Luminal ANG II is internalized as a complex with AT₁R/AT₂R heterodimers to target endoplasmic reticulum in LLC-PK₁ cells

Abstract

ANG II has many biological effects in renal physiology, particularly in Ca²⁺ handling in the regulation of fluid and solute reabsorption. It involves the systemic endocrine renin-angiotensin system (RAS), but tissue and intracrine ANG II are also known. We have shown that ANG II induces heterodimerization of its AT₁ and AT₂ receptors (AT₁R and AT₂R) to stimulate sarco(endo)plasmic reticulum Ca²⁺-ATPase (SERCA) activity. Thus, we investigated whether ANG II-AT₁R/AT₂R complex is formed and internalized, and also examined the intracellular localization of this complex to determine how its effect might be exerted on renal intracrine RAS. Living cell imaging of LLC-PK₁ cells, quantification of extracellular ANG II, and use of the receptor antagonists, losartan and PD123319, showed that ANG II is internalized with AT₁R/AT₂R heterodimers as a complex in a microtubule-dependent and clathrin-independent manner, since colchicine—but not Pitstop2—blocked this process. This result was confirmed by an increase of β-arrestin phosphorylation after ANG II treatment, clathrin-mediated endocytosis being dependent on dephosphorylation of β-arrestin. Internalized ANG II colocalized with an endoplasmic reticulum (ER) marker and increased levels of AT₁R, AT₂R, and PKCα in ER-enriched membrane fractions. This novel evidence suggests the internalization of an ANG II-AT₁/AT₂ complex to target ER, where it might trigger intracellular Ca²⁺ responses.

the renin-angiotensin system (RAS) is a key hormonal signaling system involved in the homeostasis of body salt and fluid, operating mainly at the kidney level, where the best-described active peptide, ANG II, is the main regulator of Na⁺ and water reabsorption, a function that depends on finely tuned Ca²⁺ transport in the cell (6, 15).

RAS has been classically involved in the vasoactive system, acting as an endocrine system and a local tissue paracrine system (20, 47, 66). In the endocrine system, each RAS component is synthesized in different tissues to produce the renal and extrarenal ANG II effects on arterial blood pressure. Briefly, the liver is the main source of angiotensinogen (AGT), the only known precursor in the synthesis of ANG II, which is the main biologically active peptide for RAS. To produce ANG II, kidney-derived renin cleaves AGT to produce ANG I, which is rapidly hydrolyzed by the endothelial angiotensin-converting enzyme (ACE) to ANG II. Several tissues, including the brain, cardiovascular, adipose, gonad, pancreas, placenta, adrenal gland, liver, and kidney tissues, synthesize all of the RAS components, characterizing the local tissue paracrine RAS (7, 18, 43, 44, 54, 57). The paracrine system seems to mediate the physiological and pathophysiological responses of the cell, which includes effects on proliferation, cell growth, and metabolism (reviewed in Ref. 47).

More recently, an intracrine or intracellular RAS (50) was revealed when ANG II was bound to nuclear receptors and triggered gene transcription in kidney proximal tubule cell nuclei (37) and cardiomyocytes (60). Intracrine RAS depends on an intracellular source of ANG II and synthesis of RAS components, which involves different classes of receptors in the nucleus (37) and mitochondria (1, 67).

Intracellular ANG II comes from two sources: 1) cells that synthesize all of the RAS components and 2) extracellular ANG II internalized with its receptors (66). Indeed, ANG II is known to bind to its AT₁R, the complex being internalized. A cellular response is triggered, followed by receptor recycling to the plasma membrane or endosome (23). However, the ANG II receptor AT₂R may be resistant to internalization and trafficking (27, 31).

ANG II receptors occur in organelles from different cell types, e.g., mitochondria and nuclei (2, 34). In the kidney, mitochondrial AT₂R appears to be involved in the regulation of NO formation and O₂ consumption (1). In the nucleus, AT₁R stimulates ROS, and one possible functional consequence is the modulation of a redox-sensitive signaling pathway (48). Moreover, this class of receptors is involved in Ca²⁺ homeostasis and function in the nucleus (69). Although it is plausible that ANG II binding to organelle receptors may participate in Ca²⁺ homeostasis, this has not been properly established, and it is also unknown whether intracrine RAS acts in organelles other than the mitochondria and nuclei.

The role of intracellular RAS has been explored in different tissues; in the kidney, overexpression of intracellular ANG II in a transgenic mouse increases blood pressure and thrombotic microangiopathy (51). In cardiomyocytes, intracellular levels of ANG II are 4−5-fold higher in diabetic (streptozotocin-treated) rats (56) and diabetic hypertensive patients (22) than in nondiabetic patients and were associated with gene expression, oxidative damage, apoptosis, and necrosis.

Although there is no consensus that ANG II always promotes the dimerization of receptors, several reports have demonstrated that in both endocrine and paracrine RAS, ANG II binding to any of its receptors induces their homodimerization and/or heterodimerization, which, in turn, induces different cellular responses (3, 4, 9, 19). For instance, ANG II-induced receptor heterodimerization is required to modulate different ATPase activities involved in ion reabsorption (6, 14). We have previously shown that luminal ANG II induces the heterodimerization of AT₁ and AT₂ receptors to stimulate the sarco(endo)plasmic reticulum calcium ATPase (SERCA) activity, without an effect on plasma membrane Ca²⁺-ATPase (PMCA) (19).

However, the specific effect of AT₁R/AT₂R-induced ANG II on SERCA activity remains unclear in terms of its targets. Our working hypothesis is that the luminal ANG II-AT₁R/AT₂R complex has to have an intracellular target to regulate Ca²⁺ storage in renal cells. We show here that ANG II accessing the luminal side is internalized as a complex with AT₁R/AT₂R heterodimer, and targets the ER to stimulate SERCA activity.

METHODS

Materials.

LLC-PK₁ cells and Hanks solution were purchased from American Type Culture Collection (CL-101) (Manassas, VA). Culture medium 199, FBS, trypsin solution (0.25% trypsin plus 0.04% EDTA-tetrasodium salt), ER Tracker Red (E34250) and ANG II-Alexa Fluor 488 (A13439) were obtained from Invitrogen Life Technologies (Carlsbad, CA). Fluorescent carboxyfluorescein-labeled ANG II (FAM-ANG II) (AS-60275) was obtained from Anaspec (Fremont, CA). Antibodies against p-β-arrestin (sc-16639-R, lot F0904), AT₁R (sc-57036, lot B1210), AT₂R (sc-9040, lot F2513), and PKCα (sc-208, lot E0412) were obtained from Santa Cruz Biotechnology (Santa Cruz, CA), and antibodies against Na⁺/K⁺-ATPase (A276, lot SLBN0461V), SERCA (S1439, lot SLBL0170V) and β-actin (A5441, lot 014M4759) were obtained from Sigma-Aldrich (St. Louis, MO). Nitrocellulose membranes, secondary antibodies, and ECL system were obtained from GE Healthcare (São Paulo, Brazil). Losartan was obtained from Merck Sharp & Dohme (Whitehouse Station, NJ), PD123319, protease inhibitor cocktail, and trypsin inhibitor (type II-S soybean) from Sigma-Aldrich (St. Louis, MO). Pitstop 2, recombinant AT₁R protein (ab157870, lot 1010313–759) and recombinant AT₂R protein (ab157871, lot FA261-6226) were purchased from Abcam Biochemicals (Cambridge, UK). All other reagents were of the highest purity available.

Cell culture.

Immortalized porcine proximal tubule epithelial cell line (LLC-PK₁) was cultured in medium 199 with 3% FBS in an atmosphere of 5% CO₂ in air at 37°C. Approximately 5 × 10⁵ cells were seeded in 25 cm² culture flasks and incubated until ~90% confluent (~3 days). Cells were washed in PBS at pH 7.5 (10 mM Na₂HPO₄ + 13.7 mM NaCl + 1.7 mM KH₂PO₄ + 2.7 mM KCl) and incubated with 1 ml trypsin/EDTA to detach the cells. The cells were centrifuged at 525 g for 2 min, and the pellets were resuspended in 5 ml medium 199 with 3% FBS and distributed into new flasks.

Living cell fluorescent microscopy.

Cultures of 10⁴ cells were grown in 96-well plates in medium 199 with 3% FBS in an atmosphere of 5% CO₂ in air at 37°C for 2 days. Cells were incubated with 0.1 μM FAM-ANG II in medium (200 μl) without FBS for 30 min at 37°C in a 5% CO₂ atmosphere, as previously described, with modifications (36). Cells were treated with 1 μM losartan and/or PD123319 20 min before incubation with 0.1 μM FAM-ANG II for 30 min to analyze the involvement of AT₁R/AT₂R heterodimers in ANG II internalization. Cells were washed with PBS, and the cells were observed in fluorescence microscope (ImageXpress Micro; Molecular Devices, Sunnyvale, CA). The concentration of ANG II used in this experiment intended to mimic the tubular ANG II concentration (53).

To analyze ANG II subcellular location, LLC-PK₁ cells were grown on coverslips in eight-well plates in medium with 3% FBS in an atmosphere of 5% CO₂ at 37°C for 2 days. The cells were incubated with 0.3 μM ANG II-Alexa Fluor 488 for 1 h in 500 μl medium without serum before being washed with Hanks solution and incubated with 0.1 μM ER Tracker Red for 15 min still under culture conditions. Cells were washed in Hanks solution and examined by confocal fluorescence microscopy (TCS SP8, Leica Microsystems, Exton, PA). Images were scanned under identical conditions, and prepared for presentation with PhotoShop CS6 software.

Quantification of ANG II by HPLC.

Cultures of 5 × 10⁵ cells were seeded in 25 cm² culture flasks for 3 days before being incubated with 8 µM ANG II in 2 ml PBS for 2 h in a 5% CO₂ atmosphere at 37°C or 4°C. Supernatants were collected and concentrated in speedvac (Thermo Savant, Holbrook, NY), and analyzed by HPLC (model LC10AS; Shimadzu, Kyoto, Japan), using a C-18 reverse-phase column (Rexcrom, 25 cm × 4.6 mm; Regis Technologies, Morton Grove, IL). To observe the receptors involved in ANG II internalization, cells were treated with 0.1 nM losartan or 0.1 µM PD123319 (or both) 15 min before adding 8 µM ANG II for 2 h at 37°C. To analyze the endocytic pathway, cells were incubated with 30 μM Pitstop 2 or 1 μM colchicine before incubation with 8 µM ANG II for 2 h at 37°C. This concentration of ANG II was used to ensure proper and reproducible UV detection because at <5 μM, the signal is progressively much noisier.

Vesicular membrane fraction.

Preparations enriched with vesicles derived from plasma membranes and ER from LLC-PK₁ cells were obtained after growing cells in 12 flasks (150 cm²) under both conditions (with or without 0.1 nM ANG II for 30 min). The methodology for cell fractionation was carried out following Parys et al. (46), with modifications. Briefly, cells were resuspended in 20 ml homogenization buffer containing 250 mM sucrose, 10 mM Tris·HCl (pH 7.6), 0.1 mM PMSF, 1 mg/ml trypsin inhibitor, and protease inhibitor cocktail 1:400, and then lysed with a Potter-Elvejhem homogenizer with a Teflon pestle. Differential centrifugation was used to collect the vesicles derived from the plasma and ER membranes. After recovery of the first fraction, the following were obtained from the supernatants obtained after each centrifugation step: the total homogenate was centrifuged using a Sorvall SS-34 rotor at 2,400 g for 15 min to obtain the p1 fraction. The p2 fraction was obtained from the first supernatant after 15 min at 9,800 g. The next steps used a Beckman-Coulter 70 Ti rotor to collect p3 fraction (17,400 g for 20 min), p4 (23,600 g for 20 min), p5 (36,900 g for 40 min), and finally p6 (105,000 g for 60 min). All of the pellets were resuspended in the same buffer at 1:1 vol/vol. Total protein concentration of pellets was determined by the method of Lowry et al. (39) and ranged from 3 to 6 mg/ml.

Sarco(endo)plasmic reticulum Ca²⁺-ATPase activity.

Cultures of LLC-PK₁ cells (5 × 10⁵) were grown in 25 cm² culture flasks in medium 199 supplemented with 3% FBS in an atmosphere of 5% CO₂ in air at 37°C for 3 days. Cells were starved and preincubated for 20 min at 37°C with 0.1 nM losartan, 0.1 µM PD123319 or in the presence of both antagonists. Then, 0.1 nM ANG II was added, and after 30 min at 37°C in a 5% CO₂ atmosphere, the cells were harvested and lysed as described in Ref. 19. SERCA activity was measured in the experimental conditions previously described (19) and expressed as the difference between total Ca²⁺-ATPase and that measured in the presence of 1 μM thapsigargin, the specific SERCA inhibitor.

Western blot analysis.

To analyze the involvement of β-arrestin in ANG II internalization, cells were incubated with or without 1 µM ANG II for 30 min. They were washed with PBS (pH 7.5) and incubated with medium in the absence of FBS in 5% CO₂ at 37°C. Cells were harvested and lysed in a buffer containing 250 mM sucrose, 20 mM HEPES-Tris (pH 7.0), 1 mM EDTA, and 0.15 mg/ml trypsin inhibitor, using a Potter-Elvejhem homogenizer. Cell lysates (50 and 100 μg total protein for p-β-arrestin and 40 μg for Na⁺/K⁺-ATPase, SERCA, AT₁R, AT₂R, and PKC assays) were electrophoresed in a 10% SDS-PAGE and transferred to nitrocellulose membranes. These were probed with primary antibodies, followed by incubation with a horseradish peroxidase-conjugated secondary antibody for 1 h at room temperature, and finally developed with the ECL system. Membrane stripping was done with 0.2 M glycine (pH 2.2) for 1 h at room temperature. Primary antibody incubation was carried out at room temperature for 1 h, at dilutions of anti-p-β-arrestin (1:200), β-actin (1:5,000), Na⁺/K⁺-ATPase (1:500), SERCA (1:750), AT₁R (1:200), PKCα (1:500). AT₂R was probed at 1:400 overnight at 4°C. Secondary antibodies were five times further diluted than their respective primary antibodies.

Despite concerns regarding antibodies specificity, we validated the Santa Cruz Biotechnology antibodies AT₁R and AT₂R by using adsorption experiments as described before (55), but now using samples from LLC-PK₁ cell culture. Recombinant AT₁R protein (amino acids 250–359) and AT₂R (amino acids 1–363) were used as templates in experiments to see whether they adsorb or not the antibodies against AT₁R (raised against amino acids 297–356) and AT₂R (raised against amino acids 221–363), thereby checking the specificity of each antibody. A three-fold mass of each recombinant protein was incubated with the two antibodies separately for 72 h at 4°C under gentle shaking. The resulting mixture was centrifuged at 18,000 g for 1 min, and the supernatant was diluted 1:400 with respect to the original antibody. Electrophoresis of LLC-PK₁ ER-enriched membrane preparations (obtained as described in “Vesicular Membrane Fraction”) was carried out using 40 µg of total protein in 12% acrylamide SDS-PAGE and transferred to nitrocellulose membranes. Standard molecular weight of 38 kDa was used to confirm the expected band for ANG II receptors (~40 kDa). The membranes were probed with the antibodies preincubated with recombinant AT₁R or AT₂R proteins. After stripping with 0.2 M glycine (pH 2.2), the membranes were reprobed with anti-AT₁R or anti-AT₂R antibodies alone (i.e., without the recombinant proteins) at the same dilution (1:400). Probing different membranes in parallel (without stripping) using the same antibodies (alone or preincubated) gave identical results.

Statistical analysis.

Student’s t-test was used to compare two groups, and one-way ANOVA with Newman-Keuls post hoc test was used to compare three or more groups. Statistical significance was set at P < 0.05. Data were analyzed with GraphPad Prism 5.0 software (San Diego, CA).

RESULTS

ANG II is internalized via luminal membranes of LLC-PK₁ cells with AT₁R/AT₂R heterodimers.

In LLC-PK₁ cells incubated with 0.1 μM FAM-labeled ANG II, intracellular localization of ANG II was seen after 30 min, indicating that the peptide had been internalized (Fig. 1A). If the antagonists of AT₁R (1 μM losartan) or AT₂R (1 μM PD123319) receptors were given before FAM-ANG II, they prevented the internalization of FAM-ANG II and, thereby, caused the retention of ANG II in the plasma membrane (Fig. 1, B and C). When both antagonists were added simultaneously before FAM-ANG II incubation (Fig. 1D), the result was unexpectedly the same as without the receptors’ antagonists, i.e., showing the internalization of the peptide after 30 min. Absence of FAM-ANG II was used as the negative control (Fig. 1E).

Fig. 1.

ANG II is internalized through the luminal membranes of LLC-PK₁ cells with AT₁R/AT₂R heterodimers. A: cells incubated with FAM-ANG II (0.1 μM) for 30 min at 37°C. B–D: cells treated with losartan (1 μM) (LOS) or PD123319 (1 μM) (PD), or both antagonists for 20 min before FAM-ANG II incubation. Arrows indicate ANG II intracellular localization (A, D) or plasma membrane retention (B, C). E: negative control (cells only, no FAM-ANG II).

This result was confirmed by the quantification of ANG II levels in the culture medium by HPLC. Figure 2 shows that ANG II is also internalized at the higher concentration of 8 µM—which is required for HPLC analysis—after 2 h of incubation with the cells. This internalization was again partially prevented when losartan or PD123319 was used separately, with no significant difference between them. Incubation with both antagonists did not prevent ANG II internalization, also confirming the observation in Fig. 1D.

Fig. 2.

HPLC quantification of ANG II in culture medium containing LLC-PK₁ cells preincubated or not with losartan (0.1 nM), PD123319 (0.1 μM), or with both antagonists, as shown on the abscissa, for 15 min at 37°C before adding 8 µM ANG II. The assayed material was further incubated for 2 h at 37°C. Data bars indicate means ± SE (n = 3−7). The empty bar corresponds to ANG II at zero time. Different lowercase letters above the bars indicate statistical difference between the mean values (P < 0.05).

Figure 3 shows 1) the stimulation of SERCA activity after the ANG II internalization, 2) the consequence of the partial internalization blockade by losartan or PD123319, and 3) the recovery of the stimulation profile when both antagonists were assayed together. As previously shown (19), ANG II stimulated SERCA activity by >90%, and preincubation with losartan or PD123319 totally blocked this stimulation. The return of SERCA activity to control levels matches the inhibition of ANG II internalization (∼50%). Interestingly—and in line with Fig. 2—ANG II stimulated SERCA activity to the same extent as ANG II alone when the cells were preincubated with both antagonists.

Fig. 3.

ANG II-induced stimulation of SERCA activity from LLC-PK₁ cells is preserved when the AT₁R and AT₂R antagonists are both added. Cells were preincubated with losartan (0.1 nM) (Los) or PD123319 (0.1 μM) (PD), or with both antagonists, as shown on the abscissa, for 20 min before adding 0.1 nM ANG II. The cells were further incubated at 37°C for 30 min; longer times imply loss of activation, as described in Ref. 19. Data bars indicate means ± SE (n = 3–7). Different lowercase letters above the bars indicate statistical difference between the mean values (P < 0.05). Following the Committee on Publication Ethics (COPE) guidelines, we disclose that we have previously reported internal controls using losartan or PD123319 alone (19).

ANG II-AT₁/AT₂ complex via luminal membranes of LLC-PK₁ cells is internalized through microtubule-dependent and clathrin-independent endocytic pathway.

Different methods were used to investigate the endocytic pathway involved in ANG II internalization in renal epithelial cells. Culturing cells at low temperatures is known to block membrane trafficking and inhibit receptor endocytosis (36). When 8 µM ANG II was incubated with the cells at 4°C, the peptide concentration in the medium decreased significantly (Fig. 4A). Because it is not internalized at this temperature (36), it is clear that the decrease in medium ANG II is due to binding to the membrane surface. The influence of microtubule disruption on ANG II internalization was examined by using 1 µM colchicine treatment of LLC-PK₁ cells for 2 h at 37°C; this totally prevented internalization, i.e., ANG II was endocytosed in a microtubule-dependent manner (Fig. 4B). To examine the classical β-arrestin/clathrin endocytic pathway, the cells were incubated with Pitstop 2—which prevents clathrin-mediated endocytosis—for 30 min before ANG II treatment. Pitstop 2 was ineffective in blocking ANG II endocytosis, indicating a clathrin-independent internalization mechanism (Fig. 4C). Indeed, LLC-PK₁ cell incubation with 1 μM ANG II for 30 min induced phosphorylation of β-arrestin at Ser-412 (Fig. 5). Dephosphorylation of β-arrestin is required to target clathrin-coated pits for receptor endocytosis, which confirms that this protein is not involved in the endocytic pathway of ANG II-AT₁R/AT₂R complex internalization.

Fig. 4.

Internalization of ANG II-AT₁R/AT₂R complex is a microtubule-dependent and clathrin-independent endocytic pathway. The empty column in each panel represents medium ANG II before HPLC development (zero time). A: gray bar shows that 50% of the ANG II incubated for 2 h at 37°C was internalized, while the black bar shows binding of ANG II to the luminal membranes after 2 h at 4°C, since low temperature prevents internalization (27). B: cells incubated with 1 µM colchicine (microtubule polymerization inhibitor) for 2 h before 8 µM ANG II was added followed by incubation for 2 h at 37°C. C: cells incubated with 30 µM Pitstop 2 (a clathrin inhibitor) for 30 min before 8 µM ANG II was added and incubated for 2 h at 37°C. After the last incubation, culture media were carefully collected and analyzed by HPLC. The graphs represent ANG II in the culture media. Data bars indicate means ± SE (n = 3 or 4). Different lowercase letters above the bars indicate statistical differences among the mean values (P < 0.05).

Fig. 5.

ANG II stimulates β-arrestin phosphorylation in LLC-PK₁ cells. A: cells incubated with or without 1 μM ANG II for 30 min. Cell lysates (50 and 100 μg protein) were Western blotted using a polyclonal anti-p-β-arrestin antibody (top blot). After being stripped, membranes were probed with monoclonal anti-β-actin (bottom blot). B: densitometric representation of the immunoreactive signal ratio p-β-arrestin/β-actin. Data bars indicate means ± SE using different cell lysate preparations (n = 7 or 8). Different lowercase letters indicate statistical differences (P < 0.05).

Internalized ANG II via luminal membrane of LLC-PK₁ is partially located in the endoplasmic reticulum.

Knowing that luminal ANG II specifically stimulates SERCA activity in LLC-PK₁ cells with no effect on PMCA activity (19), this suggests an intracellular ANG II effect, probably in the ER. Living-cell fluorescent imaging was used by incubating LLC-PK₁ cells with 0.3 μM ANG II-Alexa Fluor 488 for 1 h and subsequently adding 0.1 μM ER Tracker Red for 15 min. Internalized ANG II was present at high density in the ER (Fig. 6), a physical association that may culminate in the previously described activation of SERCA (19).

Fig. 6.

ANG II internalized via luminal membrane of LLC-PK₁ is localized in the ER. LLC-PK₁ cells were incubated with ANG II-Alexa Fluor 488 (0.3 μM) for 1 h at 37°C and then with ER Tracker Red (0.1 μM) for 15 min at 37°C. Living cell fluorescent imaging using confocal microscopy was obtained for ANG II-Alexa Fluor 488 (A, D), ER Tracker (B, E), and merge (C, F). A: negative control for ANG II (cells only). B, C: cells incubated with only ER Tracker, i.e., without ANG II.

AT₁ and AT₂ receptor content is increased in the endoplasmic reticulum of LLC-PK₁ cells after ANG II-AT₁R/AT₂R complex internalization.

To investigate whether the ANG II-AT₁R/AT₂R complex, rather than the peptide alone, is directed to the ER, LLC-PK₁, cells were fractionated by differential centrifugation to obtain plasma membrane and ER-enriched vesicles (Fig. 7). Western blotting was undertaken to characterize membrane vesicles using Na⁺/K⁺-ATPase and SERCA as plasma membrane and ER markers, respectively. Fractions p2–p4 contain plasma membrane, while fractions p3–p6 correspond to ER-enriched vesicles (Fig. 7A), showing higher amounts of SERCA. Therefore, p3–p6 fractions were used in Western blotting for AT₁R, AT₂R, and PKCα—the PKC isoform that is coupled to the luminal effects of ANG II in proximal tubules (15, 19)—in the absence and presence of 0.1 nM ANG II for 30 min, i.e., at a concentration that maximally activates SERCA (19). AT₁R and AT₂R content clearly increased in p4 to p6 fractions (Fig. 7B). The densitometric analysis of p5, which had the highest SERCA-positive profile and lower Na⁺/K⁺-ATPase levels, show 1) an increase of ~40 and ~100% of the AT₁R and AT₂R immunosignals, respectively, and 2) an increase of 150% in the PKCα immunosignal, compared with nontreated cells, using SERCA content as the loading control (Fig. 7C). Given concerns regarding antibody specificity (28, 32), we validated both AT₁R and AT₂R antibodies and confirmed the nonexistence of cross-reactivity between them in LLC-PK₁ cells (Fig. 8). Preadsorption of AT₁R antibody with AT₁R recombinant protein abolished the corresponding immunosignal for AT₁R (Fig. 8A, left); when the AT₂R recombinant protein was used as a template, there was no difference in the AT₁R immunosignal (Fig. 8A, right). The preadsorption of AT₂R antibody with AT₂R recombinant protein partially blocked the immunosignal (Fig. 8B, right), but there was no significant difference when the antibody was incubated with the AT₁R recombinant protein (Fig. 8B, left).

Fig. 7.

AT₁R and AT₂R and PKCα content increased in the ER of LLC-PK₁ cells after ANG II-AT₁R/AT₂R internalization. A: cell fractions p1−p6 obtained by differential centrifugation were Western blotted using monoclonal anti-Na⁺/K⁺-ATPase and anti-SERCA antibodies (upper and lower blots, respectively). B: cells incubated with or without 0.1 nM ANG II for 30 min, as indicated on the abscissa. Cell fractions p3−p6 were Western blotted using monoclonal anti-AT₁R, polyclonal anti-AT₂R, monoclonal anti-SERCA, and polyclonal anti-PKCα antibodies. C: densitometric representation of the immunoreactive signal ratio primary antibodies (AT₁R, AT₂R, or PKCα) and SERCA antibody in p5. Data bars indicate means ± SE (n = 2 or 3). *Statistically different from the control without ANG II (P < 0.05).

Fig. 8.

Specificity of the ANG II AT₁R and AT₂R antibodies. ER-enriched membrane preparations were probed against AT₁R monoclonal (A) or AT₂R polyclonal (B) antibodies with or without preincubation with recombinant AT₁R or AT₂R protein (as indicated). Preincubation of each antibody with the corresponding immunogenic peptide decreased the immunosignal, whereas preincubation with the other recombinant protein did not. 38 kDa corresponds to standard molecular weight used in all conditions.

DISCUSSION

The present study shows that ANG II is internalized as part of a complex with AT₁R/AT₂R heterodimers, with the ER being a new target organelle for intracrine RAS. This mechanism may explain the specificity of luminal ANG II to stimulate SERCA activity in LLC-PK₁ proximal tubule cells (19).

G protein-coupled receptor (GPCR) heterodimerization has been extensively studied, including the functional properties in physiological and pathological contexts (45). AT₁R/AT₂R heterodimers are not internalized after ANG II induction (49), and AT₂R is not internalized after stimulation with ANG II (31). However, both AT₁ and AT₂ receptor antagonists in our cellular proximal tubule model impeded ANG II internalization when assayed separately, indicating that AT₁R and AT₂R are involved in luminal ANG II endocytosis (Fig. 1). An explanation for this apparent discrepancy may rest on the cells used. Indeed, Porrello et al. (49) and Hein et al. (31) used HEK293, a nonpolarized human embryonic kidney cell line of epithelial origin, which does not have membrane domains with differentiated “sidedness” (asymmetry and polarity) and functions. In contrast, the LLC-PK₁ cells preserve vectorial ion transport, with well-demarcated luminal and basolateral membranes, and, therefore, with architecture similar to that observed in renal tissues in vivo. The data suggest that cell polarity, essential for renal epithelial cell physiology, is important for luminal ANG II to induce AT₁R/AT₂R heterodimerization. Immunoprecipitation assays clearly indicate that AT₁R/AT₂R heterodimers in the plasma membrane are essential for luminal ANG II internalization and SERCA stimulation in response to luminal ANG II (19). AT₁R/AT₂R heterodimers have also been described in membrane fractions from sheep proximal tubules (6) and rat renal cells (14). Unexpectedly, the combination of losartan and PD123319 did not prevent ANG II internalization in LLC-PK₁ cells, and, as will be discussed below, in the case of the same effect for AT₁R and AT₂R, this indicates that simultaneous binding of the antagonists to AT₁R/AT₂R heterodimers is mutually exclusive.

ANG II internalization was confirmed by another approach: there was a decrease in medium ANG II concentration in HPLC assays when LLC-PK₁ cells were incubated with the peptide, a phenomenon that was partially blocked when losartan or PD123319 was present (Fig. 2). At first sight, partial blockade of ANG II internalization (at 8 μM) by 0.1 nM losartan or 0.1 μM PD123319 seems to be unlikely. Both antagonists are considered low-affinity inhibitors, and their effects on ANG II internalization are, therefore, difficult to understand in terms of simple competitiveness. LLC-PK₁ cells need to be preincubated with the antagonists for 15 min at 37°C before the addition of ANG II for the blockade of internalization to occur. Thus, the reversible interaction of the antagonists with their respective receptors could be followed by irreversible inactivation in the absence of ANG II. Again, and as encountered for FAM-ANG II (Fig. 1), the simultaneous presence of the antagonists resulted in ANG II internalization at levels similar to those of controls without additions, confirming that interactions with peptide and antagonists at receptor level are the same, irrespective of the ANG II/receptors antagonists’ ratio.

The involvement of β-arrestin-/clathrin-coated pits is well known as an endocytic pathway for GPCRs (62). Internalization of the ANG II-AT₁R/AT₂R complex occurs by a clathrin-independent pathway, as shown in Fig. 4C, which is supported by the induction of β-arrestin phosphorylation by ANG II in this cell line (Fig. 5). Receptor endocytosis mediated by clathrin binding requires dephosphorylation of β-arrestin (38). Thus, since internalization of the ANG II-AT₁R/AT₂R complex in LLC-PK₁ cells does not involve the classic clathrin-dependent pathway, it may be occurring through the nonclassical receptor endocytic pathway involving the caveolar-type membrane/lipid rafts (30). Indeed, AT₁R moves to caveolin-enriched fractions through interaction with caveolin-1 following vascular smooth muscle cell exposure to ANG II (35).

Microtubules are important for membrane traffic, maintenance of epithelial cell polarity, and luminal endocytosis in polarized cell types (5, 17). GPCR interacts with tubulin, suggesting that GPCR membrane trafficking is microtubule-dependent (16), and here, we have shown that internalization of the ANG II-AT₁R/AT₂R complex in LLC-PK₁ cells is microtubule-dependent (Fig. 4B), a view supported by its complete reversal by colchicine. The high specificity of colchicine for tubulin is due to binding site in the tubulin (40), but other independent-microtubule effects of colchicine could not be ruled out, as described by Marques-da-Silva et al. (42). Furthermore, involvement of actin in internalization should be considered, since cytochalasin D blocks endocytosis of high molecular weight complexes in renal cell lines of distal origin (25).

It is tempting to suppose that a caveolar-type membrane/lipid raft microtubule-dependency is the principal mechanism of ANG II-AT₁R/AT₂R internalization. However, this mechanism itself may be not be sufficient to explain the removal of 4 nmol ANG II/ml from the medium in 2 h in the HPLC assays—crossing a 25 cm²-cell layer—because 1) there are ~120,000 AT₁R receptors per LLC-PK₁ cell (68) and 2) the 30-s time lapse required by Ca²⁺ mobilization (19). Considering that megalin in the brush-border membranes of proximal tubule cells is involved in the uptake of ANG II (24), as well as its sensitivity to colchicine (8), we speculate that a colchicine-sensitive, megalin- and microtubule-dependent mechanism participates in ANG II-AT₁R/AT₂R internalization (17, 26, 36). Moreover, a fast colchicine-sensitive PLC-mediated IP₃ release occurs in primary culture of proximal tubule cells after 30 s of luminal ANG II stimulation (52), confirming the fast response that we had observed.

The main source for renal intracellular ANG II is extracellular ANG II internalized with its receptors (65, 66). Nuclear and mitochondrial membranes also have AT₁ and AT₂ receptors responsive to intracrine ANG II (reviewed in Ref. 2). It is well established that these organelles are involved in intracellular Ca²⁺ storage (33, 41); however, the ER cannot be neglected because 1) this organelle expresses different proteins involved in Ca²⁺ storage, and 2) SERCA has a pivotal role in Ca²⁺ movements and signaling in polarized cells (10, 46). Since ANG II stimulates SERCA activity that induces Ca²⁺ mobilization (19), it is reasonable to propose that ER can also be an intracellular target of ANG II. An intracrine ANG II effect on Ca²⁺ homeostasis occurs, but these studies did not indicate the presence of ANG II receptors in the ER (13, 69). Thus, ANG II-associated signaling upon SERCA may be due to the simultaneous endocytosis of the receptors (Figs. 6 and 7).

Assuming a similar t_1/2 (~2.3 min) in LLC-PK₁ cells like Chinese hamster ovary cells and HEK293 cells (59, 61), we can estimate that just over 10% of the receptors could be internalized by the time Ca²⁺-ATPase is maximally activated. This estimation is in agreement with the pharmacological principle of the “receptor reserve” theory, which states that a full agonist (e.g., ANG II) only requires binding to a small amount of receptors to elicit its maximal effect with a great efficacy (21, 58). Therefore, a low binding level should be enough to sustain Ca²⁺ mobilization, as long as internalization continues over 30 min before this process starts to drop (19).

Live cell imaging data confirm that the ER is the target organelle of luminal ANG II, which induces internalization of AT₁R/AT₂R heterodimers (Fig. 6). The data were confirmed by Western blot analysis of the sequential fractions obtained by differential centrifugation (Fig. 7), which suggests the existence of an intracellular signaling pathway triggered by ANG II in the ER, which accounts for the specific effect of luminal ANG II on SERCA activity to enhance Ca²⁺ storage (19), a process essential for regulation of Ca²⁺-modulated fluid reabsorption in proximal tubule cells (15). The ER can form signaling intermediates of the downstream AT₁ receptors in the ANG II pathway (63). PKC, a central effector protein in ANG II signaling (15, 19), is not only activated (19), but its content is increased in the ER fraction after ANG II treatment (Fig. 7, B and C), indicating that a full preexisting G protein transduction system is enhanced in response to luminal ANG II. This point deserves special discussion because of the observation that ANG II internalization occurs at control levels when losartan and PD123319 are added together (Figs. 1 and 2). Binding of the antagonists to their corresponding target during preincubation might be mutually exclusive. In other words, with both compounds present, the dynamics of the individual AT₁R with losartan and AT₂R with PD123319 in the heterodimer is sterically hindered. As a consequence, the heterodimer permits internalization, as under control conditions without the blockers. This phenomenon is in line with the observation that ANG II-stimulated SERCA activity is maintained in the presence of both losartan and PD123319 (Fig. 3), confirming that both antagonists are ineffective in blocking heterodimerization and/or ANG II internalization when assayed together (Figs. 1 and 2). The ligand-specific conformational changes induced in ligand-binding domains, and the long-range intramolecular communications required for ANG II receptor blocking (64), cannot occur simultaneously in the heterodimer.

In conclusion, we have found new evidence for the formation and internalization of a proximal tubule luminal ANG II-AT₁R/AT₂R complex to target the ER, where it triggers specific intracellular Ca²⁺ responses. The role of ANG II-AT₁R/AT₂R signaling in physiological and pathological aspects of Ca²⁺ homeostasis in the renal system requires further investigation. GPCR heterodimers are known to trigger different signaling pathways compared with homodimers and monomers (9, 12, 45). Understanding the signaling pathways triggered by internalized ANG II heterodimer receptors directed to the ER may help the development of new therapeutic approaches in renal pathophysiological conditions in which imbalance of Ca²⁺ is a central event (11, 29).

GRANTS

This work was supported by the following agencies: the National Council for Scientific and Technological Development (CNPq), the Carlos Chagas Filho Research Support Foundation, the National Institutes of Science and Technology, and the Coordination for the Improvement of Higher Education Personnel (Brazil); the National Institute of Diabetes & Digestive & Kidney Diseases and the American Heart Association in Aid (USA) (2R01DK-067299 and 1R01DK-102429-01 award to J. L. Zhuo). F. M. Ferrão was supported by a fellowship grant from the Brazilian CNPq to support her stay at the University of Mississippi Medical Center.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

F.M.F., A.V., and J.L. conceived and designed research; F.M.F., L.H.C., H.A.D., X.C.L., J.L.Z., D.S.G., L.S.L., and J.L. performed experiments; F.M.F., L.H.C., H.A.D., X.C.L., J.L.Z., D.S.G., L.S.L., A.V., and J.L. analyzed data; F.M.F., L.H.C., H.A.D., X.C.L., J.L.Z., D.S.G., L.S.L., A.V., and J.L. interpreted results of experiments; F.M.F., L.H.C., H.A.D., X.C.L., J.L.Z., L.S.L., A.V., and J.L. prepared figures; F.M.F., L.H.C., L.S.L., A.V., and J.L. drafted manuscript; F.M.F., L.H.C., H.A.D., X.C.L., J.L.Z., D.S.G., L.S.L., A.V., and J.L. edited and revised manuscript; F.M.F., L.H.C., H.A.D., X.C.L., J.L.Z., D.S.G., L.S.L., A.V., and J.L. approved final version of manuscript.