected in large intrarenal arteries only at 15.5 days of gestation, (P)RR mRNA is already present on day 12 in the ureteric bud and at later stages in vesicles and S-shaped bodies (Contrepas et al. 2007). In newborn mice, (P)RR expression is high in epithelial cells of distal, proximal and collecting tubules and low in glomeruli and arteries (Contrepas et al. 2007). These observations in zebra fish and in the developing mouse kidney suggest that the (P)RR has functions essential for cell survival and proliferation that are unrelated to the RAS.

Details are in the caption following the image

Schematic structure of (P)RR (top) and its association with V−ATPase (bottom)
The arrow in the top diagram shows a putative cleavage site generating a truncated form of (P)RR composed of the C−terminal part of the (pro)renin receptor that remains associated with V−ATPase. The N−terminal part is the (pro)renin binding region of the (P)RR. Adapted from Nishi & Forgac (2002). Abbreviations: SP, signal peptide; TM, transmembrane domain; and Cyt, cytoplasmic domain.

Analysis of the sequence of (P)RR coding cDNA shows that the sequence coding for the transmembrane and the intracellular domain putatively associated with the V-ATPase is remarkably conserved between invertebrates and vertebrates, whereas the cDNA sequence coding for the extracellular domain responsible for renin and prorenin binding is conserved in vertebrates only (Burcklé & Bader, 2006; Bader, 2007). This leads to the postulate that the (P)RR/ATP6ap2 gene may result from the fusion of two genes: an ancient gene (corresponding to the C-terminus) coding for a protein essential for cell survival; and a more recent gene in vertebrates (corresponding to the N-terminus), which binds renin and prorenin. However, to date, we have no arguments to confirm or to deny that the role of the (P)RR in cell survival is related to V-ATPase activity.

Conclusion

The discovery of the (P)RR confirms the hypothesis of Tigerstedt and Bergman more than a century ago that renin is a hormone (Tigerstedt & Bergman, 1898). The (P)RR also endows prorenin with a function that was suspected over 25 years ago by Luetscher and co-workers in diabetic patients (Luetscher et al. 1985). Experimental studies now suggest that the (P)RR might be a major target in cardiovascular disease and in diabetes-induced organ damage. Tissue-specific knock-out of (P)RR should soon establish whether the (P)RR plays a role in cardiovascular pathologies and in diabetes and to what degree HRP exerts (P)RR-dependent effects. As mentioned recently (Luft & Weinberger, 2008), the future of renin research ‘is certain not to be dull’ and will certainly keep us extremely busy.

Appendix

Acknowledgements

G.N. has received funding from Novartis and from Institut de Recherche Servier. A.H.J.D. has received funding from Novartis.