urinary concentrating defect of adrenal...

Urinary Concentrating Defect of Adrenal Insufficiency

PERMISSIVE ROLEOFADRENALSTEROIDSONTHE HYDROOSMOTIC

RESPONSEACROSSTHE RABBIT CORTICALCOLLECTINGTUBULE

MICHAEL J. SCHWARTZand JUHA P. KOKKO, Department of Internal Medicine,University of Texas Health Science Center, Dallas, Texas 75235

A B S T RA C T Mineralo- and glucocorticoid-deficientstates, such as Addison's disease, are partly charac-terized by an inability to generate a maximally con-centrated urine. The purpose of the present studywas to develop a model of adrenal insufficiency andto determine whether changes in the intrinsic functionof the collecting duct could partly account for this con-centrating defect. Two kinds of experiments were per-formed: an assessment of the in vivo ability of adrenal-ectomized rabbits to concentrate their urine, and anexamination of the intrinsic hydroosmotic responsive-ness of in vitro perfused collecting ducts isolated fromnormal and adrenalectomized rabbits. The presentstudy demonstrates that adrenalectomized rabbits areunable to concentrate their urine maximally, and thatthe in vivo administration of either deoxycorti-costerone, 250 ,ug/kg, or dexamethasone, 50 ,g/kg,restored to or toward normal their concentrating ability.When cortical collecting tubules from adrenalec-tomized rabbits were perfused in vitro, they demon-strated a markedly blunted hydroosmotic response toantidiuretic hormone (ADH), which was corrected bythe in vitro addition of either aldosterone (50 pM) ordexamethasone (50 pM), but not progesterone (50 pM).The steroids by themselves, in the absence of ADH,had no intrinsic effect on the water permeability ofthe collecting duct. The blunted hydroosmotic re-sponse across cortical collecting tubules from adrenal-ectomized rabbits was corrected by the addition ofeither 8-bromo cyclic AMPor a potent phosphodies-terase inhibitor, 1-methyl-3-isobutylxanthine. Thepresent studies show that the cortical collectingtubules obtained from adrenalectomized rabbits donot respond normally to ADH. The poor hydroosmoticresponse to ADHwas corrected by exogenous aldos-

Portions of this work were submitted in partial fulfillmentfor the requirements of the Ph.D., Case Western ReserveUniversity, Cleveland, Ohio.

Received for publication 3 October 1979 and in revisedform 16 April 1980.

234

terone, dexamethasone, an analog of cyclic AMP, or aphosphodiesterase inhibitor. In conclusion, the presentstudies are consistent with the view that the concen-trating defect seen in adrenal insufficiency is at leastpartly the result of the absence of the permissiveeffect that adrenal steroids exert on the ADH-inducedreabsorption of water across the collecting duct. Theabsence of adrenal steroi(ds results in a diminishedrate of cyclic AMPaccumulation in the cells of thecollecting duct, either as a result of an augmentedactivity of cyclic AMPphosphodiesterase or a dimin-ished rate of cyclic AMPgeneration.

INTRODUCTION

Addisonian patients who are gluco- and mineralocorti-coid deficient are characterized by numerous derange-ments in the regulation of salt and water excretion:decreased plasma volume as a result of inappropriatenatriuresis; hyperkalemic acidosis as a result of de-creased excretion of hydrogen and potassium; inabilityto excrete normally a free water load (1-3); and de-creased ability to generate a maximally concentratedurine (4). Whereas the first three derangements havebeen examined in some detail previously, the mechanismby which urine cannot be concentrated maximally hasnot received the same scrutiny. The purpose of thepresent studies was to examine the role of the collectingduct in the inability to form maximally concentratedurine in a gluco- and mineralocorticoid-deficientstate. Specifically, the present studies were designed toexamine whether adrenal steroids by themselves in-fluence the intrinsic osmotic water impenneability of thecollecting duct, and to determine whether an alteredhydroosmotic response of the collecting duct to anti-diuretic hormone (ADH),' in the absence of adrenal

I Abbreviations used in this paper: ADH, antidiuretichormone; J,, rate of net volume flux in nl/mm min; Lp,hydraulic conductivity coefficient; MIX, 1-methyl-3-isobutyl-xanthine; PGE, prostaglandin E.

J. Clin. Invest. © The American Society for Clinical Investigation, Inc. * 0021-9738/80/08/0234/09 $1.00Volume 66 August 1980 234-242

steroidis, couild contribute to the concentrating defectseen in this setting. Fuirther, througlh the uise of in vitromiciiroperfuisioIn technii(Iies, we have examiined themodtulationi of the initrinsic hydroosmiiotic responsiv,e-nless of the cortical collecting tubule to ADHby adrenalsteroids.

METHOIDS

Adrenailectom y. Female New Zealanid white rabbitsweighinig 1.5-2.5 kg were anesthetized with anl intramiius-cular injection of 50 mg/kg ketamine plus 3.5 mg/kg proma-zine (ketaset plus; Bristol Myers Company, Syracuse, N. Y.)and adrenalectomized under microscopic control via a midllineabdominal incisioni. No accessory adrenal tissue was seen.Postoperatively, each animal vas administered a single intra-mnuscular injection of dexamethasone phosphate (hexadrolphosphate; Organon Inc., West Oranige, N. J.), 50 jug/kg,and maintained thereafter on 0.9% sodium chloride drinkingfluid, without additional exogenous steroid. The adeqIuacy ofadrenalectomy was assessed by the measuremiienit of plasmacorticosteronie conicentrations by radioimmunlliloatssavs (Endo-crine Science, Tarzana, Calif.). In seven adrenalectomizedrabbits, 2 wk after surgery the plasma corticosterone con-centration was 38±5 ng/dl.2 This was in contrast to the plasmacorticosterone concentration of 2,501+491 ng/ldl similalrly ob-tained in four normial rabbits. All adrenalectomiiized rabbitswere studied 10 d-2 wk after surgery.

Maximum urinary concenitratintg ability. 6 normall andcl 12adrenalectomized rabbits, all drinking 0.9% sodiumn chloride,were adapted to metabolic cages. Urine vas collectedunder water-equilibrated mineral oil startinig 2 h after theintramu.scular injection of 1.25 U/kg of vasopressin tannlcatein oil (pitressini; Parke, Davis & Co., Detroit, 'Mich.).After 8 h of collection, the osmolality of the urine vas deter-mined by freezing point depression (advanced osmomneter;Advanced Instruments, Inc., Needham Heights, Mass.). 1 wklater, the saime 12 adrenalectomnized rabbits vere restudiedto determine whether the replacemiient of either minieratlo- orglucocorticoid steroid hormone improved the urinary con-centrating ability. Each rabbit was given an intramuscularinjection of either deoxycorticosterone acetate in oil (Per-corten; Ciba-Geigy Corp., Pharmaceuticals Div., Summllit,N. J.), 250 jug/kg, or dexamethasone phosphate, 50 ,ug/kg;2 h later the maximumii urinary osmolality after vasopressinwas determined as before.

In vitro microperfusion. The technique of in vitro micro-perfusion used in the present study was similar to thatpreviously reported (5); only the salient features vill be de-scribed. Rabbits were decapitated, and a 1-2-mm coronalslice of kidney was placed on ice in a dish containing a solu-tion of artificial ultrafiltrate, the com)positioni of which %vasidentical to that of the solutioni bathing the outside of thetubules in vitro (see below) except that the dissectinig soltutionadditionally conitained 5 mg/mI of bovine senimii albumnin.Cortical collectinig tubules vere dissected from the slice, withspecial care being taken so that the proximal end of all tubulesbegan immnlediately below the superficial branchpoint of thecollecting duict; the distal end of the tuibule was clearly locatedwithin the cortex of the kidney.

Cortical collecting tubules were transferred to a Luicitechamber sitting on the stage of an inverted microscope andthe ends cannulated within concentric glass micropipettes.Hydraulic insulation of the ends of the tubules within the

2 The limit of detection by this assay is 20 ng/dl.

pipettes Wcas assured by the use of Sylgard 184 (Dow CorningCorp., Midland, Mich.). The solutioni used to bathe theoutside of the tubule was an artificial ultrafiltrate with thefollowing comiiposition, (mM): NaCl 115; KCl 5, NaHCO325;NaH2PO4 2.3, Na acetate 10; MgSO4 1, glucose 8; alanine5; CaCl2 1.8. No protein was present in the bathing solution.The solution used to perfuse the lumen of the tubule wasof an identical electrolyte composition to the bathing sollution,except that the concentration of NaCl was 10 mM. Thus, thefinal osmolalitv of the perfusate was 120 mosM, whereas thebathing solution osmolality was 300 mosM. These weremeasured in each experiment. In addition, the perfusatecontained 20 ACi/ml of an impermeant volume marker, 125I_iothalamate (Ahbott Laboratories, Chemical Div., NorthChicago, Ill.). All solutions were equilibrated at 37°C with95% 02-5% CO2 so that the final pH at 37°C was 7.4. In afew experimiienits, the 180 mosM1 o.smotic gradienit across thetubule was established by making the perfusate isotonicwith plasmiia anid the bathing solution hypertonic with raf-finose. Becau.se the hydroosmotic action of ADHacross thecortical collecting tubule is e(qually expressedI when the bath-ing solution is hypertonic (6), and because the results ob-tained in these tubules were identical to those in the othertubules in the same treatment group, all results are presentedtogether. All experiments were conducted at 37°C.

Because cortical collecting tubules maniifest a .small osinoticwater permeability in the absence of ADH when initiallyperfused in vitro (7), all tubules were allowed to equilibratein vitro for 60 mIi. From 60 to 90 min after commiiiiencing per-fusion, control collectionis were made to measure the osmoticwater flux before the addition of ADH. At 90 min, experi-mental collections were made continuously for collectecl fluidosmolality and radioactivity. At 37°C, after reachinig thepeak value 10-20 min after its addition, the hydroosmoticresponse to ADHdeclines gradually (8). For this reason, thevalues for Jv and L, reported here represent the peak valuesobtained from each tubule. The time course of the hydroos-motic response to ADH was identical in tubules obtainedfrom normal and adrenalectomized rabbits.

The net volume flux wvas given as:

Jv =L

where J,. = rate of net volumlle flux in nl/mm min; V1 = per-fusion rate in nI/min; V0 = collection rate in nl/min; and L= tubular length in mm. The measurement of Jv does notgive a true estimate of the osmotic water permeability ofthe collecting duct in that it does not take into account themagnitude of the driving force for volume reabsorption, theosmotic gradient. Thus, the hydraulic conductivity coefficient(Lp) was used to measure the hydroosmotic response to ADHand was derived as described (9, 10):

-ViCr cj - Co 1 (CO- Cb)C1Lp= + inI

RTA CiCbCO (Cb)2 (Ci - Cb)CO J

where Lp = hydraulic conductivity coefficient in cm/s atmx 107; R = gas constant; T = absolute temperature; A= internal tubular surface area (with an assumed internaldiameter of 20 microns);3 Cb = osmolality of the bathing solu-tion; C, = osmolality of the perfusate; and C0 = osmolalityof the collected fluid. The osmolality of the bathing, collected,and perfusate solutions were measured by freezing point de-

3 The internal diameter of tubules from normal and adrenal-ectomized rabbits were not different.

Concentration Defect of Adrenal Insufficiency 235

pression in nanoliter quantities (Clifton nanoliter osmometer;Clifton Technical Physics, Hartford, N. Y.).

Study groups. Four groups of in vitro experiments wereperformed. The experiments of group 1 were performed to de-termine whether collecting ducts from adrenalectomizedrabbits were intrinsically permeable to water in the ab-sence of ADH, and to compare the hydroosmotic responsewith ADHof collecting ducts from normal and adrenalec-tomized rabbits. Thus, control collections were made duringthe initial 90 min of perfusiofi of collecting ducts fromnormal and adrenalectomized rabbits, and the J, determinedbefore the addition of ADH. After this time, ADHat a finalconcentration of 200 ,uU/ml was added continuously to thebathing solution and the J, and Lp determined over the next120 min at 5- 10-min intervals.

Group 2 experiments were first designed to determinewhether adrenal steroids, by themselves, affected the in-trinsic water permeability of collecting ducts from adrenalec-tomized rabbits; second, whether adrenal steroids affectedthe blunted hydroosmotic responsiveness of collecting ductsfrom adrenalectomized rabbits to ADH; and third, whetherthere was any specificity of the action of adrenal steroidson the hydroosmotic response to ADH. Thus, either al-dosterone, dexamethasone, or progesterone, each at a finalconcentration of 50 pM, was singly present in the perfusingsolution from the very beginning of the experiment. Tubularfluid was collected from 60 to 90 min after the beginning oftheexperiment in the presence of one of the steroids but beforethe addition of ADH, thereby allowing an examination ofthe effect that a single steroid might alone have on theintrinsic water permeability across cortical collecting tubulesfrom adrenalectomized rabbits. At 90 min, ADHwas thenadded and the Jv and Lp were measured at continuousintervals.

The experiments of group 3 were designed to determinethe hydroosmotic response of tubules from adrenalectomizedrabbits to exogenous cyclic AMPin an effort to localize theblunted hydroosmotic response to ADH to either steps in-volved in the intracellular accumulation of cyclic AMPor,alternatively, to those steps distal to cyclic AMP(see Fig. 6).Thus, as in group 1, after control collections, an analog ofcyclic AMP, 8-bromo cyclic AMP (Sigma Chemical Co.,St. Louis, Mo.), at a final concentration of 0.1 mM, was addedcontinuously to the solution that bathed collecting ducts ofnormal and adrenalectomized rabbits, and the Lp wasdetermined.

The experiments of group 4 were designed to examinewhether cyclic AMPphosphodiesterase inhibition restoredthe hydroosmotic responsiveness of collecting ducts fromadrenalectomized rabbits to ADH. Thus, the phospho-diesterase inhibitor, 1-methyl-3-isobutylxanthine (MIX;Aldrich Chemical Co., Inc., Milwaukee, Wis.) was presentcontinuously in the bathing solution from the beginning ofthe experiment at a final concentration of 0.1 mM. Collec-tions of tubular fluid were then made before and after theaddition of 2 ,uU/ml of ADH, as in group 1, and the Jv and Lpdetermined. Cortical collecting tubules from normal andadrenalectomized rabbits were studied. In addition, a secondseries of tubules from normal rabbits were exposed only toADHat a concentration of 2 juU/ml in the absence of MIXto measure the hydroosmotic response resulting from thislower concentration of ADH.

All data in the text and figures is expressed as the mean+SEM. Paired and unpaired t tests were used to determinesignificance, which was taken as P < 0.05.

RESULTSMaximum urinary osmolality. The maximum urine

osmolality was measured in 6 normal and 12 adrenalec-

tomized rabbits to determine whether adrenal insuf-ficiency in the rabbit was associated with a concentrat-ing defect. All rabbits drank 0.9% sodium chloride andwere given 1.25 U/kg vasopressin tannate. As shown inFig. 1, adrenalectomized rabbits with a maximum uri-nary osmolality of 1,327±65 mosMhad a significantlyless (P < 0.02) concentrated urine then identicallytreated normal rabbits with a maximum urinary os-molality of 1,901+207 mosM.

To determine whether this concentrating defect wasa direct result of adrenal steroid absence or somesecondary consequence of adrenalectomy, the same 12adrenalectomized rabbits were given a single injectionof either deoxycorticosterone, 250 ,ug/kg, or dexametha-sone, 50 ,g/kg, and the maximum urine osmolalityafter vasopressin determined again. As shown in Fig. 1,

ADX vs NORMALI-0 s P<0.02 -

2500 r

2400k

2300 F2200 F

2100F

ADXvs ADX+ DEX

ADX vs ADX+ DOCA

p

I /

I',2000 F

UosmmosM 1900k

1800FH--

1700 F

1600

1500

1400

1300

1200JI -I

ADX ADX+ DOCA(-)(n012) (nu6)

ADX+ DEX (---)(n -6)

NORMAL(n =6)

FIGURE 1 The effect of adrenalectomy and the selective re-placement of mineralo- or glucocorticoid hormones on themaximum urinary concentrating ability in normal andadrenalectomized (ADX) rabbits. Maximum urinary os-molality was determined after the administration of 1.25 U/kgADH. ADX rabbits were retested after either deoxycorti-costerone (DOCA) (solid lines), 250 gg kg, or dexamethasone(DEX) (broken lines), 50 ,g/kg.

236 M. J. Schwartz and J. P. Kokko

either deoxycorticosterone (P < 0.01) or dexametha-sone (P < 0.02) significantly increased the maximumurinary osmolality to or toward normal in both groupsof rabbits (P = NS compared with normal).

Group 1. The basal J, before the addition of ADHwas not different from zero across cortical collectingtubules from both normal and adrenalectomized rab-bits (Table I; group 1, J, before). Thus, these experi-ments suggest that collecting ducts from adrenalec-tomized rabbits are not intrinsically permeable to waterbefore the addition of ADH.

To determine whether the concentrating defect seenin adrenalectomized rabbits in vivo could involve al-terations in the response of the collecting duct to ADH,cortical collecting tubules from normal and adrenalec-tomized rabbits were perfused in vitro and the hydro-osmotic response to a maximal concentration of ADHwas determined. As shown in Table I (group 1) andFig. 2, the Lp across cortical collecting tubules fromnormal rabbits after ADH was 275+39 cm/s atmx 107, a value similar to that reported by othersunder comparable experimental conditions (8). This

TABLE IHydroosmotic Response across Cortical Collecting Tubules from Normal and Adrenalectomized Rabbits

lvExperimental Collected Lp

group L Vj Osm Before* After* after PI

mm nl/min mOsm/kg H20 nl/mm/min cm/ls atmn x 10'

Group INormal + 200 ,uU/ml ADH(n = 7)

2.18 12.05 193§±0.15 ±1.05 ±11

Adrenalectomized + 200 ,uM/ml ADH(n = 9)1.89 9.11 143§

+0.13 +0.92 +6

0.03 1.74+0.02 +0.27

0.00 0.39+0.02 +0.04

Group IIAdrenalectomized + 50 pM Aldosterone + 200 ,tU/ml ADH(n = 6)

2.46 13.17 172 -0.01 1.40+0.13 +1.24 ±6 +0.02 +0.18

Adrenalectomized + 50 pM dexainethasone + 200 ,uU/ml ADH(n = 6)2.06 12.49 154 0.00 1.08

±0.07 ±0.94 +7 ±0.01 +0.25Adrenalectomized + 50 pM progesterone + 200 ,uU/ml ADH(n = 4)

2.32 12.24 140 0.01 0.56±0.29 ±0.62 ±3 ±0.01 ±0.05

Group IIINormal + 0.1 mM8-bromo cyclic AMP(n = 5)

2.22 13.07 193 0.01 2.00±0.23 ±1.07 ±+14 ±0.01 ±0.21

Adrenalectomized + 0.1 mM8-bromo cyclic AMP(n = 5)1.97 10.72 185 0.03 2.09

±0.12 ±1.00 ±+12 ±0.02 +0.34

Group IVNormal + 2 ,uU/ml ADH(n = 5)

2.14 13.94 160 -0.01 1.38±0.20 +0.38 +7 ±0.01 ±0.10

Normal + 0.1 mM1-methyl-3-isobutylxanthine + 2 ,U/ml ADH(n = 7)2.11 12.27 230 0.02 2.71

±0.08 ±0.66 + 14 ±0.01 +0.30Adrenalectomized + 0.1 mM1-methyl-3-isobutylxanthine + 2 uU/ml ADH(n = 5)

2.04 10.27 234 0.03 2.44+0.14 +0.56 +16 +0.02 ±0.23

196±27

155+32

<0.001 compared withadrenalectomized alone

<0.005 compared withadrenalectomized alone

60 NS compared with±8 adrenalectomized alone

286+29

268+50

NScompared with normal

175+23

467+74

<0.05 compared withnormal + ADHalone

503 NS compared with±133 normal + MIX + ADH

* The values reported are the maximum observed before and after the addition to the bathing solution of either ADHor

8-bromo cyclic AMP. All other agents were present from the beginning of the experiment, whereas either ADHor 8-bromocyclic AMPwas added 90 min after the beginning of the experiment.I Refers to Lp values.§ In some of these experiments the perfusate was isotonic and the bathing solution hypertonic, although the osmotic gradientwas identical to the rest of the experiments in this group. Because the results were identical, these experiments are included.

Concentration Defect of Adrenal Insufficiency

275±39

48+5

<0.0005

237

Lpcm/s atm

350rx 17

300F

250F

200F

collecting ducts from adrenalectomized rabbits toeither aldosterone, dexamethasone, or progesterone re-sulted in a J, of zero. Thus, none of the steroids testedaltered the intrinsic water impermeability of the col-lecting duct before the addition of ADH.

The results of steroid addition on the hydroosmoticresponse to ADHare shown in Table I (group 2) andFig. 3, where it can be seen that either aldosterone

gor dexamethasone significantly increased (P < 0.001,........P < 0.005, respectively) the Lp after ADH, from 48+5

to 196+27 and 155±32 cm/s atm x 107, respectively.Progesterone did not significantly increase the hydro-osmotic response to ADH, with the resultant Lp beingonly 60+8 cm/s atm x 107(P = NS).

Group 3. The locus of this "permissive" effect ofadrenal steroids on the hydroosmotic response to ADHcould reside anywhere along the pathway of the ADHresponse in the collecting duct, from the binding ofADH to its receptor on the antiluminal side of thecell (11) to the apical plasma membrane, the ultimatebarrier for water reabsorption (12). Weattempted to de-termine whether adrenal steroids were acting to in-crease the rate of intracellular accumulation of cyclicAMPor whether steroids affected some step distal tocyclic AMP accumulation through the use of apermeable analog of the nucleotide, 8-bromo cyclicAMP(see Fig. 6 in the discussion section for review

1501F

100o

50[

ADX((nu 9) tn-7 )

FIGuRE 2 The hydroosmotic response to ADHacross corti-cal collecting tubules from normal and adrenalectomized(ADX) rabbits. The L, was measured after the addition of200 ,uU/ml ADH to the solution bathing tubules in vitro.P < 0.0005.

was significantly greater (P < 0.0005) than the similarlydetermined Lp across cortical collecting tubules fromadrenalectomized rabbits after ADH, which had a valueof 48±+ 5 cm/s atm x 107.

Group 2. Although the in vivo administration ofeither deoxycorticosterone or dexamethasone cor-rected the concentrating defect in adrenalectomizedrabbits, the blunted hydroosmotic response acrosscortical collecting tubules from these rabbits could bethe result of some secondary systemic consequence ofadrenal insufficiency or the result of a direct effect ofsteroids on the epithelium of the collecting duct. Toexamine whether adrenal steroids had a direct effecton the hydroosmotic response of the epithelium ofthe collecting duct to ADH, cortical collecting tubulesfrom adrenalectomized rabbits were exposed in vitro toeither aldosterone, dexamethasone, or progesterone,each at a final concentration of 50 pM, and the netosmotic water permeability measured before and afterthe addition of ADH. As shown in Table I (group 2,J, before) before the addition of ADH, the exposure of

Lpcm/s atm

250rxo07

2001-

1501

100l

50-

ADX ADX

ALDO(n39) (n-6)

ADX ADX

DEX PROGEST(n-6) (nu4)

FIGURE 3 The effect of the in vitro addition of a singlesteroid on the hydroosmotic response across cortical collect-ing tubules from adrenalectomized (ADX) rabbits to ADH.Either aldosterone alone (ALDO), dexamethasone alone(DEX), or progesterone alone (PROGEST) was added to theperfusate at a final concentration of 50 pM; the Lp after200 ,uU/ml ADHwas measured. ADXvs. ALDO, P < 0.001;ADX vs. DEX, P < 0.005; ADX vs. PROGEST, P = NS;ALDOvs. DEX, P = NS.


of the relevant biochemical steps). The analog alonewas added to the solution that bathed cortical collect-ing tubules from normal and adrenalectomized rabbits,and the resultant hydroosmotic response was deter-mined. These experiments are presented in Table I(group 3) and Fig. 4, where it can be seen that 8-bromocyclic AMPincreased the Lp across cortical collectingtubules from normal rabbits to 286+29 cm/s atmX 107, a value similar to that seen after the addition of200 ,tU/ml ADHto tubules from normal rabbits. Underthe same experimental conditions, 8-bromo cyclicAMPincreased the Lp across cortical collecting tubulesfrom adrenalectomized rabbits to 268+50 cm/s atmX 107, a value virtually identical to that obtained fromtubules of normal rabbits.

Group 4. The results of the experiments of group 3suggest that the blunted hydroosmotic response oftubules from adrenalectomized rabbits to ADH in-volves a diminished rate of intracellular cyclic AMPaccumulation. To further investigate the possibility thatadrenal steroids act to increase the rate of intracellularcyclic AMPaccumulation, cortical collecting tubulesfrom normal and adrenalectomized rabbits were per-fused in the continual presence of a potent phospho-diesterase inhibitor, MIX, and the hydroosmotic re-

Lpcm/s atm

350r

300-

x j07

T TI250-

200-

150-

1001

50-

200 p&U/miADH

(n-9)

8-8romo-cAMP(n5)

NORMAL NORMAL+ +

200 tU/ml 8-Bromo-ADH cAMP

(n=7) (n-5)

FIGURE 4 The hydroosmotic response across cortical collect-ing tubules from normal and adrenalectomized (ADX) rab-bits after a cyclic AMPanalog. 8-bromo cyclic AMPwas addedto the solution that bathed cortical collecting tubules fromnormal and ADXrabbits, and the Lp was measured.

sponse to a submaximal concentration of ADH wasdetermined. As shown in Table I (group 4) and Fig. 5,cortical collecting tubules from normal rabbits, whenexposed to 2 ,uU/ml of ADHwithout MIX, respondedwith a Lp of 175±32 cm/s atm x 107. The additionof MIX plus 2 uU/ml ADH to tubules from normalrabbits resulted in a significantly greater Lp, 467±76cm/s atm x 107 (P < 0.05). When cortical collect-ing tubules from adrenalectomized rabbits were ex-posed to MIX plus 2 ,uU/ml ADH, the resultant Lp was503±133 cm/s atm x 107, a value that did not differfrom that obtained from tubules of normal rabbitsidentically treated (P = NS).

DISCUSSION

The present study demonstrates that the rabbit, likeman (4), dog (13), and rat (14), is unable to generate amaximally concentrated urine when rendered adre-nally insufficient. Previous studies have revealed adiminished medullary tonicity associated with adrenalinsufficiency (13, 14), and have suggested that the in-

Lpcm/s atm

650_600_550_500_450_400_350_300_250_200_150_100_50

x 107

T

T

ADX ADX NORMAL NORMAL+ +

MIX MIX(n=9) (n-5) (n-5) (n=7)

FIGURE 5 The effect of the addition of a phosphodiesteraseinhibitor on the hydroosmotic response to a subinaximalconcentration of ADH. Cortical collecting tubules fromnormal or adrenalectomized (ADX) rabbits were exposed to0.1 mMMIX in the bathing solution, and the Lp after the addi-tion of 2 AU/ml ADH to the bath was measured. Corticalcollecting tubules from normal rabbits exposed only to 2 ,LU/mlADH, or from ADX rabbits exposed to 200 ,uU/ml ADH, areshown for comparison. ADX vs. ADX+ MIX, P < 0.001;NORMALvs. NORMAL+ MIX, P < 0.05; NORMALvs. ADX,P < 0.0005; ADX+ MIX vs. NORMAL+ MIX, P = NS.


ability to concentrate maximally was the result of adiminished driving force for water reabsorption sec-ondary to a steroid-dependent decrease in solute trans-port by the thick ascending limb of Henle. Whether infact the thick ascending limb of Henle is physio-logically responsive to adrenal steroids is unclear atthis time. Using the activity of the tricarboxylic acidcycle enzyme, citrate synthase, as a marker for the ac-tion of aldosterone (15), we have recently demon-strated that the activity of the enzyme in the corticalthick ascending limb of Henle is unchanged byadrenalectomy or the addition of exogenous al-dosterone, whereas in the cortical collecting tubule,adrenalectomy is associated with a diminution of citratesynthase activity, and aldosterone treatment with asignificant increase in activity (16).

The present study reveals that cortical collectingtubules from adrenalectomized rabbits are not os-motically "leaky" before the addition of ADH(TableI). Thus, we find no evidence to support the conceptsuggested from in vivo micropuncture studies (17) thatthe inability of adrenalectomized animals to excreteexcess water is the result of a steroid-dependent loss ofthe basal impermeability of the distal nephron. Theprincipal difference between these earlier studies andthe present work is that the present study was con-ducted in vitro, where a direct examination of the in-trinsic function of the tubular epithelium, isolatedfrom all systemic influences, is possible.

The present study does reveal that cortical collect-ing tubules from adrenalectomized rabbits demonstratea markedly blunted hydroosmotic response to ADH(Fig. 2). This diminution in the hydroosmotic effec-tiveness of ADH across the collecting duct couldpartly account for the inability to generate a maximallyconcentrated urine in adrenal insufficiency. Whenadrenalectomized rabbits were given either deoxy-corticosterone or dexamethasone in vivo they wereable to increase significantly the maximum osmolalityof their urine after exogenous ADH (Fig. 1). Whencortical collecting tubules isolated from adrenalec-tomized rabbits were exposed to either aldosterone ordexamethasone, but not progesterone, they were ableto respond to ADHwith a significant increase in waterreabsorption (Fig. 3). None of the steroids alone hadany effect on the intrinsic water permeability of collect-ing ducts from adrenalectomized rabbits (Table I,group 2, J, before) in that the net water movement be-fore the addition of ADHwas zero. Thus, as originallydefined by Ingle (18), these effects of adrenal steroidsare permissive in that they have no effect themselveson the water permeability of the collecting duct butact to augment the response to a second hormone,ADH. This permissive effect of adrenal steroids wasexpressed after 90 min of exposure and at a concentra-tion of 50 pM, suggesting that it represents a physio-

logically significant effect. Although this data does notindicate whether the in vivo administration of eitheradrenal steroid acted solely on the collecting duct invivo, it does suggest that at least part of the concen-trating defect seen in adrenalectomized animals is re-ferable to an inability of the collecting duct to reabsorbwater under the influence of ADH.

In an effort to explore the mechanism(s) involved inthe permissive effect of adrenal steroids on the hydro-osmotic response to ADH, cortical collecting tubulesfrom normal and adrenalectomized rabbits were ex-posed to 0.1 mM8-bromo cyclic AMPin vitro and theresultant Lp determined (Fig. 4). This concentration ofthe cyclic AMPanalog has been shown previously toinduce maximal water reabsorption across cortical col-lecting tubules from normal rabbits studied under con-ditions similar to those used in the present study (8).In the present study, 8-bromo cyclic AMPinduced aLp across tubules from normal rabbits, comparable withthat induced by a maximum concentration of ADH.When cortical collecting tubules from adrenalec-tomized rabbits were exposed to the same concentra-tion of 8-bromo cyclic AMP, the resultant Lp wasvirtually identical to that observed across tubules fromnormal rabbits. This suggests that the inability of col-lecting ducts to respond to ADH in the absence ofadrenal steroids is the result of a diminished rate ofintracellular cyclic AMPaccumulation rather than adefect at some site distal to cyclic AMP(Fig. 6).

To further explore the mechanism(s) of the permis-sive effect of adrenal steroids on the action of ADH,the effect of the addition of 0.1 mMof a potent phos-phodiesterase inhibitor, MIX, on the hydroosmotic

Blood Collecting Duct I Lumen

FIGURE 6 A summary of our current understanding of thehormonal control of the hydroosmotic response to ADHin themammalian collecting duct.


response to a submaximal (2 ,uU/ml) concentration ofADHwas determined (Fig. 5). This concentration ofMIX has been shown to maximally inhibit rabbit renalmedullary phosphodiesterase (19). As reported byothers (19), the present study demonstrates that theaddition of MIX increased the hydroosmotic responseof cortical-collecting tubules from normal rabbits to asubmaximal concentration of ADH (Fig. 5, right), afinding consistent with phosphodiesterase inhibition.When cortical collecting tubules from adrenalec-tomized rabbits were perfused under identicalconditions, MIX plus 2 ,uU/ml ADH, the resultinghydroosmotic response was not different from that ob-served across tubules of normal rabbits (Fig. 5, left).

The results of the present study are comparable withthose reported previously in the toad urinary bladder,an epithelium often studied as a model of the mam-malian collecting duct. Whenthe toad urinary bladderwas rendered steroid-deplete, it became hydroos-motically unresponsive to ADH(20). The addition ofvery low concentrations of either a mineralo- orglucocorticoid hormone in vitro restored the hydro-osmotic responsiveness to ADH. As in the presentstudy, mineralo- and glucocorticoid steroid hormoneswere essentially equipotent in their permissive effects.The addition of a phosphodiesterase inhibitor to thesteroid-depleted toad urinary bladder restored thehydroosmotic responsiveness to ADH, and when theactual activity of phosphodiesterase was measured incells scraped from steroid-depleted toad urinary blad-ders, it was elevated (21). The addition of either al-dosterone or dexamethasone to the toad urinary bladderwas associated with a diminution in phosphodiesteraseactivity, and the permissive effect of adrenal steroidswas dependent on RNA transcription and proteinsynthesis.

Fig. 6 represents our present understanding of thehormonal control of ADH-mediated water reabsorptionin the mammalian collecting duct cell.

We interpret the results depicted in Figs. 4 and 5to be consistent with the hypothesis that steps distalto cyclic AMPare not rate limiting in the absence ofsteroids and that at least one site of action of adrenalsteroids in exerting their permissive effects on thehydroosmotic response to ADH involves the rate ofcyclic AMPaccumulation. Because the cyclic AMPanalog mimicked the action of ADH in tubules fromnormal rabbits (Fig. 4), its efficacy in inducing waterreabsorption across cortical collecting tubules fromadrenalectomized rabbits suggests that adrenal steroidsact to increase the rate of endogenous cyclic AMPaccumulation. As shown in Fig. 6, this could involveeither an increased rate of generation by adenylcyclase or a diminution in the rate of cyclic AMPdegradation by phosphodiesterase (or both).

It is not possible to distinguish between these two

possibilities from the data displayed in Fig. 4 in thatthe cyclic AMPanalog not only acts to bypass thosesteps involved in endogenous cyclic AMPgenerationbut, because the analog is resistant to hydrolysis byphosphodiesterase (22), could as well suggest an in-creased activity of endogenous cyclic AMPdegradationin the absence of adrenal steroids. The experimentsof Fig. 5 show that the addition of a phosphodiesteraseinhibitor restored to normal the hydroosmotic responseof tubules from adrenalectomized rabbits, a findingconsistent with at least one effect of adrenal steroidsbeing the attenuation of the activity of phosphodies-terase. Such a conclusion would require the directdetermination of cyclic AMP-dependent phospho-diesterase activity in single segments of corticalcollecting tubule taken from normal and adrenalec-tomized rabbits. Unfortunately such an assay is notavailable for isolated tubule segments, and the nextbest set of experiments were therefore performed usingthe most specific inhibitor of phosphodiesterase avail-able, MIX. It is of interest, however, that whole kidneyphosphodiesterase activity is greater in adrenalecto-mized than normal rats, and that the in vivo administra-tion of glucocorticoid hormone returns whole kidneyphosphodiesterase activity to normal levels (23).

In addition to a postulated effect of adrenal steroidsin modulating phosphodiesterase activity, the resultsdepicted in Fig. 5 are also compatible with the pos-sibility that adrenal steroids act to increase the rate ofcyclic AMPgeneration in response to ADH. This couldbe mediated, for example, by diminishing the rate ofendogenous prostaglandin E (PGE) biosynthesis. Inthe toad urinary bladder, adrenal steroid depletion isassociated with an increased rate of endogenous PGEsynthesis that is restored to normal by low concentra-tions of either aldosterone or dexamethasone (24). PGEhas been shown to inhibit the hydroosmotic action ofADH across isolated perfused cortical collectingtubules (25), partly by a direct inhibition of adenylcyclase (26). As shown in Fig. 6, ADH itself acts tostimulate the endogenous synthesis of PGEin both thetoad urinary bladder (24) and rabbit renal medullaryinterstitial cell cultures (27), and the biosynthetic en-zyme for PGE, cyclooxygenase, has been identified incortical collecting duct cells (28, 29). A direct demon-stration of PGE synthesis by the cortical collectingduct has not been reported, and whether the collectingduct, like the toad bladder, responds to steroid deple-tion with an increase in the rate of PGE synthesisis unknown.

Whereas the physiologic and clinical significance ofthe present studies is conjectural, the following con-clusions seem warranted: the isosthenuria of Addison'sdisease is the result of derangement of many physio-logical regulatory processes; the inability to concen-trate the urine maximally in Addison's disease is partly


the result of an inability of the collecting duct torespond to ADHwith an increase in water reabsorp-tion; this loss of responsiveness is the result of a directeffect of adrenal steroids on the epithelium of the col-lecting duct and partly involves a diminished rate ofcyclic AMPaccumulation.

ACKNOWLEDGMENTS

The authors gratefully acknowledge the excellent technicalassistance of Sharon Bracey, Susan Corona, and JamesHarrison.

This work was supported in part by National Institute ofArthritis and Metabolic Diseases Research grant 1-R01-AM-14677 and U. S. Public Health Service Training grant5-T32-AM-07257.

REFERENCES

1. Robinson, F. J., M. H. Power, and E. J. Kepler. 1941. Twonew procedures to assist in the recognition and exclusionof Addison's disease: a preliminary report. Proc. StaffMeet. Mayo Clin. 16: 577-586.

2. Boykin, J., A. deTorrent6, A. Erickson, G. Robertson, andR. W. Schrier. 1978. Role of plasma vasopressin in im-paired water excretion of glucocorticoid deficiency.

J. Clin. Invest. 62: 738-744.3. Green, H. H., A. R. Harrington, and H. Valtin. 1970. On

the role of antidiuretic hormone in the inhibition of acutewater diuresis in adrenal insufficiency and the effects ofgluco- and mineralocorticoids in reversing the inhibition.

J. Clin. Invest. 49: 1724-1736.4. Willson, D. M., and F. W. Sunderman. 1939. Studies in

serum electrolytes. XII. The effect of water restriction ina patient with Addison's disease receiving sodium chlor-ide. J. Clin. Invest. 18: 35-43.

5. Stokes, J. B., and J. P. Kokko. 1977. Inhibition of sodiumtransport by prostaglandin E2 across the isolated, per-fused rabbit collecting tubule. J. Clin. Invest. 59:1099-1104.

6. Schafer, J. A., and T. E. Andreoli. 1972. The effect ofantidiuretic hormone on solute flows in mammalian col-lecting tubules. J. Clin. Invest. 51: 1279-1286.

7. Grantham, J. J., and M. B. Burg. 1966. Effect of vasopres-sin and cyclic AMP on the permeability of isolatedcollecting ducts. Am. J. Physiol. 211: 255-259.

8. Fine, L., D. Schlondorff, W. Trizna, R. M. Gilbert, andN. S. Bricker. 1978. Functional profile of the uremicnephron. Impaired water permeability and adenylcyclase responsiveness of the cortical collecting tubuleto vasopressin.J. Clin. Invest. 61: 1519-1527.

9. DuBois, R., A. Vemiory, and M. Abramow. 1976. Com-putation of the osmotic water permeability of perfusedtubule segments. Kidney Int. 10: 478-479.

10. Grantham, J. J., J. M. Irish, and D. A. Hall. 1978. Studiesof isolated renal tubules in vitro. Annu. Rev. Physiol.40: 249-277.

11. Bockaert, J., C. Roy, R. Rajerison, and S. Jard. 1973. Spe-cific binding of 3H-lysine-vasopressin to pig kidneyplasma membranes. J. Biol. Chem. 248: 5922-5931.

12. Grantham, J. J., C. E. Ganote, M. B. Burg, and J. Orloff.

1969. Paths of transtubular water flow in isolated renalcollecting tubules. J. Cell. Biol. 41: 562-576.

13. Cooke, R. C., and R. W. Steenburg. 1973. Effects ofaldosterone and cortisol on the renal concentratingmechanism. J. Lab. Clin. Med. 82: 784-792.

14. Sigler, M. H., J. N. Forrest, and J. R. Elkington. 1965.Renal concentrating ability in the adrenalectomized rat.Clin. Sci. (Oxf.). 28: 29-37.

15. Law, P. Y., and I. S. Edelman. 1978. Induction of citratesynthase by aldosterone in the rat kidney.J. Membr. Biol.41: 41-64.

16. Marver, D., and M. J. Schwartz. 1980. Identification ofmineralocorticoid target sites in the isolated rabbitcortical nephron. Proc. Natl. Acad. Sci. U. S. A. 77:In press.

17. Stolte, H., J. P. Brecht, M. Wiederholt, and K. Hierholzer.1968. Einflub von adrenalektomie und glucocorticoidenauf die wasserpermeabilitat corticaler nephronabschnitteder ratteniere. Pfluegers Arch. 299: 99-127.

18. Ingle, D. J. 1952. The role of the adrenal cortex in homeo-stasis. J. Endocrinol. 8: XXII.

19. Hall, D. A., L. D. Barnes, and T. P. Dousa. 1977. CyclicAMP in action of antidiuretic hormone: effects ofexogenous cyclic AMPand its new analogue. Am. J.Physiol. 232: F368-F376.

20. Stoff, J. S., J. S. Handler, and J. Orloff. 1972. The effect ofaldosterone on the accumulation of adenosine 3',5'-cyclic monophosphate in toad bladder epithelial cells inresponse to vasopressin and theophylline. Proc. Natl.Acad. Sci. U. S. A. 69: 805-808.

21. Stoff, J. S., J. S. Handler, A. S. Preston, and J. Orloff. 1973.The effect of aldosterone on cyclic nucleotide phospho-diesterase activity in toad urinary bladder. Life Sci. 13:545-552.

22. Muneyama, K., R. J. Bauer, D. A. Shuman, R. K. Robins,and L. N. Simon. 1971. Chemical synthesis and biologicalactivity of 8-substituted adenosine 3',5'-cyclic monophos-phate derivatives. Biochemistry. 10: 2390-2395.

23. Senft, G., S. Schultz, K. Munskee, and M. Hoffinann. 1968.Effects of glucocorticoids and insulin on 3',5'-adeno-sine monophosphate phosphodiesterase activity inadrenalectomized rats. Diabetologia. 4: 330-335.

24. Zusman, R. M., H. R. Keiser, and J. S. Handler. 1978.Effect of adrenal steroids on vasopressin-stimulated PGEsynthesis and water flow.Am.J. Physiol. 234: F532-F540.

25. Grantham, J. J., and J. Orloff. 1968. Effect of prosta-glandin El on the permeability response of the isolatedcollecting tubule to vasopressin, adenosine 3',5'-mono-phosphate and theophylline. J. Clin. Invest. 47: 1154-1161.

26. Beck, N. P., T. Kaneko, U. Zor, J. B. Field, and B. B. Davis.1971. Effects of vasopressin and prostaglandin E, on theadenyl cyclase-cyclic 3',5'-adenosine monophosphatesystem on the renal medulla of the rat. J. Clin. Invest.50: 2461-2465.

27. Zusman, R. M., and H. R. Keiser. 1977. Prostaglandinbiosynthesis by rabbit renomedullary interstitial cells intissue culture. J. Clin. Invest. 60: 215-223.

28. Bohman, S. 0. 1977. Demonstration of prostaglandinsynthesis in collecting duct cells and other cell types ofthe rabbit renal medulla. Prostaglandins. 14: 729-744.

29. Smith, W. L., and T. G. Bell. 1978. Immunohistochemicallocalization of the prostaglandin-forming cyclooxygenasein renal cortex. Am. J. Physiol. 235: F451-F457.


urinary concentrating defect of adrenal...

Documents