ing the guanylyl cyclase A. ... at room temperature for 10 min in phosphate-buffered saline incubated for 18 h at 4°C. The .... staining (Figure 2c). (72.5±11.9 ...
Nephrol Dial Transplant (1998) 13: 2529–2535
Nephrology Dialysis Transplantation
Original Article
The renal natriuretic peptide urodilatin is present in human kidney Monika Herten, Wolfgang Lenz, Rupert Gerzer and Christian Drummer Institut fu¨r Luft- und Raumfahrtmedizin, Deutsches Forschungszentrum fu¨r Luft und Raumfahrt (DLR), Ko¨ln, Germany
Abstract Background. The natriuretic peptide urodilatin was first isolated from human urine and may be one of the important mediators of natriuresis, while the atrial natriuretic peptide a-ANP, the circulating member of the family, rather seems to play a role in cardiovascular regulation. Although the renal expression of the common propeptide for urodilatin and a-ANP has been detected in rat and pig, it is not yet shown that urodilatin is synthesized in human kidney. Methods. Immunhistochemically we localized urodilatin with an urodilatin-antibody, which does not cross-react with a-ANP, the ANP propeptide, brain natriuretic peptide (BNP), and C-type natriuretic peptide (CNP). Radiolabelled urodilatin binding was examined autoradiographically. Results. We could demonstrate that urodilatin is present in human distal tubular cells, in which we could also localize propeptide immunoreactivity. The glomeruli, the cells of the proximal tubule, and the collecting duct did not show any urodilatin immunoreactivity. In human kidney homogenate the urodilatin content was 4600±520 fmol/g protein (n=6). The renal concentration of BNP and CNP, mainly localized in the distal tubule, was much lower at 40±8 and 400±50 fmol/g protein (n=6) respectively. Furthermore the autoradiographic examinations showed that radiolabelled urodilatin binds to natriuretic peptide receptors in the glomeruli, blood vessels, and collecting ducts. Conclusions. Our data suggest that urodilatin may be the predominant representative of natriuretic peptides in human kidneys. Urodilatin being synthesized in the distal tubular region may be transported as a paracrine factor to the collecting duct, where it exerts its suppressing effect on the sodium reabsorption by stimulating the guanylyl cyclase A. Key words: autoradiography; immunhistochemistry; natriuretic peptides
Correspondence and offprint requests to: Dr Christian Drummer, Institut fu¨r Luft- und Raumfahrtmedizin, Deutsches Forschungszentrum fu¨r Luft und Raumfahrt (DLR), D-51170 Ko¨ln, Germany.
Introduction Urodilatin, a member of the family of natriuretic peptides, was isolated from human urine and identified as a 32-amino-acid peptide containing the amino acid sequence 95 to 126 of the cardiac ANP prohormone [1]. It is composed of the entire sequence of a-ANP (ANP 99–126) plus a four-amino-acid N-terminal extension. The fact that it does not circulate in the human blood [2] suggests that the peptide is derived from the human kidney. In rat [3] and human tissues [4] an ANP-immunoreactive signal has been localized in distal tubule cells and intercalated cells of the collecting duct [5] by immunohistochemical methods. The ANP prohormone, which is also regarded to be the prohormone of urodilatin, was localized in the distal tubules of the rat kidney [6 ] whereas Greenwald et al. [7] and Lee et al. [8] confirmed its gene expression in rat tissue. Salt loading even increased the expression of ANP mRNA in rats [8]. Urodilatin has not yet been isolated from any species other than human. Most urodilatin measurements published in recent years have been obtained using a polyclonal antibody raised in rabbit against the N-terminal part of the human amino-acid sequence, which does not cross-react with a-ANP [2]. However, the supposed porcine urodilatin, which, based on the porcine ANP-propeptide sequence, differs from human urodilatin in only one N-terminal amino-acid, crossreacts with our anti-human-urodilatin antibody. By applying this antibody in pigs Feller et al. could show that urodilatin is present in the renal tubules and the collecting ducts [9]. However, the observation that urodilatin is detectable in and derived from human kidney is still lacking. Therefore the aim of this study was to examine the cellular localization and concentration of urodilatin in human kidney. In order to verify whether urodilatin is of renal origin, we tried to co-localize the ANPprohormone in human kidney tissue using antibodies against pro-ANP 26–55 and pro-ANP 56–92. In autoradiographic investigations on rat kidneys with [125I ]a-ANP it was shown that ANP binds to glomeruli, vascular structures and to inner medullary sites [10]. In human adrenal tissue as well as in GC-A transfected cells, urodilatin displayed similar binding
© 1998 European Renal Association–European Dialysis and Transplant Association
2530
properties as ANP and stimulated GC-A activity [11] increasing the level of intracellular cGMP [12]. Furthermore urodilatin has also been shown to displace labelled [125I ]a-ANP from binding sites in rat kidneys [13,4] and to inhibit sodium and chloride reabsorption in the rat medullary collecting duct [15]. However the specific binding of [125I ]urodilatin in humans has not been examined in situ so far. Therefore we examined the binding of [125I ]urodilatin to its renal receptors using autoradiographical methods in kidney tissue sections, where the morphological structures were well retained. We also investigated the contents and localization of human brain natriuretic peptide (h-BNP-32) and human C-type natriuretic peptide (h-CNP-22) in the human kidney in order to understand the relative occurrence of urodilatin amongst other natriuretic peptides.
Subjects and methods Immunohistochemistry Immunohistochemistry was performed on tissue sections of six normal kidneys (tumour nephrectomies) using the peroxidase–antiperoxidase (PAP) technique on material which was fixed in 2%/2% (w/v) paraformaldehyde/glutaraldehyde in phosphate buffer (0.05 M, pH 7.4) and embedded in paraffin. The specific antisera were applied and the sections were incubated for 18 h at 4°C. The sections were treated with the secondary antiserum, horse-radish-peroxidase-conjugated goat anti-rabbit IgG (Nordic Laboratories, Tilburg, NL) in 1530 dilution for 1 h and afterwards incubated for 30 min with the peroxidase antiperoxidase antibody IgG (Nordic Laboratories, Tilburg, NL) in a 15300 dilution. As substrate for the horse-radish–peroxidase diaminobenzidine was used (1.7 mM with 0.0075% H O in 50 mM Tris buffer). The 2 2 sections were counterstained with Mayer’s haemalum (5%) and embedded in DPX mountant (Fluka, Buchs, CH ).
Antisera All specific antisera were raised in rabbit. The urodilatin antiserum has been characterized for radioimmunoassay by Drummer et al. [2] and does not show any cross-reactivity with a-ANP or the ANP prohormone (ANP 1–126). The BNP and CNP antisera were obtained from Phoenix Pharmaceuticals Inc., Mountain View, CA, USA. The BNP antiserum showed no cross-reactivity with a-ANP, urodilatin, ANP 1–126, or CNP; the CNP antiserum had no crossreactivity with a-ANP, urodilatin and BNP as determined by radioimmunoassay. Pro-ANP 26–55 and 56–92 antisera were purchased from Peninsula Laboratories Europe, Ltd, St Helens, UK. The pro-ANP 26–55 antiserum showed no cross-reactivity with pro-ANP 56–92, a-ANP, b-ANP, BNP and CNP, but 100% cross-reactivity with ANP 1–126. The pro-ANP 56–92 antiserum did not display any crossreactivity with a-ANP, but 100% with ANP 1–126 (data provided by Peninsula).
Specific controls Preincubation of the diluted antisera with 80 mM of the respective synthetic antigen abolished the immunoreactive
M. Herten et al.
staining in tissue sections. All antigens were purchased from Peninsula. The sections were also incubated with normal rabbit serum (Sigma) instead of specific antisera. Under these conditions no staining was observed.
Human kidney homogenates Fresh renal tissue samples (6–8 g) from six surgically removed human kidneys were homogenized under liquid nitrogen using mortar and pestle and suspended in ice-cold 0.2 M acetic acid. After sonification (2×20 s) the material was centrifuged at 60 000 g for 20 min and the supernatants were lyophilized. Protein was precipitated by addition of ethanol. The supernatant was lyophilized and reconstituted in RIA buffer. The protein content of the extracts was determined by the method of Lowry.
Immunoassays The urodilatin assay was performed as described by Drummer et al. [2]. The BNP and CNP radioimmunoassays were purchased from Phoenix Pharmaceuticals. For the determination the kidney homogenate had to be diluted fiveand 15-fold respectively.
Autoradiography Immediately after removal, the kidney material was frozen in freezing methylbutane suspended in liquid nitrogen. Cryostat sections (7 mm) were cut at −20°C and preincubated at room temperature for 10 min in phosphate-buffered saline (10 mM phosphate, pH 7.4 containing 0.2% bovine serum albumin, 0.6 mM phenylmethylsulphonyl fluoride, 1 mM benzamidine, 0.1 mM aprotinin, 0.1 mM pepstatin A). [125I ]urodilatin (2000 mCi/nmol, 170–340 nM ) in 10 mM phosphate buffer pH 7.4 containing 10 mM phosphoramidon to block neutral endopeptidase (EC 3.4.24.11) activity was added to the tissue in the absence or presence of 200 nM unlabelled urodilatin and incubated for 15 min at room temperature. After washing three times in ice-cold buffer/ distilled water the tissue was fixed with 2% (v/v) glutaraldehyde in phosphate buffer (50 mM, pH 7.4) for 15 min in order to promote cross-linking of the ligand with the receptor. The slides were washed three times, dried under a cold airstream and dipped in liquid photo-emulsion (Ilford K5) at 43°C. The slides were stored in the dark at 4°C for 18–30 days, developed in Ilford Phenisol developer for 5 min at 20°C, fixed, stained with 0.1% toluidine blue and mounted in synthetic medium (DPX ).
Results The immunohistochemical experiments with the urodilatin-specific antiserum demonstrate that the urodilatin immunoreactivity in human kidneys is confined to the distal tubular cells (Figure 1a). No immunostaining was observed in glomeruli, the renal vasculature, or in the proximal tubular segments. In order to further characterize the renal structures, we used an antiactin antibody to identify renal blood vessels. Alkaline phosphatase, the marker enzyme of the brush-border region of the proximal tubular cells, was used to distinguish between proximal and distal tubules. Figure 1b shows
Urodilatin in human kidney
2531
Fig. 1. (a, b) Immunoreactive urodilatin in distal tubule cells of the human kidney stained with antiserum against urodilatin in a concentration of 151500 using the PAP method. G, glomerulum; PT, proximal tubule; DT, distal tubule; bar=50 mm.
at a higher magnification that the positive staining of distal tubular cells was clearly opposing to the nonstained proximal tubule cells. ANP immunoreactivity occurred in the distal segment of the nephron (data not shown). The application of the antisera against the ANP propeptides 26–55 ( Figure 2a) and 56–92 ( Figure 2b) resulted in a positive immunostaining of distal tubular segments. Using non-immunized rabbit control serum instead of the antisera did not produce any immunostaining ( Figure 2c). Figure 3a shows that BNP immunoreactivity in the human kidney is mainly localized in distal tubular segments. The distribution of CNP immunoreactive tubules (Figure 3b) was quite similar to that of BNP. Positive CNP immunostaining was found predominantly in distal tubules. No immunostaining of BNP and CNP was observed in glomeruli or the renal vasculature. To quantify urodilatin and other natriuretic peptides in the human kidney, the homogenate was assayed in specific radioimmunoassays. Urodilatin immuno-
reactivity was present in the human kidney. As seen in the right part of Figure 4, the average urodilatin concentration of six human kidneys was found to be 4673±526 fmol/g protein (n=6, mean±SEM ) (561.3±63.2 ranging from 441 to 792 fmol/g wet weight). The content of ANP immunoreactivity was measured in the same samples with two different assay systems and was found to be 1876±212 fmol/g protein (225.3±25.4 fmol/g wet weight) for the Nichols ANP assay (ir ANP-Ni) and 604±99 fmol/g protein (72.5±11.9 fmol/g wet weight) for the Amersham assay (ir ANP-Am) respectively. In the right part of Figure 4 the values for the other natriuretic peptides BNP and CNP are shown, which were 38.7±7.7 fmol/g protein (4.6±0.9 fmol/g wet weight) and 398.1±54.2 fmol/g protein (47.8±6.5 fmol/g wet weight) respectively. The content of BNP and CNP were at least one order of magnitude lower than these of urodilatin. The autoradiographic investigations ( Figure 5) show the binding of [125I ]urodilatin to the glomeruli and to vascular structures in the kidney cortex region as well as to some tubular structures. Non-specific binding
2532
M. Herten et al.
Fig. 2. (a) Immunoreactive pro-ANP 26–55 in the human kidney stained with antiserum against pro-ANP 26–55 in a concentration of 15750. (b) Immunoreactive pro-ANP 56–92 in the human kidney stained with antiserum against pro-ANP 56–92 in a concentration of 15750. (c) Human kidney stained with normal goat serum in a concentration of 15300. The PAP method was used in all pictures. G, glomerulus; PT, proximal tubule; DT, distal tubule; bar=100 mm.
was determined in the presence of 200 nM unlabelled urodilatin and did not show any signal.
Discussion The present data show for the first time that the renal natriuretic peptide urodilatin is present in human
kidney. The distribution of urodilatin which was discovered, isolated, and characterized from human urine, has not been studied in human kidney so far. In our investigation we localized urodilatin in human distal tubular cells by immunohistochemistry as well as by quantitative analysis of kidney homogenate using an urodilatin antibody which does not cross-react with ANP.
Urodilatin in human kidney
2533
Fig. 3. (a) Immunoreactive BNP in the human kidney stained with antiserum against BNP in a concentration of 151000. (b) Immunoreactive CNP in the human kidney stained with antiserum against CNP in a concentration of 15300. The PAP method was used in both pictures. G, glomerulus; PT, proximal tubule; DT, distal tubule; bar=100 mm.
Fig. 4. Contents of natriuretic peptides (NP) in human kidneys. NP immunoreactivity was determined with specific RIAs. Each bar represents the mean±SEM (n=6).
In agreement with recent data showing a sodiumchloride-stimulated release of urodilatin from a human embryonic kidney cell line [Lenz et al., unpublished ], the present findings strongly support the conclusion derived from many examinations in humans [16–18] and dogs [19,20], in which the close correlation between urodilatin and sodium excretion is demonstrated under various measurements in which bodyfluid regulation is stimulated. The renal natriuretic peptide urodilatin may therefore be considered as one of the major regulators of sodium excretion in man.
Fig. 5. [125I ]urodilatin binding in human kidney sections. Autoradiographic development followed a 30-day exposure to the photographic emulsion. The receptor binding sites are represented by the small black autoradiographic silver grains in the emulsion, counterstaining of the nuclei with toluidine blue. G, glomerulus; TS, tubular structure; BV, blood vessel; bar=100 mm.
The urodilatin immunoreactive signal was detected in distal tubular cells, while other parts of the human nephron like the glomeruli, the blood vessels, and the alkaline-phosphatase-positive proximal tubules were urodilatin negative. Additionally our data show that the prohormone is also localized in the distal tubular
2534
segments of the human kidney. The co-localization of urodilatin and prohormone in the distal part of the nephron suggests that the renal natriuretic peptide is indeed synthesized in these human kidney cells. In parallel to the urodilatin-specific measurements we also localized ANP immunoreactivity in human kidney tissue slices and in homogenate. For that purpose we used several commercially available antibodies that detect comparable amounts of ANP in human plasma (data not shown). Because polyclonal antibodies are in most instances directed against the ring structure between amino acids 105 and 121 of ANP and urodilatin, they are cross reacting with urodilatin. Both ANP antibodies labelled the distal tubules, comparable to the urodilatin and ANP propetide antibodies. These data suggest that the urodilatin and the ANP-antibodies most probably detected the 32-amino-acid peptide. In addition to the renal natriuretic peptide urodilatin, other natriuretic peptides such as BNP and CNP are also present in the human kidney, where they occur predominantly in distal tubular cells. Compared to urodilatin, their respective content in the kidney homogenate is much lower, suggesting that their relative role may be of lower relevance than that of urodilatin. In rat kidney tissue and rat cell culture the renal expression of CNP has been demonstrated [21] and also the renal expression of its specific receptor, the guanylyl cyclase B (GC-B) was shown previously [22]. In the vascular endothelium and the smooth-muscle layer, CNP is synthesized and appears to have a relaxing and antiproliferative effect through activation of GC-B in a paracrine mode [23]. Although the physiological role of the CNP/GC-B system in the kidney is not fully understood, the CNP/GC-B system might also play a similar paracrine role, despite its low tissue content, possibly interacting with the urodilatin–GC-A system, where it might be involved in the transtubular transport of electrolytes. BNP was localized immunohistochemically and its concentration determined in kidney homogenate, whether it has a physiological role in the kidney metabolism has to be further elucidated. In the kidney homogenates we found an urodilatin concentration of 4.6 pmol/g protein. The homogenate concentration of ANP were measured with two of the mentioned ANP antibodies that should detect ANP and urodilatin on an equimolar base. However, the values differed considerably suggesting that the measured tissue ANP concentrations depend on the chosen antibody. In the literature, different data exist concerning the ANP content in human kidney, ranging between 1 pmol/g [24] and 500–1200 pmol/g protein [4]. Our ANP-data are with 1.9 pmol/g protein for the Nichols assay within the range measured by Totsune et al. [24]. The differences in the ANP-immunoreactive data could be due the use of different antibodies. It is also possible that depending on the various tissue extractions there is a decrease in the measurable ANP content caused by the damaged ring structure of the molecule. It is known that some ANP antisera recognize the intact ANP-molecule as well as the open ring form [25], which is derived by cleavage by endopeptid-
M. Herten et al.
ase E.C 3.4.24.11. [26 ]. Most antisera, however, need an intact ring [27], which could explain the differences in the detected values for ANP by our ANP-antibodies as well as those described by other authors [4,24]. Although the specific binding of urodilatin has already been shown in cell and membrane preparations [13,14], the specific binding of radiolabelled urodilatin in human kidney tissue sections has not been shown so far. In the present study we show for the first time the binding of [125I ]urodilatin to the natriuretic peptide receptors in human kidney tissue. As expected and already shown for [125I ]ANP [10] we observed urodilatin binding to the receptors in the glomeruli, most probably clearance receptors, which prevent any circulating ANP to reach the GC-A in the collecting duct. Urodilatin binding was also evident in the actinpositive stained renal vascular system. This corresponds with data from rat kidney which were labelled with [125I ]ANP [28]. Furthermore we observed [125I ]urodilatin binding to tubular structures with a large diameter in the human kidney, which might be collecting ducts. Already in the 80s the distribution of bioactive natriuretic peptide receptors along the nephron was examined and shown to be exclusively present in the collecting ducts [12]. Therefore our data indicate that the urodilatin-binding sites represent most probably the receptor domain of this guanylyl cyclase in the collecting duct, which is supposed to be the physiological target for urodilatin binding. In this investigation we could show that urodilatin is present in human kidney homogenate and localized in the distal tubule cells. The propeptide is also present in distal tubule cells, confirming the renal origin of urodilatin. Our results emphasize that urodilatin is synthesized in the distal human nephron, may be luminally released and stimulate the guanylyl cyclase (GC-A) in the collecting ducts, thereby promoting sodium excretion. Acknowledgements. This study was supported by a research grant from the DFG (Dr-301/1–1). We thank the DLR programme directorate ‘Raumfahrt’ for continuous support. We also thank Prof. Bru¨hl and Dr Albers, Bonn, and Prof. Marx, Cologne, for kindly providing the kidney material from tumour nephrectomies and Dr Bechtel, Munich, Dr Meyer zum Gottesberge, Du¨sseldorf and Dr Dircksen, Bonn for the support with the immunohistochemistry.
References 1. Schulz-Knappe P, Forssmann K, Herbst F, Hock D, Pipkorn R, Forssmann WG. Isolation and structural analysis of ‘urodilatin’, a new peptide of the cardiodilatin-(ANP)-family, extracted from human urine. Klin Wochenschr 1988; 66: 752–759 2. Drummer C, Fiedler F, Bub A et al. Development and application of a urodilatin (CDD/ANP-95–126)-specific radioimmunoassay. Eur J Physiol 1993; 423: 372–377 3. McKenzie JC, Tanaka I, Misono K, Inagami T. Immunocytochemical localization of atrial natriuretic factor in the kidney, adrenal medulla, pituary and atrium of the rat. J Histochem Cytochem 1985; 33: 828–832 4. Figueroa CD, Lewis HM, MacIver AG, Mackenzie JC, Bhoola KD. Cellular localization of atrial natriuretic factor in the human kidney. Nephrol Dial Transplant 1990; 5: 25–31 5. Ramirez G, Saba, SR, Dietz J, Vesely DL. Immunocytochemical
Urodilatin in human kidney
6. 7. 8. 9. 10.
11.
12.
13. 14. 15. 16.
localization of proANF 1–30, proANF 31–67 and atrial natriuretic factor in the kidney. Kidney Int 1992; 41: 334–341 Forssmann WG, Nokihara K, Gagelmann M et al. The heart is the center of a new endocrine, paracrine and neuroendocrine system. Arch Histol Cytol 1989; 52: 293–315 Greenwald JE, Ritter D, Tetens E, Rotwein PS. Renal expression of the gene for atrial natriuretic factor. Am J Physiol 1992; 263: F974–978 Lee YJ, Shin SJ, Tan MS, Hsieh TJ, Tsai JH. Increased renal atrial natriuretic peptide synthesis in rats with deoxycorticosterone acetate–salt treatment. Am J Physiol 1996; 271: F779–789 Feller SM, Gagelmann M, Forssmann WG. Urodilatin. a newly described member of the ANP family. Trends Pharmacol Sci 1989; 10: 93–94 Rutherford RAD, Wharton J, Needleman P, Polak JM. Autoradiographic discrimination of brain and atrial natriuretic peptide-binding sites in the rat kidney. J Biol Chem 1990; 266: 5819–5826 Heim J-M, Sing S, Fu¨lle HJ, Gerzer R. Urodilatin and b-ANF. Comparison of a cloned ANF-sensitive guanylate cyclase (GC-A) with particulate guanylate cyclase from adrenal cortex. Naunyn Schmiedebergs Arch Pharmacol 1992; 345: 64–70 Tremblay J, Gerzer R, Vinay P, Pang SC, Be´liveau R, Hamet P. The increase of cGMP by atrial natriuretic factor correlates with the distribution of particulate guanylate cyclase. FEBS Lett 1985; 181: 17–22 Valentin JP, Sechi LA, Qiu C, Schambelan M, Humphreys MH. Urodilatin binds to and activates renal receptors for atrial natriuretic peptide. Hypertension 1993; 21: 432–438 Saxenhofer H, Fitzgibbon WR, Paul R. Urodilatin. binding properties and stimulation of cGMP generation in rat kidney cells. Am J Physiol 1993; 264: F267–273 Gunning ME, Otuechere G, Zeidel ML. Mechanism of urodilatin (ANP ) inhibition of Na+ transport in rabbit inner medul95–125 lary collecting duct (IMDC ) cells. J Am Soc Nephrol 1991; 2: 402 Drummer C, Franck W, Heer M, Forssmann WG, Gerzer, Goetz K. Postprandial natriuresis in humans. further evidence that urodilatin, not ANP, modulates sodium excretion. Am J Physiol 1996; 270: F301–310
2535 17. Drummer C, Gerzer R, Heer M et al. Effects of an acute saline infusion on fluid and electrolyte metabolism in humans. Am J Physiol 1992; 262: F744–754 18. Heer M, Drummer C, Baisch F, Gerzer R. Long term evaluations of dietary sodium produce parallel increase in the renal excretion of urodilatin and sodium. Eur J Physiol 1993; 425: 390–394 19. Goetz KL, Drummer C, Zhu JL, Leadley R, Fiedler F, Gerzer R. Evidence that urodilatin, rather than ANP, regulates renal sodium excretion. J Am Soc Nephrol 1990; 1: 867–874 20. Emmeluth C, Drummer C, Gerzer R, Bie P. Roles of cephalic Na+ concentrations and urodilatin in control of renal Na+ excretion. Am J Physiol 1992; 262: F513–516 21. Dean AD, Greenwald JE. Localization and regulation of C-type natriuretic peptide (CNP) in the kidney. J Am Soc Nephrol 1993; 4: 436 22. Canaan-Ku¨hl S, Jamison RL, Myers BD, Pratt RE. Identification of ‘B’ receptor for natriuretic peptide in human kidney. Endocrinology 1992; 130: 550–552 23. Cahill PA, Hassid A. Clearance receptor-binding atrial natriuretic peptides inhibit mitogenesis and proliferation of rat aortic smooth muscle cells. Biochem Biophys Res Comm 1991; 179: 1606–1613 24. Totsune K, Takahashi K, Murakami O et al. Natriuretic peptides in the human kidney. Hypertension 1994; 24: 758–762 25. Wilkins MR, Settle SL, Stockmann PT, Needleman P. Maximizing the natriuretic effect of endogenous atriopeptin in a rat model of heart failure. Proc Natl Acad Sci USA 1990; 87: 6465–6469 26. Gagelmann M, Hock D, Forssmann W-G. Urodilatin (CDD/ ANP-95–126) is not biologically inactivated by a peptidase from dog kidney cortex membranes in contrast to atrial natriuretic peptide-cardiodilatin (a-ANP/CDD-99–126). FEBS Lett 1988; 233: 249–254 27. Rosmalen FMA, Tan ACITL, Tan HS, Benraad TJ. A sensitive radioimmunoassay of ANP in human plasma, using a tracer with an immobilized glycouril agent. Clin Chim Acta 1987; 165: 331–340 28. Chai SY, Sexton PM, Allen AM, Figdor R, Mendelsohn AO. In vitro autoradiographic localization of ANP receptors in rat kidney and adrenal gland. Am J Physiol 1986; 250: F753–757 Received for publication: 3.2.98 Accepted in revised form: 3.6.98