Pendrin localizes to the adrenal medulla and modulates catecholamine release

Pendrin localizes to the adrenal medulla and modulates catecholamine release

Yoskaly Lazo-Fernandez,1 Greti Aguilera,2 Truyen D. Pham,1 Annie Y. Park,1 William H. Beierwaltes,3

Roy L. Sutliff,1,4 Jill W. Verlander,5 Karel Pacak,6 Adeboye O. Osunkoya,7 Carla L. Ellis,7

Young Hee Kim,1 Gregory L. Shipley,8 Brandi M. Wynne,1 Robert S. Hoover,1,4,11 Shurjo K. Sen,9

Paul M. Plotsky,10 and Susan M. Wall1,11

1Department of Medicine, Emory University School of Medicine, Atlanta, Georgia; 2Section on Endocrine Physiology,Developmental Endocrinology Branch, Eunice Kennedy Shriver National Institute of Child Health and Human Development,National Institutes of Health, Bethesda, Maryland; 3Hypertension and Vascular Research Division, Henry Ford Hospital andWayne State School of Medicine, Detroit, Michigan; 4Atlanta Veterans Affairs Hospital, Atlanta, Georgia; 5Department ofMedicine, University of Florida, Gainesville, Florida; 6Program in Reproductive and Adult Endocrinology, Eunice KennedyShriver National Institute of Child Health and Human Development, National Institutes of Health, Bethesda, Maryland;7Department of Pathology, Emory University School of Medicine, Atlanta, Georgia; 8Department of Integrative Biology andPharmacology, University of Texas Medical School at Houston, Houston, Texas; 9Cardiovascular Disease Section, andNational Human Genome Research Institute, National Institutes of Health, Bethesda, Maryland; 10Department of Psychiatry,Emory University School of Medicine, Atlanta, Georgia; and 11Department of Physiology, Emory University School ofMedicine, Atlanta, Georgia

Submitted 26 January 2015; accepted in final form 9 July 2015

Lazo-Fernandez Y, Aguilera G, Pham TD, Park AY, Beier-waltes WH, Sutliff RL, Verlander JW, Pacak K, Osunkoya AO,Ellis CL, Kim YH, Shipley GL, Wynne BM, Hoover RS, SenSK, Plotsky PM, Wall SM. Pendrin localizes to the adrenalmedulla and modulates catecholamine release. Am J Physiol En-docrinol Metab 309: E534 –E545, 2015. First published July 14,2015; doi:10.1152/ajpendo.00035.2015.—Pendrin (Slc26a4) is aCl�/HCO3

� exchanger expressed in renal intercalated cells and me-diates renal Cl� absorption. With pendrin gene ablation, blood pres-sure and vascular volume fall, which increases plasma renin concen-tration. However, serum aldosterone does not significantly increase inpendrin-null mice, suggesting that pendrin regulates adrenal zonaglomerulosa aldosterone production. Therefore, we examined pendrinexpression in the adrenal gland using PCR, immunoblots, and immu-nohistochemistry. Pendrin protein was detected in adrenal lysatesfrom wild-type but not pendrin-null mice. However, immunohisto-chemistry and qPCR of microdissected adrenal zones showed thatpendrin was expressed in the adrenal medulla, rather than in cortex.Within the adrenal medulla, pendrin localizes to both epinephrine- andnorepinephrine-producing chromaffin cells. Therefore, we examinedplasma catecholamine concentration and blood pressure in wild-typeand pendrin-null mice under basal conditions and then after 5 and 20min of immobilization stress. Under basal conditions, blood pressurewas lower in the mutant than in the wild-type mice, although epi-nephrine and norepinephrine concentrations were similar. Catechol-amine concentration and blood pressure increased markedly in bothgroups with stress. With 20 min of immobilization stress, epinephrineand norepinephrine concentrations increased more in pendrin-nullthan in wild-type mice, although stress produced a similar increase inblood pressure in both groups. We conclude that pendrin is expressedin the adrenal medulla, where it blunts stress-induced catecholaminerelease.

pendrin; chloride; epinephrine; norepinephrine; dopamine

PENDRIN, encoded by Slc26a4, is an aldosterone-sensitive, elec-troneutral, Na�-independent Cl�/HCO3

� exchanger, ex-pressed in the apical regions of type B and Non-A, non-B

intercalated cells (2, 15, 29, 32, 41, 42), which are minority celltypes found within the renal cortex. Aldosterone increasespendrin-mediated Cl� absorption and HCO3

� secretion, whichcontributes to the pressor response that follows steroid hor-mone administration (39). During dietary NaCl restriction,pendrin-null (Slc26a�/�, Pds�/�) mice excrete more Na� andCl� and have lower blood pressure than wild-type mice (16,40). As such, pendrin gene ablation reduces blood pressure (16,23) due to the fall in circulating vascular volume, whichfollows the limited ability of these mutant mice to fullyconserve urinary Na� and Cl�.

Plasma renin concentration increases with pendrin geneablation (16, 40), as expected when circulating vascular vol-ume and blood pressure fall. However, pendrin gene ablationdoes not increase circulating aldosterone in proportion to theincrease in circulating plasma renin concentration (16, 40, 42).The fall in the ratio of aldosterone to renin observed withpendrin gene ablation suggests that the adrenal zona glomeru-losa of these mutant mice has a lower sensitivity to angiotensinII and thereby synthesizes less aldosterone in response to thispeptide hormone. These data raise the possibility that pendrinis expressed in the adrenal cortex and has a role in modulatingadrenal glomerulosa cell responsiveness. The aim of this studywas to determine whether pendrin is expressed in the adrenalgland and whether it modulates adrenal function. The resultsshow that pendrin is expressed in the adrenal gland but unex-pectedly localizes to the adrenal medulla rather than the adre-nal cortex, where it plays a role in restraining catecholamineproduction during stress.

METHODS

Animal Treatment

Animal and animal procedures. Wild-type and pendrin-null mice(10) on a 129 S6SvEvTac background were bred in the EmoryUniversity vivarium. In all experiments, age- and sex-matched micewere compared. Unless otherwise stated, mice received a balancedNaCl-replete diet given as a gel (0.8 meq/day NaCl) (39) for 3 daysprior to and during the experimental period. Male Sprague-Dawleyrats, weighing 100–200 g, were purchased from Harlan.

Address for reprint requests and other correspondence: S. M. Wall, RenalDivision, W.M.B. Rm. 338, 1639 Pierce Dr., NE, Atlanta, GA 30322 (e-mail:[email protected]).

Am J Physiol Endocrinol Metab 309: E534–E545, 2015.First published July 14, 2015; doi:10.1152/ajpendo.00035.2015.

http://www.ajpendo.orgE534

Blood pressure measurements. Blood pressure was measured inconscious mice by telemetry before, during, and after immobilizationstress, using methods we have reported previously (16, 23). Meanarterial pressure (MAP) was calculated from the systolic and diastolicpressures, using the relationship (systolic � 2� diastolic)/3.

Treatment Models

Treatment 1: effect of NaCl intake on mouse renal and adrenalpendrin protein abundance. For 7 days prior to their being euthanized,mice were ration-fed a gelled diet (27.0% sodium-deficient Harlan TD90228 rodent chow, 72.1% water, 0.75% agar), which gave 0.03meq/day NaCl. Mice received the gel diet alone or supplemented withNaCl to give 1.4 meq/day NaCl.

Treatment 2: effect of KCl intake on mouse renal and adrenalpendrin protein abundance. For 7 days prior to their being euthanized,mice were ration-fed a gelled diet (27.0% Harlan TD 90228 rodentchow, 72.1% water, 0.75% agar, 1.2% Equal sweetener) supple-mented with NaCl and KCl to give (in meq/day) 0.33 Na�, 0.37 Cl�,and 0.67 K� (control) or 0.33 Na�, 4.07 Cl�, and 4.36 K� (highKCl).

Treatment 3: effect of KHCO3 intake on mouse renal and adrenalpendrin protein abundance. For 7 days prior to their being euthanized,mice were ration-fed a gelled diet (27.0% Harlan TD 90228 rodentchow, 72.1% water, 0.75% agar, 1.2% Equal sweetener) supple-mented with NaCl and KHCO3 to give (in meq/day) 0.33 Na�, 0.37Cl�, and 0.67 K� (control) or 0.33 Na�, 0.37 Cl�, and 4.36 K� (highKHCO3).

Treatment 4: pendrin immunostaining in rat adrenal gland follow-ing a NaCl-deficient diet. Male Sprague-Dawley rats (100–200 g)were fed the NaCl-deficient diet as pellets (Harlan TD 90228 rodentchow) and drank water ad libitum for 1 wk and were then euthanizedand perfusion- fixed in situ.

Treatment 5: effect of pendrin gene ablation on adrenal tyrosinehydroxylase abundance. For 7 days prior to their being euthanized,wild-type and pendrin-null mice were ration-fed a gelled diet (27.0%Zeigler Brother’s rodent chow #53881300, 72.1% water, 0.75% agar),supplemented with NaCl to give mice 0.8 meq/day NaCl.

Placement of Chronic Indwelling Catheters in the Jugular Vein andBlood Collection Before and After Immobilization Stress

Mice were anesthetized with 1–2.5% isoflurane/100% O2 andplaced on a heating pad in the supine position. The hair over the neckwas shaved and then swabbed with betadine (povidone iodine) alter-nating with 70% alcohol and then draped with sterile towels. Under astereomicroscope, a 6- to 8-mm incision was made to expose �5 mmof the right external jugular vein, which was freed from the surround-ing tissue by blunt dissection. Two silk 3-0 sutures were placedunderneath the vein. The cephalic suture was tied around the vein tointerrupt blood flow, whereas the caudal suture remained untied. Thebeveled tip of a 2-Fr silicone catheter (Dow corning CN 508-001)connected to PE-10 tubing was inserted into the jugular vein throughan incision made between the two ligatures and then advanced 0.7 to1 cm until the tip reached the right atrium. Catheter placement wasconfirmed by blood return through the catheter. If blood withdrawalwas poor, the catheter was repositioned. Blood was returned in warm,sterile, heparinized saline (30 U/ml heparin). The caudal suture wastied around the cannulated vein to secure the catheter. The mouse wasthen repositioned on its right side, and a sharpened stainless steelguide cannula (18-G, 1.2 � 40 mm) was tunneled subcutaneously tothe nape of the neck and pushed through the skin behind the ears. Thecatheter was then rapidly detached from the syringe, fitted through theguide cannula, exteriorized at the other end, and secured with anadditional suture. The mouse was returned to the supine position.Cefazolin (20 mg/kg body wt sc) was given 2 h before and immedi-ately after (30 U/ml) catheter insertion and then locked. The syringewas detached, and the exteriorized tubing was cut �1.5 cm from the

skin and bent at a 180° angle. Silicone rubber tubing was slipped overthe crimp to prevent air entry. The incision was closed with tissueadhesive (Vetbond, no. 1469; 3M Animal Care Products) andswabbed with betadine (povidone iodine). Over the next 2–3 days, themouse was allowed to recover. Catheters were flushed daily withheparinized saline and then locked with heparinized saline (30 U/ml)containing cefazolin (4 mg/ml). To perform blood collections, thecatheter was connected to 25 cm of polyurethane tubing (PU-033, SAIUSA) that was passed through the lid of the cage and connected to a1-ml syringe filled with sterile, heparinized saline. For the hour priorto blood sampling, the mouse was undisturbed. A 130- to 150-�l basalblood sample was taken while disturbance of the mouse was mini-mized. To generate an acute stress response, the mouse was removedfrom the cage and placed on a board in the prone position andimmobilized with paper adhesive tape (immobilization stress). Bloodsamples (130–150 �l) were then taken after 5 and 20 min ofimmobilization stress and the mouse was then euthanized. After eachblood collection, the mouse was slowly infused with 130–150 �l ofnormal saline warmed to 37°C. Blood was collected in EDTA-coatedtubes (Microtainer CN #365973) containing 0.1% sodium metabisul-fite and centrifuged at 2,000 g for 15 min. The resulting plasma wasstored at �80°C until assayed.

Immunoblot Analysis

Immunoblots were performed of kidney and adrenal lysates, similarto methods reported previously (16, 19). One kidney and both adrenalglands were harvested from each mouse and placed in an ice-cooledbuffer (0.3 M sucrose, 25 mM imidazole, pH 7.2, containing 1�Roche Complete Protease Inhibitor Cocktail). Tissue was immediatelyhomogenized using an Omni THQ Tissue Homogenizer (Omni Inter-national) and then centrifuged at 1,000 g for 15 min at 4°C. Totalprotein was measured in the supernatant using the DC Protein AssayKit (Bio-Rad, Hercules, CA) and then dissolved in Laemmli buffer.Aliquots containing equal amounts of protein were separated bySDS-PAGE on 8.5% acrylamide gels and then electroblotted to PVDFmembranes (Immobilon; Millipore, Bedford, MA). Blots wereblocked with Odyssey Blocking Buffer (LI-COR Biosciences) follow-ing the manufacturer’s instructions and then incubated with primaryantibody overnight at 4°C followed by incubation for 2 h at roomtemperature with Alexa fluor 680-linked anti-rabbit IgG (MolecularProbes). Pendrin protein was detected using a rabbit anti-rat pendrinantibody described previously (20). Tyrosine hydroxylase was de-tected using a rabbit anti-rat tyrosine hydroxylase antibody (Abcam,Cambridge, MA #ab112). In some experiments, equal protein loadingand transfer were confirmed by labeling the same blot for �-tubulin(SIGMA T3559). To correct for differences in protein loading be-tween lanes, membranes were Coomassie stained as reported previ-ously (44). Signals were visualized with an Odyssey Infrared ImagingSystem (LI-COR Biosciences). Immunoblot and Coomassie banddensities were quantified using the software program Image J (NIH,available at http://rsb.info.nih.gov/). Pendrin band density was nor-malized to the density of band on the Coomassie membrane with thesame mobility.

Immunohistochemistry

We used an anti-pendrin antibody raised in rabbits that recognizesthe COOH-terminal 29 amino acids of the rat pendrin sequence (20).For double-labeling experiments, we used a rabbit anti-tyrosine hy-droxylase (Abcam, ab112) diluted 1:20,000 and a rabbit anti-�-phenylethanolamine N-methyltransferase (�-PNMT) antibody, di-luted 1:10,000 to 1:20,000, that was raised in rabbits and recognizesamino acids 44–58 of the mouse �-PNMT sequence (26) (a generousgift of Dr. James Powers).

Kidneys and adrenal glands from mice and rats were fixed in situand embedded in paraffin or polyester wax [polyethylene glycol 400distearate (Polysciences, Warrington, PA) and 10% 1-hexadecanol] as

E535PENDRIN LOCALIZES TO THE ADRENAL MEDULLA

AJP-Endocrinol Metab • doi:10.1152/ajpendo.00035.2015 • www.ajpendo.org

described previously (24). Pendrin immunoreactivity was detectedusing immunoperoxidase procedures. Blocking was done with 3%H2O2 in methanol for 30 min, followed by protein blocking using 1%bovine serum albumin, 0.2% gelatin, 0.05% saponin solution. Theanti-pendrin antibody was diluted in PBS (1:1,000). Sections wererinsed with PBS supplemented with 0.1% BSA, 0.05% saponin, and0.2% gelatin, and labeling was visualized with horseradish peroxi-dase-conjugated goat anti-rabbit secondary antibody (1:200, DAKO).Sections that labeled pendrin were visualized by diaminobenzidine(DAB) staining. Sections were washed with distilled water and coun-terstained with hematoxylin. In experiments using polyester wax-embedded tissue, the anti-pendrin primary antibody was diluted up to1:50,000. Endogenous peroxidase activity was blocked by incubatingthe sections in 3% H2O2 in distilled water for 45 min. The sectionswere blocked for 15 min with Serum-Free Protein Block (DakoCytomation) and then incubated at 4°C overnight with primary anti-body. The sections were washed in PBS and incubated for 30 min withpolymer-linked, peroxidase-conjugated goat anti-rabbit IgG (VectorImmPRESS; Vector Laboratories, Burlingame, CA), again washedwith PBS and then exposed to a DAB substrate kit (Vector) for 5 min.The sections were washed in distilled water, dehydrated with gradedethanols and xylene, mounted, and observed by light microscopy.

Double immunolabeling was done using sequential immunoperox-idase procedures in polyester wax-embedded tissue, as describedpreviously (18). Tissue sections were labeled with the anti-pendrinantibody. After the DAB reaction, sections were washed in PBS andthen blocked using 3% H2O2 in methanol. A second immunolabelingprocedure was done on the same sections using the �-PNMT or thetyrosine hydroxylase antibodies as the primary antibody and VectorSG (Vector Laboratories, Burlingame, CA) for the peroxidase sub-strate, which produces a blue reaction product easily distinguishedfrom the DAB brown reaction product. Sections were then washedwith glass-distilled water, dehydrated with graded ethanols and xy-lene, mounted, and observed by light microscopy.

Identification of Pendrin mRNA in Human Tissues

Publicly accessible data from comprehensive transcriptomic pro-filing of human liver, kidney, and adrenal obtained using RNA-seqwas downloaded from the Illumina Human Body Map 2.0 project(http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?accGSE30611).Fastq files from the sequencing output were aligned to the hg18version of the human genome, and the resulting BAM files wereconverted to bigWig format and loaded into a custom session on theUCSC Genome Browser (www.genome.ucsc.edu).

Rat Adrenal Dissection and mRNA Preparation

Adrenal glands were rapidly removed and cleaned of surroundingfat. The capsule containing the zona glomerulosa was separated from

the rest of the cortex and medulla by making a small incision in thecapsule and squeezing the inner zones from the bottom to the top ona tissue paper humidified in PBS between the index finger and thumb.The capsule was immediately frozen, and the inner zones were placedon an ice-filled Petri dish. After sectioning about 1 mm of cortex attwo opposed sides, inner zones were bisected and flipped 90°, and themedulla was punched out using a 2-mm diameter needle. Tissues wereimmediately frozen in a 1.5-ml microtube on dry ice.

Total RNA was prepared by homogenizing tissue in 1 ml of TRIzolreagent (Invitrogen, Hopkinton, MA), using a hand-held motorizedhomogenizer (Bio-Gen Pro200; Pro Scientific, Oxford, CT), for 5 s,setting 2, followed by purification using an RNeasy minikit andcolumn DNase digestion (Qiagen, Valencia, CA) (39, 41). Aliquots oftotal RNA, �700 ng for capsule and inner cortical zones and 300 ngfor the medulla, were subjected to DNase 1 treatment in solution toeliminate remaining traces of genomic DNA contamination. RNase-free DNase I (10 U/�l; Roche Applied Science, Indianapolis, IN) wasdiluted 10-fold in 1� Jumpstart PCR buffer. One hundred microlitersof total RNA in nuclease-free H2O at 50–100 ng/�l was treated with1 U DNase I and 1 mM MgCl2 at 37°C for 30 min, exactly followedby 75°C for precisely 10 min, cooled on ice, and stored at �80°C.

Quantitative PCR in Rat Adrenal Gland

Quantitative (q)PCR assays were performed as described previ-ously (39, 41, 42). Target-specific qPCR quantitative assays weredeveloped using Beacon Designer or AlleleID software (PremierBiosoft, Palo Alto, CA) based on the latest refseq sequence informa-tion available at the time from the NCBI. Real-time qPCR assayinformation is provided in Table 1. cDNA was synthesized in a 5-�lvolume within a 384-well plate by the addition of 3 �l/well RT mastermix consisting of 400 nM assay-specific reverse primer (from theqPCR assay), 500 �M deoxynucleotides, 1� Affinityscript buffer,and 1 U/�l Affinityscript reverse transcriptase (Agilent Technologies,Santa Clara, CA) for �RTase reactions or without for the �Rtasereactions, to a 384-well plate (Applied Biosystems) and followed bya 2-�l volume of sample (50 ng/�l). Each sample was assayed intriplicate wells plus a control well without reverse transcriptase(NAC) in the master mix to access DNA contamination level. Eachplate also contained an assay-specific sDNA standard (synthetic PCRamplicon oligo) spanning a 5-log template concentration range (200to 2E6 copies) in duplicate and two no-template PCR control wells(NTC). The oligo standards were added into RT master mix withreverse transcriptase. Each plate was covered with Biofilm A (Bio-Rad) and incubated in a DYAD (384-well block) thermocycler (Bio-Rad) for 30 min at 50°C followed by 72°C for 10 min.

The complete PCR master mix (15 �l/well) was added directly tothe 5-�l cDNA volume in the 384-well qPCR plate. Final concentra-tions for the PCR were 400 nM forward and reverse primers (IDT,

Table 1. Primers employed

Gene Name Slc26A4 TH RNA18S5 Actb

Protein Pendrin Tyrosine hydroxylase None �-ActinSpecies rat Rat Human RatAccession no. NM_019214 NM_012740 NR_003286 NM_031144Synonyms Pds The RN18S1 ActxForward Primer 911� CAGTCCCGATTCCTATAG 1281� CAAGATCAAACCTACCAG 1335� CGGCTTAATTTGACTCAACAC 1032� TACTGCCCTGGCTCCTAGCReverse Primer 988� AATTTGCTTCCAAGTTGG 1366� ATACGAGAGGCATAGTTC 1401� ATCAATCTGTCAATCCTGTCC 1113� AGAGCCACCAATCCACACAProbe 940� FAM-ACAATTATCGCCACCG

CCA-BHQ11338� FAM-CCTTGGCGTCATTGAAGCTCTC-BHQ1

1359� FAM-AAACCTCACCCGGCCCG-BHQ1

1060� FAM-ATCAAGATCATTGCTCCTCCTGAGCGC-BHQ1

Amplicon Length,bases

78 86 67 82

PCR Efficiency, % 100 97 99 98LOQ, copies 20 20 2000 20Exon Boundary 7/8 13/14 N/A 5/6Splice Variants All detected None None All detectedm-Fold None None None NoneBLAST Specific Specific Specific Specific*

LOQ, limit of quantification. *Might detect �-actin.

E536 PENDRIN LOCALIZES TO THE ADRENAL MEDULLA


Coralville, IA), 100 nM fluorogenic probe (Biosearch Technologies,Petaluma, CA), 5 mM MgCl2, and 200 mM deoxynucleotides, 1�JumpStart PCR buffer, 150 nM SuperROX dye (Biosearch Technol-ogies) and 0.25 U JumpStart Taq polymerase per reaction (Sigma, St.Louis, MO). RT master mixes, all RNA samples, and DNA oligostandards were pipetted with a Tecan Genesis RSP 100 roboticworkstation (Tecan US, Research Triangle Park, NC); PCR mastermixes were pipetted utilizing a Biomek 2000 robotic workstation(Beckman, Fullerton, CA). Each assembled plate was then coveredwith optically clear film (Applied Biosystems) and run in a 7900real-time qPCR instrument using the following cycling conditions:95°C for 2 min, followed by 40 cycles of 95°C, 12 s, and 60°C for 30s. The resulting data were analyzed using SDS 2.3 (7900) software(Applied Biosystems) with FAM reporter and ROX as the referencedye. Synthetic PAGE-purified DNA oligonucleotides were used asstandards (sDNA) and encompassed at least the entire 5=-3= PCRamplicon for the assay (Sigma-Genosys, The Woodlands, TX). Eacholigo standard was diluted in 100 ng/�l E. coli tRNA-H2O (MolecularBiology Grade, RNase/DNase free; Roche Applied Science) andspanned a 5-log range in 10-fold decrements starting at 0.8 pg/reaction. Copy numbers were calculated based on the amount ofsDNA used and amplicon length. In vitro transcribed RNA ampliconstandards (sRNA) and sDNA standards have the same PCR efficiencyand sensitivity when the reactions are performed as described abovewith PCR amplicons of fewer than 100 bases in length (G. L. Shipley,unpublished observations).

The final data were normalized to the geometric mean of 18S rRNAand �-actin values based on the observed low variability of thesetranscripts among biological groups (geNorm algorithm). Due to thehigh abundance of the 18S ribosomal RNA, a second set of experi-mental samples, diluted 100-fold, was made in 100 ng/�l E. colitRNA-H2O for each sample to bring the Cq values in line withabundant mRNAs. The final data are presented as the molecules ofunknown transcript/geometric mean of molecules of 18S rRNA and�-actin � 100, extrapolated from their respective standard curves(Vandesompele et al, Genome Biology, 2002- http://genomebiology-.com/2002/3/7/research/0034.1). The final data are expressed as apercentage of the geometric mean of the normalizer transcripts (6, 38).Quantities for unknown samples were obtained from an internalassay-specific standard curve run on each 384-well plate, as describedabove. Samples with data points below the limit of quantification(LOQ) for the assay were not considered in the data analysis. Simi-larly, samples with DNA levels (�RTase) within 6 cycles (Cq) of themean of the transcript signals (�RTase) were not considered in thedata analysis.

Measurement of Catecholamines, Renin, Aldosterone, Electrolytes,and Arterial Blood Gases

Plasma levels of epinephrine and norepinephrine were quantified in50 �l of plasma by use of reverse-phase liquid chromatography withelectrochemical detection after partial purification by adsorption ofcatechols onto alumina (11). Plasma corticosterone levels were mea-sured in 10 �l of plasma, using the DetecX Corticosterone Chemilu-minescent Immunoassay Kit (Arbor Assays, Ann Arbor, MI) accord-ing to the manufacturer’s instructions. Plasma renin and serum aldo-sterone concentrations were measured as described previously (40).Aldosterone measurements were performed at the the CardiovascularPharmacology Research Laboratory, University of Iowa College ofPharmacy. Serum electrolytes were measured using an iSTAT System(Abbot Point of Care, Princeton, NJ).

All procedures were approved by Emory University and the Na-tional Institute of Child Health and Human Development (NICHD)Animal Care and Use Committees.

Statistics

Data are presented as means � SE. Each n used in the statisticalanalysis represents data from separate animals. To test for statisticalsignificance between two groups, an unpaired Student’s t-test wasused. The criterion for statistical significance was P 0.05.

RESULTS

Pendrin Gene Ablation Reduces the Aldosterone/Renin Ratio

Following moderate dietary NaCl restriction, we observedthat pendrin gene ablation increased circulating renin con-centration without changing serum aldosterone (16). Furtherexperiments examined plasma aldosterone and renin con-centration in pendrin-null and wild-type mice following ahigh K� diet (see Treatment 2, METHODS), which stimulatesaldosterone release through a mechanism distinct from thatof NaCl restriction. As shown (Table 2), pendrin geneablation reduced the ratio of aldosterone to renin. Sinceplasma renin concentration was either the same or higher,while serum aldosterone concentration was the same orlower in pendrin-null relative to wild-type mice given a highK� diet, the reduced aldosterone/renin ratio observed withpendrin gene ablation is likely due to increased plasma reninconcentration, reduced serum aldosterone concentration, orboth. We conclude that, in response to dietary NaCl restric-tion (16) and K� loading (Table 2), pendrin gene ablationincreases circulating renin concentration more than aldoste-rone concentration.

Pendrin Is Expressed in Human Adrenal Gland

The low aldosterone/renin ratio observed in pendrin-nullmice following high dietary K� intake suggests that pendrin isexpressed in the adrenal glomerulosa and is necessary for amaximal aldosterone response. To explore this possibility, wemined publicly accessible transcriptomic data (Illumina Hu-man Body Map 2.0 project) to determine whether pendrinmRNA is present in human adrenal gland (Fig. 1). As shown,pendrin mRNA is expressed in human kidney but not in liver,as expected on the basis of previous reports (29). Significantly,pendrin transcript was also detected in human adrenal gland atlevels comparable to that found in human kidney.

Pendrin Localizes to Chromaffin Cells Within MouseAdrenal Medulla

To determine whether pendrin protein is expressed in themouse adrenal gland, we examined pendrin protein abundance

Table 2. Effect of pendrin gene ablation on serumelectrolytes and aldosterone and plasma renin concentrationin mice consuming a KCl-rich diet

Wild Type Pendrin Null

Na� (meq/l) 134 � 6.2 (n 4) 138 � 0.3 (n 3)K� (meq/l) 3.7 � 0.5 (n 4) 5.6 � 0.5 (n 3)*Cl� (meq/l) 97 � 3 (n 4) 97 � 0 (n 3)Aldosterone (nM) 49.4 � 5.7 (n 7) 52.6 � 6.9 (n 6)Renin concentration (�g Ang

I/ml/h) 1.02 � 0.15 (n 5) 2.97 � 1.14 (n 6)Aldosterone/renin ratio (nM/�g

Ang I/ml/h) 43.1 � 1.4 (n 5) 24.1 � 4.6 (n 6)*

*P 0.05.



by immunoblot in adrenal lysates taken from wild-type andpendrin-null mice. Kidney lysates from wild-type and pendrin-null mice were employed as positive and negative controls. Asshown (Fig. 2), we observed a band that migrated at 110 kDa,the expected mobility of pendrin (17), in renal and adrenallysates from wild-type but not from pendrin-null mice.

To determine the site of pendrin expression within theadrenal gland, pendrin immunolabeling was examined inmouse adrenal gland sections. As shown (Fig. 3A), pendrinlabeling was not observed in the adrenal cortex as we expectedbut rather in the adrenal medulla. No pendrin labeling wasobserved in the adrenal medulla of pendrin-null mice (Fig. 3B),which confirmed immunostaining specificity. The intensity andsubcellular distribution of pendrin immunoreactivity variedamong cells in the adrenal medulla (Fig. 3, C and D). Manycells contained a single locus of intense immunoreactivity,whereas others had weaker, more diffuse cytoplasmic label.Occasional cells had accentuated immunoreactivity at the pe-riphery of the cells, consistent with plasma membrane distri-bution (Fig. 3D).

Further experiments explored the cell type(s) within theadrenal medulla that express(es) pendrin. To do so, tyrosinehydroxylase, the rate-limiting enzyme in catecholamine bio-synthesis, was used as a chromaffin cell marker (7). Figure 3E

shows that pendrin immunoreactivity was present almost ex-clusively in cells that were positive for tyrosine hydroxylase.Therefore, pendrin is expressed primarily within chromaffincells of the adrenal medulla. However, occasional pendrin-positive, tyrosine hydroxylase-negative cells were observed(Fig. 3E). The identity of these cells is unclear but mayrepresent entrapped cortical cells, sustentacular cells, endothe-lial cells, or neurons. To further explore the cell subtype(s) thatexpress(es) pendrin, we used an enzyme that converts norepi-nephrine to epinephrine (�-PNMT), as a marker of epineph-rine-producing chromaffin cells (30). As shown (Fig. 3F),pendrin label was present in cells that were both positive andnegative for �-PNMT. Therefore, pendrin localizes to bothepinephrine- and norepinephrine-producing chromaffin cellswithin the adrenal medulla.

Pendrin Protein and mRNA Are Expressed in the RatAdrenal Medulla

Additional studies explored whether pendrin expression inthe adrenal medulla is unique to mouse or whether it isobserved in another species. Because dietary NaCl restrictionincreases pendrin abundance in kidney (27, 42), we examinedpendrin labeling in adrenal gland sections taken from NaCl-restricted rats (Fig. 4, Treatment 4). As shown, in rat adrenalgland pendrin labeling was also observed in the medulla, butnot in the cortex.

To confirm pendrin expression in the adrenal medulla, weexamined pendrin mRNA levels in microdissected samplestaken from the rat adrenal medulla, capsule/glomerulosa,and zona fasciculata. Tyrosine hydroxylase mRNA was usedas a marker of adrenal medullary chromaffin tissue. Asshown (Fig. 5), pendrin mRNA was highly expressed in therat adrenal medulla, with much lower levels in other regionsof the rat adrenal gland. Since similar low levels of pendrinand tyrosine hydroxylase mRNA were detected in samplestaken from the zona fasciculata the pendrin mRNA detectedin these samples was more likely due to contamination by

Fig. 1. Pendrin (Slc26a4) mRNA is expressed in humanadrenal gland. UCSC Genome Browser tracks (gray bars atleft) show, from top to bottom, the chromosomal location ofSLC26A4 (pendrin) on chromosome 7 along with scale barand transcript structure. Thick blue blocks denote exons; thinlines connecting these are introns. Expression in kidney,adrenal gland, and liver are shown. The latter 3 expressiontracks are RNA-Seq read data converted to UCSC GenomeBrowser bigWig format and are all scaled to common y-axisheight of 100. Slc26a4 was detected in kidney and adrenalgland but not in liver. RNA-Seq data used to create these 3tracks were derived from the Illumina Human Body Map 2.0project and can be found at http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?accGSE30611.

Fig. 2. Pendrin is expressed in mouse adrenal gland. Pendrin band density wasexamined by immunoblot of mouse adrenal and kidney lysates. As shown, theanti-pendrin antibody detected a band that migrates at �115 kDa, the expectedmobility of pendrin, in adrenal and kidney lysates from wild-type, but notpendrin-null mice. �-Tubulin immunoreactivity was employed as a loadingcontrol. All lysates were run on the same gel. Dashed line indicates a region onthe gel where lanes were excluded.



chromaffin tissue than from pendrin mRNA actually ex-pressed in this region.

Regulation of Pendrin Protein Abundance Differs in Kidneyand Adrenal Gland

Because aldosterone increases renal pendrin abundance (39),further experiments examined changes in pendrin abundance inkidney and adrenal gland in mice following increased intake ofK� (Treatments 2 and 3) or NaCl (Treatment 1), both of whichreduced pendrin protein abundance in kidney (27, 34, 42).Pendrin abundance was semiquantified by immunoblot of kid-ney (Fig. 6) and adrenal gland lysates (Fig. 7) taken from micein each treatment group. As shown, pendrin abundance inkidney fell with increased intake of NaCl or of K�, whethergiven as KCl or KHCO3. However, adrenal gland pendrinabundance was unaffected by changes in dietary intake of NaClor K�. We conclude that the regulation of pendrin proteinabundance differs in kidney and in adrenal gland.

Pendrin Gene Ablation Increases Stress-Induced Epinephrineand Norepinephrine Release

Because pendrin is expressed in epinephrine- and norepineph-rine-producing chromaffin cells, we were prompted to measureserum catecholamine concentration in wild-type and pendrin-nullmice bearing chronic indwelling catheters. Circulating catechol-amine and corticosterone concentrations were quantified underbasal conditions and then after 5 and 20 min of immobilizationstress. As shown (Fig. 8), basal epinephrine and norepinephrinelevels were similar in pendrin-null and wild-type mice. Circulat-ing epinephrine and norepinephrine increased dramatically in boththe wild-type and pendrin-null mice within the first 5 min ofimmobilization stress. However, after 20 min of immobilizationstress epinephrine and norepinephrine levels were higher in pen-drin-null than in wild-type mice. Differences in the production ofdihydroxyphenylglycerol (DHPG), a metabolite of norepineph-rine, did not reach statistical significance. Moreover, there were no

Fig. 3. Pendrin is expressed in chromaffin cells ofthe mouse adrenal medulla. Pendrin immunolabel-ing was observed in the adrenal medulla of wild-type (A), but not of pendrin-null mice (B). C: pen-drin immunolabel in adrenal medulla of wild-typemice. D: inset from C at higher magnification. Openarrows mark areas of peripheral immunoreactivityconsistent with plasma membrane labeling; closedarrows show cells with intense foci of immunola-beling within the cell cytoplasm. E: pendrin (brown)and tyrosine hydroxlase (TH; blue) double labeling.Virtually all pendrin-positive cells are also positivefor TH. Occasional cells are positive for pendrin,but TH negative (arrows). F: pendrin (brown) and�-phenylethanolamine N-methyltransferase (�-PNMT; blue) double labeling. Pendrin immunore-activity is present in both �-PNMT-positive (openarrows) and �-PNMT-negative cells (closed ar-rows). None of these images showed pendrin immu-nolabel in the adrenal cortex.



differences between wild-type and pendrin-null mice in circulat-ing dopamine or the dopamine metabolite 3,4-dihydroxyphenyla-cetic acid (DOPAC) either under basal conditions or followingimmobilization stress.

Because the adrenal gland produces corticosterone, we alsoexplored the effect of pendrin gene ablation on circulatingcorticosterone levels under both basal conditions and then afterimmobilization stress (Fig. 9). As shown, corticosterone con-centrations were similar in pendrin-null and wild-type miceunder basal conditions and increased with stress in bothgroups. However, after 5 min of immobilization stress, plasmacorticosterone concentration was higher in pendrin-null than inwild-type mice. We conclude that pendrin gene ablation in-creases stress-induced release of epinephrine, norepinephrine,and corticosterone, but not of dopamine.

Pendrin Gene Ablation Does Not Change TyrosineHydroxylase Protein Abundance

Because pendrin gene ablation increased the stress-inducedrelease of catecholamines, we asked whether this occurs fromincreased abundance of tyrosine hydroxylase, the rate-limitingenzyme in the synthesis of epinephrine by the adrenal medulla.As shown (Fig. 10), tyrosine hydroxylase total protein abun-

Fig. 4. Pendrin is expressed in rat adrenal medulla. Sections of rat adrenalgland were labeled for pendrin. Top: sections at low magnification of ratadrenal gland labeled for pendrin. Bottom: area within the box at highermagnification. Arrows indicate pendrin immunolabel. As shown, pendrinlabeling was observed in rat adrenal medulla but not cortex.

Fig. 5. Slc26a4 (pendrin) mRNA is ex-pressed in rat adrenal medulla. Slc26a4 tem-plate molecules (mRNA) were quantified insamples taken from rat adrenal capsule, fas-ciculata, and medulla by punch biopsy. THmRNA was quantified in the same samplesas a marker of adrenal medullary chromaffintissue. As shown, pendrin (Slc26a4) mRNAwas much higher in adrenal medulla relativeto fasciculata or glomerulosa. Groups werecompared using one-way ANOVA with aHolm-Sidak posttest.

Fig. 6. Renal pendrin protein abundance decreases when dietary NaCl or K�

intake increases. Pendrin protein abundance was quantified by immunoblot ofkidney lysates taken from mice given a high (1.4 meq/day) or low (0.03meq/day) dietary intake of NaCl (HS or LS, Treatment 1). In other experi-ments, pendrin abundance was quantified in lysates taken from mice thatreceived 0.67 or 4.36 meq/day K� given as KCl or KHCO3 (Treatments 2 and3). As shown, increasing dietary NaCl or K� intake reduced pendrin banddensity. Groups were compared with a 2-tailed, Student’s t-test. *P 0.05.



dance was similar in adrenal glands from unstressed wild-typeand pendrin-null mice. We conclude that changes in tyrosinehydroxylase protein abundance cannot explain the increasedstress-induced release of catecholamines that follows pendringene ablation.

Immobilization Stress Increases Blood Pressure in Pendrin-Null and Wild-Type Mice

Because pendrin gene ablation modulates stress-inducedcatecholamine release, we asked whether pendrin gene ablationalso modulates the pressor response to immobilization stress.To answer this question, we examined the effect of immobili-zation stress on heart rate and blood pressure in the wild-typeand mutant mice (Fig. 11). Systolic, diastolic, and MAP bloodpressures were lower in pendrin-null than in wild-type miceunder basal conditions (Fig. 11, A–C), consistent with previousobservations (16, 23). With immobilization stress, MAP in-

creased rapidly by �16–26 mmHg in both wild-type andpendrin-null mice (Fig. 11E), although blood pressure re-mained lower in the pendrin-null than in wild-type mice after20 min of immobilization stress (Fig. 11D). However, 30 minafter relief of stress, MAP was similar in wild-type and pen-drin-null mice. Pendrin-null and wild-type mice had similarheart rates under basal and stimulated conditions (Fig. 11,F–H). We conclude that although immobilization stress pro-duces a similar increment in blood pressure in the pendrin-nulland in wild-type mice, after termination of stress, blood pres-sure returns to basal levels more quickly in wild-type than inpendrin-null mice.

DISCUSSION

The aim of this study was to determine whether pendrin isexpressed in the adrenal gland and whether it modulates

Fig. 7. Adrenal pendrin protein abundance does not fall with increased dietaryK� or NaCl intake. Pendrin protein abundance was quantified by immunoblotof adrenal lysates taken from mice given a HS or LS or K� (Treatments 1–3).Pendrin band density was unaffected by changes in dietary NaCl or K�.Groups were compared with an unpaired, 2-tailed, Student’s t-test.

Fig. 8. Pendrin gene ablation increasesstress-induced catecholamine release. Bloodwas sampled for epinephrine and norepi-nephrine, the norepinephrine metabolite di-hydroxyphenylglycerol (DHPG), and for do-pamine and the dopamine metabolite 3,4-dihydroxyphenylacetic acid (DOPAC),before and 5 and 20 min after initiation ofimmobilization stress. Values were com-pared between pendrin-null and wild-typemice at each of these time points using anunpaired, 2-tailed, Student’s t-test. After 20min of immobilization stress, epinephrineand norepinephrine levels were significantlyhigher in pendrin-null than in wild-typemice. Wild-type, n 6; pendrin-null, n 5.*P 0.05.

Fig. 9. Pendrin gene ablation increases stress-induced release of corticosterone.Blood was sampled for corticosterone before and 5 and 20 min after initiationof stress immobilization. Corticosterone values were compared between wild-type and pendrin-null mice at each of these time points using an unpairedStudent’s 2-tailed t-test. After 5 min of immobilization stress, corticosteronelevels were higher in pendrin-null than in wild-type mice. Wild-type, n 8;pendrin-null, n 6. * P 0.05.



adrenal function. Our results show that pendrin is expressed inchromaffin cells of the adrenal medulla, where it modulatesstress-induced catecholamine release. While stilbene-sensi-tive Na�-independent Cl�/HCO3

� exchange has been ob-

served in the rat adrenal cortex (33), our data indicate thatthis Cl�/HCO3

� exchange process is not mediated by pen-drin, since pendrin is expressed at very low levels in ratadrenal cortex and since pendrin is relatively resistant tostilbene inhibitors (8, 28).

Although we did not detect pendrin expression in the adrenalcortex, we observed that pendrin gene ablation increased thestress-induced release of corticosterone, a hormone producedin the adrenal cortex. The exact mechanism by which pendrinmodulates stress-induced release of corticosterone remainsunclear, although pendrin gene ablation likely increases stress-induced corticosterone release through the action of cat-echolamines. In this regard, peripheral catecholamines stimu-late the the hypothalamus-pituitary-adrenal axis to increase

Fig. 10. Pendrin gene ablation does not change adrenal TH protein abundance.TH protein abundance was quantified by immunoblot of adrenal lysates takenfrom wild-type (WT) and pendrin-null (KO) mice following 5 days of theNaCl-replete gelled diet described in Treatment 5. TH band density wasunaffected by pendrin gene ablation. Groups were compared with an unpaired,2-tailed, Student’s t-test.

Fig. 11. Effect of pendrin gene ablation on stress-induced changes in blood pressure. Systolic (A), diastolic (B), mean arterial pressure (MAP, C), and heart rate(F) were measured by telemetry in conscious wild-type (n 5) and pendrin-null (n 7) mice before, during 20 min of immobilization stress, and for 30 minafter termination of stress (Recovery). Under basal conditions, blood pressure was lower in pendrin-null than in wild-type mice (A–D). MAP (C and D) increasedin wild-type and pendrin-null mice with immobilization stress. However, the increment in blood pressure that followed immobilization stress was similar inwild-type and mutant mice (E). As such, blood pressure remained lower in pendrin-null than in wild-type mice after 20 min of stress (D). Thirty minutes afterrelief of stress (Recovery), systolic, diastolic, and MAP were similar in pendrin-null and wild-type mice (A–D). As such, blood pressure returned to basal levels30 min after relief of stress in wild-type mice but remained elevated relative to baseline in pendrin-null mice (E). Both absolute heart rate and stress-inducedincrement in heart rate were similar in wild-type and pendrin-null mice under these treatment conditions (D–F).



adrenal cortical corticosterone release (5, 37). Moreover, epi-nephrine directly stimulates corticosterone production by theadrenal cortex (9). However, because the time courses for thependrin-dependent, stress-induced increment in corticosteroneand catecholamine concentrations are not identical, pendrinmay alter stress-induced catecholamine release independent ofepinephrine and norepinephrine.

Because pendrin is expressed in chromaffin cells within theadrenal medulla, we were prompted to explore the role ofpendrin in adrenal medullary function. The time courses forepinephrine and norepinephrine responses to immobilizationstress that we observed in wild-type mice are consistent withtime courses reported previously in rats (22). In the rat and inwild-type mice, norepinephrine peaked 5 min after the initia-tion of immobilization stress and then remained stable for thenext 15 min, whereas epinephrine concentration continued torise over 20 min of immobilization stress (22). In pendrin-nullmice, however, both catecholamines continued to increase over20 min of immobilization stress. Moreover, after 20 min ofstress, levels of both catecholamines were significantly higherin pendrin-null than in wild-type mice. The enhanced catechol-amine response to immobilization stress that we observed withpendrin gene ablation suggests that the presence of pendrinblunts the responsiveness of the adrenal medulla to stress.Whether pendrin modulates chronic, stress-induced changes incatecholamine release remains to be determined.

Since immobilization stress fails to increase circulating epi-nephrine concentration following adrenalectomy (22), epi-nephrine is produced predominantly in the adrenal gland (1).Therefore, the enhanced epinephrine production observed inpendrin-null mice is most likely of adrenal origin. Whilenorepinephrine is also produced in the adrenal medulla, nerveterminals are the primary site of norepinephrine synthesis andrelease into the circulation (22). The recent demonstration ofpendrin mRNA expression in brain (http://mouse.brain-ma-p.org/experiment/show/70302117) raises the possibility thatablation of a central modulatory action in pendrin-null micemay contribute to the enhanced catecholamine response tostress. Whether pendrin is expressed in central nuclei control-ling sympathoadrenal activity or in noradrenergic nerve termi-nals remains to be determined.

How pendrin modulates catecholamine release is unclear.Since pendrin mediates Cl� uptake in kidney and since pendrinis probably expressed on the plasma membrane of adrenalmedullary chromaffin cells, pendrin may modulate catechol-amine release in the adrenal medulla through changes in Cl�

uptake. If so, pendrin gene ablation might stimulate catechol-amine release by reducing Cl� uptake in chromaffin tissue.However, previous studies have shown an opposite effect ofCl� uptake on epinephrine secretion by chromaffin cells (25).When chromaffin cell Cl� uptake is reduced through extracel-lular Cl� removal, epinephrine release falls. Therefore, themechanism by which pendrin modulates epinephrine releaseremains to be determined. While catecholamine release may bemodulated specifically by adrenal medullary pendrin expres-sion, we cannot exclude the possibility that pendrin geneablation modulates stress-induced catecholamine releasethrough an indirect, extra-adrenal effect of pendrin gene abla-tion.

The present study is the first to show that pendrin geneablation enhances stress-induced release of catecholamines,

presumably through enhanced activation of the sympatheticnervous system. Epinephrine and norepinephrine can bothincrease blood pressure, although epinephrine raises bloodpressure only at high concentrations (12). While it is wellestablished that the sympathetic nervous system is critical inacute blood pressure regulation (3), its role in chronic bloodpressure regulation is less well established (3). Epinephrineand norepinephrine also have different effects on heart rate,since epinephrine increases whereas norepinephrine decreasesheart rate, due to activation of baroreceptors and the vagusnerve (12).

Pendrin gene ablation reduces blood pressure, whereas micethat overexpress pendrin have profound salt-sensitive hyper-tension (13). Since pendrin gene ablation increases catechol-amine release in response to immobilization stress, we predictthat pendrin overexpression will attenuate stress-induced cat-echolamine release. In this regard, some people with salt-sensitive hypertension have a blunted stress-induced release ofnorepinephrine, which is consistent with our predictions in thismouse model of salt-sensitive hypertension (43). Why cate-cholamine release is suppressed in salt-sensitive hypertensionis unclear, but it is thought to mitigate the increase in bloodpressure observed with increased dietary salt intake (43).

In addition to enhanced stress-induced catecholamine re-lease, pendrin-null mice have enhanced vascular smooth mus-cle contractility in response to �-agonists, such as phenyleph-rine. (35). In particular, we observed that pendrin gene ablationincreases the maximum force of contraction/cross-sectionalarea in the thoracic aorta in response to phenylephrine (35).Since epinephrine and norepinephrine are also �-agonists,enhanced contractile force in response to these catecholaminesis expected in vascular smooth muscle of pendrin-null mice,which might change the stress-induced pressor response.

We (16) observed previously that following dietary NaClrestriction the increment in serum aldosterone relative to reninis blunted in pendrin-null mice. The present study demon-strates that the aldosterone/renin ratio is also lower in pendrin-null relative to wild-type mice when aldosterone production isstimulated through increased dietary K� intake. The decline inthis ratio could occur because plasma renin concentrationincreases, because plasma aldosterone concentration falls, orboth. Since circulating aldosterone concentration is very highin both wild-type and pendrin-null mice after a high-K� diet, itis unlikely that pendrin gene ablation significantly impairsaldosterone secretion. More likely, pendrin gene ablation re-duces the aldosterone/renin ratio by increasing plasma reninconcentration rather than by reducing aldosterone. Increasingdietary K� increases serum K� concentration while reducingplasma renin concentration or activity (4, 31). Because serumK� is higher in pendrin-null than in wild-type mice followinga high-K� diet (Table 1), circulating renin concentrationshould be either unchanged or reduced in the mutant mice.Since plasma renin concentration is the same or increased inthe pendrin-null mice, pendrin gene ablation may reduce theratio of aldosterone to renin by limiting the ability to suppressrenin release rather than limiting the ability to produce aldo-sterone. Why pendrin gene ablation increases plasma reninconcentration, however, remains to be established. The en-hanced stress-induced epinephrine and norepinephrine releaseobserved in pendrin-null mice might stimulate renin production(14, 21, 36). However, other mechanisms are possible.



We conclude that pendrin is expressed in the adrenal me-dulla, where it modulates stress-induced release of cat-echolamines. The mechanism by which pendrin changes cate-cholamine release remains to be determined.

ACKNOWLEGMENTS

We thank Drs. James F. Powers and Arthur S. Tischler for their helpfulsuggestions and for providing the �-PNMT antibody. We also thank TanishaThomas at the University of Florida College of Medicine Electron MicroscopyCore for expert technical assistance with tissue processing and immunohisto-chemical experiments.

GRANTS

This study was supported by National Institute of Diabetes and Digestive andKidney Diseases grants DK-46493 (to S. M. Wall), DK-085097 (to R. S. Hoover),and T32 DK-07656 (to Y. Lazo-Fernandez, B. M. Wynne,. and A. Y. Park). Thiswork was also supported by the National Institute of Child Health and HumanDevelopment, Intramural Research Program (G. Aguilera and K. Pacak).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

Author contributions: Y.L.-F., G.A., T.D.P., A.Y.P., W.H.B., R.L.S.,J.W.V., K.P., A.O.O., C.L.E., Y.H.K., G.L.S., and B.M.W. performed exper-iments; Y.L.-F., G.A., W.H.B., R.L.S., J.W.V., K.P., C.L.E., Y.H.K., G.L.S.,B.M.W., S.K.S., P.M.P., and S.M.W. analyzed data; Y.L.-F., G.A., R.L.S.,J.W.V., K.P., Y.H.K., G.L.S., S.K.S., P.M.P., and S.M.W. interpreted resultsof experiments; Y.L.-F., G.A., T.D.P., A.Y.P., R.L.S., J.W.V., Y.H.K.,B.M.W., S.K.S., and S.M.W. prepared figures; Y.L.-F., G.A., Y.H.K., G.L.S.,R.S.H., P.M.P., and S.M.W. edited and revised manuscript; Y.L.-F., G.A.,T.D.P., A.Y.P., W.H.B., R.L.S., J.W.V., K.P., A.O.O., C.L.E., Y.H.K., G.L.S.,B.M.W., R.S.H., S.K.S., P.M.P., and S.M.W. approved final version ofmanuscript; G.A., J.W.V., A.O.O., G.L.S., P.M.P., and S.M.W. conception anddesign of research; G.L.S. and S.M.W. drafted manuscript.

REFERENCES

1. Alberts B, Bray D, Lewis J, Raff M, Robert K, Watson JD. MolecularBiology of the Cell. New York: Garland, 1983.

2. Amlal H, Petrovic S, Xu J, Wang Z, Sun X, Barone S, Soleimani M.Deletion of the anion exchanger Slc26a4 (pendrin) decreases the apicalCl�/HCO3

� exchanger activity and imipairs bicarbonate secretion in thekidney collecting duct. Am J Physiol Cell Physiol 299: C33–C41, 2010.

3. Biaggioni I. Sympathetic control of the circulation in hypertension:lessons from autonomic disorders. Curr Opin Nephrol Hypertens 12:175–180, 2003.

4. Brunner HR, Baer L, Sealey JE, Ledingham JGG, Laragh JH. Theinfluence of potassium administration and of potassium deprivation onplasma renin in normal and hypertensive subjects. J Clin Invest 49:2128–2138, 1970.

5. Bugajski J, Turon M, Gadek-Michalska A, Borycz JA. Catecholamin-ergic regulation of the hypothalamic-pituitary-adrenocortical activity. JPhysiol Pharmacol 42: 93–103, 1991.

6. Bustin SA, Benes V, Garson JA, Hellemans J, Huggett J, Kubista M,Mueller R, Nolan T, Pfaffl MW, Shipley GL, Vandesompele J, Wit-twer CT. The MIQE guidelines: minimum information for publication ofquantitative real-time PCR experiments. Clin Chem 55: 611–622, 2009.

7. Dagerlind A, Goldstein M, Hokfelt T. Most ganglion cells in the ratadrenal medulla are noradrenergic. Neuroreport 1: 137–140, 1990.

8. Emmons C. Transport characteristics of the apical anion exchanger ofrabbit cortical collecting duct �-cells. Am J Physiol Renal Physiol 276:F635–F643, 1999.

9. Engeland WC, Gann DS. Splanchnic nerve stimulation modulates steroidsecretion in hypophysectomized dogs. Neuroendocrinology 50: 124–131,1989.

10. Everett LA, Belyantseva IA, Noben-Trauth K, Cantos R, Chen A,Thakkar SI, Hoogstraten-Miller SL, Kachar B, Wu DK, Green ED.Targeted disruption of mouse Pds provides insight about the inner-eardefects encountered in Pendred syndrome. Hum Mol Genet 10: 153–161,2001.

11. Holmes C, Eisenhofer G, Goldstein DS. Improved assay for plasmadihydroxyphenylacetic acid and other catechols using high-performanceliquid chromatography with electrochemical detection. J Chromatog BBiomed Appl 653: 131–138, 1994.

12. Innes IR, Nickerson M. The Pharmacological Basis of Therapeutics.New York: Macmillan, 1975.

13. Jacques T, Picard N, Miller RL, Riemondy KA, Houillier P, Sohet F,Ramakrishnan SK, Busst CJ, Jayat M, Corniere N, Hassan H, Aron-son PS, Hennings JC, Hubner CA, Nelson RD, Chambrey R, EladariD. Overexpression of pendrin in intercalated cells produces chloride-sensitive hypertension. J Am Soc Nephrol 4: 1104–1113, 2013.

14. Johnson MD, Fahri ER, Troen BR, Barger AC. Plasma epinephrine andcontrol of plasma renin activity: possible extrarenal mechanisms. Am JPhysiol Heart Circ Physiol 236: H854–H859, 1979.

15. Kim YH, Kwon TH, Frische S, Kim J, Tisher CC, Madsen KM,Nielsen S. Immunocytochemical localization of pendrin in intercalatedcell subtypes in rat and mouse kidney. Am J Physiol Renal Physiol 283:F744–F754, 2002.

16. Kim YH, Pech V, Spencer KB, Beierwaltes WH, Everett LA, GreenED, Shin WK, Verlander JW, Sutliff RL, Wall SM. Reduced ENaCexpression contributes to the lower blood pressure observed in pendrin-null mice. Am J Physiol Renal Physiol 293: F1314–F1324, 2007.

17. Kim YH, Pham TD, Zheng W, Hong S, Baylis C, Pech V, BeierwaltesWH, Farley DB, Braverman LE, Verlander JW, Wall SM. Role ofpendrin in iodide balance: going with the flow. Am J Physiol Renal Physiol297: F1069–F1079, 2009.

18. Kim YH, Verlander JW, Matthews SW, Kurtz I, Shin WK, WeinerID, Everett LA, Green ED, Nielsen S, Wall SM. Intercalated cellH�/OH� transporter expression is reduced in Slc26a4 null mice. Am JPhysiol Renal Physiol 289: F1262–F1272, 2005.

19. Klein JD, Martin CF, Kent KJ, Sands JM. Protein kinase C-� mediateshypertonicity-stimulated increase in urea transporter phosphorylation inthe inner medullary collecting duct. Am J Physiol Renal Physiol 302:F1098–F1103, 2012.

20. Knauf F, Yang CL, Thomson RB, Mentone SA, Giebisch G, AronsonPS. Identification of a chloride-formate exchanger expressed on the brushborder membrane of renal proximal tubule cells. Proc Natl Acad Sci USA98: 9425–9430, 2001.

21. Kolloch RE, Kruse HJ, Friedrich R, Ruppert M, Overlack A, StumpeKO. Role of epinephrine-induced hypokalemia in the regulation of reninand aldosterone in humans. J Lab Clin Med 127: 50–56, 1996.

22. Kvetnansky R, Sun CL, Lake CR, Thoa N, Torda T, Kopin IJ. Effectof handling and forced immobilization on rat plasma levels of epinephrine,norepinephrine, and dopamine-beta-hydroxylase. Endocrinology 103:1868–1874, 1978.

23. Pech V, Pham TD, Hong S, Weinstein AM, Spencer KB, Duke BJ,Walp E, Kim YH, Sutliff RL, Bao HF, Eaton DC, Wall SM. Pendrinmodulates ENaC function by changing luminal HCO3

�. J Am Soc Nephrol21: 1928–1941, 2010.

24. Pech V, Thumova M, Kim YH, Agazatian D, Hummler E, Rossier BC,Weinstein AM, Nanami M, Wall SM. ENaC inhibition stimulates Cl�

secretion in the mouse cortical collecting duct through an NKCC1-dependent mechanism. Am J Physiol Renal Physiol 303: F45–F55, 2012.

25. Pollard HB, Pazoles CJ, Creutz CE, Scott JH, Zinder O, Hotchkiss A.An osmotic mechanism for exocytosis from dissociated chromaffin cells.J Biol Chem 259: 1114–1121, 1984.

26. Powers JF, Brachold JM, Ehsani SA, Tischler AS. Up-regulation of retby reserpine in the adult rat adrenal medulla. Neuroscience 132: 605–612,2005.

27. Quentin F, Chambrey R, Trinh-Trang-Tan MM, Fysekidis M, Cam-billau M, Paillard M, Aronson PS, Eladari D. The Cl�/HCO3

� ex-changer pendrin in the rat kidney is regulated in response to chronicalterations in chloride balance. Am J Physiol Renal Physiol 287: F1179–F1188, 2004.

28. Reimold FR, Heneghan JF, Stewart AK, Zelikovic I, Vandorpe DH,Shmukler BE, Alper SL. Pendrin function and regulation in Xenopusoocytes. Cell Physiol Biochem 28: 435–450, 2011.

29. Royaux IE, Wall SM, Karniski LP, Everett LA, Suzuki K, KnepperMA, Green ED. Pendrin, encoded by the pendred syndrome gene, residesin the apical region of renal intercalated cells and mediates bicarbonatesecretion. Proc Natl Acad Sci USA 98: 4221–4226, 2001.

30. Schinner S, Bornstein SR. Cortical-chromaffin cell interactions in theadrenal gland. Endocr Pathol 16: 91–98, 2005.



31. Sealey JE, Clark I, Bull MB, Laragh JH. Potassium balance and thecontrol of renin secretion. J Clin Invest 49: 2119–2127, 1970.

32. Shcheynikov N, Yang D, Wang Y, Zeng W, Karniski LP, So I, WallSM, Muallem S. Slc26a4 functions as an electroneutral Cl-/I-/HCO3-exchanger: role of Slc26a4 and Slc26a6 in I- and HCO3- secretion and inregulation of CFTR in the parotid duct. J Physiol 586: 3814–3824, 2008.

33. Shepherd RM, Williams GH, Quinn SJ. Regulation of intracellular pHin single rat zona glomerulosa cells. Am J Physiol Cell Physiol 262:C182–C190, 1992.

34. Shibata S, Rinehart J, Zhang J, Moeckel G, Castaneda-Bueno M,Stiegler AL, Boggon TJ, Gamba G, Lifton RP. Mineralocorticoidreceptor phosphorylation regulates ligand binding and renal response tovolume depletion and hyperkalemia. Cell Metab 18: 660–671, 2013.

35. Sutliff RL, Walp ER, Kim YH, Walker LA, El-Ali AM, Ma JJ, BonsallR, Ramosevac S, Eaton DC, Verlander JW, Hansen L, Gleason RL,Pham TD, Hong S, Pech V, Wall SM. Contractile force is enhanced inaortas from pendrin-null mice due to stimulation of angiotensin II-dependent signaling. PLos One 9: e105101, 2014.

36. Takagi M, Atarashi K, Matsuoka H, Sugimoto T. A biphasic effect ofnoradrenaline on renin release from rat juxtaglomerular cells in vitro ismediated by alpha 1- and beta-adrenoceptors. J Endocrinol 132: 133–140,1992.

37. Timmermans W, Xiong H, Hoogenraad CC, Krugers HJ. Stress andexcitatory synapses: from health to disease. Neuroscience 248: 626–636,2013.

38. Vandesompele J, De Preter K, Pattyn F, Poppe B, Van Roy N, DePaepe A, Speleman F. Accurate normalization of real-time quantitativeRT-PCR data by geometric averaging of multiple internal control genes.Genome Biol 3: RESEARCH0034, 2002.

39. Verlander JW, Hassell KA, Royaux IE, Glapion DM, Wang ME,Everett LA, Green ED, Wall SM. Deoxycorticosterone upregulates PDS(Slc26a4) in mouse kidney: role of pendrin in mineralocorticoid-inducedhypertension. Hypertension 42: 356–362, 2003.

40. Verlander JW, Kim YH, Shin W, Pham TD, Hassell KA, BeierwaltesWH, Green ED, Everett L, Matthews SW, Wall SM. Dietary Cl�

restriction upregulates pendrin expression within the apical plasma mem-brane of type B intercalated cells. Am J Physiol Renal Physiol 291:F833–F839, 2006.

41. Wall SM, Hassell KA, Royaux IE, Green ED, Chang JY, Shipley GL,Verlander JW. Localization of pendrin in mouse kidney. Am J PhysiolRenal Physiol 284: F229–F241, 2003.

42. Wall SM, Kim YH, Stanley L, Glapion DM, Everett LA, Green ED,Verlander JW. NaCl restriction upregulates renal Slc26a4 through sub-cellular redistribution: role in Cl� conservation. Hypertension 44: 982–987, 2004.

43. Weber CS, Thayer JF, Rudat M, Sharma AM, Perschel FH, BuchholzK, Deter HC. Salt-sensitive men show reduced heart rate variability,lower norepinephrine and enhanced cortisol during mental stress. J HumHypertens 22: 423–431, 2008.

44. Welinder C, Ekblad L. Coomassie staining as loading control in westernblot analysis. J Proteome Res 10: 1416–1419, 2011.



Documents

Pendrin localizes to the adrenal medulla and modulates catecholamine release