Renal Physiology

Angiotensin IV AT4-receptor system in the rat kidney

Rajash K. Handa, Luke T. Krebs, Joseph W. Harding, Shelly E. Handa


Angiotensin IV, {[des-Asp1,Arg2]ANG II or ANG-(3—8)}, has been shown to preferentially bind to a novel angiotensin binding site (AT4receptor). The cellular location and function of this receptor in the rat kidney is unknown. Autoradiography localized AT4 receptors to the cell body and apical membrane of convoluted and straight proximal tubules in the cortex and outer stripe of the outer medulla. ANG IV (0.1 pM-1 μM) elicited a concentration-dependent decrease in transcellular Na+ transport (as measured by proximal tubule O2 consumption rates) in fresh suspensions of control or nystatin-stimulated (bypasses rate-limiting step of apical Na+entry) rat proximal tubules. The inhibitory effect of 1 pM ANG IV was unaltered by either 1 μM losartan (AT1-receptor antagonist) or 1 μM PD-123319 (AT2-receptor antagonist) and yet was abolished by 1 μM divalinal-ANG IV (AT4-receptor antagonist) or ouabain pretreatment. These results demonstrate that the kidney AT4-receptor system is localized to the proximal tubule and suggests that one potential biological role of this system is in the regulation of Na+ transport by inhibiting a ouabain-sensitive component of Na+-K+-adenosinetriphosphatase activity in the rat.

  • oxygen consumption
  • sodium transport
  • proximal tubule
  • autoradiography

the renin-angiotensin system is composed of a cascade of biochemical reactions involving the generation and processing of the decapeptide, angiotensin (ANG) I, and is one of the basic homeostatic mechanisms for maintaining the internal environment of the organism. The general consensus has been that the octapeptide, ANG II, and its shorter fragment [des-Asp1]ANG II (ANG III) were the only biologically active products of the renin-angiotensin system, having a vast range of actions throughout the body, including potent vasoconstrictor and dipsogenic effects, neuromodulatory activity (central and peripheral nervous system), endocrine actions (e.g., stimulating aldosterone and vasopressin secretion), regulating epithelia transport (most notably increasing renal proximal Na+, Cl, and HCO3 reabsorption), and influencing cell remodeling (hyperplasia and hypertrophy) (23). Two types of the ANG II receptor have been cloned and termed AT1 (with at least AT1A, AT1B, and AT1C subtypes) and AT2 (proposed subtypes). Most of the functional responses to ANG II and ANG III in the brain and periphery have been ascribed to AT1-receptor activation. However, recent studies (41, 43) suggest that AT2-receptor stimulation may counteract AT1-mediated events in processes, such as blood pressure control, growth, and dipsogenesis.

There is now mounting evidence that processing of ANG I to fragments smaller than ANG II and ANG III [e.g., ANG-(1—7) and ANG IV] can produce a receptor-mediated biological response (17, 38,46). A new angiotensin binding site, distinct from ANG II type AT1 and AT2 receptors, has been pharmacologically described and demonstrates high specificity and affinity for the hexapeptide, ANG IV (42). This novel binding site has been designated AT4 and, to date, has been shown to be heavily distributed in the brain, spinal cord, aorta, heart, lung, uterus, colon, prostate, adrenals, bladder, kidney, vascular smooth muscle, and endothelial cells of several species (human, monkey, bovine, porcine, horse, sheep, cat, rabbit, rat, and guinea pig) (32, 34, 42, 46). ANG IV can be formed from ANG II and ANG III by the action of aminopeptidases (20, 25) and potentially from ANG I-(3—10) by an angiotensin-converting enzyme-dependent pathway (5,14). The hexapeptide appears to have weak classic ANG II activities, such as effects on blood pressure and thirst, and yet possesses unique biological attributes potentially important in memory, cell growth, and cardiovascular control (16, 25, 35, 46). Although the kidney contains high-affinity ANG IV binding sites, the functional significance of this receptor system is unknown (8, 10, 11, 21). The hexapeptide has been reported to inhibit (26) or have no effect (31) on renal renin secretion and yet appears to be a potent potentiator of renal cortical blood flow, which was shown to be mediated by a novel ANG IV receptor and linked to the stimulation of nitric oxide (6, 7, 42). The actions of ANG IV on other aspects of renal physiology are unknown. The aim of the present study was 1) to provide evidence for the presence, distribution, and specificity of AT4-receptor sites in the rat kidney, 2) to determine the possible influence of ANG IV on proximal epithelial Na+ transport,3) to pharmacologically characterize the receptor involved in mediating the tubular action of ANG IV, and4) to gain some insight into the mechanism(s) by which ANG IV exerts its action on cellular Na+ transport.


Autoradiographic studies.

Male rats were anesthetized with an intraperitoneal injection of equithesin or pentobarbital sodium, and the kidneys were perfused with phosphate-buffered saline (PBS, pH 7.4 at room temperature) in vivo. The kidneys were then removed, frozen in isopentane at −20°C, and stored at −70°C until sectioned. Autoradiographic analysis of rat kidney binding was performed using 20-μm tissue sections mounted on gelatin-coated slides. Initially, sections were preincubated for 30 min in isotonic buffer [150 mM NaCl, 50 mM tris(hydroxymethyl)aminomethane, 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 bovine serum albumin (BSA) at pH 7.4] at room temperature and then incubated in isotonic buffer containing 0.4 nM labeled ligand with or without 10 μM displacers for 25 min, rinsed with 3 × 2 min isotonic buffer washes, dried, and exposed to X-ray film (Kodak 5B5 in Wolf cassettes, stored at −70°C for 24–48 h, and then developed with Kodak D19). The incubation buffer contained an excess of aminopeptidase and carboxypeptidase inhibitors and ion chelators to prevent the metabolism of the radiolabeled probe and its binding to ANG IV-degrading enzyme proteins.

For emulsion-coated autoradiography, kidneys were initially perfused with PBS followed by 2% paraformaldehyde and 0.5% glutaraldehyde and finally with 20% sucrose in vivo. The kidneys were frozen, sectioned (12 μm), radiolabeled (see above procedure), and dried for several days. The radiolabeled sections were then postfixed with paraformaldehyde vapors at 80°C for 2 h, dehydrated by immersion in graded ethanols (50%-100% ethanol for 5 min), defatted in xylene (10 min), rehydrated in an inverse series of ethanols followed by distilled water, and then allowed to air dry. Slides were then uniformly coated with warm Kodak NTB-2 emulsion in a dark room, air dried for 3 h, and stored overnight at room temperature in desiccant-containing light-proof slide boxes followed by storage at −70°C for 7–20 days. After exposure, the slides were developed in Kodak D-19, rinsed in distilled water, fixed in Ektaflo (Kodak), and counterstained with hematoxylin and eosin. Sections were examined using both light- and dark-field microscopy.

Isolation of rat proximal tubules.

A suspension of cortical proximal tubules was obtained by a previously described method (17). All buffers and solutions used for the isolation of proximal tubules had a pH of 7.40 and an osmolality of 295 mosmol/kgH2O and were equilibrated with 95% O2-5% CO2. In brief, two male rats weighing 350–400 g were anesthetized with ketamine (50 mg/kg im) and xylazine (10 mg/kg im), and their kidneys were perfused at a constant pressure of ∼125 mmHg with Krebs-Henseleit buffer (KHB) at a temperature of 37°C. After a complete blood washout, perfusion was continued with the KHB solution supplemented with 1 mg/ml collagenase, 0.67 mg/ml hyaluronidase, and 0.67 mg/ml BSA. After enzyme perfusion for ∼5 min, the kidneys were excised, and the cortex was removed, minced, and incubated in the enzyme-supplemented KHB for 20 min at 37°C and aerated with 95% O2-5% CO2. The tubule suspension was subsequently washed three times by centrifugation (50g for 2 min), followed by resuspension with ice-cold KHB, and then washed three times with ice-cold Ca2+-free KHB to reduce clumping during the isolation step for proximal tubules. The tubule solution was suspended in 45% isosmotic Percoll solution, and a band of proximal tubules was isolated by centrifugation at 4°C for 10 min at 19,430g. The proximal tubule fraction was washed three times to remove the Percoll, divided equally between several test tubes, and then stored on ice in ice-cold KHB until O2 consumption measurements were performed.

Determination of tissue O2 consumption rates.

Suspensions of proximal tubules were incubated for 10 min at 37°C in a shaker bath and aerated with 95% O2-5% CO2 before O2 consumption rate (Qo 2) measurements. A 100-μl aliquot of proximal tubules was placed in a thermoregulated 2-ml chamber containing 1.85 ml KHB, which was then sealed, and Qo 2 was measured polarographically with a Clarke oxygen electrode. We have previously reported that receptor-mediated Qo 2stimulatory and inhibitory pathways remain intact after rat proximal tubule isolation procedure (17). The measurement of Qo 2 can be used as a direct reflection of Na+-K+-adenosinetriphosphatase (ATPase) activity and Na+transport because of the tight coupling between Na+-K+-ATPase activity and mitochondrial oxidative phosphorylation in the proximal tubule (30). In addition, ouabain-suppressible Qo 2provides an index of active transport (30).

In the studies that used angiotensin-receptor antagonists, the blockers were added to both the proximal tubule suspension (incubated for 10 min as described above) and the chamber. Other drugs [angiotensin peptides, fenoldapam, platelet-activating factor (PAF), 5 mM nystatin, and 5 mM ouabain] were added in 25-μl boluses into the tubule-containing chamber via its injection port. The O2 tension in the closed chamber was recorded as a function of time, and the resulting slope indicated the Qo 2, which was calculated as a function of tubular protein content as measured by the method of Lowry et al. (29). To minimize the variability in basal Qo 2 from different tubule preparations (range 17–39 nmol O2 ⋅ min−1 ⋅ mg protein−1), the effect of ANG IV and other receptor agonist treatments were expressed as a percent change from basal or nystatin-stimulated Qo 2. All drug solutions were prepared fresh daily, and their molar concentrations indicate the final concentrations achieved in the chamber.


We received gifts of fenoldapam (SKF-82526) from Smith Kline & Beecham, losartan (DuP-753) from Du Pont/Merck Pharmaceuticals, and PD-123177 and PD-123319 from Parke-Davis.d-[Val1]ANG IV, [Nle1]ANG IV, and divalinal-ANG IV {[Val1,3ψ(CH2NH)1–2,3–4]ANG IV previously known as WSU-1291} were prepared in the laboratory of J. W. Harding. ANG IV, ANG II, PAF (l-α-phosphatidylcholine, β-acetyl-γ-O-hexadecyl), nystatin, and other reagents were purchased from Sigma.


All values are presented as means ± SE. Multiple groups were analyzed by one- or two-way analysis of variance and the post hoc Student-Newman-Keuls test (Crunch Interactive Statistical Package or SigmaStat). Differences between means were taken to be significant at the 5% level.


Autoradiographic studies.

Figure 1 demonstrates the in vitro autoradiographic localization of AT4-receptor sites in the rat kidney. There was a moderate density of diffuse125I-ANG IV binding over the entire cortex, with an especially high density of125I-ANG IV binding localized to the outer stripe of the outer medulla. No binding was found in the inner stripe of the outer medulla, inner medulla, or papilla (Fig.1 A). The125I-ANG IV binding was displaced by both unlabeled ANG IV (Fig. 1 B) and the putative AT4-receptor antagonist, divalinal-ANG IV (Fig.1 C), but not by the AT1-receptor antagonist, losartan (Fig. 1 D), or the AT2-receptor antagonist, PD-123177 (Fig. 1 E). To confirm the specificity of divalinal-ANG IV to bind exclusively to the renal AT4 receptor, we also examined the distribution of 125I-divalinal-ANG IV binding sites. The localization of125I-divalinal-ANG IV (Fig.1 F) was identical to that of125I-ANG IV and was displaced by both ANG IV and divalinal-ANG IV (Fig. 1,G andH, respectively). There was no cross displacement of the125I-divalinal-ANG IV binding with losartan (Fig. 1 I) or PD-123177 (Fig. 1 J).

Fig. 1.

In vitro autoradiographic localization of AT4-receptor sites in rat kidney. Panels on left show total125I-ANG IV binding (A), +ANG IV (B), +divalinal-ANG IV (C), +losartan (D), and +PD-123177 (E). Panels onright show total125I-divalinal-ANG IV binding (F), +ANG IV (G), +divalinal-ANG IV (H), +losartan (I), and +PD-123177 (J). Concentration of125I-labeled peptides and unlabeled angiotensin-receptor agonist or antagonists were 0.4 nM and 10 μM, respectively (autoradiography performed byLTK).

Emulsion autoradiography revealed that the AT4 receptor was localized to proximal tubules in the cortex and outer stripe of the outer medulla (Figs. 2 and 3). The receptor was present on both the proximal tubule cell body as well as within its lumen. Staining sequential sections with periodic acid-Schiff reagent indicated that the binding observed within the dilated tubule lumen was due to receptors located on the apical brush border (not shown). The highest density of silver grains were observed in straight proximal tubules located in the outer stripe of the outer medulla and extending throughout the medullary rays. The density of binding sites was less in cortical convoluted proximal tubules and absent in descending thin limbs of Henle, ascending thin and thick limbs of Henle, distal convoluted tubules, and cortical and medullary collecting ducts. Despite the appearance of binding sites in the glomerulus (Fig. 3 B), we did not consistently find a difference between binding in superficial, midcortical, juxtamedullary glomeruli, and background (e.g., Fig.3 H, bottom left). A similar distribution of kidney AT4 binding sites was observed in sections treated with both a combination of losartan and PD-123177.

Fig. 2.

Emulsion autoradiography of rat kidney demonstrated AT4-receptor sites on cell body and in lumen of proximal tubules in outer stripe of outer medulla. Total 125I-divalinal-ANG IV binding: bright-field (A) and dark-field (B) photomicrographs. Nonspecific binding in presence of divalinal-ANG IV: dark-field (C) photomicrograph. Arrows show a single straight proximal tubule, ×100 magnification; mr, medullary ray; isom, inner stripe of outer medulla.

Fig. 3.

Emulsion autoradiography localized125I-divalinal-ANG IV binding sites to specific renal cell types. AT4 receptors on apical brush border of convoluted proximal tubule emerging from a superficial glomerulus: bright (A)- and dark-field (B) photomicrographs, ×400 magnification. Straight proximal tubule located in midcortical region of a medullary ray: bright (C)- and dark-field (D) photomicrographs, ×250 magnification. Transition of straight proximal tubule into descending thin limb of Henle’s loop in outer stripe of outer medulla: bright (E)- and dark-field (F) photomicrographs, binding in presence of losartan and PD-123177, ×250 magnification. Distal convoluted tubule entering a collecting duct in outer cortex: bright (G)- and dark-field (H) photomicrographs, ×250 magnification. Arrows in dark-field photomicrographs indicate location of binding sites; glom, glomerulus; pct, proximal convoluted tubule; pst, proximal straight tubule; dtl, descending thin limb of Henle’s loop; dct, distal convoluted tubule; ccd, cortical collecting duct.

Tissue Qo2concentration-response curve to ANG IV.

Because the results of the autoradiography study were consistent with the localization of AT4 receptors on proximal tubular structures, we examined its effect on proximal tubule Na+ transport. In preliminary studies (Fig. 4), we found that ANG IV inhibited Qo 2 in a concentration-dependent fashion in proximal tubules, where the movement of Na+ across the apical membrane and into the cell was rate limiting (control group). Treatment of proximal tubules with nystatin (Na+ ionophore) bypasses the rate-limiting step of apical Na+entry and permits extracellular Na+ to freely enter the cell and intracellular K+ to exit the cell, which accelerates basolateral Na+-K+-ATPase activity, causing ∼60% increase in tubule Qo 2. Under nystatin-stimulated conditions, ANG IV also concentration dependently inhibited Qo 2, suggesting that at least one action of ANG IV was to inhibit Na+ transport across the basolateral membrane. The concentration-response curves to ANG IV (10 fM to 1 nM) in control and nystatin-stimulated tubules were superimposable, with a threshold dose for ANG IV biological activity of 100 fM. Although the curves appeared to diverge at higher concentrations of ANG IV (>1 nM), the two concentration-response curves were not significantly different from each other. These results indicate that ANG IV has direct actions on the proximal tubule epithelium to inhibit energy-dependent Na+ transport. All future studies were conducted on nystatin-stimulated proximal tubules to allow us to examine changes in Qo 2independent of possible confounding ANG IV actions on apical Na+ entry into the cell, and we employed a single ANG IV dose of 1 pM, which was on the linear portion of the concentration-response curve.

Fig. 4.

Concentration-dependent reduction in proximal tubule O2 consumption rate (Qo 2) induced by ANG IV in control (○) or nystatin-stimulated (•) proximal tubules. Each point represents mean of 3–11 separate measurements. Comparisons within and across groups were assessed using a 2-way factorial analysis of variance (ANOVA).P values for group (G), concentration (C), and interaction (GC) are shown. Interaction was not statistically significant, indicating that effect of ANG IV treatment on proximal tubule Qo 2did not differ between groups.

Pharmacological characterization of angiotensin-receptor subtype.

To demonstrate that the inhibitory action of ANG IV was indeed a receptor-mediated effect, both active and inactive isomers of ANG IV were employed. As shown in Fig. 5, Qo 2 in nystatin-treated tubules was inhibited 21% (P < 0.001) by 1 pM ANG IV. Substitution of the l-valine in the NH2-terminal position 1 of ANG IV with d-valine was without biological activity. This finding indicated the presence of a stereospecific ANG IV binding site on proximal tubules.

Fig. 5.

Effect of levorotary (l,n = 7) and dextrorotary (d,n = 6) stereoisomers of valine in position 1 of ANG IV peptide on proximal tubule Qo 2 (each 1 pM). * P < 0.001 using an unpaired Student’s t-test.n, No. of separate measurements.

The subtype of angiotensin receptor involved in the action of ANG IV on Qo 2 was examined using selective angiotensin-receptor subtype antagonists. The incubation concentration of the receptor antagonists employed in the present study did not exhibit partial agonist activity, since basal Qo 2measurement of proximal tubules incubated with the angiotensin-receptor antagonists were similar to untreated control rates. In addition, injection of angiotensin-receptor antagonists into chambers containing nystatin-stimulated proximal tubules did not significantly alter Qo 2(vehicle: −0.1 ± 1.4%, n = 7; 1 μM losartan: −1.9 ± 1.2%,n = 6; 1 μM PD-123319: −1.0 ± 3.4%, n = 7; 1 μM divalinal-ANG IV: −3.2 ± 3.1%,n = 19). The inhibition of Na+ transport by ANG IV was not altered by preincubation of tubules with either losartan or PD-123319. In contrast, incubation with divalinal-ANG IV abolished the inhibitory action of ANG IV on proximal tubule Qo 2 (Fig.6).

Fig. 6.

Effect of 1 μM losartan (n = 8), 1 μM PD-123319 (n = 7), or 1 μM divalinal-ANG IV (n = 6) on reduction of proximal tubule Qo 2 induced by 1 pM ANG IV (n = 11). * P < 0.01 from all other groups using 1-way ANOVA and Student-Newman-Keuls test.

Specificity of divalinal-ANG IV as an AT4-receptor antagonist.

Preincubation of proximal tubules with divalinal-ANG IV did not significantly alter basal Qo 2(control: 26.7 ± 1.3 nmol O2 ⋅ min−1 ⋅ mg−1,n = 34; divalinal-ANG IV: 24.8 ± 0.9 nmol O2 ⋅ min−1 ⋅ mg−1,n = 42) or the effect of nystatin to increase QO2 (control: 69 ± 6%, n = 34; divalinal-ANG IV: 63 ± 5%, n = 42) by enhancing Na+-K+-ATPase activity. As shown in Fig. 7, divalinal-ANG IV abolished the receptor-mediated inhibition of nystatin-stimulated Qo 2 by ANG IV or [Nle1]ANG IV (high-affinity binding analog of ANG IV; Ref. 39) and yet did not interfere with the inhibitory actions of fenoldapam (dopamine DA1-receptor agonist), PAF (lipid-receptor agonist), or ANG II (AT1- and AT2-receptor agonist). The reduction in Qo 2 by dopamine-, lipid-, and ANG II-receptor agonists could be attenuated or abolished by preincubating proximal tubules with the receptor antagonists, SCH-23390, BN-52021, and [Sar1,Thr8]ANG II, respectively (not shown).

Fig. 7.

Effect of 1 pM ANG IV (control: n = 4; divalinal-ANG IV: n = 6), 1 pM [Nle1]ANG IV (control:n = 12; divalinal-ANG IV:n = 12), 1 μM fenoldapam (Fen: control, n = 5; divalinal-ANG IV,n = 5), 1 nM platelet-activating factor (PAF; control: n = 8; divalinal-ANG IV: n = 8) and 1 pM ANG II (control: n = 5; divalinal-ANG IV:n = 11) on tissue Qo 2 in control and 1 μM divalinal-ANG IV-treated proximal tubules. * P < 0.001 from corresponding control value using a unpaired Student’st-test.

Interaction of ANG IV and ANG II: role of Na+-K+-ATPase.

Having determined that both ANG IV and ANG II reduce Qo 2 of proximal tubules that are maximally transporting Na+ and act through different angiotensin-receptor subtypes (AT4and non-AT4 receptor, respectively), we examined whether their actions were additive and whether the angiotensin peptide congeners inhibited Qo 2 through a single effector pathway. The reduction in nystatin-stimulated Qo 2 of 20% by either ANG IV or ANG II (both at submaximal concentrations of 1 pM, which lay in middle of their respective concentration-response curves) was similar to that observed when the angiotensin peptides were administered simultaneously. The magnitude of the ouabain-inhibitable component of proximal tubule Qo 2 is shown for comparison (Fig. 8). One interpretation of this finding is that there is cross talk between the two receptor signaling systems that leads to a functional interaction to regulate the reduction in energy-dependent cell Na+ transport. ANG IV and ANG II were shown not to influence mitochondrial oxidative phosphorylation activity, since they did not alter uncoupled mitochondrial Qo 2 induced by 5 μM carbonyl cyanidep-trifluoromethoxyphenylhydrazone (FCCP, an oxidative phosphorylation uncoupler, not shown). Addition of 5 mM ouabain (Na+-K+-ATPase inhibitor) to tubules reduced nystatin-stimulated Qo 2 by 50%. Under these ouabain-treated conditions, administration of ANG IV or ANG II at 1 pM caused no further reduction in Qo 2 (Table1). Therefore the suppression of Qo 2 by ANG IV or ANG II appears to involve a ouabain-inhibitable component of Na+-K+-ATPase activity.

Fig. 8.

Effect of ANG IV (1 pM, n = 14) and ANG II (1 pM, n = 9) as well as their combined interaction (each 1 pM, n = 11) on proximal tubule Qo 2. Magnitude of inhibitory response of ouabain (5 mM,n = 9) is shown for comparison.

View this table:
Table 1.

ANG IV and ANG II inhibits a ouabainsuppressible component of proximal tubule O2 consumption


Autoradiograms of kidney 125I-ANG IV binding sites demonstrated a dense labeling of the outer stripe of the outer medulla and diffuse labeling of the entire cortex. The binding was unaffected by specific antagonists of the ANG II-type AT1 or AT2 receptor and yet was completely displaced by unlabeled ANG IV or the putative AT4-receptor antagonist, divalinal-ANG IV. These results extend our previous study indicating that kidney ANG IV binding protein and AT1 and AT2 receptors are distinct (21). Furthermore, the distribution of125I-ANG IV and125I-divalinal-ANG IV binding sites were identical and had similar displacement characteristics to selective angiotensin-receptor antagonists, lending support to divalinal-ANG IV being a specific ligand for the ANG IV type AT4 receptor. The pattern of125I-peptide labeling was consistent with receptor sites being present in tubular structures. This was confirmed by emulsion autoradiography that localized ANG IV binding sites to both the microvilli and cell bodies of convoluted proximal tubules throughout the cortex, with a higher density and similar distribution of sites present on the straight proximal tubules originating from superficial and midcortical glomeruli (located in medullary rays) and the straight proximal tubules of juxtamedullary glomeruli (located in outer stripe of outer medulla). This distribution of ANG IV binding sites suggests targeting of the rat kidney AT4 receptor to both proximal apical and basolateral membranes, which would be in agreement with the presence of high-affinity ANG IV binding sites in isolated rabbit proximal apical and basolateral membranes (10). Kidney AT4 receptors have also been identified in the monkey and guinea pig (46), rabbit proximal tubule (10, 11), opossum proximal tubule (11), Madin-Darby bovine kidney epithelial cells (19), and gerbil and human proximal tubule (Handa, unpublished observations). Although we found no evidence of AT4 receptors localized to either the glomerulus or the cortical or medullary collecting duct, high levels of ANG IV binding sites have been reported in both rat mesangial cells and human collecting duct cell membranes (2, 8). Together, these reports strongly suggest a conserved and presumably functional role of the ANG IV AT4-receptor system in kidney physiology across mammalian species.

We then examined whether ANG IV may influence proximal tubule Na+ transport because of the location of the ANG IV binding sites on proximal tubule cell bodies and microvilli, the critical role of the convoluted and straight proximal tubule in Na+ reabsorption, and the known role of ANG II as a powerful controller of proximal tubule Na+ reabsorption (22). Using primary cultures of rat proximal tubules and measuring tissue Qo 2 as an on-line, integrated index of transcellular Na+ transport, we found that ANG IV inhibited energy-dependent Na+transport in both control proximal tubules and tubules treated with nystatin (Na+ ionophore that allows extracellular Na+ to bypass rate-limiting step of apical Na+entry into cell and maximally stimulates basolateral Na+-K+-ATPase pump activity). The inhibitory action of ANG IV was not observed in ouabain (Na+-K+-ATPase inhibitor)- or FCCP (mitochondrial oxidative phosphorylation uncoupler)-treated tubules. Together, these results suggest that one site of ANG IV action was at the basolateral membrane to inhibit energy-dependent Na+ transport by reducing ouabain-sensitive Na+-K+-ATPase activity.

ANG IV has been shown to have reasonable efficacy, but low affinity, to a number of known angiotensin-receptor systems, including the ANG II AT1-type receptor to elicit vasoconstriction in the rat aorta, pulmonary, mesenteric, hindlimb, and renal vascular beds (13, 28, 37), feline hindquarter, and mesenteric vascular beds (5, 14), and the ANG II AT2-type receptor to cause kinin-mediated nitric oxide release from isolated canine coronary vessels (40). Our results suggest that ANG IV binds with high affinity to a novel proximal tubule receptor to elicit a biological response because 1) a stereospecific receptor protein for biological activity was present,2) inhibitory effect of ANG IV on Na+ transport was not mediated by losartan-sensitive AT1-type or PD-123319-sensitive AT2-type receptors, and 3) biological activity was abolished by divalinal-ANG IV, a putative AT4-receptor antagonist. Divalinal-ANG IV is a partial nonpeptide of ANG IV with a valine substituted for isoleucine in position 3 and isostere bonds incorporated between the 1–2 and 3–4 amino acids. These modifications provide both stability and metabolic resistance to the peptide as well as receptor antagonist activity (27). The specificity of divalinal-ANG IV as an AT4-receptor antagonist was demonstrated by its ability to displace125I-ANG IV and125I-divalinal-ANG IV binding to rat kidney slices, whereas losartan or PD-123319 or both were without effect. On a functional basis, divalinal-ANG IV blocked the response to ANG IV and [Nle1]ANG IV (AT4-receptor agonists) and yet did not influence receptor systems activated by fenoldapam (dopamine DA1-receptor agonist), PAF (lipid-receptor agonist), or ANG II (AT1- and AT2-receptor agonist). The dopaminergic and lipidergic receptor systems were selected because both inhibit proximal tubule Na+reabsorption (12, 18). The ANG II-receptor system was chosen as a result of our finding that low concentrations of ANG II caused a receptor-mediated inhibition of Na+ transport in proximal tubules with maximally elevated Na+-K+-ATPase activity, most likely reflecting a decrease in maximal reaction rate (V max) of the enzyme. This differs from reports that ANG II did not change (1, 4) or increased (15) theV max of Na+-K+-ATPase in rat proximal tubules, suggesting complex regulation of enzyme activity by ANG II. Similar concentrations of ANG II under non-V maxconditions can stimulate Na+reabsorption in proximal tubules primarily by decreasing the Michaelis constant of Na+ for Na+-K+-ATPase (1, 4, 22). Membrane binding studies confirmed that divalinal-ANG IV does not displace125I-[Sar1,Ile8]ANG II binding to AT1 receptors in the PD-123177-treated rat liver or to AT2 receptors in the losartan-treated rat adrenal medulla (not shown). To date, all studies using divalinal-ANG IV have shown it to be a specific antagonist of the AT4-receptor system (6, 25, 27, present study).

The intracellular signaling mechanisms of the ANG IV AT4-receptor system are presently unknown. ANG IV produced a sustained increase in intracellular Ca2+ and inositol phosphates in vascular smooth muscle cells that was distinct from the transient spike evoked by ANG II, suggesting different mechanisms in mobilizing Ca2+ (9). Activation of these vascular AT4 receptors may result in the stimulation of nitric oxide biosynthesis (7). The AT4 receptor has been shown not to be coupled to G proteins, and its affinity is generally unchanged (45,46) or, at least in one case, increased by sulfhydryl-reducing agents (24). In contrast, Dulin et al. (11) reported that ANG IV caused a transient, dihydropyridine-sensitive, increase in intracellular Ca2+ which did not result from phosphoinositol metabolism in the opossum OK7A proximal tubule cell line. The ANG IV binding site appeared to be G protein linked and was strongly inhibited by a sulfhydryl-reducing agent (11). When alternative renal transporting epithelia were used, the ANG IV binding site in human collecting duct cells and bovine kidney epithelial cells did not exhibit classical G protein coupling, unaltered by sulfhydryl-reducing agents (8, 19), and mobilized intracellular Ca2+ only at low concentrations (19). The hexapeptide appears to have no or minimal effect on guanosine 3′,5′-cyclic monophosphate and adenosine 3′,5′-cyclic monophosphate production in renal epithelial cells (8, 11). Some of these apparent contradictions may be due to the existence of AT4 receptor subtypes and the receptor having multiple intracellular signaling pathways that are cell specific.

Our results suggest that one potential role of the renal ANG IV AT4-receptor system may be in the handling of Na+ by the kidney. Supporting this contention is the demonstration that1) ANG IV binding sites are present in Na+ transporting epithelia, such as the renal convoluted and straight proximal tubule and collecting duct (8, 11, present study),2) stimulation of proximal tubule AT4 receptors caused a decrease in energy-dependent transcellular Na+transport (present study), 3) anesthetized rats infused intrarenally with ANG IV can elicit a natriuresis and diuresis in the absence of changes in total renal blood flow or blood pressure (Handa, unpublished observations), and4) in Wistar-Kyoto rats, a high-salt diet for 2 wk resulted in a 28% increase in the concentration of renal ANG IV binding sites (K. L. Grove and C. F. Deschepper, personal communication). Together, these findings provide a basis for further investigation into the physiological significance of this novel receptor system in tubular function.

The renal actions of ANG IV do not appear to be limited to the nephron. Ardaillou and Chansel (2) have recently found high levels of AT4 receptors in rat mesangial membranes and reported that ANG IV inhibited the rat mesangial cell contractile response to ANG II, suggesting the possibility that the ANG IV AT4-receptor system may influence glomerular function, e.g., permeability, filtration, and growth. Although our results do not support the presence of AT4 receptors in the glomerulus, autoradiographic techniques may not provide the sensitivity necessary to detect a low density of AT4-receptor populations. Alternatively, growing glomerular mesangial cells in vitro may stimulate the expression of AT4receptors. Infusion of ANG IV into the renal artery of anesthetized rats has been reported to increase cortical blood flow that could be prevented by AT4-receptor blockade or reversed to a cortical vasoconstriction after nitric oxide synthesis inhibition, implying that ANG IV’s predominant effect was an AT4 receptor-mediated stimulation of nitric oxide that masked an underlying renal vasoconstrictor action of ANG IV (6, 7, 42). However, infusing ANG IV into the renal artery of anesthetized rats caused a biphasic decrease in total renal blood flow that was abolished by AT1-receptor blockade without revealing a vasodilatory action of ANG IV (44). Similarly, intravenous bolus injections of ANG IV caused only an AT1 receptor-mediated renal vasoconstriction in the conscious rat (13). Clearly, additional studies are needed to resolve the role of ANG IV in renal vascular function, including the conclusive demonstration of the existence and localization of AT4 receptors in the renal vascular bed, the influence of vasomotor tone in ANG IV’s renal blood flow response, and whether there is cross talk between the AT1- and AT4-receptor systems.

Studies have also shown the ANG IV-receptor system to be present in many central and peripheral tissues (42, 46) and to be potentially involved in many diverse regulatory functions, including memory retrieval (46), growth of cardiac, endothelial and neuronal cells (3,16, 35), blood flow regulation in the brain (20, 27, 36), angiogenesis, wound healing, and thrombolysis (16, 25, 45). The wide distribution and many functions of the angiotensin AT4-receptor system suggest that it may be an important homeostatic system in the regulation of the internal environment. Despite ANG IV being an agonist with high affinity for the AT4 receptor, it has yet to be determined whether the hexapeptide is the natural endogenous ligand for the receptor. A recent study has suggested that LVV-hemorphin-7, generated by proteolytic processing of globin precursors, may be an endogenous ligand for the AT4 receptor in the brain (33). However, it is clear that the ANG IV molecule provides a tool to elucidate physiological and pharmacological properties of this novel receptor system. The recent development of divalinal-ANG IV as an AT4-receptor antagonist will also greatly aid in exploring the relative importance of this endogenous AT4-receptor system.

In summary, the results from autoradiographic and functional studies complement each other and demonstrate that rat proximal tubules contain the AT4-receptor system and that this system can potentially regulate proximal tubule Na+ transport. ANG IV reduced energy-dependent Na+ transport by acting exclusively through a non-AT1, non-AT2 angiotensin receptor that could be blocked by the selective AT4-receptor antagonist, divalinal-ANG IV, and inhibited a ouabain-suppressible component of Na+-K+-ATPase activity.


We thank Drs. Catherine M. Ulibarri and J. Lindsay Oaks, Jr., for their assistance and use of facilities for the emulsion autoradiographic studies.


  • Address for reprint requests: R. K. Handa, Dept. of Veterinary and Comparative Anatomy, Pharmacology, and Physiology, College of Veterinary Medicine, Washington State University, Pullman, WA 99164–6520.

  • Funding for this work was provided by the Washington Affiliate of the American Heart Association (WA-94–510) to R. K. Handa.


View Abstract