Resistance to hypertension mediated by intercalated cells of the collecting duct

Abstract

The renal collecting duct (CD), as the terminal segment of the nephron, is responsible for the final adjustments to the amount of sodium excreted in urine. While angiotensin II modulates reabsorptive functions of the CD, the contribution of these actions to physiological homeostasis is not clear. To examine this question, we generated mice with cell-specific deletion of AT_1A receptors from the CD. Elimination of AT_1A receptors from both principal and intercalated cells (CDKO mice) had no effect on blood pressures at baseline or during successive feeding of low- or high-salt diets. In contrast, the severity of hypertension caused by chronic infusion of angiotensin II was paradoxically exaggerated in CDKO mice compared with controls. In wild-type mice, angiotensin II induced robust expression of cyclooxygenase-2 (COX-2) in renal medulla, primarily localized to intercalated cells. Upregulation of COX-2 was diminished in CDKO mice, resulting in reduced generation of vasodilator prostanoids. This impaired expression of COX-2 has physiological consequences, since administration of a specific COX-2 inhibitor to CDKO and control mice during angiotensin II infusion equalized their blood pressures. Stimulation of COX-2 was also triggered by exposure of isolated preparations of medullary CDs to angiotensin II. Deletion of AT_1A receptors from principal cells alone did not affect angiotensin II-dependent COX2 stimulation, implicating intercalated cells as the main source of COX2 in this setting. These findings suggest a novel paracrine role for the intercalated cell to attenuate the severity of hypertension. Strategies for preserving or augmenting this pathway may have value for improving the management of hypertension.

Figure 1. Hoxb7-Cre expression in principal and intercalated collecting duct cells.

Representative kidney cross-sections ofHoxb7-Cre⁺ mT/mGmice. (A and B) Green fluorescence indicates the presence of Hoxb7-Cre expression, whereas red fluorescence indicates the absence of Hoxb7-Cre expression (original magnification, [A] ×4 and [B] ×40). (C) Double immunofluorescence with either GFP and AQP2 or GFP and V-ATPase confirms that Hoxb7-Cre is expressed in principal and intercalated cells of the collecting duct in the kidney (original magnification, ×20). (D) AT_1A receptor mRNA expression is significantly reduced in isolated collecting ducts of CDKO compared with control mice (**P < 0.01 vs. control; n = 6–8). Unpaired Student’s t test was used. (E) Baseline blood pressures did not differ between CDKO and control mice (n = 10). Blood pressures significantly decrease and increases to a similar extend on low-salt and high-salt diets in CDKO and control mice (*P < 0.05; **P < 0.01 vs. baseline; n = 10). ANOVA followed by Bonferroni’s multiple comparisons post-hoc test was used to compare the differences between the groups.

Figure 2. Exaggerated blood pressure response to angiotensin II in mice lacking AT1 receptors from the collecting duct (CDKO mice).

(A) Chronic angiotensin II infusion (1,000 ng/kg/min) caused a greater hypertensive blood pressure response in CDKO compared with control mice (MAP: 163 ± 3 vs. 151 ± 3 mmHg, *P < 0.001 vs. angiotensin II control; n = 16). (B) In accordance with the exaggerated blood pressure response, cumulative positive sodium balance was significantly increased in CDKO compared with control mice during the first 5 days of angiotensin II infusion (0.75 ± 0.07 vs. 0.50 ± 0.04 mmol; *P < 0.02 vs. angiotensin II control; n = 6). ANOVA followed by Bonferroni’s multiple comparison post-hoc test was used to compare the differences between the groups.

Figure 3. NO does not modify the exaggerated blood pressure response to angiotensin II in mice lacking AT₁ receptors from the collecting duct (CDKO mice).

(A) During the first 5 days of angiotensin II infusion, cumulative urinary nitrate/nitrite levels did not differ significantly between CDKO and control mice (102.7 ± 26.30 vs. 69.34 ± 27.68 mM/mg/dl creatinine; P = 0.12; n = 6). Unpaired Student’s t test was used to compare the various groups. (B) The exaggerated blood pressure response to angiotensin II is not due to impaired NO generation, as chronic administration of l-NAME (20 mg/kg/d) did not decrease the blood pressure difference between CDKO and control mice (n = 6). ANOVA followed by Bonferroni’s multiple comparison post-hoc test was used to compare the differences between the groups.

Figure 4. Deletion of collecting duct–specific AT_1A receptors modifies medullary COX-2 expression.

(A) After 14 days of angiotensin infusion, medullary COX-2 mRNA expression was significantly increased in CDKO (1.0 ± 0.1 vs. 5.5 ± 0.5 AU; *P < 0.001; n = 6–11) and control mice (1.0 ± 0.2 vs. 10.7 ± 1.8 AU; *P < 0.001; n = 6–11) compared with untreated mice. Moreover, the increase in medullary COX-2 mRNA expression was significantly attenuated in CDKO compared with control mice (10.7 ± 1.8 vs. 5.5 ± 0.5 AU; ^#P < 0.01; n = 11). Unpaired Student’s t test was used to compare the various groups. (B) Likewise, medullary COX-2 protein levels were significantly reduced in angiotensin II– treated CDKO compared with control mice (0.5 ± 0.1 vs. 1.0 ± 0.3; *P < 0.05; n = 6). Unpaired Student’s t test was used. Lanes were run on the same gel but were noncontiguous. See complete unedited blots in the supplemental material. Results are representative of 3 independent experiments. (C) In order to determine the cellular source of COX-2, immunocytochemistry was performed in the medulla of untreated and angiotensin II–treated control mice. In untreated mice, representative confocal laser scanning microscopy showed minimal COX-2 staining. By contrast, in angiotensin II–treated control mice, COX-2 staining was apparent and colocalized almost completely with V-ATPase, a marker for intercalated cells (original magnification, ×20).

Figure 5. Densitometric analysis of COX-2 in isolated inner medullary collecting ducts of 129/SvEv, Agtr1a^–/–, PCKO, and CDKO mice.

(A) Angiotensin II (Ang II) increases COX-2 expression in isolated collecting ducts of 129/SvEv control mice in a dose-dependent manner (10^–8 M: 1.8 ± 0.5 COX-2/GAPDH [fold increase], *P < 0.05 vs. control; n =5; 10^–7 M: 2.3 ± 0.7 COX-2/GAPDH [fold increase], P < 0.05 vs. control; n =5). Paired Student’s t test was used to compare the different groups. (B) Angiotensin II (10^–7 M) failed to increase COX-2 protein in isolated collecting ducts of Agtr1a^–/– mice. (C and D) Angiotensin II (10^–7 M) increases COX-2 protein levels by about 2-fold in isolated collecting ducts of mice lacking the AT_1A receptor only in principal cells (1.7 ± 0.2 vs. 1.0 ± 0.4; *P < 0.05; n = 5) but failed to induce COX-2 expression in CDKO mice. Unpaired Student’s t test was performed to compare the various groups. (E and F) Angiotensin II (10^–7 M) significantly induced COX-2 expression in MDCK cells (1.00 ± 0.10 [0 hours]; 1.30 ± 0.20 [1 hour]; 3.70 ± 0.63 [3 hours]; 6.38 ± 0.60 [6 hours]; 1.90 ± 0.76 [12 hours]; 1.14 ± 0.26 [24 hours]; *P < 0.05, **P < 0.01; n = 4) and significantly increased prostaglandin E₂ (PGE₂) generation measured in the supernatant of the stimulated MDCK cells (272 ± 53 [0 hours]; 4,397 ± 288 [6 hours]; 4,752 ± 209 [12 hours]; 4,310 ± 157 [24 hours]; ***P < 0.001; n = 4). ANOVA followed by Bonferroni’s multiple comparison post-hoc test was used to compare the differences between the groups. See complete unedited blots in the supplemental material.

Figure 6. AT_1A receptors in the collecting duct stimulate vasodilator prostanoids.

After chronic angiotensin II infusion, urinary metabolites of the vasodilator prostanoids (A) PGE₂ and (B) prostacyclin (6-keto PGF_1a) were significantly reduced in CDKO mice compared with control mice (PEG₂: 2,395 ± 441 vs. 5,852 ± 1,726 pg/24-hour urine; **P < 0.01 vs. angiotensin II–treated controls, n = 10–11; prostacyclin 6-keto PGF_1a: 1,889 ± 386 vs. 5,125 ± 1,374 pg/24-hour urine; *P < 0.05 vs. angiotensin II–treated controls, n = 10–11). (C) In contrast, urinary thromboxane (TXB₂) excretion did not differ between CDKO and control mice (3,233 ± 839 vs. 4,451 ± 1,011 pg/24-hour urine, n = 10–11). Unpaired Student’s t test was performed to compare the different groups.

Figure 7. Effects of COX-2 inhibition on angiotensin II–dependent hypertension.

After 1 week of chronic angiotensin II treatment, CDKO and control mice were co-treated with a specific COX-2 inhibitor (rofecoxib 10 mg/kg/d). After 1 week of angiotensin II infusion, MAP was significantly higher in CDKO compared with control mice (156 ± 4 vs. 142 ± 6 mmHg; *P < 0.05; n = 10). Co-treatment with rofecoxib (10 mg/kg) caused a significant increase in blood pressure in both groups and abolished the difference in blood pressure between CDKO and control mice (163 ± 6 vs. 161 ± 5 mmHg; P = 0.70; n = 10). ANOVA followed by Bonferroni’s multiple comparisons post-hoc test was used to compare the differences between the groups.