Severe Hyperaldosteronism in Neonatal Task3Potassium Channel Knockout Mice Is Associated WithActivation of the Intraadrenal Renin-AngiotensinSystem

Sascha Bandulik,* Philipp Tauber,* David Penton, Frank Schweda, Ines Tegtmeier,Christina Sterner, Enzo Lalli, Florian Lesage, Michaela Hartmann,Jacques Barhanin, and Richard Warth

Departments of Medical Cell Biology (S.B., P.T., D.P., I.T., C.S., R.W.), and Physiology (F.S.), University ofRegensburg, 93053 Regensburg, Germany; Institut de Pharmacologie Moléculaire et Cellulaire (D.P.,E.L., F.L.), Centre National de la Recherche Scientifique (CNRS), and Université de Nice Sophia Antipolis,06560 Valbonne, France; Division of Pediatric Endocrinology and Diabetology (M.H.), Steroid Researchand Mass Spectrometry Unit, Center of Child and Adolescent Medicine, Justus Liebig University, 35385Gießen, Germany; Transport Ionique Aspects Normaux et Pathologiques (J.B.), CNRS, and Université deNice Sophia Antipolis, UMR6097, 06108 Nice Cedex, France; and Laboratories of Excellence Ion ChannelScience and Therapeutics (F.L., J.B.), CNRS UMR7275 Université de Nice Sophia Antipolis, F-06560Valbonne, France

Task3 K� channels are highly expressed in the adrenal cortex and contribute to the angiotensin IIand K� sensitivity of aldosterone-producing glomerulosa cells. Adult Task3�/� mice display a par-tially autonomous aldosterone secretion, subclinical hyperaldosteronism, and salt-sensitive hyper-tension. Here, we investigated the age dependence of the adrenal phenotype of Task3�/� mice.Compared with adults, newborn Task3�/� mice displayed a severe adrenal phenotype with stronglyincreased plasma levels of aldosterone, corticosterone, and progesterone. This adrenocortical dys-function was accompanied by a modified gene expression profile. The most strongly up-regulatedgene was the protease renin. Real-time PCR corroborated the strong increase in adrenal reninexpression, and immunofluorescence revealed renin-expressing cells in the zona fasciculata. To-gether with additional factors, activation of the local adrenal renin system is probably causative forthe severely disturbed steroid hormone secretion of neonatal Task3�/� mice. The changes in geneexpression patterns of neonatal Task3�/� mice could also be relevant for other forms ofhyperaldosteronism. (Endocrinology 154: 2712–2722, 2013)

The function of potassium channels in adrenal zona glo-merulosa cells is pivotal for the control of aldosterone

secretion. A very high K� conductance in glomerulosacells leads to a hyperpolarized membrane potential closeto the K� equilibrium potential. Thereby, even a smallincrease of plasma K� or inhibition of the K� conductanceby angiotensin II (Ang II) result in membrane depolariza-tion, which activates voltage-sensitive Ca2� channels. To-gether with a release of Ca2� from intracellular stores, thisleads to an increase of the cytosolic Ca2� concentrationthat activates downstream signaling cascades and ulti-

mately aldosterone synthesis (1). Improperly high aldo-sterone secretion can cause salt and water retention and isone of the major factors for the pathogenesis of arterialhypertension (2, 3).

In addition to KCNJ5 (4–9) potassium channels, mem-bers of 2P-domain potassium channel family (KCNK)have been the focus of research because KCNK2 (Trek1),KCNK3 (Task1; Task: TWIK-related acid-sensitive K�

channel), and KCNK9 (Task3) are strongly expressed inadrenal glands and regulated by Ang II and ACTH, twokey stimuli of adrenocortical cells (10). The physiologic

ISSN Print 0013-7227 ISSN Online 1945-7170Printed in U.S.A.Copyright © 2013 by The Endocrine SocietyReceived January 28, 2013. Accepted May 13, 2013.First Published Online May 22, 2013

* S.B. and P.T. contributed equally to this work.Abbreviations: Ang II, angiotensin II; Cry, cryptochrome; Cyp11b2, name of the aldoste-rone synthase gene; KCNK3, name of the TASK1 gene; KCNK9, name of the TASK3 gene;ss, sense strand; Task, TWIK-related acid-sensitive K� channel; TWIK, tandem of P domainsin a weak inward rectifying K� channel.

G L U C O C O R T I C O I D S - C R H - A C T H - A D R E N A L

2712 endo.endojournals.org Endocrinology, August 2013, 154(8):2712–2722 doi: 10.1210/en.2013-1101

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 August 2014. at 00:03 For personal use only. No other uses without permission. . All rights reserved.

relevance of Task1 and Task3 channels has been high-lighted in mouse models: Female Task1�/� mice displayeddepolarization of adrenocortical cells and a remarkableectopic expression of aldosterone synthase in the zonafasciculata of the adrenal cortex. This mislocalizationcaused a diet-independent, low-renin, and glucocorticoid-remediable form of hyperaldosteronism. Male Task1�/�

mice exhibited the same phenotype only before puberty,but recovered at adult stage when testosterone-dependentmechanisms restored normal zonation of aldosterone syn-thase (11). Task1/Task3 double-knockout male miceshowed low renin hyperaldosteronism without ectopic ex-pression of aldosterone synthase in the zona fasciculata(12). Although glomerulosa cells were strongly depolar-ized, these mice could partially adapt their aldosteronesecretion to a high or low Na� intake.

The different phenotype in single Task1�/� andTask1�/�/Task3�/� double-mutant mice pointed to dif-ferent functions of Task1 and Task3 channels in adreno-cortical cells. In fact, Task3�/� mice didn’t show ectopicexpression of aldosterone synthase but, nevertheless, astrong increase of the plasma aldosterone/renin ratio anda salt-sensitive hypertension (13, 14). Adrenal cells fromTask3�/� mice were strongly depolarized and showed analtered electrical and Ca2� response to K� and Ang II (13).Consequently, a fraction of aldosterone secretion inTask3�/� mice was autonomous and independent fromAng II. Interestingly, a compensatory decrease of reninlargely normalized plasma aldosterone levels of Task3�/�

mice under normal salt diet. The importance of Task3K�-channels for human adrenal gland function was un-derlined by the finding that common TASK3 gene varia-tions are associated with the risk for hyperaldosteronismand hypertension in humans (15).

The aim of the present study was to gain further insightsinto the specific function of Task3 in adrenal glands byinvestigating the adrenal phenotype of Task3�/� miceduring postnatal development. In contrast to adult mice,which compensated the autonomous aldosterone secre-tion under normal salt diet, newborn Task3�/� mice pre-sented a strong hyperaldosteronism. In addition, the new-born Task3�/� mice had increased plasma corticosteroneand progesterone concentrations, pointing to a generaldisturbance of adrenal function. To disclose factors thatpossibly contribute to the severity of the newborn pheno-type, a genechip analysis of developing adrenal glands wasperformed. Renin mRNA and protein expression, as wellas renin activity, were found to be markedly increased inadrenal glands of newborn Task3�/� mice, whereasplasma renin was not. These findings suggested that acti-vation of the local adrenal gland renin-angiotensin systemaggravated the phenotype in newborn Task3�/� mice.

Materials and Methods

Animal experimentsTask3�/� mice were generated as described elsewhere (16).

The animals were backcrossed for 11 generations into the C57/Bl/6J genetic background. The animals had free access to foodand water. Age-matched mice were used in all experiments. Datawere obtained from both sexes, if not stated otherwise. The ex-perimental protocols were approved by the local councils foranimal care and were conducted according to the German andFrench laws for animal care and the NIH Guide for the Care andUse of Laboratory Animals. Mouse anesthesia was carried outwith 1.5%–3% isoflurane (Baxter Deutschland GmbH, Unter-schleißheim, Germany) or with a mixture of ketamine(120 �g/gbody weight; WDT, Garbsen, Germany) and xylazine (8 �g/gbody weight; Serumwerk, Bernburg, Germany).

Steroid hormone measurements in plasma samplesand adrenal tissue lysates

Blood samples from adult anesthetized mice were taken fromthe retroorbital venous plexus. Blood samples from neonatalanesthetized mice were obtained lethally by incision of the rightatrium. Blood was collected in heparinized capillaries. Plasmawas separated by centrifugation in a hematocrit centrifuge andstored at �20°C until measurements. Specific ELISAs (from IBL,Hamburg, Germany) were used to determine the concentrationsof aldosterone, corticosterone, progesterone, and pregnenolonein appropriate dilutions of plasma. Before the measurement ofaldosterone concentration in tissue lysates, both adrenal glandswere isolated and homogenized in PBS with 10% BSA, afterwhich steroids were extracted using ethylacetate (Sigma-AldrichGmbH, Taufkirchen, Germany). The organic phase was dried byevaporation under nitrogen, and the steroid containing precip-itate was dissolved in aldosterone-free serum. As negative con-trols, aldosterone content of tissue samples from liver and kidneywas measured; they showed less than 5% of aldosterone-specificactivity per wet weight compared with adrenal samples.

Immunofluorescence stainingAdult anesthetized mice were killed by replacement of blood

by 0.9% NaCl solution containing 10 I.U./mL heparin via acatheter placed into the abdominal aorta. In neonatal mice,blood was replaced via a syringe and needle placed into the leftventricle of the heart. For tissue fixation, mice were then perfusedwith a fixative containing (mM): 3% paraformaldehyde, 100sucrose, 90 NaCl, 15 K2HPO4, 1 EGTA, 2 MgCl2, pH 7.4. Ad-renals and adjacent kidneys were removed, incubated in 17%sucrose solution for 30 minutes, and frozen in isomethylbutane(�30°C). Cryosections (5 �m) were mounted on Polysine slides(Kindler, Freiburg, Germany). Before incubation with the pri-mary antibodies, sections were incubated in sodium dodecyl sul-fate 0.1% (5 min) to unmask epitopes and rinsed again with PBS.Blocking of unspecific binding sites was done in PBS (pH 7.4)containing 0.04% Triton X-100 and 5% BSA (Sigma-AldrichGmbH, Taufkirchen, Germany). Primary and secondary anti-bodies were diluted in blocking solution containing 0.5% in-stead of 5% BSA. Primary antibodies were applied overnight at4°C: a monoclonal rat antirenin antibody (R&D Systems, Min-neapolis, Minnesota), a polyclonal rabbit anti-Dab2 antibody(Santa Cruz Biotechnology, Santa Cruz, California), a poly-

doi: 10.1210/en.2013-1101 endo.endojournals.org 2713

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 August 2014. at 00:03 For personal use only. No other uses without permission. . All rights reserved.

clonal rabbit antialdosterone synthase antibody (kindly pro-vided by Dr. Celso Gomez Sanchez [Refs. 16 and 17]) and apolyclonal rabbit anti-Task3 antibody (Alomone Labs, Jerusa-lem, Israel). As secondary antibodies, appropriate CY2- and Al-exa555-labeled antibodies were used (Life Technologies GmbH,Darmstadt, Germany).

Renin concentration in plasma samples and intissue lysates

For the measurement of the plasma renin concentration ofmice, blood samples were taken via puncture of the heart andcentrifuged in heparinized capillaries. Adrenal gland and kidneytissues were homogenized on ice in lysate buffer containing: 5%Glycerin, 10 mM EDTA (pH 8), 0.1 mM phenylmethylsulfonyl-fluoride, and 0.1 mM 4-(2-aminoethyl)-benzensulfonylfluoride.Tissue lysates were centrifuged for 5 minutes at 14 000 � g (4°C)to obtain a clear supernatant for renin measurements. Plasmaand tissue lysate samples were incubated for 1.5 hours at 37°Cwith plasma of bilaterally nephrectomized male rats as reninsubstrate. The production of Ang I (ng/mL/h) was measured by125I RIA (Byk and DiaSorin Diagnostics, Sulzfeld, Germany) todetermine the renin concentration. Values obtained from tissuelysates were normalized against protein concentrations mea-sured by a standard Bradford assay (Biorad, München,Germany).

RNA isolationTotal RNA from adrenal glands was isolated using a column-

based kit optimized for the purification from small amounts oftissue according to the manual (RNeasy Micro Kit, Qiagen,Hilden, Germany). The RNA concentration was quantified witha photometer (Nanodrop, PEQLAB Biotechnologie GmbH, Er-langen, Germany). Integrity of the RNA used for genechip mi-croarray assay was verified on an Agilent 2100 Bioanalyzer withthe EukaryoteTotal RNA Nano chip (Agilent Technologies, Bö-blingen, Germany). Quality of the RNA used for real-time RT-PCR was tested by agarose electrophoresis.

Quantitative real-time RT-PCRReverse transcription with M-MLV-RT (Promega GmbH,

Mannheim, Germany) and random primers (Fermentas GmbH,St. Leon-Rot, Germany) was done using 1 �g total RNA to gen-erate single-stranded cDNA. Relevant contamination withgenomic DNA was excluded by negative control reactions with-out the reverse transcriptase enzyme (�RT). Real-time PCR ofcDNA samples was performed on a LightCycler 480 device(Roche, Basel, Switzerland) using specific and, wherever appli-cable, intron-spanning primers, and a Sybr Green mastermix(Roche). Target gene expression levels in adrenal glands fromTask3�/� and Task3�/� mice were quantified relative to �-actinexpression under consideration of PCR efficiencies calculated onthe basis of standard dilution curves. The level of �-actin ex-pression in adrenal glands did not alter in the experimental con-ditions. The specificity of PCR amplifications was verified byagarose electrophoresis and melting curve analysis. Primer se-quences (Life Technologies GmbH, Darmstadt, Germany) arelisted in Supplemental Table 1 published on The Endocrine So-ciety’s Journals Online web site at http://endo.endojournals.org.

Genechip microarray assaySample preparation for microarray hybridization was carried

out as described in the Ambion WT Expression Kit Protocol (LifeTechnologies, Carlsbad, California) and the Affymetrix WT Ter-minal Labeling and Hybridization User Manual (Affymetrix,Inc., Santa Clara, California). In brief, 300 ng of total adrenalRNA from 1- and 12 day-old Task3�/� and Task3�/� male mice,respectively, were used to generate double-stranded cDNA. Sub-sequently synthesized cRNA was purified and reverse tran-scribed into sense-strand (ss) cDNA, whereby unnatural dUTPresidues were incorporated. Purified ss cDNA was fragmentedusing a combination of uracil DNA glycosylase and apurinic/apyrimidinic endonuclease 1 followed by a terminal labelingwith biotin. Fragmented and labeled ss cDNA (2.8 �g) was hy-bridized to an Affymetrix Mouse Gene 1.1 ST Array Plate. Forhybridization, washing, staining, and scanning an AffymetrixGeneTitan system was used. Sample processing was performedat an Affymetrix Service Provider and Core Facility: “KFB-Cen-ter of Excellence for Fluorescent Bioanalytics” (Regensburg,Germany; www.kfb-regensburg.de).

Microarray data analysisMicroarray data were obtained from 4 biological replicates

per group. Summarized probe set signals were calculated by us-ing the RMA (18) algorithm with the Affymetrix GeneChip Ex-pression Console Software. After exporting into Microsoft Ex-cel, average signal values, comparison fold changes, andsignificance P values were calculated. Probe sets with a foldchange �2-fold and a Student’s t test P � .05 were considered assignificantly regulated. A short list of regulated genes with a foldchange �3-fold is shown in the main manuscript; a complete listcan be found in the Supplemental Data Sheet. Regulation ofgenes that are presumably relevant for adrenal function was ver-ified with quantitative real-time PCR in biological replicates.

StatisticsThe numbers of experiments stated in the figures refer to the

number of animals that have been studied to calculate meanvalues � SEM (biological replicates). For the gene-chip microar-ray assay only male mice were used. All other data were collectedfrom male and female mice and pooled to calculate mean values,because there were no significant sex-dependent differences. In-dividual mice were studied to generate data corresponding toeach time point along the development because blood samplingfrom neonatal mice and removal of the adrenal glands werelethal experiments. Accordingly, unpaired Student’s t test wasused to calculate the level of significance. Differences betweenthe genotypes for individual time points were considered sig-nificant if P � .05.

Results

Localization of Task3 and Cyp11b2 in adrenalglands of newborn mice

Newborn C57/Bl6 mice express the Task3 protein inthe plasma membrane of cells in the whole adrenal cortex,with a higher level in aldosterone-producing glomerulosacells in the outer zone (Figure 1A). In contrast to adult

2714 Bandulik et al Hyperaldosteronism in Task3�/� Mice Endocrinology, August 2013, 154(8):2712–2722

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 August 2014. at 00:03 For personal use only. No other uses without permission. . All rights reserved.

mice, which showed a broader distribution for Task3 inmales, the localization of Task3 was independent fromsex before puberty. Similar to adult mice (13), the lo-calization of aldosterone synthase (Cyp11b2) was re-stricted to glomerulosa cells in newborn mice of bothgenotypes (Figure 1B).

Newborn Task3�/� mice have a severe adrenalphenotype

Despite the preserved physiologic localization ofCyp11b2, the concentrations of aldosterone, corticoste-rone, and progesterone in the plasma of newbornTask3�/� mice were strongly increased compared withwild-type mice. Plasma Na� was increased in newbornTask3�/� mice, and hematocrit was slightly (but not sig-nificantly) lower (Supplemental Figure 1). During post-natal development, there was a progressive decrease ofaldosterone, corticosterone, and progesterone in both ge-notypes. This decrease was more pronounced in Task3�/�

mice, leading to the normalization of the plasma levels ofaldosterone, corticosterone, and progesterone at the adultstage (Figure 2, A–C). Plasma pregnenolone showed an

age-dependent decrease in both genotypes and was lowerin 12-day-old Task3�/� mice compared with wild type(Figure 2D).

The mRNA expression of rate-limiting enzymes forproduction of aldosterone (Cyp11b2) and progesterone(Hsd3b6) was increased in newborn Task3�/� mice (Fig-ure 3, A and C). Cyp11b2 expression fell to normal levelsin adult mice, whereas expression of Hsd3b6 remainedhigher in adult Task3�/� compared with wild-type mice.The corticosterone-producing enzyme (Cyp11b1) wasequal in both genotypes at any age (Figure 3B).

Abnormal adrenocortical expression of renin inTask3�/� newborn mice

A genechip analysis of the adrenal mRNA expressionpattern during the postnatal development was performedto identify factors that are associated with the adrenalphenotype of neonatal mice, or with the compensation ofthe Task3 deletion in adult mice. Among the genes withmodified expression levels, renin showed the strongest up-regulation of mRNA expression in adrenal glands of

Figure 1. Task3 (A) and Cyp11b2 (aldosterone synthase) (B)Immunofluorescence staining in adrenal glands from newborn mice.Task3 is mainly expressed in zona glomerulosa cells (Zg) of the adrenalcortex. Task3�/� mice have a normal localization of Cyp11b2 in zonaglomerulosa as it was observed in Task3�/� mice. Zf, zona fasciculata.

Figure 2. Plasma levels of steroid hormones in 1-day, 12-day, andadult Task3�/� and Task3�/� mice. Plasma aldosterone (A) andprogesterone (C) concentrations were increased in 1-day and 12-day-old Task3�/� mice, but normalized in adult mice. Plasmacorticosterone (B) was higher in 1-day-old Task3�/� mice. Plasmapregnenolone (D) was not increased in Task3�/� mice compared withTask3�/� mice. Numbers in brackets represent experiments per group.*, P � .05 comparing genotypes at individual age.

doi: 10.1210/en.2013-1101 endo.endojournals.org 2715

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 August 2014. at 00:03 For personal use only. No other uses without permission. . All rights reserved.

Task3�/� newborns, whereas the expression returned tonormal levels in 12-day-old mice (Table 1 and Supple-mental Table 2).

The differential expression of renin was verified by real-time PCR in adrenal gland samples from 1-day, 12-day,and adult Task3�/� or Task3�/� mice (Figure 4A). Notonly the renin mRNA expression, but also the renin con-centration in extracts of adrenal tissue, measured by thecleavage of angiotensinogen to angiotensin I, was in-creased in newborn Task3�/� mice (Figure 4B). Immuno-fluorescence revealed a strong labeling of renin in zonafasciculata cells of Task3�/� mice that was not detectablein wild-type mice. The physiologic localization of renin injuxtaglomerular cells of the kidney was not altered by thedeletion of Task3 (Figure 4C).

Renal and plasma renin levels in Task3�/� miceNext, we quantitatively tested whether the up-regula-

tion of renin expression in Task3�/� mice is specific for theadrenal gland or part of a generalized increase of reninexpression that involves also other renin-producing tis-sues such as the kidneys. The renin concentration in theplasma, as well as in kidney lysates, was compared be-tween 1- and 12-day-old Task3�/� and wild-type mice.Plasma renin concentrations were not different inTask3�/� compared with wild-type mice (Figure 5A).Renin concentrations of the kidney were even decreased in1-day-old Task3�/� mice, possibly representing a com-pensatory effect due to the hyperaldosteronism. At 12days of age, renal renin concentration was not differentbetween the genotypes (Figure 5B). The plasma aldoste-rone to renin ratio (ARR) was significantly increased inneonatal Task3�/� mice, indicating an autonomous aldo-sterone secretion. (Supplemental Figure 2).

Adrenal phenotype of Task3�/�

mice appears first during latefetal development

In order to test how the adrenalphenotype of Task3�/� mice is re-lated to adrenal development, aldo-sterone content in extracted adrenallysates was measured at days 17 and19 of gestation as well as 1 day afterbirth (Figure 5C). The aldosteronecontent was not different at day 17but significantly increased in Task3�/�

mice at day 19 of gestation. Obvi-ously, the disturbance of adrenalfunction develops before birth.However, the difference in aldoste-rone of adrenal lysates was morepronounced in 1-day-old mice.

Further candidate genes related to the Task3�/�

adrenal phenotypeSeveral genes that were differentially expressed accord-

ing to our genechip analysis were previously described inthe literature as direct or indirect effectors of adrenal func-tion. Those genes, in particular, might be relevant for thephenotype of Task3�/� mice, and therefore, their expres-sion was verified by real-time PCR in adrenal gland sam-ples from 1-day, 12-day, and adult mice (Table 2).

The following factors known to stimulate aldosteronesecretion were transcriptionally up-regulated in neonatalTask3�/� mice: Nr4a2 (Nurr1), a transcription factorknown to promote aldosterone synthase (Cyp11b2) ex-pression) (19, 20); Gadd45b, a factor induced by Ang IIand enhancing the Ang II response (1, 21, 22); Gal (23–27)and Tac1 (28–30), factors known to stimulate corticoste-rone and aldosterone secretion; and the activin A receptorIC that could regulate adrenal development and stimulatealdosterone secretion via its ligand activin A (31, 32).

Other factors are probably involved in limiting hyper-aldosteronism and represent ways for compensation:down-regulated mRNA expression of Nnt, which pro-duces NADPH and is needed for hydroxylating reactionsof all Cyp-enzymes (33, 34); down-regulated mRNA ex-pression of the store-operated Ca2�-channel Trpc5, whichis possibly involved in the generation of the Ca2� signalafter Ang II stimulation (35, 36); increased expression ofthe somatostatin receptor Sstr2, which could inhibit al-dosterone synthesis via binding of somatostatin (37, 38);and increased expression of Dkk3, a factor that probablyreduces aldosterone secretion in male Task1�/� knockoutmice (39).

Figure 3. Adrenal gene expression levels of steroidogenic enzymes in 1-day, 12-day, and adultTask3�/� and Task3�/� mice as measured by real-time PCR. The mRNA expression levels ofCyp11b2 (A) and Hsd3b6 (C) that catalyze the final steps for aldosterone and progesteronesynthesis, respectively, were increased in 1-day and 12-day-old Task3�/� mice. Hsd3b6expression was high at all points in time tested. The expression of the corticosterone-producingenzyme Cyp11b1 (B) was not different between genotypes. Numbers in brackets representexperiments per group. *, P � .05 comparing genotypes at individual age.

2716 Bandulik et al Hyperaldosteronism in Task3�/� Mice Endocrinology, August 2013, 154(8):2712–2722

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 August 2014. at 00:03 For personal use only. No other uses without permission. . All rights reserved.

Discussion

In this study, we examined the age dependence of the ad-renal phenotype of Task3�/� mice. Compared withadults, newborn Task3�/� mice displayed a severe adrenalphenotype with strongly increased plasma levels of aldo-sterone, corticosterone, and progesterone. This adreno-cortical dysfunction was accompanied by a modified geneexpression profile. The most strongly up-regulated geneexpression in adrenals from Task3�/� mice was found forthe protease renin. Together with additional factors, ac-tivation of the local adrenal renin system is probably caus-

ative for the disturbed steroid hormone secretion of neo-natal Task3�/� mice.

Age-dependent disturbance of adrenal steroidhormone synthesis in Task3�/� mice

Adult Task3�/� mice show partially autonomous al-dosterone secretion and, as a consequence, develop salt-sensitive hypertension (13, 14). Task3�/� mice, in con-trast to Task1�/� mice, did not show ectopic expression ofthe aldosterone synthase in zona fasciculata cells. Inter-estingly, the adrenal phenotype of male Task1�/� micewas strictly age-dependent with aberrant synthesis of

Table 1. Differential mRNA Expression in Adrenal Glands from Task3�/� and Task3�/� Mice

Symbol DescriptionFold Change(Task3�/� vs Task3�/�) P Value

mRNA Expression in Adrenal Glands from 1-Day-Old MiceRen1 Renin 1 structural 9.6 1.1E-04A2m �-2-Macroglobulin 6.3 1.7E-02Ecel1 Endothelin converting enzyme-like 1 5.8 5.0E-03Lypd3 Ly6/Plaur domain containing 3 5.3 2.1E-02Nefl Neurofilament, light polypeptide 4.2 7.2E-03Gm10002 Predicted gene 10002 3.9 1.1E-02St6galnac5 ST6 (�-N-acetyl-neuraminyl-2,3-�-galactosyl-1,3)-N-

acetylgalactosaminide �-2,6-sialyltransferase 53.8 2.4E-03

Sv2c Synaptic vesicle glycoprotein 2c 3.6 2.0E-04Slc16a3 Solute carrier family 16 (monocarboxylic acid

transporters), member 33.6 1.6E-02

Hcn1 Hyperpolarization-activated, cyclic nucleotide-gated K� 1 3.2 4.2E-05Sez6l Seizure-related 6 homolog like 3.2 2.3E-04Sncg Synuclein, � 3.1 4.6E-02Slc6a17 Solute carrier family 6 (neurotransmitter transporter), member 17 3.0 1.1E-04Ctse Cathepsin E �5.0 6.5E-03Car1 Carbonic anhydrase 1 �4.5 4.0E-02Kel Kell blood group �4.0 4.6E-02Cldn13 Claudin 13 �4.0 1.7E-02Rhag Rhesus blood group-associated A glycoprotein �3.7 1.6E-02Rhd Rh blood group, D antigen �3.6 2.3E-02Trpc5 Transient receptor potential cation channel, subfamily C,

member 5�3.5 2.3E-04

Hemgn Hemogen �3.4 4.1E-02Slc14a1 Solute carrier family 14 (urea transporter), member 1 �3.3 1.3E-02Spna1 Spectrin � 1 �3.2 2.1E-02Gypa Glycophorin A �3.2 3.8E-02Tspo2 Translocator protein 2 �3.1 1.7E-02Htr1b 5-Hydroxytryptamine (serotonin) receptor 1B �3.1 3.5E-05

mRNA Expression in Adrenal Glands from 12-Day-Old MiceLcn2 Lipocalin 2 10.9 4.9E-02Gm10002 Predicted gene 10002 4.7 8.8E-04Tac1 Tachykinin 1 4.5 1.1E-04Ctse Cathepsin E 4.5 3.7E-02Tm4sf4 Transmembrane 4 superfamily member 4 4.4 5.0E-03Gal Galanin 4.3 7.4E-05Hp Haptoglobin 3.4 3.8E-02Calb1 Calbindin 1 3.1 5.3E-03Sv2c Synaptic vesicle glycoprotein 2c 3.0 3.8E-045730407I07Rik RIKEN cDNA 5730407I07 gene �4.8 3.8E-06Htr1b 5-Hydroxytryptamine (serotonin) receptor 1B �3.1 4.5E-03

Genechip analysis was performed using adrenal mRNA samples from 1- and 12-day-old mice (n � 4 per group and genotype). The fold changewas calculated as the ratio of the mean values. Genes with a positive or negative fold change �3 and a P � 0.05 comparing Task3�/� and Task3�/

� mice at each age are shown. A complete list (fold change � 2) can be found in the Supplemental data.

doi: 10.1210/en.2013-1101 endo.endojournals.org 2717

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 August 2014. at 00:03 For personal use only. No other uses without permission. . All rights reserved.

aldosterone in fasciculata cells before puberty and res-toration of normal zona glomerulosa-restricted aldo-sterone synthesis after puberty (11). Therefore, in thepresent study we examined age-dependent variations ofthe adrenal phenotype of Task3�/� mice. In contrast tothe relatively mild hyperaldosteronism of adult animals(13, 14), newborn Task3�/� mice showed severehyperaldosteronism.

Expression of the aldosterone synthase (Cyp11b2)mRNA in glomerulosa cells is considered as the limitingstep for chronic stimulation of aldosterone synthesis.Cyp11b2 mRNA expression was increased in 1- and 12-day-old Task3�/� mice, but the physiologic localization inglomerulosa cells was preserved. Therefore, the strong hy-peraldosteronism of newborn Task3�/� mice is caused byhypersecretion in glomerulosa cells and not by ectopicexpression of Cyp11b2 in fasciculata cells. The hyperal-dosteronism phenotype of young Task3�/� mice is remi-niscent of the phenotype of adult male double-knockoutTask1/Task3 mice (12), which display hyperaldosteron-ism with normotopic Cyp11b2 expression.

The increase in hormone secretion in newbornTask3�/� mice was not restricted to aldosterone; secretionof corticosterone and progesterone was also strongly in-

creased. In fasciculata cells, the last step in the synthesis ofcorticosterone is catalyzed by the 11ß-hydroxylase (en-coded by the Cyp11b1 gene). Surprisingly, newbornTask3�/� mice expressed normal amounts of Cyp11b1mRNA, leaving the question open by which mechanismsplasma corticosterone was increased to very high levels. Inprinciple, two mechanisms are likely: 1) increased avail-ability of the substrate (11-deoxycorticosterone) of 11ß-hydroxylase or 2) posttranscriptional stimulation of11ß-hydroxylase activity. 11-Deoxycorticosterone is pro-duced by the Cyp21a1 using progesterone as a substrateand, subsequently, 11-deoxycorticosterone is trans-formed to aldosterone in glomerulosa cells or to cortico-sterone in fasciculata cells. The expression level ofCyp21a1 was not significantly increased in Task3�/� mice(data not shown). However, newborn Task3�/� mice hadvery high levels of progesterone that probably give rise tohigh levels of 11-deoxycorticosterone and, ultimately, in-creased levels of corticosterone and aldosterone.

Increased expression of Hsd3b6 yields highprogesterone and aldosterone

The relevance of high progesterone as a predisposingfactor for hyperaldosteronism has been highlighted by a

Figure 4. Renin mRNA expression (A), renin concentration (B), and localization of renin (C) in adrenal glands of Task3�/� and Task3�/� mice.Renin mRNA-expression (A) and adrenal renin concentration (panel B, measured as formation of angiotensin I) were increased in 1-day-oldTask3�/� mice, but normal in older mice. Numbers in brackets represent experiments per group. *, P � .05 comparing genotypes at individualage. C, Immunofluorescence staining of adrenal glands from 1-day-old mice with specific antibodies against renin (green) and the zonaglomerulosa (Zg) cell marker Dab2 (disabled homolog 2 (52); red); nuclear staining with HOE33342 in blue. Upper panel gives an overview of thewhole adrenal gland (a) adjacent to the kidney (k). Lower panel shows the same adrenal glands with higher magnification. Adrenal renin stainingcould only be observed in zona fasciculata (Zf) cells of Task3�/� mice. Typical renal renin staining (*) was visible in both genotypes.

2718 Bandulik et al Hyperaldosteronism in Task3�/� Mice Endocrinology, August 2013, 154(8):2712–2722

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 August 2014. at 00:03 For personal use only. No other uses without permission. . All rights reserved.

recent study (40): knockout mice for the circadian coreclock components Cryptochrome-1 (Cry1) and Crypto-chrome-2 (Cry2) (Cry-null mice) exhibited hyperaldoste-

ronism and salt-sensitive hypertension due to abnormallyhigh Hsd3b6 mRNA expression. Hsd3b6 is transcription-ally controlled by the circadian clock and encodes the“hydroxy-�-5-steroid dehydrogenase, 3 �- and steroid�-isomerase 6”, an enzyme synthesizing progesterone spe-cifically in glomerulosa cells (40). Interestingly, mRNAexpression of Hsd3b6 was also enhanced in Task3�/�

mice and likely contributes to high progesterone and hy-peraldosteronism phenotype of young mice and the salt-sensitive hyperaldosteronism of adult Task3�/� mice.

Hyperaldosteronism of newborn Task3�/� mice:sign of a developmental deficit?

In mice, fetal and early postnatal adrenal developmentis very dynamic. Adrenal Task3 expression is very strongduring fetal development whereas Task1 expression ishardly detectable (41). In wild-type mice aldosterone lev-els are age dependent with a peak around birth and de-crease in the first weeks of life (42). Therefore, the adrenalphenotype of neonatal Task3�/� mice could reflect retar-dation or broadening of the normal embryonic peak ofaldosterone secretion up to the early postnatal period orhypersecretion of aldosterone with preserved physiologicage dependence. Analysis of the aldosterone content ofembryonic adrenal glands revealed that wild-type micehave a peak of aldosterone content at embryonic day E19followed by a decrease at postnatal day P1. In Task3�/�

mice, the aldosterone peak at day E19 was preserved butmuch higher and the decrease at P1 was attenuated. Ap-parently, the hyperaldosteronism of embryonic and neo-natal Task3�/� mice is caused by pathologic hyperstimu-lation of hormone secretion rather than by time shift of thephysiologic adrenal development.

Table 2. Quantitative Real-Time PCR of Adrenal cDNA from 1-Day (1d) and 12-Day (12d)-Old, and Adult Task3�/�

and Task3�/� Mice

Symbol Description 1 d 12 d Adult Reference Citation

Tac1 Tachykinin 1 5.5a 5.3a 1.7 28–30Gadd45b Growth arrest and DNA-damage-inducible 45 � 4.2a 0.5 1.1 1,21,22Nr4a2 Nuclear receptor subfamily 4, group A, member 2 3.4a 0.9 1.5 19,20Gal Galanin 2.9a 9.0a 1.0 23–27Fst Follistatin 2.8a 2.7 0.8 53–55Dkk3 Dickkopf homolog 3 (Xenopus laevis) 2.5a 2.7a 1.9a 39,56Acvr1c Activin A receptor, type IC 2.0a 1.8a 1.0 31,32Nnt Nicotinamide nucleotide transhydrogenase 0.4a 0.5a 0.4 33,34Htr1b 5-Hydroxytryptamine (serotonin) receptor 1B 0.5a 0.3a 0.6 57–59Trpc5 Transient receptor potential cation channel,

subfamily C, member 50.5a 0.6 0.2a 35,36

Sstr2 Somatostatin receptor 2 0.9 3.6a 1.5 37,38

Specific mRNA expression was normalized against ß-actin. Data are shown as relative expression levels of Task3�/� compared with Task3�/� miceof the respective age. The fold change was calculated as the ratio of the mean values (n � 6–19 per group and genotype).a P � .05 comparing genotypes at individual age. Expression levels of renin, Cyp11b2, Cyp11b1, and Hsd3b6 are shown separately in Figures 3and 4 as bar graphs.

Figure 5. Plasma (A) and renal (B) renin concentration in 1-day and 12-day-old Task3�/� and Task3�/� mice. Aldosterone content in adrenal glands (C)during late fetal development. Plasma renin concentration was not differentbetween the genotypes. In 1-day-old Task3�/� mice, renal reninconcentration was lower than in wild-type mice. Aldosterone content peradrenal gland was higher in Task3�/� mice at day 19 of gestation (E19) and 1day after birth. In panels A and B, number of experiments per group was n �6. In panel C, numbers of experiments per group are shown in brackets. *,P � .05 comparing genotypes at individual age.

doi: 10.1210/en.2013-1101 endo.endojournals.org 2719

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 August 2014. at 00:03 For personal use only. No other uses without permission. . All rights reserved.

Relation between Task3 K� channel function andadrenal renin expression

The protease renin stimulates adrenal aldosterone pro-duction via generation of angiotensin and, thereby, controlssalt- and fluid homeostasis and arterial blood pressure. Jux-taglomerular cells of the kidney are the major source of plas-matic (systemic) renin. In addition, local renin-angiotensinsystems have been found in a variety of tissues including theadrenal gland (43). These local renin-angiotensin systemsinfluence tissue functions and might contribute to the sys-temic renin-angiotensin system (43). In adult rat adrenalglands, renin is mainly expressed in glomerulosa cells, whereits secretion can be stimulated by Ang II or high K� concen-trations (44). Moreover, the renin-angiotensin system isstimulated after nephrectomy and probably contributes tothe increase in plasma aldosterone under these conditions(45, 46). Renin was also detected in human adrenal glandsand in the adrenocarcinoma cell line NCI-H295R (47, 48).In mice, adrenal renin is expressed during fetal developmentwith peak values around embryonic day E16 and largelydisappears after birth (49, 50). However, mice lacking the

gene for aldosterone synthase show astrong induction of adrenal renin ex-pressioneven in theadult stage (51). Inour Affymetrix chip analysis of 1-day-old (P1)Task3�/� mice, the reningeneexpression showed the strongest up-regulation (9-fold) compared withwild-type animals. Systemic levels ofrenin in P1 mice, however, were notchanged,andrenal expressionof reninwasevensuppressed,probably reflect-ingcompensation.Thereafter, adrenalrenin expression dropped in 12-day-old (P12) and almost disappeared inadult knockout mice. Surprisingly,immunofluorescence of P1 adrenalglands revealed that renin was mainlyexpressed in fasciculata cells and notin glomerulosa cells, where Task3showed its strongest expression. Per-haps, soluble factors secreted bystronglydepolarizedglomerulosacellsof Task3�/� mice transmitted abnor-mal paracrine signals to fasciculatacells leading to renin expression. Forexample, higher expression of the se-creted protein follistatin in Task3�/�

mice could be part of a modifiedsignaling between glomerulosa andfasciculata cells. In addition, loss ofTask3 function in fasciculata cell (al-

though the expression is not very high) could have triggeredpathologic renin expression. The fasciculata-derived renin,in turn, probably led to the paracrine stimulation of aldo-sterone secretion in glomerulosa cells via local angiotensinformation (Figure 6).

Other factors possibly modulating theage-dependent adrenal phenotype of Task3�/� mice

Chip-based gene expression analysis disclosed changesin several genes that have already been implicated in thecontrol of aldosterone secretion as well as many novelfactors that have not yet been linked to the regulation ofadrenal hormone production. Further work is needed totest which of those factors are involved in promoting thepathologic phenotype as well as in the compensation ob-served in adult animals.

Conclusion

The hypersecretion of aldosterone, corticosterone, andprogesterone in neonatal Task3�/� mice underscores the

Figure 6. Proposed mechanisms underlying the disturbed hormone secretion in neonatalTask3�/� mice. A, In Task3�/� mice, adrenal glomerulosa cells (Zg) are hyperpolarized underresting conditions due to a very high K� conductance. Plasma renin stimulates the formation ofAng II leading to the inhibition of Task1 and Task3 K� channels, and therefore to membranedepolarization. Voltage-activated Ca2� channels promote an increase of cytosolic Ca2�, whichinduces a signaling cascade, and finally, via the transcription factor Nr4a2, stimulates theexpression of aldosterone synthase. The expression of the Hsd3b6 encoded enzyme enablessynthesis of progesterone, which is converted to the substrate for aldosterone synthase reaction(11-deoxycorticosterone, data not shown). The final step for corticosterone synthesis infasciculata cells (Zf) depends on expression of Cyp11b1. Task1 and, to a lesser extent, Task3channels are also expressed in fasciculata cells. B, The hyperactivation of the Ca2�-dependentsignaling cascade in glomerulosa cells of Task3�/� mice, which leads to hyperaldosteronism, ispresumably driven by multiple factors: 1) depolarization (depol.) by a reduced K� conductancedue to Task3 deletion, 2) increased Ang II formation by the action of local adrenal renin, and 3)higher substrate supply by up-regulation of Hsd3b6. Higher progesterone levels could alsoamplify corticosterone synthesis in fasciculata cells. The mechanisms inducing the reninexpression in zona fasciculata are still unclear and could involve: 1) paracrine action of solublefactors from glomerulosa cells or 2) an altered differentiation from the glomerulosa to fasciculatacell type. Even though Task3 is weakly expressed in fasciculata cells, the deletion could alsodirectly affect the function of these cells, eg, by alteration of the membrane potential.

2720 Bandulik et al Hyperaldosteronism in Task3�/� Mice Endocrinology, August 2013, 154(8):2712–2722

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 August 2014. at 00:03 For personal use only. No other uses without permission. . All rights reserved.

role of this K� channel as determinant of adrenocorticalhormone secretion. Gene expression profiling revealedtranscriptional changes of a variety of factors that are partof the complex networks regulating aldosterone secretion.Among them, the pathologic high activity of the local ad-renal renin-angiotensin system as well as increasedHsd3b6 mRNA expression possibly contribute to the in-appropriately high aldosterone secretion. Abnormal localrenin expression could also be relevant for the develop-ment of primary hyperaldosteronism in a subset of humanpatients.

Acknowledgments

Address all correspondence and requests for reprints to:Dr. Sascha Bandulik, Medizinische Zellbiologie, Universita-etsstrasse 31, 93053 Regensburg, Germany. E-mail: sascha.bandulik@ur.de.

This work was supported by the Deutsche Forschungsge-meinschaft (FOR1086 to R.W. and S.B.); the European Sectionof Aldosterone Council (ESAC) (to J.B.); the French Agence Na-tionale pour la Recherche (ANR) BeyondTASKs grant (to E.L.and J.B.); LabEx Ion channel Science and Therapeutics grant(ANR-11-LABX-0015–01, to F.L. and J.B.); and the CNRSProject for International Scientific Cooperation (PICS) (to J.B.).

Disclosure Summary: The authors have nothing to disclose.

References

1. Spat A, Hunyady L. Control of aldosterone secretion: a model forconvergence in cellular signaling pathways. Physiol Rev. 2004;84:489–539.

2. Hannemann A, Wallaschofski H. Prevalence of primary aldosteron-ism in patient’s cohorts and in population-based studies–a review ofthe current literature. Horm Metab Res. 2012;44:157–162.

3. Funder JW, Carey RM, Fardella C, et al. Case detection, diagnosis,and treatment of patients with primary aldosteronism: an endocrinesociety clinical practice guideline. J Clin Endocrinol Metab. 2008;93:3266–3281.

4. Choi M, Scholl UI, Yue P, et al. K� channel mutations in adrenalaldosterone-producing adenomas and hereditary hypertension. Sci-ence. 2011;331:768–772.

5. Zennaro MC, Jeunemaitre X. Mutations in KCNJ5 gene cause hy-peraldosteronism. Circ Res. 2011;108:1417–1418.

6. Mulatero P, Tauber P, Zennaro MC, et al. KCNJ5 mutations inEuropean families with nonglucocorticoid remediable familial hy-peraldosteronism. Hypertension. 2012;59:235–240.

7. Funder JW. The genetics of primary aldosteronism: chapter two.Hypertension. 2012;59:537–538.

8. Charmandari E, Sertedaki A, Kino T, et al. A novel point mutationin the KCNJ5 gene causing primary hyperaldosteronism and early-onset autosomal dominant hypertension. J Clin Endocrinol Metab.2012;97:E1532–E1539.

9. Murthy M, Azizan EA, Brown MJ, O’Shaughnessy KM. Charac-terization of a novel somatic KCNJ5 mutation delI157 in an aldo-sterone-producing adenoma. J Hypertens. 2012;30:1827–1833.

10. Bandulik S, Penton D, Barhanin J, Warth R. TASK1 and TASK3

potassium channels: determinants of aldosterone secretion and ad-renocortical zonation. Horm Metab Res. 2010;42:450–457.

11. Heitzmann D, Derand R, Jungbauer S, et al. Invalidation of TASK1potassium channels disrupts adrenal gland zonation and mineralo-corticoid homeostasis. EMBO J. 2008;27:179–187.

12. Davies LA, Hu C, Guagliardo NA, et al. TASK channel deletion inmice causes primary hyperaldosteronism. Proc Natl Acad Sci USA.2008;105:2203–2208.

13. Penton D, Bandulik S, Schweda F, et al. Task3 potassium channelgene invalidation causes low renin and salt-sensitive arterial hyper-tension. Endocrinology. 2012;153:4740–4748.

14. Guagliardo NA, Yao J, Hu C, et al. TASK-3 channel deletion in micerecapitulates low-renin essential hypertension. Hypertension. 2012;59:999–1005.

15. Jung J, Barrett PQ, Eckert GJ, et al. Variations in the potassiumchannel genes KCNK3 and KCNK9 in relation to blood pressureand aldosterone production: an exploratory study. J Clin Endocri-nol Metab. 2012;97:E2160–E2167.

16. Guyon A, Tardy MP, Rovere C, Nahon JL, Barhanin J, Lesage F.Glucose inhibition persists in hypothalamic neurons lacking tan-dem-pore K� channels. J Neurosci. 2009;29:2528–2533.

17. Wotus C, Levay-Young BK, Rogers LM, Gomez-Sanchez CE, Enge-land WC. Development of adrenal zonation in fetal rats defined byexpression of aldosterone synthase and 11�-hydroxylase. Endocri-nology. 1998;139:4397–4403.

18. Irizarry RA, Hobbs B, Collin F, et al. Exploration, normalization,and summaries of high density oligonucleotide array probe leveldata. Biostatistics. 2003;4:249–264.

19. Bassett MH, Suzuki T, Sasano H, White PC, Rainey WE. The or-phan nuclear receptors NURR1 and NGFIB regulate adrenal aldo-sterone production. Mol Endocrinol. 2004;18:279–290.

20. Nogueira EF, Rainey WE. Regulation of aldosterone synthase byactivator transcription factor/cAMP response element-binding pro-tein family members. Endocrinology. 2010;151:1060–1070.

21. Romero DG, Plonczynski M, Vergara GR, Gomez-Sanchez EP, Go-mez-Sanchez CE. Angiotensin II early regulated genes in H295Rhuman adrenocortical cells. Physiol Genomics. 2004;19:106–116.

22. Kodama S, Negishi M. Pregnane X receptor PXR activates theGADD45� gene, eliciting the p38 MAPK signal and cell migration.J Biol Chem. 2011;286:3570–3578.

23. Andreis PG, Malendowicz LK, Rebuffat P, Spinazzi R, ZiolkowskaA, Nussdorfer GG. Galanin enhances corticosterone secretion fromdispersed rat adrenocortical cells through the activation of GAL-R1and GAL-R2 receptors coupled to the adenylate cyclase-dependentsignaling cascade. Int J Mol Med. 2007;19:149–155.

24. Belloni AS, Malendowicz LK, Rucinski M, Guidolin D, NussdorferGG. Galanin stimulates cortisol secretion from human adrenocor-tical cells through the activation of galanin receptor subtype 1 cou-pled to the adenylate cyclase-dependent signaling cascade. Int J MolMed. 2007;20:859–864.

25. Bauer FE, Hacker GW, Terenghi G, Adrian TE, Polak JM, BloomSR. Localization and molecular forms of galanin in human adrenals:elevated levels in pheochromocytomas. J Clin Endocrinol Metab.1986;63:1372–1378.

26. Malendowicz LK, Nussdorfer GG, Nowak KW, Mazzocchi G. Thepossible involvement of galanin in the modulation of the function ofrat pituitary-adrenocortical axis under basal and stressful condi-tions. Endocr Res. 1994;20:307–317.

27. Mazzocchi G, Malendowicz LK, Rebuffat P, Nussdorfer GG. Effectsof galanin on the secretory activity of the rat adrenal cortex: in vivoand in vitro studies. Res Exp Med (Berl). 1992;192:373–381.

28. Whitworth EJ, Kosti O, Renshaw D, Hinson JP. Adrenal neuropep-tides: regulation and interaction with ACTH and other adrenal reg-ulators. Microsc Res Tech. 2003;61:259–267.

29. Neri G, Andreis PG, Nussdorfer GG. Effects of neuropeptide-Y andsubstance-P on the secretory activity of dispersed zona-glomerulosacells of rat adrenal gland. Neuropeptides. 1990;17:121–125.

doi: 10.1210/en.2013-1101 endo.endojournals.org 2721

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 August 2014. at 00:03 For personal use only. No other uses without permission. . All rights reserved.

30. Nussdorfer GG, Malendowicz LK. Role of tachykinins in the reg-ulation of the hypothalamo-pituitary-adrenal axis. Peptides. 1998;19:949–968.

31. Beuschlein F, Looyenga BD, Bleasdale SE, et al. Activin inducesx-zone apoptosis that inhibits luteinizing hormone-dependent ad-renocortical tumor formation in inhibin-deficient mice. Mol CellBiol. 2003;23:3951–3964.

32. Vanttinen T, Liu J, Kuulasmaa T, Kivinen P, Voutilainen R. Ex-pression of activin/inhibin signaling components in the human ad-renal gland and the effects of activins and inhibins on adrenocorticalsteroidogenesis and apoptosis. J Endocrinol. 2003;178:479–489.

33. Spat A, Fulop L, Szanda G. The role of mitochondrial Ca(2�) andNAD(P)H in the control of aldosterone secretion. Cell Calcium.2012;52:64–72.

34. Meimaridou E, Kowalczyk J, Guasti L, et al. Mutations in NNTencoding nicotinamide nucleotide transhydrogenase cause familialglucocorticoid deficiency. Nat Genet. 2012;44:740–742.

35. Liao Y, Plummer NW, George MD, Abramowitz J, Zhu MX, Birn-baumer L. A role for Orai in TRPC-mediated Ca2� entry suggeststhat a TRPC:Orai complex may mediate store and receptor operatedCa2� entry. Proc Natl Acad Sci USA. 2009;106:3202–3206.

36. Hong C, Kim J, Jeon JP, et al. Gs cascade regulates canonical tran-sient receptor potential 5 (TRPC5) through cAMP mediated intra-cellular Ca2� release and ion channel trafficking. Biochem BiophysRes Commun. 2012;421:105–111.

37. O’Carroll AM. Localization of messenger ribonucleic acids for so-matostatin receptor subtypes (sstr1–5) in the rat adrenal gland.J Histochem Cytochem. 2003;51:55–60.

38. Ziegler CG, Brown JW, Schally AV, et al. Expression of neuropep-tide hormone receptors in human adrenal tumors and cell lines:antiproliferative effects of peptide analogues. Proc Natl Acad SciUSA. 2009;106:15879–15884.

39. El Wakil A, Bandulik S, Guy N, et al. Dkk3 is a component of thegenetic circuitry regulating aldosterone biosynthesis in the adrenalcortex. Hum Mol Genet. 2012;21:4922–4929.

40. Doi M, Takahashi Y, Komatsu R, et al. Salt-sensitive hypertensionin circadian clock-deficient Cry-null mice involves dysregulated ad-renal Hsd3b6. Nat Med. 2010;16:67–74.

41. Aller MI, Wisden W. Changes in expression of some two-pore do-main potassium channel genes (KCNK) in selected brain regions ofdeveloping mice. Neuroscience. 2008;151:1154–1172.

42. Dalle M, Giry J, Gay M, Delost P. Perinatal changes in plasma andadrenal corticosterone and aldosterone concentrations in the mouse.J Endocrinol. 1978;76:303–309.

43. Paul M, Poyan Mehr A, Kreutz R. Physiology of local renin-angio-tensin systems. Physiol Rev. 2006;86:747–803.

44. Doi Y, Atarashi K, Franco-Saenz R, Mulrow PJ. Effect of changes in

sodium or potassium balance, and nephrectomy, on adrenal reninand aldosterone concentrations. Hypertension. 1984;6:I124–I129.

45. Peters J, Obermuller N, Woyth A, et al.Losartan and angiotensin IIinhibit aldosterone production in anephric rats via different actionson the intraadrenal renin-angiotensin system. Endocrinology. 1999;140:675–682.

46. Inagami T, Mizuno K, Naruse M, et al. Active and inactive renin inthe adrenal. Am J Hypertens. 1989;2:311–319.

47. Naruse M, Sussman CR, Naruse K, Jackson RV, Inagami T. Reninexists in human adrenal tissue. J Clin Endocrinol Metab. 1983;57:482–487.

48. Hilbers U, Peters J, Bornstein SR, et al. Local renin-angiotensinsystem is involved in K�-induced aldosterone secretion from humanadrenocortical NCI-H295 cells. Hypertension. 1999;33:1025–1030.

49. Kon Y, Hashimoto Y, Kitagawa H, Sugimura M, Murakami K.Renin immunohistochemistry in the adrenal gland of the mousefetus and neonate. Anat Rec. 1990;227:124–131.

50. Jones CA, Sigmund CD, McGowan RA, Kane-Haas CM, Gross KW.Expression of murine renin genes during fetal development. MolEndocrinol. 1990;4:375–383.

51. Lee G, Makhanova N, Caron K, et al. Homeostatic responses in theadrenal cortex to the absence of aldosterone in mice. Endocrinology.2005;146:2650–2656.

52. Romero DG, Yanes LL, de Rodriguez AF, et al. Disabled-2 is ex-pressed in adrenal zona glomerulosa and is involved in aldosteronesecretion. Endocrinology. 2007;148:2644–2652.

53. Suzuki J, Otsuka F, Inagaki K, Takeda M, Ogura T, Makino H.Novel action of activin and bone morphogenetic protein in regulat-ing aldosterone production by human adrenocortical cells. Endo-crinology. 2004;145:639–649.

54. Kogawa K, Ogawa K, Hayashi Y, Nakamura T, Titani K, Sugino H.Immunohistochemical localization of follistatin in rat tissues. En-docrinol Jpn. 1991;38:383–391.

55. Sahut-Barnola I, de Joussineau C, Val P, et al. Cushing’s syndromeand fetal features resurgence in adrenal cortex-specific Prkar1aknockout mice. PLoS Genet. 2010;6:e1000980.

56. Chen M, Hornsby PJ. Adenovirus-delivered DKK3/WNT4 and ste-roidogenesis in primary cultures of adrenocortical cells. HormMetab Res. 2006;38:549–555.

57. Lenglet S, Louiset E, Delarue C, Vaudry H, Contesse V. Activationof 5-HT(7) receptor in rat glomerulosa cells is associated with anincrease in adenylyl cyclase activity and calcium influx through T-type calcium channels. Endocrinology. 2002;143:1748–1760.

58. Lefebvre H, Compagnon P, Contesse V, et al. Production and me-tabolism of serotonin (5-HT) by the human adrenal cortex: para-crine stimulation of aldosterone secretion by 5-HT. J Clin Endocri-nol Metab. 2001;86:5001–5007.

59. Watts SW, Morrison SF, Davis RP, Barman SM. Serotonin andblood pressure regulation. Pharmacol Rev. 2012;64:359–388.

2722 Bandulik et al Hyperaldosteronism in Task3�/� Mice Endocrinology, August 2013, 154(8):2712–2722

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 August 2014. at 00:03 For personal use only. No other uses without permission. . All rights reserved.