INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

PHARMACOLOGICAL AND MOLECULAR biology approaches have identified a number of angiotensin receptor subtypes that have high affinity for angiotensin-(1---8) (ANG II) and its NH₂- and COOH-terminal deleted peptide fragments. These include the AT₁ and AT₂ receptors (and their subtypes) that have high affinity and specificity for both ANG II and angiotensin-(2---8) (known as ANG III), the AT_(1---7) receptor that has high affinity and specificity for angiotensin-(1---7), and the AT₄ receptor that has high affinity and specificity for angiotensin-(3---8) (known as ANG IV) (1, 11, 36, 40).

The AT₁ and AT₂ receptors have been identified throughout the kidney, with the AT₁ receptor largely responsible for mediating the actions of ANG II on renal microcirculation, tubular transport processes, and cell growth, with a counterregulatory role proposed for the AT₂ receptor (1, 4). Although the results of many studies support the existence of an AT_(1---7) receptor (26, 36, 38), such a receptor has not yet been identified in the kidney. Nonetheless, a growing number of functional studies have proposed that such a renal AT_(1---7) receptor(s) may be involved in regulating kidney hydroelectrolyte balance (36). Receptor autoradiography has demonstrated the expression of AT₄ receptors in the rat kidney (17, 25) that are localized to the apical membrane and cell body of proximal convoluted and straight tubules within the cortex and outer stripe of the outer medulla (24). In addition, the results from radioligand binding studies have alluded to the possible presence of AT₄ receptors on apical and basolateral membranes derived from freshly isolated rabbit cortical tubules (12). All other published studies have employed established kidney cell lines, which are representative of various segments of the nephron, or culture-specific kidney cell types to demonstrate localized expression of biologically functional AT₄ receptors in rat mesangial cells, opossum proximal tubule cells, porcine proximal tubule cells, human proximal tubule cells, human collecting duct cells, rabbit collecting duct cells, and bovine distal tubule/collecting duct cells (6, 7, 13, 23; see also citations in Ref. 22). Collectively, the results of these studies suggest that the AT₄ receptor can be expressed in proximal and distal segments of the nephron and can activate several intracellular signaling systems to elicit a biological response.

However, there is the possibility that in vitro culturing of cells may allow the expression of AT₄ receptors that are not normally found in vivo. Cultured rat glomerular mesangial cells contain abundant AT₄ receptors (6), whereas freshly isolated rat glomeruli (6) or glomeruli within normal rat kidney sections (24) do not appear to express AT₄ receptors. Therefore, it is unclear whether the expression of biologically functional AT₄ receptors in specific nephron cell types under culture conditions are representative of that found in the normal kidney. Consequently, the aims of the present study were threefold: first, to determine whether the expression of AT₄ receptors in the normal kidney is a common finding across species; second, to localize and compare the kidney distribution of the AT₄ receptor with that of other angiotensin receptors [e.g., AT₁ and AT_(1---7) receptor] and to determine whether the cellular localization of the AT₄ receptor was consistent with findings in cell culture systems; and third, to determine whether the expression of renal AT₄ receptors, and other identifiable kidney angiotensin receptor subtypes, could be altered by physiological, pharmacological, and pathological factors.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Autoradiographic studies. The kidneys from various adult mammalian [male rat (Sprague Dawley and Wistar, 270-500 g), male rabbit (New Zealand, 2.5-3.5 kg), male mongolian gerbil (meriones unguiculatus, 49 days old), male mouse (C57BL/6, ~2 mo old), male guinea pig (Hartley)] and avian [male house sparrow (passer-domesticus), female chicken (gallus gallus), and male turkey (meleagris gallopavo)] species were removed, frozen in isopentane at 25°C, and stored at 70 or 90°C until sectioned. Autoradiographic analysis of radioiodinated peptide binding to tissues was performed using 20-µm kidney sections mounted on gelatin-coated slides. Initially, sections were preincubated for 30 min in isotonic buffer [150 mM NaCl, 50 mM Tris, 50 µM Plummer's inhibitor (carboxypeptidase inhibitor), 20 µM bestatin (aminopeptidase inhibitor), 5 mM EDTA, 1.5 mM 1,10-phenanthroline (divalent ion chelators), and 0.1% heat-treated BSA at pH 7.4] at room temperature and then incubated in isotonic buffer containing ~0.4 nM ¹²⁵I-labeled divalinal-ANG IV (AT₄ receptor ligand), ~0.4 nM ¹²⁵I-[Sar¹, Ile⁸]ANG II or ¹²⁵I-[Sar¹, Thr⁸]ANG II (AT₁/AT₂ receptor ligands) with or without 10 µM displacers (unless stated otherwise) for 30 min. For the binding of ~0.8 nM ¹²⁵I-ANG-(1---7) [AT_(1---7) receptor ligand] to kidney sections, the incubation buffer described above was modified to contain 1 mM EDTA, and supplements were added to the isotonic buffer consisting of 10 µM amastatin (aminopeptidase inhibitor), 1 mM phenylmethylsulfonyl fluoride (serine protease inhibitor), 1 mM captopril (angiotensin-converting enzyme inhibitor), and 5 mM MgCl₂. All radiolabeled kidney sections were then rinsed with 3 × 2-min isotonic buffer washes, dried, and exposed to X-ray film (Kodak OMAT AR film in Wolf cassettes, stored at 90°C for 1-31 days, and then developed with Kodak D-19).

Pathophysiological rat kidney models. Male Wistar rats (~270 g) were individually housed in metabolic cages with unrestricted access to food and water, and on day 6 two rats were subjected to one of the following experimental protocols. For group 1, water diuresis was induced by 10% D-glucose in the drinking water with no access to food pellets for 4 days. For group 2, animals had daily intraperitoneal injections of furosemide (100 mg · kg¹ · day¹) and access to a high-electrolyte drinking solution (8 g/l NaCl and 1 g/l KCl) for 8 days. For group 3, animals consumed N-nitro-L-arginine methyl ester (L-NAME; ~100 mg · kg¹ · day¹) in the drinking water for 20 days. For group 4, a single intravenous injection of puromycin aminonucleoside (PAN; ~80 mg/kg) and were monitored for 10 days. Group 5 rats underwent bilateral ureteral obstruction (BUO) for 24 h. Group 6 comprised untreated rats monitored for 20 days. In other control experiments, we found that one rat subjected to a sham BUO (untied ligatures placed loosely around both ureters and then left in the abdomen) for 24 h, or one rat given daily intraperitoneal injections of saline for 4 days, had a similar density and distribution of kidney AT₄ and AT₁ receptors as in untreated rats monitored for 20 days. Daily measurements were made of body weight, fluid intake, urine volume output, urinary Na⁺ output, urine osmolality, and protein content for several of the experimental groups. A nonquantitative measurement of protein content in the daily output of urine (and in the bladder at the termination of the experiment) was performed by mixing urine with 5% sulfosalicylic acid solution (1:1, vol/vol) and visually ranking the degree of protein precipitation: 0 (clear solution), +1 (cloudy solution), +2 (cloudy solution with small white precipitates), +3 (cloudy solution with larger white precipitates), and +4 (large lumpy white precipitates). At the end of the experiment, the rats were anesthetized with pentobarbital sodium, urine was collected from the bladder, and then the kidneys were briefly perfused with warm PBS to remove blood. The kidneys were then excised, frozen, and prepared for autoradiographic studies as described above, except that 1,10-phenanthroline was omitted from the incubation buffer solution. All animals were subjected to institutional guidelines for the care and use of research animals.

Iodination of angiotensin peptides. All angiotensin-related peptides were monoiodinated using a previously described chloramine T method (23).

Drugs and materials. We received gifts of losartan (DuP 753) from DuPont/Merck Pharmaceuticals (Wilmington, DE), PD-123319 from Parke-Davis Pharmaceuticals (Ann Arbor, MI), divalinal-ANG IV, [Nle¹-leual³]ANG IV and [Nle¹]ANG IV from Dr. Joseph W. Harding (Washington State University, Pullman, WA), and other angiotensin-related peptides were purchased from Bachem (Torrance, CA) and Sigma (St. Louis, MO). Kidney autoradiograms were quantified and analyzed by the MCID-M1 image-analysis system (version 4.2, rev. 2.0, Imaging Research, Brock University, St. Catharines, ONT). Kidneys shown in figures were imaged by either MCID, MagnaFire Camera Imaging software (version 1.0, rev. A from Olympus America, Melville, NY) or a UMAX Astra 1220P scanner.

Statistics. All values represent means ± SE. Multiple groups were analyzed by one-way ANOVA with a post hoc Dunnett's test. Significance between two groups was analyzed by an unpaired Student's t-test. Differences between means were taken to be significant at the 5% level.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

All results shown in the present study are from kidneys that were not exposed to preservative agents before sectioning. We have found that treating kidneys with paraformaldehyde, gluteraldehyde, or several other preservative agents at the time of tissue harvesting can dramatically alter the distribution and specificity of ¹²⁵I peptide binding; the effect of which varies according to tissue species as well as the ¹²⁵I ligand employed to detect the receptor site. Kidneys from rats and gerbils were either quickly excised and frozen (similar to that employed for all other species) or perfused with PBS before being excised and frozen and produced the same patterns of ¹²⁵I peptide binding to renal structures.

Distribution of AT₄ receptors in the kidney. ¹²⁵I divalinal-ANG IV binding to mammalian kidney structures could be displaced by an excess of nonradioactive divalinal-ANG IV or [Nle¹-leual³]ANG IV (AT₄ receptor antagonists), ANG IV or [Nle¹]ANG IV (AT₄ receptor agonists), but not by a cocktail of receptor antagonists consisting of losartan (AT₁ receptor antagonist), PD-123319 (AT₂ receptor antagonist), and D-alanine⁷ (D-Ala⁷) ANG-(1---7) [AT_(1---7) receptor antagonist]. These results support previous findings that ¹²⁵I-divalinal-ANG IV binds to the renal AT₄ receptor (22-24).

View larger version (61K): [in this window] [in a new window]   Fig. 1.   Pseudocolor representation of angiotensin binding in mammalian kidneys. 125I-divalinal-ANG IV binding in kidney sections exposed for 7-40 h and 48 h-7 days, respectively (left and middle). Right column shows [125I-Sar1, Ile8]ANG II binding sites in kidney sections exposed for 7-20 days. Kidney sections were from the rat (A-C), mouse (D-F), guinea pig (G-I), rabbit (J-L), and gerbil (M). Pseudocolor calibration strip shows the relative density of 125I peptide binding. Because little or no nonspecific binding was found, only total binding is shown. Computer software programs allowed kidney images to be cut and transferred onto a blue background that approximated background binding.

View larger version (157K): [in this window] [in a new window]   Fig. 2.   Location of AT4 receptors in the rat kidney: proximal convoluted and straight tubules in the cortex (darkfield; A), proximal straight tubules in the outer stripe of the outer medulla (darkfield; B), and vascular bundles in the inner stripe of the outer medulla (C and D show light- and darkfield microphotographs, respectively); mouse kidney: proximal convoluted and straight tubules in the cortex (darkfield; E), proximal straight tubules in the transition of cortex to outer stripe of the outer medulla (F and G show light- and darkfield microphotographs, respectively), and proximal straight tubules in the transition of outer stripe of the outer medulla to the inner stripe (darkfield; H); guinea pig kidney: proximal convoluted and straight tubules in the cortex (darkfield; I), all cell types present in the outer medulla (darkfield; J), collecting tubules in the inner medulla (K and L show light- and darkfield microphotographs, respectively), and interstitium (most likely interstitial cells) of the papilla (M and N show light- and darkfield microphotographs, respectively); gerbil kidney: proximal convoluted and straight tubules in the cortex (darkfield; O) and proximal straight tubules in the outer stripe of the outer medulla (darkfield; P). Arrows show an absence of binding in cortical glomeruli.

View larger version (104K): [in this window] [in a new window]   Fig. 3.   Location of AT4 receptors in the rabbit kidney (light- and darkfield microphotographs; A and B, respectively), proximal convoluted and straight tubules in the cortex (light- and darkfield microphotographs; C and D, respectively), proximal straight tubules in a cortical medullary ray (light- and darkfield microphotographs; E and F, respectively); absence of receptors in a cortical artery and ascending arteriole (light- and darkfield microphotographs; G and H, respectively); no receptor expression in arcuate artery, and abundant expression in proximal straight tubules in the outer stripe of the outer medulla (light- and darkfield microphotographs; I and J, respectively), and faint signal in collecting tubules in the inner medulla (light- and darkfield microphotographs; K and L, respectively). Glomeruli shown in C and D (middle right), E and F (top left and top right corners), and G and H (top right corner) did not appear to express AT4 receptors.

View larger version (118K): [in this window] [in a new window]   Fig. 4.   Binding of 125I-divalinal-ANG IV to the rabbit renal artery (A-C), renal vein (D-F), and ureter (G-I). Left and middle columns show light- and darkfield microphotographs of total 125I-divalinal-ANG IV binding, respectively, whereas the right column shows darkfield microphotographs of nonspecific binding (in the presence of 10 µM divalinal-ANG IV).

View larger version (98K): [in this window] [in a new window]   Fig. 5.   Pseudocolor representation of total 125I-divalinal-ANG IV binding and nonspecific binding (in the presence of 10 µM divalinal-ANG IV) in the turkey (A and B, respectively), chicken (C and D, respectively), and sparrow (E and F, respectively) kidney. Pseudocolor calibration bar (right) indicates density of divalinal-ANG IV binding (fmol/g) corresponding to colors in images.

Distribution of AT₁ receptors in the kidney. Figure 1 demonstrates the autoradiographic localization of AT₁ receptors in several mammalian kidneys. We found that ¹²⁵I-[Sar¹, Ile⁸]ANG II binding to kidney structures could be abolished by excess nonradioactive [Sar¹, Ile⁸]ANG II (AT₁/AT₂ receptor antagonist), or [Sar¹]ANG II (metabolically resistant analog of ANG II and an AT₁/AT₂ receptor agonist). ¹²⁵I-[Sar¹, Ile⁸]ANG II binding throughout the kidney of all mammals was substantially reduced or abolished by losartan but not significantly affected by PD-123319 or D-Ala⁷ ANG-(1---7). Other investigators have reported that ¹²⁵I-[Sar¹, Ile⁸]ANG II binding to kidney structures is unaffected by AT₄ receptor ligands (25). In additional experiments, we found that ¹²⁵I-[Sar¹, Thr⁸]ANG II binding [total binding in the presence of divalinal-ANG IV, PD-123319 and D-Ala⁷ ANG-(1---7), each 1 µM] to rat kidneys was abolished by the further addition of 10 µM losartan to the incubation buffer. These results suggest that the majority of ¹²⁵I-[Sar¹, Ile⁸]ANG II and ¹²⁵I-[Sar¹, Thr⁸]ANG II binding to kidney structures reflects the presence of losartan-sensitive AT₁ receptors.

Distribution of putative AT_(1-7) receptor binding sites in the kidney. Tissue sections labeled with ¹²⁵I-ANG-(1---7) were exposed to either Kodak OMAT film for 31 days at 90°C or Kodak NTB2 emulsion for 14 wk at 90°C. The results obtained from both autoradiographic techniques revealed that the total (in the presence of divalinal-ANG IV and PD-123319, each 1 µM) and nonspecific [in the presence of divalinal-ANG IV, PD-123319, and 10 µM ANG-(1---7)] density of ¹²⁵I-ANG-(1---7) binding was similar in kidneys obtained from the rat (normal, sham, or subjected to BUO), mouse, gerbil (postnatal day 1, 5, 10, and 49), rabbit, sparrow, chicken, and turkey. A lack of specific binding was also found in the normal rat and rabbit kidney using 0.4 nM ¹²⁵I-[D-Ala⁷]ANG-(1---7) ± 10 µM [D-Ala⁷] ANG-(1---7). In addition, we have found that 10 µM [D-Ala⁷]ANG-(1---7) did not significantly compete with ¹²⁵I-[Sar¹, Ile⁸]ANG II binding in the rat (control or treated with glucose water, L-NAME, PAN, or furosemide), mouse, and rabbit kidney. Some of the results from rat kidney are shown in Fig. 6.

View larger version (110K): [in this window] [in a new window]   Fig. 6.   Total and nonspecific [in presence of 10 µM ANG-(1---7)] binding of 0.4 nM 125I-ANG-(1---7) for 30 min to rat kidneys briefly perfused with PBS at room temperature, followed by PBS plus 20% sucrose at room temperature; kidneys were removed and placed in ice-cold PBS containing 20% sucrose for ~3 h and then frozen (A and B, respectively; exposed to X-ray film for 16 days); total (in presence of divalinal-ANG IV and PD-123319, each 1 µM) and nonspecific [in presence of divalinal-ANG IV, PD-123319, and 10 µM ANG-(1---7)] binding of 0.8 nM 125I-ANG-(1---7) for 30 min to nonperfused rat kidneys (C and D, respectively; exposed to X-ray film for 31 days); incubation was performed with a modified isotonic buffer as described in METHODS; total and nonspecific [in presence of 10 µM [D-Ala7]ANG-(1---7)] binding of 0.4 nM 125I-[D-Ala7]ANG-(1---7) for 90 min to rat kidneys briefly perfused with PBS at room temperature (E and F, respectively; exposed to X-ray film for 16 days); total binding and competition by 10 µM [D-Ala7]ANG-(1---7) of 0.4 nM 125I-[Sar1, Ile8]ANG II incubated for 30 min to rat kidneys briefly perfused with PBS at 37°C (G and H, respectively; exposed to X-ray film for 24 days) are shown; isotonic incubation buffer did not contain 1,10-phenanthroline.

Metabolism of the radiolabeled ligand during the incubation assay. A total of 1-5 sets of slides (each set contained 10 slides) was incubated for 30 min each in the ¹²⁵I-peptide-containing buffer. In those experiments, where five sets of slides were incubated in the ¹²⁵I-peptide-containing buffer, we found no significant metabolism of ¹²⁵I-divalinal-ANG IV, ~6% metabolism of ¹²⁵I-Sar derivatives of ANG II, and ~20% metabolism of ¹²⁵I-ANG-(1---7). With regard to the latter ¹²⁵I-peptide, each set of slides contained a similar variety of tissue sections and produced similar autoradiographic results regardless of the order of incubation. These findings suggest that the absence of specific, high-affinity ¹²⁵I-ANG-(1---7) binding sites in the kidney was not due to substantial metabolism of the radioactive ligand, as the heptapeptide was likely degraded ~4% during each 30-min period of incubation with kidney section-mounted slides.

Density of kidney angiotensin receptors. As shown in Figs. 1 and 5, and graphically represented in Fig. 7, densitometric analysis of angiotensin receptor sites in the mammalian and avian kidney revealed that the average kidney density of AT₄ receptors was highest in the turkey and lowest in the mouse. Furthermore, the AT₄ receptor was several-fold more abundantly expressed than the AT₁ receptor in mammalian kidneys, with the one exception being the mouse kidney.

View larger version (28K): [in this window] [in a new window]   Fig. 7.   Density of divalinal-ANG IV-sensitive 125I-divalinal-ANG IV binding (AT4 receptors) and losartan-sensitive 125I-[Sar1, Ile8]ANG II binding (AT1 receptors) in mammalian and avian kidneys. nd, not determined.

Pathophysiological effects on kidney angiotensin receptors. All rats (with the exception of L-NAME-treated animals on the day 20 of treatment) tolerated the various treatments well. Body weight, water intake, urinary water and sodium excretion, urinary osmolality and protein content for several of the treatment groups are depicted in Fig. 8. As expected, rats given 10% glucose in their drinking water and no access to food had a progressive fall in body weight over the 4 days of treatment. These rats were successfully undergoing a water diuresis, as suggested by the findings of a markedly greater fluid intake and urinary fluid excretion, decreased urinary sodium excretion, and an extremely low urinary osmolality compared with the control group. Furosemide-treated rats had a higher fluid intake, a greater diuresis and natriuresis, and a reduced urinary osmolality compared with control rats. Rats consuming L-NAME (114 ± 12 mg · kg¹ · day¹) in their drinking water developed rapid hypertension (tail-cuff blood pressure measurement in control rats: 92 ± 1 mmHg vs. L-NAME-treated rats: 209 ± 5 mmHg on the day 18 of treatment), diuresis, reduced urinary osmolality, and proteinuria. There was no weight gain in L-NAME-treated rats throughout the treatment period. These nitric oxide-deficient rats appeared to be normal up until day 20 of treatment at which time they showed visible signs of distress, with a loss of body weight that necessitated the immediate harvesting of their kidneys (the kidneys from one rat were removed 10 min after death). Rats that were injected with PAN had a reduced urinary osmolality as well as a significant amount of protein in their urine. Metabolism measurements were not performed on rats subjected to BUO.

View larger version (31K): [in this window] [in a new window]   Fig. 8.   Measurements performed on rats in metabolic cages. A: body wt (open and shaded bars show body wt at the beginning and end of the experiment, respectively). B: water intake. C: urinary fluid excretion rate. D: absolute sodium excretion rate. E: urinary osmolality. Open and solid bars show 24-h urine and bladder urine values, respectively. F: qualitative index of urinary protein levels. *P < 0.05 from control values using 1-way ANOVA and Dunnett's test. Data were obtained from 2 animals/group.

View larger version (30K): [in this window] [in a new window]   Fig. 9.   Effect of treatment on the density of AT4 receptors (divalinal-ANG IV-sensitive 125I-divalinal-ANG IV binding) and AT1 receptors (losartan-sensitive 125I-[Sar1, Ile8]ANG II binding or losartan-sensitive 125I-[Sar1, Thr8]ANG II binding) in the rat kidney. *P < 0.05 from control values using 1-way ANOVA and Dunnett's test. For bilateral ureteral obstruction (BUO) group, *P < 0.05 from sham BUO values using an unpaired Student's t-test. Data were obtained from multiple kidney sections from 2 rats/group.

View larger version (145K): [in this window] [in a new window]   Fig. 10.   Distribution of kidney AT1 receptors (losartan-sensitive 125I-[Sar1, Ile8]ANG II binding) and AT4 receptors (divalinal-ANG IV-sensitive 125I-divalinal-ANG IV binding) in control rats (A and C, respectively) and rats drinking 10% glucose water for 4 days (B and D, respectively).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

A major finding of the present study was that AT₄ receptors were present in the normal kidneys of all mammalian and avian species examined. This suggests that the kidney AT₄ receptor is an evolutionarily conserved receptor and is likely of physiological importance.

Recently, we localized the rat AT₄ receptor to the apical and cell bodies of proximal convoluted tubules in the cortex as well as proximal straight tubules within cortical medullary rays and the outer stripe of the outer medulla (24). The present study extends the distribution of mammalian kidney AT₄ receptors to include cell types present within the inner medulla and papilla. In addition, we believe that our results are the first to describe the localization and distribution of AT₄ receptors in the normal mouse, gerbil, and rabbit kidney, as well as the avian kidney. Our results also show a more extensive distribution of AT₄ receptors in the guinea pig kidney than previously reported (40), with a high receptor density localized to the proximal straight tubules of the outer stripe of the outer medulla. Despite being unable to determine all cell types in the mammalian inner medulla and papilla that possess AT₄ receptors (due to the very low receptor density in these regions), we were able to identify AT₄ receptors on vascular bundles in the rat outer medulla; collecting tubules and interstitial cells in the guinea pig inner medulla and papilla, respectively; and the collecting tubules in the rabbit inner medulla.

AT₁ receptors were detected throughout the cortex and medulla of all mammalian kidneys, with an especially high density of binding associated with cortical glomeruli and vasa recta bundles transversing the inner stripe of the outer medulla, with less binding found in the interglomerular region (corresponding largely to proximal convoluted tubules) and the interbundle area of the inner stripe of the outer medulla (composed largely of thick ascending limbs and collecting ducts). This distribution of AT₁ receptors in the kidney is in general agreement with previous studies (33).

The avian angiotensin receptor has been cloned in the chicken and turkey and shows ~75% homology with the mammalian AT₁ receptor, as well as a markedly different pharmacology from its mammalian counterpart (30). High-affinity ANG II binding sites could not be detected in turkey kidney membranes (5), whereas a hybridization signal for angiotensin receptor mRNA was clearly detected in chicken kidney glomeruli and renal artery and was barely distinguishable from the background in proximal tubules (30). Our results show abundant expression of high-affinity AT₄ receptors throughout the turkey, chicken, and sparrow kidney and further support the notion that the AT₄ receptor may be the predominant angiotensin receptor subtype in the kidney.

A growing number of investigators have shown functional effects of ANG-(1---7) in the rat kidney that would be consistent with the presence of a renal binding site(s) that has high affinity for ANG-(1---7) (36). Indeed, several studies have successfully shown the existence of ANG-(1---7) binding sites with affinity for ANG-(1---7), [D-Ala⁷]ANG-(1---7), [Sar¹, Ile⁸]ANG II, [Sar¹, Thr⁸]ANG II, and ANG II in nonrenal cell types and tissues (26, 36, 38). However, we have found a lack of specific binding of ¹²⁵I-ANG-(1---7) [AT_(1---7) receptor agonist] or ¹²⁵I-[D-Ala⁷]ANG-(1---7) [AT_(1---7) antagonist] in the rat kidney, using several modified incubation buffers (19, present study), and that some of the biological actions of ANG-(1---7) and [D-Ala⁷]ANG-(1---7) in the kidney can potentially be mediated by AT₄ receptors (19-21). Previously, we reported that 10 µM ANG-(1---7) competed for ~70% of losartan-sensitive, [Sar¹, Thr⁸]ANG II-sensitive ¹²⁵I-ANG II binding to all rat kidney structures and assumed that this was largely due to the heptapeptide's known affinity for the AT₁ receptor (19). However, this result cannot exclude the possibility that ANG-(1---7) may have also recognized a small population of AT_(1---7) receptors that were sensitive to blockade by losartan, ANG II, and its analogs. This quandary is partially addressed by the finding that 10 µM [D-Ala⁷]ANG-(1---7) [high-affinity AT_(1---7) antagonist with a lower affinity for the AT₁ and AT₂ receptors than ANG-(1---7) (21)] did not significantly compete with ¹²⁵I-[Sar¹, Ile⁸]ANG II binding to the rat, mouse, and rabbit kidney (present study). Therefore, our findings suggest that high-affinity AT_(1---7) receptors are either absent or of such low density in the kidney that they are difficult to detect by conventional receptor autoradiographic techniques. Because tissue preparation, receptor ligand specificity, incubation buffer composition, and conditions can dramatically influence the ability to detect receptor sites, a negative finding should be interpreted with some caution (as is the case in all of science).

Few studies have examined the location and function of the ANG IV-AT₄ receptor system in the normal kidney. Reports that ANG IV acts on rat kidney AT₄ receptors to increase cortical blood flow (9) appears incompatible with those reporting only a concentration-dependent, AT₁ receptor-mediated decrease in rat kidney blood flow (14). We did not detect AT₄ receptors in the rat renal artery and its smaller cortical branches, although a faint signal was observed over medullary vascular bundles that contain ascending and descending vasa recta as well as thin loops of Henle. In contrast, AT₄ receptors were readily demonstrated in the rabbit renal artery and vein but not in smaller cortical blood vessels, suggesting a heterogeneous distribution of receptors along the rabbit renal vascular tree. The role of the ANG IV-AT₄ receptor system in renal vascular function could perhaps be resolved with the rabbit kidney being best examined. The AT₄ receptor is not expressed in the glomeruli of the normal rat kidney (24) or freshly isolated rat glomeruli (6). Our results extend this observation to kidney glomeruli of other mammalian species, as well as several rat kidney pathophysiological models that impact on glomerular function. A consistent finding in the present study was that mammalian kidney AT₄ receptors were heavily associated with tubular epithelia involved in solute transport processes, especially the proximal tubule. Activation of the tubular AT₄ receptor system has been shown to decrease active proximal tubule solute transport (24), increase proximal tubule intracellular Na⁺ concentration (22), and increase whole kidney urinary sodium and water excretion that is independent of sympathetic innervation (18, 22). Interestingly, Wistar-Kyoto rats fed a high-salt diet (conditions in which the kidney attempts to excrete excess sodium in the urine) had an increase in proximal tubule AT₄ receptor density (17). Studies in proximal tubule and collecting duct epithelial cell culture systems have also implicated the AT₄ receptor system in regulating plasminogen-activating inhibitor type 1 (PAI-1) expression (15), mitogen-activated protein (MAP) kinase activity (20, 22), focal adhesion kinase (FAK) and paxillin activity (7), and intracellular Ca²⁺ levels (13, 22, 23). Our results support these findings in tubular epithelial cell culture systems in that AT₄ receptors can be expressed in both proximal and distal segments of the nephron. The distribution of functional AT₄ receptors on an array of kidney structures and cell types, which are linked to multiple intracellular signaling systems, suggests that we are only beginning to explore the many possible actions of the AT₄ receptor system on kidney physiology.

Because the natural endogenous ligand(s) of the kidney AT₄ receptor is unknown, we have attempted to identify treatments that result in a marked alteration in kidney AT₄ receptor expression, in the hope that they may provide potentially interesting rat models for future exploration of the role of the AT₄ receptor system in kidney function. Although each treatment will likely produce complex changes in many kidney systems, we point out potential factors that may contribute to the observed change in receptor expression.

The presence of a functional countercurrent multiplier system in the renal medulla allows the generation of an interstitial osmotic gradient that increases from the corticomedullary junction to the tip of the papilla (32). Osmolality is known to regulate receptor and enzyme systems in cell types contained within the inner medulla (39, 42). Consequently, we examined whether the very low density of AT₄ receptors found in the inner medulla of the rat kidney may be related to the medulla's high osmolality. Rats fed glucose water excreted a hyposmolar urine on day 2 of treatment, which was likely due to both a reduction in the circulating level of vasopressin (not an unreasonable expectation given that fluid intake was ~7 times greater in glucose-fed rats compared with control rats), leading to a reduction in water being absorbed from the collecting ducts (32), as well as a progressive dissipation of the corticomedullary osmolal gradient, with dramatic decreases in medulla and papilla osmolytes and osmolality, during prolonged maintenance of water diuresis (2, 3, 41). The density and distribution of AT₄ receptors in the kidneys of such animals were not dramatically altered, whereas AT₁ receptors were additionally expressed throughout the inner medulla. This novel finding leads us to hypothesize that tissue osmolality may be a key factor in regulating binding and/or expression of the AT₁ receptor, but not of the AT₄ receptor, in the rat kidney. Support for this contention comes from a recent publication which demonstrated that decreasing the osmolality of a vasopressin-free medium surrounding cultured rat renomedullary interstitial cells resulted in an upregulation of AT₁ receptor binding sites (31). In vivo measurements of the corticomedullary osmotic gradient, as well as the determination of the threshold and duration of a reduced medullary osmolality to influence AT₁ receptor binding, will be critical in determining whether this hypothesis has merit. The functional consequence and mechanism(s) by which tissue osmolality regulates kidney AT₁ receptor expression, and presumably ANG II actions in the kidney medulla, are presently unknown.

We have also attempted to pharmacologically reduce the osmotic gradient in the rat kidney by daily injections of furosemide, a loop diuretic whose prime action is to inhibit the electroneutral Na⁺-K⁺-2Cl cotransport system in the thick ascending limb of the loop of Henle. Rats were fed an electrolyte-enriched solution in an attempt to maintain a positive Na⁺ and K⁺ balance and to reduce the degree of contraction of the extracellular fluid volume resulting from the furosemide-induced loss in urinary water and electrolyte excretion. We found that daily furosemide treatment did not affect kidney AT₁ receptor distribution, which could be argued as opposing our hypothesis that low tissue osmolality may upregulate medullary AT₁ receptor expression. Previous studies have shown that furosemide treatment resulting in the excretion of an isosmotic urine (37) had an almost identical reduction in medullary osmolality as rats infused with glucose and undergoing a hypotonic water diuresis (3). In the present study, furosemide-treated rats did not achieve the expected excretion of an isosmotic urine (daily urine osmolality fluctuated between 996 and 725 mosmol/kgH₂O during the last 6 days), suggesting that the corticomedullary osmotic gradient had not been ablated and that likely higher tonicity remained within the medulla than in glucose-water fed rats undergoing a hypotonic water diuresis. Coupled with the fact that there was a tendency for a reduction in kidney AT₁ receptor binding [presumably as the overall consequence of elevated ANG II levels having opposing effects on AT₁ receptor expression in glomeruli, proximal tubules, and other renal cell types (8)], this may explain our finding of an unchanged distribution of kidney AT₁ receptors in the furosemide-treated group. Measurements of tissue osmolality will help clarify whether our assumption that medullary osmolality is higher in intraperitoneal furosemide-treated rats than glucose-water fed rats is indeed correct. Unexpectedly, we found that furosemide treatment resulted in a profound decrease in kidney AT₄ receptor density. Further studies are needed to elucidate whether the mechanism(s) responsible for the decrease in AT₄ receptor expression is related to furosemide's action on membrane transporters and/or secondary to altered hormone status.

Administration of PAN to rats induced an intense proteinuria that is generally regarded to be due to selective injury of glomerular epithelial cells (16). This experimental rat model of nephrotic syndrome is also associated with attributes of tubulointerstitial nephritis and interstitial fibrosis (27). Chronic inhibition of nitric oxide biosynthesis with L-NAME caused severe systemic hypertension and proteinuria and has also been known to be accompanied by renal fibrosis (28). Only L-NAME-treated rats were associated with a decrease in kidney AT₁ receptor density, and this is a finding supported by the recent report that glomerular AT₁ receptor mRNA levels were reduced during chronic L-NAME treatment in Dahl-Iwai salt-sensitive rats (29), perhaps as a consequence of a locally activated renin-angiotensin system (28, 29). This decrease in kidney angiotensin receptor expression may be an attempt to limit the known role of the renal AT₁ receptor system in the progression of renal damage induced by chronic L-NAME treatment (28). Both PAN and L-NAME treatments demonstrated a preferential and marked decrease in kidney AT₄ receptor density. We had previously observed a similar reduced kidney AT₄ receptor density in adult male Sprague-Dawley rats progressively treated with 30, 60, and 100 mg · kg¹ · day¹ L-NAME over a 4-wk period (Grove KL and Deschnepper CF, unpublished observations; kidneys were kindly provided by Drs. Grove and Deschnepper). Recent studies have implicated the AT₄ receptor system as playing a role in the regulation of renal tissue architecture by virtue of its ability to stimulate the synthesis of PAI-1, a fibrogenic molecule, and phosphorylate FAK/paxillin signaling proteins (part of the focal adhesion complex that is involved in cell/extracellular matrix remodeling) in endothelial, mesangial, and epithelial cells of the kidney (7, 15, 35). Consequently, it would be of interest to examine whether the preferential decrease in kidney AT₄ receptor expression (limiting the role of the AT₄ receptor system?) in PAN- and L-NAME-induced renal failure may potentially be in response to the renal fibrogenic process.

BUO of short duration (24 h) is known to impair glomerular hemodynamics and tubular reabsorptive function with only minimal ultrastructural changes of the parenchyma (10). We found that there was a selective reduction in kidney AT₁ receptor density in rats after 24 h of ureteral obstruction. This finding is in agreement with previous studies in adult rats that have proposed that the downregulation in kidney AT₁ receptor expression during acute ureteral obstruction represents a negative feedback loop response to increased local ANG II production (34). This would, in turn, suggest that such increases in the intrarenal level of ANG II per se do not regulate kidney AT₄ receptor expression, at least within a 24-h time period.

We did not perform saturation binding isotherm experiments and, therefore, our results do not reveal whether the alteration in receptor density, in various models of renal dysfunction, was due to a change in total receptor number (downregulation and/or loss due to epithelial cell necrosis) or receptor affinity. Regardless, our results have identified a number of treatments that can selectively influence the density and/or distribution of AT₄ and AT₁ receptors in the kidney. Both the functional and mechanistic implication(s) of these changes in angiotensin receptor expression in renal pathophysiology can now be explored.

In conclusion, our results demonstrate the presence of AT₄ receptors on a wide array of cell types within the mammalian and avian kidney, thus providing some confidence that previous findings in kidney cell culture systems are likely to be applicable to the normal kidney. Furthermore, a number of interventions were identified that appear to selectively regulate AT₄ and AT₁ receptor expression in the kidney.