Expression of a dominant negative PKA mutation in the kidney elicits a diabetes insipidus phenotype

Abstract

PKA plays a critical role in water excretion through regulation of the production and action of the antidiuretic hormone arginine vasopressin (AVP). The AVP prohormone is produced in the hypothalamus, where its transcription is regulated by cAMP. Once released into the circulation, AVP stimulates antidiuresis through activation of vasopressin 2 receptors in renal principal cells. Vasopressin 2 receptor activation increases cAMP and activates PKA, which, in turn, phosphorylates aquaporin (AQP)2, triggering apical membrane accumulation, increased collecting duct permeability, and water reabsorption. We used single-minded homolog 1 (Sim1)-Cre recombinase-mediated expression of a dominant negative PKA regulatory subunit (RIαB) to disrupt kinase activity in vivo and assess the role of PKA in fluid homeostasis. RIαB expression gave rise to marked polydipsia and polyuria; however, neither hypothalamic Avp mRNA expression nor urinary AVP levels were attenuated, indicating a primary physiological effect on the kidney. RIαB mice displayed a marked deficit in urinary concentrating ability and greatly reduced levels of AQP2 and phospho-AQP2. Dehydration induced Aqp2 mRNA in the kidney of both control and RIαB-expressing mice, but AQP2 protein levels were still reduced in RIαB-expressing mutants, and mice were unable to fully concentrate their urine and conserve water. We conclude that partial PKA inhibition in the kidney leads to posttranslational effects that reduce AQP2 protein levels and interfere with apical membrane localization. These findings demonstrate a distinct physiological role for PKA signaling in both short- and long-term regulation of AQP2 and characterize a novel mouse model of diabetes insipidus.

Keywords: aquaporin-2; arginine vasopressin; cAMP; principal cell; protein kinase A.

Fig. 1.

Phenotypic analysis after Sim1-Cre-mediated expression of a dominant negative PKA regulatory subunit (RIαB). A: COOH-terminal exons of Prkar1a gene showing activation of the RIαB mutation (G324D) after Cre-mediated recombination. RIαB-OFF mice are heterozygous for Prkar1a in all tissues. B: breeding strategy and generation of experimental animals. C–G: average daily food intake (C), reproductive fat pad weight (D), body weight (E), average water intake per hour measured over a 5-day period (F), and urine output per hour measured over a 4-h period (in G) in control [wild-type (WT) and RIαB-OFF] and RIαB-ON mice (n = 5–8 mice/group). *P < 0.05; **P < 0.01; ***P < 0.001. Values were measured in 8-wk-old male and female mice and represent means ± SE.

Fig. 2.

Isolation of cell type-specific transcripts from tissues where Sim1-Cre is active. A: Sim1-Cre mice were crossed to the ROSA26R reporter mouse line, and cryosections from the hypothalamus and kidney were stained with X-gal to reveal β-galactosidase expression. B: schematic depicting purification of ribosome-bound mRNA through RiboTag immunoprecipitation (IP). Sim1-Cre mice were crossed to the RiboTag mouse, and the hypothalamus or kidneys were homogenized as previously described (51, 52). Protein A/G magnetic bead/anti-hemagglutinin (HA) antibody complexes were used to immunoprecipitate polyribosomes and bound mRNAs for gene expression analysis of Sim1-positive cells in each tissue. C: representative quantitative RT-PCR analysis demonstrating enrichment of Sim1 mRNA in IPs from both the hypothalamus (left) and kidney (right). Transcripts analyzed included 2′,3′-cyclic nucleotide 3′-phosphodiesterase (Cnp), β-actin (Actb), Prkar1a, melanocortin-4 receptor (Mc4r), arginine vasopressin (Avp), Slc26a4, Slc14a2, ribosomal protein S18 (Rps18), and aquaporin 2 (Aqp2) (see Table 1 for details). Enrichment shown is the ratio of IP mRNA versus input mRNA from the crude tissue homogenate.

Fig. 3.

Avp mRNA expression and circulating peptide levels. A: colocalization of AVP and HA after Sim1-Cre mice were crossed to the RiboTag mouse. 3V, third ventricle. B: total PKA activity in the presence or absence of 5 μM cAMP (n = 3 mice/group). Homogenates were prepared as described from punches that were taken from coronal sections containing paraventricular nucleus (PVN) neurons. C: total hypothalamic expression of Avp mRNA in control and RIαB-ON mice with ad libitum water consumption or after 24-h dehydration (n = 4 mice/group). *P < 0.05. D: urinary AVP levels with ad libitum water consumption, as assessed by radioimmunoassay (n = 4 mice/group).

Fig. 4.

Effect of RIαB mutation on AQP2 localization and expression. A: Sim1-Cre-driven expression of tdTomato fluorescent protein and colocalization with AQP2 in the inner medulla of the kidney. B: AQP2 localization in principal cells of RIαB-OFF and RIαB-ON mice. C: representative Western blots and quantitation of total AQP2 protein levels in the inner medulla. Quantitation shows both glycosylated (Gly-AQP2) and nonglycosylated (NG-AQP2) bands. β-Actin was used as a loading control. n = 4. ***P < 0.001. D: total Aqp2 mRNA levels in isolated inner medullas from control and RIαB-ON mice (n = 4). Actb transcript levels were used to ensure equivalent loading of RNA. E: total inner medulla PKA activity in the presence or absence of 5 μM cAMP. n = 4. ***P < 0.001. F: representative Western blot and quantitation of PKA regulatory and catalytic subunit expression in the inner medulla of control (WT and RIαB-OFF) and RIαB-ON mice. n = 3. *P < 0.05; **P < 0.01; ***P < 0.001.

Fig. 5.

Effect of desmopressin (DDAVP) administration on urinary concentration and AQP2 phosphorylation and trafficking. A: mice were water loaded to suppress endogenous AVP levels and, after 1 h, injected with vehicle or DDAVP (0.1 μg/kg). Urine was collected for 2 h, and osmolality was assessed as described in experimental procedures. n = 4. *P < 0.05; #P < 0.001. B: in response to DDAVP injection, urinary cAMP excretion was assessed by ELISA in control and RIαB-ON mice (n = 3). C: control and RIαB-ON mice were water loaded as in A and euthanized 20 min after vehicle or DDAVP injection, and AQP2 localization was assessed by immunohistochemistry. D: additional groups of mice were treated as in C, and inner medullas were microdissected and dounce homogenized in RIPA buffer for Western blot analysis. Representative Western blots and quantitation of total AQP2 and phosphorylated (P-)AQP2 (Ser²⁵⁶) are shown. n = 3. *P < 0.05; **P < 0.01. β-Actin was used as a loading control.

Fig. 6.

Effect of 24-h dehydration on urinary concentration and AQP2 phosphorylation and trafficking. A: urine osmolality in control and RIαB-ON mice with ad libitum water consumption or after 24-h dehydration. n = 6–8. **P < 0.01; #P < 0.001. B: quantitative RT-PCR analysis of Aqp2 and Aqp3 expression in control and RIαB-ON mice with or without 24-h dehydration (n = 4). Actb transcript levels were used as a control. C, top: control and RIαB-ON mice were subjected to 24-h dehydration, and total AQP2 localization was assessed by immunohistochemistry. Bottom, low-magnification image of AQP2 expression in RIαB-OFF and RIαB-ON mice subjected to 24-h dehydration. D: representative Western blots and quantitation of total AQP2 and P-AQP2 (Ser²⁵⁶) levels in control and RIαB-ON mice with ad libitum water consumption or after 24-h dehydration. β-Actin was used as a loading control. n = 3. *P < 0.05; **P < 0.01; #P < 0.001.

Fig. 7.

Effect of genetic deletion of RII subunits on urinary concentration. A: immunohistochemistry demonstrating colocalization of PKA RII subunits and AQP2 in the inner medulla. B: urine osmolality in WT mice and mice lacking RIIα or RIIβ regulatory subunits with free access to water or after 24-h dehydration (n = 5).