Renal tubular responsiveness to atrial natriuretic peptide in sodium retaining chronic caval dogs A possible role for kinins and luminal actions of the peptide

Renal tubular responsiveness to atrial

natriuretic peptide in sodium-retaining chronic

caval dogs. A possible role for kinins and

luminal actions of the peptide.

L Legault, … , E Maher, D Farber

J Clin Invest. 1992;90(4):1425-1435. https://doi.org/10.1172/JCI116009.

60% of chronic caval dogs with ascites did not respond to atrial natriuretic peptide (ANP) (75 ng.kg-1.min-1) with a natriuresis (TIVC-NR; delta UNaV = 2 +/- 0.8 mu eq/min) whereas the remaining 40% responded normally (TIVC-R; delta UNaV = 216 +/- 50 mu eq/min). Since proximal tubule neutral endopeptidase 24:11 (NEP) destroys most of intrarenal luminal ANP and kinins, we attempted to convert TIVC-NR into TIVC-R by providing NEP inhibition with SQ 28603 at 30 mg/kg. This potent and specific NEP inhibitor produced a natriuresis when administered alone to nine TIVC-NR dogs (delta UNaV = 67 +/- 2 mu eq/min) and permitted a natriuresis in the presence of ANP (delta UNaV = 97 +/- 18 mu eq/min). A natriuretic response to ANP could also be induced in TIVC-NR dogs by providing renal arterial bradykinin or intravenous captopril, a kininase inhibitor. Urodilatin, a natriuretic peptide not destroyed by intrarenal NEP was without effect in TIVC-NR dogs but increased UNaV when given to TIVC-R and normal dogs. Providing bradykinin to TIVC-NR now permitted an increment in delta UNaV (62 mu eq/min) when urodilatin was reinfused. TIVC-R dogs could be converted into TIVC-NTIVC-R by pretreating with a specific bradykinin

antagonist before infusing ANP. We conclude that TIVC-NR dogs are deficient in intrarenal kinins but are converted to responding dogs after NEP inhibition because […]

Research Article

Find the latest version:

Renal Tubular Responsiveness

to Atrial Natriuretic

Peptide

in

Sodium-retaining

Chronic Caval

Dogs

APossible Role for Kinins and Luminal Actionsof thePeptide

Louis Legault, Peter Cemacek, Mortimer Levy,ElizabethMaher,andDavidFarber

Departments ofPhysiologyandMedicine,McGillUniversity, and the DivisionsofNephrologyand ClinicalBiochemistry, RoyalVictoriaHospital, Montreal, Quebec, Canada H3GI Y6

Abstract

60% of chroniccavaldogswith ascites didnotrespondtoatrial natriuretic peptide(ANP)(75 ng-

kg-'

*min-')witha

natri-uresis(TIVC-NR;AUNaV=2±0.8

,geq/min)

whereas the

re-maining 40% responded normally(TIVC-R;

AUNaV

=216±50

Ateq/min).

Since proximal tubule neutralendopeptidase24:11 (NEP) destroysmostof intrarenal luminalANP andkinins,we

attemptedto convertTIVC-NR into TIVC-RbyprovidingNEP inhibition withSQ 28603at30mg/kg.Thispotentand specific

NEPinhibitor producedanatriuresiswhenadministered alone

tonineTIVC-NR dogs

_(AUN.V

=67±2

geq/min)

and

permit-ted a natriuresis in the presence ofANP

(AUNaV

= 97±18

leq/min).

A natriureticresponse to ANP could also be

in-duced inTIVC-NR dogs byprovidingrenalarterialbradykinin

orintravenouscaptopril,akininaseinhibitor.Urodilatin,a

na-triuretic peptidenotdestroyed by intrarenalNEPwaswithout effect in TIVC-NR dogs but increased UNV when given to

TIVC-R and normal dogs.Providing bradykinintoTIVC-NR

now permitted an increment in

_AUN.V

(62 ,eq/min) when urodilatinwasreinfused. TIVC-Rdogscould beconvertedinto TIVC-NR by pretreating withaspecific bradykinin antagonist

before infusing ANP. We conclude that TIVC-NR dogs are

deficient in intrarenal kinins butareconverted toresponding

dogs afterNEPinhibition because of increased kinindelivery

totheinner medullary collectingduct. (J. Clin. Invest. 1992.

90:1425-1435.) Key words: neutral endopeptidase* ascites

-sodium excretion*cGMP

Introduction

In ourlaboratory, all normal dogs uniformly respond with a brisk natriuresis to an infusion of atrial natriuretic peptide

This work was presented in part as a poster at the Annual American Society ofNephrology Meeting, Baltimore, MD, 17-20 November 1991, and appeared in abstract form ( 1991. J. Am. Soc. Nephrol. 2:406. [Abstr.]).

Address correspondence to Dr. Mortimer Levy, Department of Physiology,Room 1228, McGill University, 3655 Drummond Street, Montreal, Quebec, Canada H3G 1 Y6. Dr. Legault's present address is Nephrology Division,H6pital St. Luc, and the Department of Medi-cine, UniversitedeMontreal,Montreal, Quebec.

Receivedfor publication 18February1992 and inrevisedform1 May 1992.

(ANP),' whereas edematous dogs usually show a

heteroge-neousresponse. About half of allsodium-retainingchronic

ca-val(TIVC) or cirrhotic dogswith ascites will respondto an ANPinfusionwithabrisk natriuresis whereas theremaining

half will shownonatriuresiswhatsoever ( 1, 2).Thisspectrum

ofheterogeneousresponse has also beenreportedfor cirrhotic rats(3),cirrhoticmen(4),andsodium-retainingbile

duct-li-gateddogswithoutascites(5).Suchaheterogeneousresponse to ANP appears to be unique to this peptide, since we have

demonstrated that TIVC "nonresponders" willrespond

nor-mally to various diuretics active in different nephron seg-ments(6).

Thelack oftubularsensitivitytoANP in - 50% of

sodium-retainingTIVCdogswith ascites appears to be transient and

therefore functionalin nature, since when thesenonresponders

return tosodiumbalance (though maintainingtheir ascites),a normal natriuretic response to ANP reappears ( 1 ). Despite intensive investigation (1, 2, 7)wehavesofarbeen unableto detectphysiologicalvariables thatdiscriminate between natri-ureticrespondersandnonrespondersamong TIVCorcirrhotic

dogs. All of these experimental animals, no matter their re-sponsetoANP, appearphysiologicallyequivalent. Moreover,

in a recent series of experiments (8), wedeliberately attempted

toconvertTIVC-respondingdogs into nonresponders and vice versaby manipulating theintrarenal environment(e.g.,

cate-cholamineand angiotensininfusions,angiotensin and

adrener-gic blockade, adenosine receptor antagonism, etc.) with variouspharmacological infusions calculatedtoeither attenu-ate or promote ANP effects. These attempts proved unsuc-cessful.

Neutralendopeptidase24:1 l(NEP24:1 1 )within the brush

border of theproximal convoluted tubulesserves as a major

routefordegradation offiltered ANP (9). Recently,evidence

has been adduced thatendopeptidase inhibitionmay promote anatriuresis in animal models of sodium retentionshowingan attenuated response to the natriuretic effects of ANP (1O, 11).

In the present study, we haveinvestigatedthepossibleroleof

excessive endopeptidase degradation of ANP as a possible

causefor the tubularinsensitivity tothispeptidein a popula-tion ofTIVC dogswith urinary sodium retention andascites

unresponsive topharmacological infusionsofthis potent natri-uretic agent. Because kinins may also be degraded by this

en-1.Abbreviationsused inthis paper: ABP, arterial blood pressure; ANP, atrial natriuretic peptide;BKA,bradykininreceptorantagonist; CVP, central venous pressure; FE, fractional excretion; iANP, immunoreac-tiveANP;IMCD,inner medullary collecting duct; NEP,neutral endo-peptidase;NEP 24:1 1, neutralendopeptidase24:1 1; PAH, para-amin-ohippurate; TIVC, chroniccavaldog; NR, not responding;R, respond-ing; UNaV, urinary sodium excretion.

J.Clin. Invest.

C) The American Society for Clinical Investigation, Inc. 002 1-9738/92/10/1425/1 1 $2.00

zymesystem( 12),weexaminedthe roleofkinin availability as apossible modulator ofthenatriureticeffects ofANP.

Methods

Atotalof47chronic conditioned dogs of either sex survived the proce-dureofthoracotomy and partial constriction of the supradiaphragma-tic vena cava. These dogs were studied both in the alert, unanesthetized state oranesthetized, depending on the experimental protocols. Studies were also performed on 20 acute, unconditioned dogs of either sex selectedforgood health. They were also studied in the anesthetized or unanesthetized state as required. The preparation ofthe TIVC dogs was carriedoutundersterile conditions, usingsodium thiopentone intrave-nousanaesthetic (20 mg/kg) as previously described in great detail for this laboratory ( 13). The animals were fully recovered, mobile, and eatingby the first postoperative day. The dogs received a standard chow dietcontaining45 meq Na+/d. All postoperative care was supervised bysenior veterinariansfrom the McGill Animal Resources Centre, and all components of the study protocols received approval from the Uni-versity Animal Utilization Ethics Committee. These dogs developed detectableascites usuallywithin5-9 d and were studied shortly thereaf-ter.Wheneither TIVC or acute dogs were studied standing quietly in a Pavlovsling, the following procedures were used: All dogs were sedated with an atropine-xylazine mixture given intramuscularly (5). Urine wascollected by a standard washout technique through a Foley cath-eterinserted into the bladder. All infusions were given intravenously through polyethylene catheters placed by directvenipuncture in the saphenous or antecubital veins. Inulin and para-aminohippurate (PAH) weregivenat0.5 ml/minthrough a PE50 catheter placed in oneantecubitalveintomeasureGFR and renal plasma flow, respec-tively. Allinfusionsweregivenwith constant-speed infusionpumps. Clearanceperiodswere 2 10 min in duration and were performed by theconstant-infusion technique(13).Bloodwassampledatthe mid-pointofeachperiodfromaPE190or205catheterplacedin the abdom-inal venacavafromasaphenousvenipuncture.Atleast three clearance periodsweretaken ineachexperimental phaseandaveraged.Where arterial blood pressure(ABP)and centralvenouspressure(CVP)were

tobemeasured, appropriatecatheterswereplacedin therightcarotid artery andjugular veinseveraldaysbefore theexperimentunder thio-pentoneanesthesiaandprotectedwithafelt collar aftersubcutaneous tunneling.ABP wasmeasuredbymercury manometry and CVPwas

measured by saline manometry, with thezeroreferencepointin each casefixedatheartlevel.

Whendogswerestudiedin theanaesthetizedstate,sodium pento-barbital(25mg/kg, intravenously)wasused for induction with small supplementalamountsgivenasrequiredthroughoutthestudy.Urine wascollectedthroughaFoley catheter,except for those studies where theleftrenal arterywasselectivelyperfused,in whichcasetheureters

werecannulateddirectlythrougharetroperitoneal approachtoavoid spillingascites. In normaldogswithoutascites,theureterswere

cannu-latedthroughalowerabdominalmidline incision. Blood for inulin and PAHlevelsweretaken fromanarterial cannulaplacedinonefemoral artery.Anesthetized dogswereintubatedand ventilated withaHarvard DogRespirator (HarvardApparatus, South NatickMA).When the left kidneywas tobeperfuseda26-gcurved needlewasinserted into the arterybytechniquespreviouslydescribed and held inplacewithseveral drops of adhesive(Permabond 910;PermabondInternational,

Engle-wood, NJ) ( 13).

Severalprotocolswereused in thesestudies,asfollows:

(a) Nine normaldogswerestudiedstandingquietlyinaPavlov sling.After control clearanceperiods,ANPwasinfusedat75ng/kgper min intravenous. Aftera 10-minwaiting period,anothersetof clear-anceperiodswerecollected. Plasma for inulin andPAHwerecollected atthemidpointof each urine collectionwhereasplasmaforANPlevels wascarefully collectedaspreviously described ( 1, 2)atthemidpointof eachexperimental phase.Aftera90-min recoveryperiod,repeat

clear-ancecollectionswere_again taken, andtheNEP 24:1 1 inhibitorSQ

28603was injectedin adose of 30 mg/kgintravenous in 20 ml of

isotonic NaHCO3 (provided by Bristol-Myers Squibb Research Insti-tute[Princeton, NJ];N-2-mercaptomethyl-I-oxo-3-phenylpropylB al-anine [SQ 28603] is a highly specific antagonist of NEP 24:11 and is thought to have only weak actions on other peptidases contained within the proximal convoluted tubule brush border). After a 10-min wait, clearance studies were repeated. ANP was then reinfused at the previous dose level and, after another10-minwaiting period, a final set of threeurinecollections were made.

(b) This protocol was also followed in nine TIVC dogs who were deemed nonresponders to an infusion of ANP

_(AUN.V

< 20

Aeq/min).

(c) The protocol was also followed for seven TIVC responders who showed a brisk natriuresis to ANP (AUNaV > 20

,eq/min).

(d) The following protocols were employed in additional groups of responding TIVC dogs:(i) After the initial ANP infusion and a 90-min recovery period, aprotinin was given intravenously as a bolus of 50,000 kallikrein inhibitor units (KIU) and then infused as a constant infusion of 10,000 KIU/min to blunt the generation of intrarenal kinins. Aftera 10-min wait, clearances were taken, ANP was then reinfused at pre-vious dose levels and after another 10-minwaiting period, another set of clearance periods were taken (n=5).(ii) A similar protocol as in (i), but aspecific antagonist of the bradykinin receptor(D-Argo, Hyp3, thi5, D-Phe7, thi8)bradykinin (BKA) (IAF Biochem International, Inc., Montreal, Canada) was administered into the left renal artery at 15

_,g/kg

permin after the initial ANP infusion. In separate pilot stud-ies, it was determined that this dose administered to three normal dogs prevented the increased renal plasma flow and natriuresis of brady-kinin (3 _gg/kg per min) given into the left renal artery. The right kidney was used as a control. After a set of clearances, a second ANP infusion was given intravenously while the BKA was still being infused into the left renal artery (n =4).

(e) In additional groups of nonresponding TIVC dogs, the follow-ing protocols were employed:(i) In four TIVC dogs unresponsive to an initial intravenous infusion of ANP, urodilatin was infused at 100 ng/ kgpermin i.v. after a 1-hrecovery period. After a10-minwait, three clearance collections were taken. In an additional three dogs, the intra-venous urodilatin was administered along with bradykinin given into theleft renal artery at 3_Ag/kgpermin.(ii) Four TIVC nonresponders studied in the anesthetized state were given the initial intravenous ANP infusion of 75 ng/kg per min while isotonic saline at 0.5ml/minwas infused through the left renal arterial catheter. After a 1-h recovery period, the renal arterial infusion was switched to bradykinin 3pg/kg

per mindelivered at 0.5 ml/min.After obtaining three urine collec-tions after a 10-minwaiting period, the intravenous ANP infusion was readministered concurrent with the bradykinin. After a10-min stabili-zation period, another set of urine collections were obtained. (iii) In four nonresponding TIVC dogs, bradykinin at 3 gg/kgpermin was administered into the left renal artery after an intravenous infusion of 8-Br-cGMP ( 16_jg/kgper min) had failed to initiateanatriuretic re-sponse. Although the renal arterial bradykinin was being infused, this second messenger analogue for ANPwasreadministered intravenously in an identical dose. (iv) In four nonresponding TIVC dogs we admin-istered captopril intravenously 20

Atg/kg

per min.This dose has been shown to prevent angiotensin II generation ( 14). After three clearance periods,ANP wasreadministered concurrent with the captopril anda second set of clearances taken.Inthesestudies, the captopril was being used as an inhibitor of intrarenal kininaseII.

Inadditional groups of normal dogs, serving as controls, we admin-istered urodilatin 100 ng/kg per min (n =_{3), captopril, and ANP as}

above (n= _3).

Inulin was measured by an anthrone technique and PAH was mea-sured byanautoanalyzer technique.Thesemethodshavepreviously beendescribed in detail for this laboratory ( 13). Sodiumin theurine and plasma was measured by flame photometry ( 13). Plasma protein and hematocrit were measured by techniques previously described ( 13).ImmunoreactiveANP wasanalyzed in plasma and urine by tech-niquespreviouslydescribed in great detail( 1,2).Extraction ofsamples

fluoracetic acid elution. Recoveryofsynthetic human ANP (1-28)

addedtoplasmawas 75%. Intra-assaycoefficient ofvariationwasfrom 6to11%,dependingonlevels ofhormonebeing assayed, and interas-saycoefficientof variationwasbetween10 and 11%.

Statisticalsignificancewastaken at the 5%probabilitylevel. Group meansbetweenexperimentalgroupswereanalyzedby theunpairedt test, andwithin groupsvalues wereevaluated withatwo-wayanalysis

of variance forrepeatedmeasuresor apairedttest asrequired. Dataare presentedasmean±SE.

Results

Atotalof47 TIVC dogswereexaminedinthisstudy.Ofthese, 28 were identified as natriuretic nonresponders to ANP

(AUN.V

<20

,ueq/min)

and 19were identified asnatriuretic

responders

(AUNYV

220,ueq/min).In addition, westudied 20 normalcontrol animalsfedasimilar saltdiet. Fig. 1 summa-rizes the spectrum of natriuretic responseto ANP(75 ng/kg

per minintravenously) in all threepopulationsofdogs. The

profileof AUNaV was similar forboth controland TIVC

re-sponding dogs whereas the TIVC nonresponders averaged

an increment in urinary sodium excretion of only 2±0.8

,ueq/min.

Table Isummarizessomefeaturesof these TIVCdogsafter

they haddeveloped urinarysodiumretention and ascites some

5-9 d aftersurgery. All dogswere ingood health witheasily detectable volumes ofascites. The dogs ineachTIVCgroup

werequite similar, differingonly in the natriuretic responseto

ANP and inbaselineplasma iANP levels. The observationof a

lowerbaselinelevel for plasma iANP in TIVCnonresponders

has been aninconsistentfinding inour laboratory(1, 2,5-7), and willbecommenteduponfurther in theDiscussionsection. Endopeptidase inhibition. TableIIsummarizestheeffects of administering endopeptidaseinhibitorto nine normaldogs. When ANP alonewasinfused, the subsequentnatriuresiswas

* (655)

350

300

250

AUNAV

gEq/min

200

150-

100-50

-0

(N=20)

Normal

(N=28) TIVC (NR)

* ₍₇₂₇₎ *₍₅₁₉₎ *₍₃₇₁₎

*

*0

(N=19)

TIVC(R)

Figure1. Thechange in urinary sodium excretion

(AUN.V)

from control levels afteranintravenous infusion of ANP (75ng/kgper min) in normaldogs,caval dogs unresponsiveto ANP [ TIVC(NR)], and caval dogs responsiveto ANP [TIVC(R)].Natriuretic re-sponderswerethosewho had a

_AUN.V

> 20

,eq/min.

TableI. _{BaselineData in TIVC Dogs}Comparing Natriuretic

Responders(n= _19)toNonresponders (n =₂₈₎

Responders Nonresponders

Bodyweight (kg) 16.7±1.3 16.7±1.1

ABP(mmHg)

106±8

104±7

CVP(cm

H20)

2.1±0.9

2.9±0.6

GFR

(ml/min) 45±2.9 47±2.7

CpAH

(ml/min)

124±8

119±6

V* (ml/min) 2.6±0.13

2.3±0.11

U[Nal+I

(meq/liter) 8±2

12±3

Baseline

UNaV1I

(,geq/min)

22±7

26±4

AUN.V

(post-ANP infusion)

(;teq/min)

211±50

3±0.8*

Hematocrit(%) 41.7±2.4

45.6±1.8

Plasma iANP(pg/ml) 91±21

38±12*

Postinfusionplasma

iANP(pg/ml)

1,458±123

1,853+312

*_{P<0.05. $ V,urine flowrate.} §_{UNa, urinary sodium} concentra-tion. 11UNaV, urinary sodiumexcretion.

associated with a small but

significant

increment in urinary

ANPexcretion,butamajor reductionin thefractional

excre-tion ofthe

peptide.

Thiswasdueto anenormous

capacity

to

eitherdegradethefiltered

peptide

orbindittosilent receptors. Indeed, ifoneassumes that thecirculating peptidewasfreely filtered, only 30pg/min ofafiltered load of138,780 pg/min escaped

degradation

withinthekidney. (Wewillusetheterm

"degrade"'

withinthetext as asingle description for

enzymatic

degradation

andbindingtosilent C receptors.Thoughdataare

lacking concerningANPtubular

reabsorption

orsecretion,the

equation

usedin TableII

[degraded

ANP load=_{filtered load}

-_{excreted load]seems a}reasonable assumption.Totheextent

there istubular secretionorreabsorption ofthepeptideis the

extentthis

equation

isan

approximation

and would besimilar

toanetclearance valueencompassingboth thesetubular

pro-cesses.)Theinfusion ofthe NEPinhibitorwasassociated witha

modest increment in

UN.V,

unassociatedwithanychange in ABP or GFR.

Plasma

levels of iANP increased slightly but

significantlyandtheurinary excretion ofANProsemarkedly. Thefractional excretion ofthepeptidenowrosedramatically

frombaselinelevelsasANPdegradation declined.When ANP

was reinfused, the natriuretic response was magnified

com-paredwith the initialnatriuresis. TheAfor

_UN.V

initiallywas

234±63

,teq/min,

whereas for the second ANP infusic ithe

change(comparedwith recoveryphase)was

352±39

teq,

min

(P<0.05). ThismagnificationoftheANP-induced

Atriuresis

wasassociated withanincrementin GFR not

difi

-ent from

the initial AGFR (6vs. 10 ml/min, NS). Plasma iANP levels

more thandoubled(+129%) comparedwith the firstinfusion andtheurinary excretion ofANP rosedramatically.Fractional excretion (FE) ofANP increased still further from the NEP

inhibitor only phaseby

0.7%.

ThedegradedANPloadrose in

absoluteterms, butfellinfractional terms comparedwith the

initialANPinfusion.

In fiveseparate normal control dogs, studied aspart ofa

separate protocol,theeffect of vehicle for the NEPinhibitor, i.e.,20 mlof isotonic NaHCO3,wasgivento test theeffecton

00

0.011

Table 11.Endopeptidase Inhibitionin Nine ControlDogs

Control ANP, Recovery SQ28603* SQ28603+ANP2

V(ml/min) 2.5±0.11 4.0±0.42- 3.2±0.40 3.4±0.43 6.2±0.63'

UNaV

(Aeq/min)

44±15 278±72§ 45±17

94±32"1

397±50'

GFR(ml/min) 54±5.8 60±6.9§ 51±5 50±6 61±7'

CpAH (ml/min)

145±14 168±18§ 128±11 117±6 145±13'

ABP(mmHg) 141±5 124±6§ 131±5 134±4.6 121±6'

Hematocrit(%) 50±1.5 56±2' 52±2 52±1.2 55±1.5'

Plasma iANP(pg/ml) 49±7 2,313±341' 75±9.7 126±1911 5,300±8391

UrineiANP(pg/ml) 2.7±0.6 7.4±1§ 3±0.3 80±1811 2,668±484'

UVANP (pg/min)

6.6±3.0 30±10' 9.6±2 274±2411

_16,382±250'

FE ANP(%)t 0.25±0.09 0.02±0.009§ 0.25±0.10 4.40±0.711 5.1±0.6'

Degraded ANP load(pg/min) 2,640+420 138,750±18,000§ 3,815±994 6,026±744"1 306,918±82,680' Forabbreviationssee textand Table I. *SQ 28603 = NEP24:11inhibitor. *F.E. =fractional excretion. §P<0.05 compared withcontrol phase. 11P<0.05compared withrecoveryphase. '_{P<0.05compared with SQ28603 phase. Degradedload=filtered}load-excreted load.

UNaV. This infusionwaswithouteffect(36±8vs. 39±11 ueq/

min,

NS).A similar lack of effectwas observed inthree

re-sponding and nonrere-sponding TIVC dogs, respectively.

NonrespondingTIVCdogs.TheNEPinhibitorwas admin-isteredtonine nonresponding TIVC dogs with sodium

reten-tionand ascites, and these dataare summarized in TableIII.

Theprovision of the Squibb compound caused a significant natriuresis in all dogsthat was unassociated withany

incre-mentin GFR, renal perfusion, ABP,orplasma levels of iANP. Though urinary excretion ofANPtendedto rise, this change

wasnotsignificantnor wastheincrement inFEofANP.The

riseinurinary excretion ofANP maybephysiologically

signifi-cant,however,since therateofurinary excretion ofANP dur-ing theinfusion ofthe NEP inhibitorjustbarely escaped statis-ticalsignificance.WhenANP wasreinfused in thepresenceof

NEPinhibition,UNaV increasedstillfurthersothatAUNaV of 96±18,ueq/ min(comparedwithrecoveryphase)clearly

repre-sentedasignificant natriureticeffect whencomparedwith the initialresponse(AUNV=3+0.8

,ueq/min,

P<0.05). Plasma levelsofiANP increasedby 201% compared with the initial infusion, and thefractionalexcretion ofANPincreased

dramat-icallyto 17.7±7.9%.

RespondingTIVCdogs.Table IVsummarizesthe data for

therespondingTIVC dogs. Provision ofNEPinhibition also

spontaneously increased UNaV by - 46±7 ueq/min (P

<0.05) and also magnified the ANP-induced natriuretic re-sponse(AUNaV = _280±43 compared with 211±50

_,eq/min

initially,P <0.05).Themodest natriuretic effect of NEP

inhi-bition alone occurred witha significant incrementin plasma

iANP and with significant increments intheurinary excretion

of thispeptide. When ANP wasreinfused in the presenceof

NEPinhibition, plasma levels increasedby 287% and FE ofthe

peptideincreasedto9±2.6%.

Nonresponding TIVC dogs: further studies. Because of the putative role that NEP 24:11 may play in kinin catabolism

(12),weexaminedthepossible role of kinins inthenatriuretic

response toANP. Fig. 2 summarizesthe dataobtained from fournonrespondingdogs when bradykininwasinfused intothe

left renal artery before a reinfusion ofANP. Though

brady-kinin tendedtoincrease urine flowandUNaV in theinfused kidney,thechangeswere notsignificant. WhenANP was rein-fused in thepresence of thekinin, the nonrespondingTIVC dogwasconverted intoaresponder(AUNV = 114+1.9,ueq/

mincompared with recoveryphase [P <

0.05]).

Theinitial

TableIII.Endopeptidase Inhibitionin NineNonrespondingTIVC Dogs

Control ANP, Recovery SQ28603* SQ28603+ANP2

V(ml/min) 2.28±0.11 2.39±0.14 2.20±0.04 2.37±0.11 2.47±0.11

UNaV

(Aeq/min)

2.6±0.37 4.4±0.08 3.3±0.08 70±17* 100±8§

GFR(ml/min) 47±2.7 51±4.7 40±5.7 42±6.3 39±4.9

CPAH

(ml/min) 119±6 141±20 120±13 123±11 113±9

ABP(mmHg) 104±7 92±8* 95±6 92±9 88±7

Hematocrit(%) 45.6±1.8 48±2.1 45±3 45±5 46.6±4

Plasma iANP(pg/ml) 38±11.9 1,853±312* 33±6.3 40±7.2 5,581±1,588§

UrineiANP(pg/ml) 6.5±1.9 13.5±4.2* 8±3.3 25±10* 13,857+6,275§

UVANP(pg/min) 16.6±5 35.6±12* 20±9 59±22

_{38,022+18,555§}

FE ANP(%)* 2.3±0.6 0.04±0.01* 2.1±0.8 7.9±4.3 17.7±7.9§

FilteredANPload(pg/min) 1,284+458 103,545±31,180* 1,323+269 1,469+200 217,000±86,323§

DegradedANP(pg/min) 1,267±460 103,510±31,176* 1,305+266 1,410+215 202,600±85,554§

Forabbreviationssee textand Table I. *P<0.05compared withrecoveryphase. *P<0.05comparedwith controlphase. §P<0.05

com-paredwithSQ28603 phase.

Table IV.EndopeptidaseInhibitioninSevenResponding TIVC Dogs

Control ANP, Recovery SQ28603* SQ28603+ANP2

V_(ml/min) 2.6±0.13 4.2±0.4* 3.5±0.3 3.4±0.5 4.8±0.4§

UNaV

(Aeq/min)

22±7 238±54* 21±7 68±19* 305±116§

GFR(ml/min) 45±2.9 54±8.7* 44±6 49±7.2 57±9.5§

CpAH (ml/min)

124±8 135±17 119±13 121±13 136±18

ABP(mmHg) 106±8 94±6* 103±7 104±5 94±8

Hematocrit(%) 41.7±2.4 48±2.1* 49.7±2.6 48.3±2.8 48.4±2.4

Plasma iANP(pg/ml) 91±20.7 1,458±123* 107±22 204±63* 5,645±714§

Urine iANP(pg/mi) 20±10 15±6 7.5±2.5 190±87$ 5,261±1,788§

UVANp (pg/min)

30±16 109±49 23±8 474±277* _{35,299±14,000§}

FE ANP (%)* 0.94±0.6 0.18±0.09 1.13±0.43 3.7±1* 9±2.6§

FilteredANP(pg/mmn) 5,352±948 83,725±17,680* 4,575±1,237 9,802±3,039 327,166±103,403§

Degraded ANP(pg/min) 5,322±957 83,616±17,675* 4,552±1,239 9,328±2,822 291,966±93,302§

Forabbreviationssee textandTable I. *P<0.05compared withcontrolphase. *_P<0.05compared withrecoveryphase. § P<0.05 com-paredwithSQ28603phase.

AUNaV withANPhad been 3.4±3.1

geq/min.

Fortheright

controlkidney,the initial

AUNV

was5+0.4

_,geq/min

and with

theANPreinfusionwas8±0.9

,geq/min.

This valuewasnot

changed from the initialoneandwassignificantly (P<0.05)

less than that observed in the experimental controlateral

kidney.

InfournonrespondingTIVCdogs,weadministered

capto-pril intravenouslyin anattemptto augmentintrarenalkinin

availability. Fig. 3 summarizes these data. Though captopril waswithouteffectonUNaV, GFR,or ABP(though therewas a

tendency forABP todecrease),thereinfusionofANPinthe

presenceof captoprilincreasedUNaV from 3.7±1.3to48±11

,geq/min

(P<0.05). Theprevious AUNaV had been 9

,eq/

min.Thisnatriuresis occurreddespiteabloodpressure value that wassignificantly lessinthis final phase than recordedin theinitialcontrol phase. When administered tothree normal

dogs,captoprilhadnoeffecton

_AUN.V

(280vs.262 ueq/min,

respectively) afteranANPinfusion.

InfournonrespondingTIVC dogs, thenatriureticpeptide urodilatin wasdeliveredintravenouslyat 100ng/kgpermin.

Fig. 4 A summarizesthese data. Urodilatinwasunableto

in-duceanatriuresisin thesedogs, althoughwhengiventothree normaldogsor torespondingTIVC dogs, therewas aprompt andsignificant natriuresis

_(AUNYV

=230±8 and 206±11

Aeq/

min,respectively).

140

-120

-Urinary

100-sodium excretion 80

-pEq/min

60-40

-20-

t---

i

K

Fig. 4 Billustrates in aseparate group ofthree dogs, the effect of

prior

administrationofbradykininto

nonresponding

TIVC

dogs

receiving urodilatin. The kinin now permitsthe urodilatinto inducea natriuresis in the left kidney

_(AUN.V

=₆₃

_ueq/min),

_{whereas the urodilatin reinfusion remains}

without effectonthecontrolright kidney.

Finally, the effect ofkinins onthe renal response tothe

second messenger of ANP (cGMP) was examined in four

nonrespondingTIVCdogswith ascites. These dataaregiven in

Fig. 5. As previously demonstrated (8), in doses that cause markednatriuresisinboth normal and responding TIVCdogs,

8-Br-cGMP is without effect in nonresponding TIVC

dogs.

The administration of bradykinin, however, to one kidney,

nowpermitstheinduction ofanatriureticresponsewitha sec-ond infusion of8-Br-cGMP.

110 -100

-ABP

mmHg901

80-70

-48

-GFR mlmin

46- 44-42

-60

-45

-UNAV gEq/min30

-15

- 0-Control ANP1 Recovery Bradykinin Bradykinin

+ANP2 Figure 2. The natriuretic effect of ANP in four TIVC nonresponders whenbradykinin is being infused into the left renal artery at 3 ug/kg per min. *P<0.05 compared with previous phase. *, Left kidney;o, right kidney.

I/

I

4

tt

+

~

1

Control ANP1 Recovery Captopril

imv. Captopril+ANP2

OP

N=4

®Z

N=3

* Left kidney

0Right kidney

Control ANP1 Recovery Urodilatin Control Urodilatin

*80

70

60

-50

*40

-30

-

₂₀

Recovery Bradyk.Bradyk.

Urodilation

UNAV

j±Eq/min

Figure 4. Urodilatin(100 ng/kgpermin) administered

toTIVC nonresponders without (A)andwith brady-kinin being infused intothe left renalartery(B).

RespondingTIVCdogs: further studies.Tofurthertestthe

hypothesis that kininsmaybe playingarole in the natriuretic

responsetoANPin TIVC dogs,weperformedtwo setsof

stud-ies. Inone, weadministered aprotinintofiveresponders, and inanotherset,weadministeredaspecific bradykinin

antago-nisttofourresponding dogs.

Fig.6 summarizes the data with aprotinin. Although

with-outeffectonsodium excretionorurineflow,aprotinincaused

significant blunting of the natriuretic response (AUNaV

299±69vs. 92±26 ueq/min,P<0.05)when ANPwas

rein-fusedasecondtime. This agentbyitselfwaswithout effecton

GFR, ABP, orrenalplasmaflow.

When theBKAwasadministeredintothe leftrenalartery

of fourTIVCresponders,itwasabletodramaticallyattenuate

the natriuretic responsetoANP in the experimental kidney. The natriuresis of the contralateral right kidney reappeared withanANP reinfusion andwasunchangedtothatobserved

when the dogs were initially exposed to intravenous ANP.

These dataaresummarized in Fig.7.

InthreeseparateTIVCdogsresponsiveto the natriuretic effectsofANP,the BKAwasadministered into the leftrenal

arterybefore intravenously infusing the NEPcompound only. A secondANPinfusionwasnotgiven. The initial natriuretic

responsestoANPhad been AUNaV= 52±5 (left kidney)and

49±4.7

_,geq/min

(right kidney). When theNEPinhibitorwas

administered 90 min after cessation of the ANP infusion, AUNaVfortheright kidneywas31±2.6

Aeq/min,

but inthe left

kidney receiving theBKA,the AUNaV declinedto5±0.9

_,eq/

min compared witharecoveryperiod (NS). Thus, theBKA wasableto attenuatethe natriuresis inducedby NEPinhibition

alone in the absence ofanexogenousANPinfusion.

Finally,weexaminedthepossibility that bradykinin,

capto-pril, aprotinin, or BKA caused increments in

UVANP

in the

dosesemployedwheninfused intonormaldogs.Thesedataare

100

-90

-80

-70

-60

-UNAV

gEq/min

50

40

-30

20

10

O

-*

Control

8

Br-c.

GMP

16

pg/Kg/min

i.v.

Recovery

Bradykinin

Bradykinin

+

8 Br-c. GMP

i.v.

Figure 5. The effect of

pro-viding bradykinininto the

leftrenal arteryof four TIVC

nonrespondersreceiving

in-travenous8-Br-cGMP.*P

<0.05comparedwith

pre-viousphase. *,Leftkidney;

o,right kidney.

1430 L.Legault,P.Cernacek,M._Levy,E.Maher,andD. Farber

UNAV

4

-1

E _{30 l}

U ₂₃

0' 400

~.300

0

fLx

200

E

gW100

0

Control ANP Recovery Aprotinin Aprotinin +ANP2

Figure 6. Theeffectofaprotininonfive TIVCdogsresponsivetothe natriureticeffectsofANP.Aprotininattenuatestheincrement in urineflowandsodium excretionpreviouslyobserved with ANP. *P <0.05comparedwithprevious phase.

giveninTable V. Noneoftheseagents isresponsible for aug-mentedANPdeliverytothe inner medullarycollectingducts (IMCD).

Discussion

In recent studies, including the present investigation, con-ductedoveraninterval of several years ( 1, 2, 5-8), this labora-tory has consistently demonstratedaheterogeneous natriuretic response to infusionsofpharmacological doses of ANP (50-175 ng/kg per min) administered to sodium-retaining dogs with ascites. Approximately half of the animals respond with anincrement in urinarysodiumexcretion notdifferentfrom normalcontrols whereas the remaining half fail to demonstrate any natriuretic response. Because all of our studies, including

160

-140

-UNAV

,uEq/min

120

-100

-the present one,have demonstratednodifference in postinfu-sion plasma levels ofANP(1,2) and becauserecentstudies in suspensions of IMCD cells isolated and prepared from TIVC

responders and nonresponders have failedtodemonstrate dif-ferences inANPreceptor density andaffinityand in ANP-in-duced generation ofcGMP(8), it seemsreasonable to

con-clude that the absent natriuretic response to ANPin

nonre-sponders cannot be due to deficiencies in ANPavailability, binding,orsecond messengergenerationatbasolateral recep-torsites.

Lack of response to ANP in TIVC nonresponders could

thereforebe due to(i)biologicallyinactive

peptide,

(ii)a

re-ductionin sodiumdeliverytotheIMCD, (iii) inhibitory effects ofneurohumoral stimuli overriding the natriureticeffects of

ANP,

(iv)

aproblem with post-cGMPsignal transduction, or (v)differences in the luminal delivery ofANPand/orother

factorsto the IMCD. Inthis regard,Gerbes et al. (15) have recently demonstrated in cirrhoticrats that clearance "C"

re-ceptors involved in degradation of ANP are markedly in-creasedwithin theglomerulus comparedwith controls. They

speculated thatasaresultdiminishedtubulardelivery ofANP

could be afactorproducing insensitivitytothepeptideinthis

experimental model.

Since identicallotsofpeptidewereemployed forboth

re-spondingandunrespondingTIVCdogs,the firstpossibilitycan

beeliminated. Moreover, in the presentstudy, ANPgivento

nonresponders could cause ahypotensive effect whilenot

pro-ducing a natriuretic effect (Table III), and in other studies

couldraiseGFR while notaugmenting urinary sodium

excre-tion ( 1, 2). We are thus dealing with a true dissociation of biological effects.

Though micropuncture studies have not been performed in anyofour studies, we have demonstrated that filtered sodium load, 24-h sodium excretion, baseline sodium excretion in acuteclearancestudies, and baseline urinary sodium concen-tration have been equivalent for both canine populations (1,

2). Where we have measuredUK/UNa+UKratios, a measure for distaldelivery and cation exchange ( 13, 16), we also have not detected differences between natriuretic responders and nonresponders. Thus, it seems unlikely that reduced distal

de-Recovery

BKA

BKA +

ANP2

Figure7. The effect of de-liveringaspecific bradykinin antagonist into the left renal artery of four TIVC re-sponders.*P<0.05 com-pared withpreviousphase. , Leftkidney;o,right kidney.

80

-60

-40

-20

-0'

Table 5. Influence of Various Agents onUrinaryANP Excretion(n= 6)

Agent ControlUVWp PostagentUVNP

pg/min

Bradykinin(LRA)* 32±12.7 15.8±3

Captopril(i.v.)t 17±8 24.6±4

Aprotinin (i.v.) 11.6±4 15.4±3

Bradykinin antagonist(LRA) 16±2.3 18.5±4.6

*_{LRA,agent delivered into left renal artery. *i.v.,agent delivered}

intravenously.

livery ofsodiumto the IMCDisa problem. Moreover, in a recentstudy,we havedemonstratedthatamiloride, adiuretic which inhibits Na'conductive pathwaysresponsiveto ANP,

will induceanormalnatriuresis inTIVC dogsunresponsiveto ANPinfusions (6). Itthusseemsunlikely that in our experi-mental modelinsufficientdistal delivery of sodiumcould be a

determinant for tubular unresponsiveness to ANP in TIVC nonresponders.It must beconceded, however, that Moraliet

al. (17) recently demonstrated in 10 cirrhotic patients with

ascites unresponsive to an ANP infusion that mannitol in-duced modest natriuretic responses to this peptide in six of

thesepatients.Intheremaining four, however, neither manni-tol normannitol plusANP incombinationwas sufficientto

increaseurinary sodium excretion.

In anextensive seriesofstudiesin TIVC dogs,wehave been unable to show differences between responders and

nonre-sponders for plasmavolume, GFR,renalperfusion, papillary

plasmaflow,aldosterone,endothelin,andvasopressinplasma levels(1,2, 5-8) and have ruledout arolefor catecholamines, angiotensin,lowbloodpressure,renalnerves,andadenosine in

overridingandbluntinganANP-induced natriuresisin

nonre-sponders. Althoughwehavenotyetexcluded withcertaintya

problem with signal transduction distaltocGMPgeneration,

thedatafromthepresentstudysuggeststronglythatANP

and/

orkinindeliverytothe IMCDmaybeofgreatimportanceas a

determinantin thenatriureticresponsetoinfusedANP,atleast inTIVCnonresponders.

SinceNEP24:11 within the brush border of theproximal

tubule prevents thelargestpartof filteredANPfromreaching

theIMCD,theurinary excretion ofANPwas

minimal, varying

from 6to30 pg/min inour

experimental

animals(Tables

II-IV). SinceANPisknowntoexert apotentinhibitoryeffecton

conductivesodiumtransportwhenpresentonlyontheluminal

side of the IMCD(

18),

receptorsmust

presumably

existatthis site. Moreover, immunocytochemical evidenceconfirms that withahigh degree ofprobabilityANP receptors existon the luminal aspect of both the cortical collecting duct and the

IMCD( 19).AsWilkinset al. ( 1)recentlyspeculated, these receptors,receivingasmallbutfairlyconstantamount ofnatri-ureticpeptideandprotected from "upordown"regulationby varying plasma ANP levels, may play an important role in

determiningsodiumhandling bythecollectingduct.Indeed,in

our studies, theadministration ofa NEPinhibitorto TIVC nonresponders was ableto elicita modest natriuresis in this group independent ofany change to GFR, C..., ABP, or

plasma level of iANP (TableIII).Although urinary

concentra-tionsof ANP increased markedly, UVAN or FE of ANP just

barely escaped statistical significance while increasing. In the remaining groups of dogs, a natriuretic effect after administra-tion of the NEP inhibitor was associated with increments in

plasma iANP, as well as significantincrements in

_UVAN,

and FE of ANP. Several groups of investigators (11, 20) have now

suggestedthat intraluminal delivery of ANP to the IMCD (as

opposed to the basolateral surface)may be critical in

determin-ingthe magnitude of the natriuretic response.

The observation in ourexperiments that a natriuretic effect

in TIVC dogs after administration of NEPinhibitor could

oc-curin association with increased ANP urinary excretion, but

also with a rise in plasma iANP levels (controls and TIVC

responders), as well as the observation that thebaseline ANP

excretion was not less in TIVC nonresponders than other

groups, or indeed was evenhigher during the initial ANP infu-sionthanincontrol dogs, suggested that distal delivery of ANP was not necessarily correlated to sodium excretion in our TIVC dogs. The observation that administering NEP inhibitors to

TIVC nonresponders was a potent method to induce tubular

responsivenesstoANP in animals previously completely

unre-sponsive tothis peptide indeed suggested to us that excessive degradation of ANP by proximal tubular NEP 24:1 1 might be the cause ofsuch unresponsiveness. Several laboratories ( 10,

11, 20, 21 ), working with dogs, humans, and rats, have now

demonstrated thatinhibitionof NEP 24:1 1 with various com-poundswill indeed magnify the natriuretic response to ANP, even though changes in plasma ANP levels after such

inhibi-tion arevariable.

Several observations in our laboratory suggest, however, that augmentedavailabilityof ANP may not be the entire cause

for this increased natriureticeffect. Urodilatin is a peptide simi-lar to ANP, first isolated from human urine bySchulz-Knappe

et al. (22) anddemonstratingpotent vascular and natriuretic

effects. Indeed, evidence has been presented that urodilatin may be that member of the ANP family primarily responsible

for the regulation ofurinary sodium excretion (23). Compared with ANP, it isNH2terminally extended by fouramino acids

and isthought to be produced within the kidney and to be resistant to the degradative effects of NEP 24:11 (23). If

urodi-latinwasinfusedinto TIVC nonresponders, and excessive

deg-radation ofANP by NEP 24:1 1 was an important cause for the lack ofnatriuresis,then urodilatin escaping enzymatic break-down should now beassociatedwith a significant natriuresis. Althoughthisoccurred in normal dogs and TIVC responders,

this peptide was without effect in TIVC nonresponders. In-deed, anatriuretic effectwas not observed until the dogs were

pretreated witharenal arterial infusion of bradykinin (Fig. 4).

Giventhat NEP 24:1 1 can hydrolyze other peptides viz kinins, neurotensin, endothelin, etc. ( 12), the possibility exists that thedelivery ofkinins may be a determining factor for the tubu-lar refractorinessto ANP in TIVC nonresponders. This idea receives support from the following observations. First, when

captopril,an inhibitor of kininase andangiotensin-converting

enzyme wasinfused into TIVC nonresponders, a natriuretic

effectto asecondinfusion ofANP was observed (Fig. 3). No

magnificationof a natriuretic effect was observed when capto-pril wasgivento normal controls.We have recently

demon-strated thatangiotensin infusedinto TIVC responders will not convert them intononrespondersandsimilarly theprovision ofanangiotensin antagonist, saralasin,doesnot convertTIVC

nonresponders intorespondersafter infusion ofANP(8). Cap-topril probablyachieveditseffect, therefore,byincreasingthe availability of intrarenal kinins in TIVC nonresponders. The absenceofaneffect innormalcontrolssuggeststhat there exists

anintrarenal availability of kinins beyondsomecriticallevel. Theobservation that UNaVdidnotincrease following

intrave-nouscaptopril alone despiteanunchanged GFR(Fig. 3) sug-gests thatthis agentdidnot produce itspermissive effect on

ANPbyaugmenteddistalNa' deliverytotheIMCD. Second,thoughwithouteffectonrenal functionorsodium excretion in the dosesemployed,bradykinin providedtoTIVC nonrespondersnow permitteda natriureticresponse toANP where nonepreviously could beelicited(Fig. 2). Evenmore

compellingwerethe observations thataprotinincould blunt thenatriureticeffect of ANP in TIVCresponders(Fig. 6).This

agentisanonspecific polyvalentserineproteaseinhibitor

capa-bleofattenuatingtheeffect of severalproteases,including glan-dular kallikrein, but its physiological effects have generally been ascribedto its abilityto inhibit kinin generation (12). Thoughwithouteffectonkidney functioninTIVCresponders whenadministered alone, therewas adramaticdecrement in

AUNaV after ANP administration (299±69 vs. 92±26 ,teq/ min,P<0.05) whenaprotininwasadministered concurrently.

It has previously been reported that a specific BKA can

blunt themanyphysiological functions attributabletokinins (24,25). Smitsetal.(24) have demonstrated in anaesthetized

ratsthat thepotentiation of the natriuretic effectsofANPby

NEPinhibitorswascompletely abolished by the identicalBKA

used inourpresentstudies. Theseinvestigators concluded,

par-ticularlysinceNEPinhibition didnot causeplasmaANPlevels

to rise(whengivenaloneorconcurrently withanANP infu-sion),that potentiation ofthe natriuretic effectsofANPby

NEP inhibition involved intrarenal accumulation of

brady-kinin.

In ourstudiesweobservedasimilar phenomenon but had sufficient NEPinhibitortostudyonlythreedogs in thisway. When the BKA wasgiven intoonerenal arteryof TIVC

re-sponders, therewas noeffectonUNaV. When theNEP

inhibi-torin usual doseswasgiven intravenously,therewas no

incre-mentinAUNaVfortheexperimental kidney (5±0.9

_,eq/min)

whereas the AUNaV for the control kidney was significant (31±2.6

,ueq/min).

Wealsodemonstrated however that when theBKA wasgiventorespondingTIVC dogsreceiving exoge-nousANP,thenatriuresis from the experimental kidneywas

severelyblunted whereas thatofthe controlkidney continued

unabated(Fig. 7). Thus, inourhands,antagonismof brady-kininreceptors wasalso capableof inhibitinganatriuretic ef-fect when onlyexogenous ANP wasinfused,without

simulta-neous NEPinhibition.

This phenomenonhaspreviouslybeenreported by Sybertz etal. (24). Theseinvestigators administeredtorats the same

bradykinin antagonistused inourstudies. They observed that bradykinin playedapermissiveroleforthenatriuretic effects of ANPsincethe BKAinhibitedthenatriureticresponse to both NEPinhibitorsand exogenous ANPitself. Unlikeour present

studies,theseauthorsfound that theBKAitselfwould decrease bothurine flowandurinary sodium excretion, suggesting

per-haps a moreimportant role for kinins in regulating tubular

handling of sodiuminthis species than in dogs.Ofinterest as well, was that theBKAdid not abolish theantihypertensive effect ofANPwithorwithoutsimultaneousNEPinhibitors.

The observations obtained inthe present studies would indi-catethattheavailabilityofintraluminalkinins may be a critical

modulating factor for the natriuretic effects of ANP, and that bothpeptidesmustbepresentforanatriureticresponse tobe

obtained afteranexogenousANP infusion orthedeliveryof

"extra" ANPtothe IMCD from the proximal tubule after NEP inhibition. Merely augmenting distal ANP deliveryper se ap-pears tobe insufficient fora natriuresisto occur. Our studies with 8-Br-cGMPsupportthis idea. Studiesperformedoncell

suspensions prepared from the IMCD of TIVC and normal

dogsindicate that theabilitytogeneratecGMP is equivalent between these animals when exposedtoANP(8). Neverthe-less,ourpresentstudyaswellasprevious studies (8) indicate that TIVCnonresponders,but notnormaldogs orTIVC

re-sponders, are refractory to the natriuretic effects of infused cGMP. Yet whenbradykininissupplied,TIVC nonresponders

are nowabletomountanatriureticresponsetoinfused 8-Br-cGMP.These datasuggestthat thepresenceof kininsmaybe critical for ANP-inducedsignaltransduction. Procedural diffi-culties inmeasuringurinarykininsprecluded obtainingsuch data in thepresentstudy,butclearlysomeofthese experiments

must berepeated whileassayingurinaryandperhaps plasma

kinin levels.

Itmust also beemphasizedthat despitewhatever

kinin-ANPinteractionsmaybeoccurringwithin thecollectingduct

system,otherfactorsmaybeoperativein TIVC-NRdogs (e.g.,

limited distal Na' delivery, humoral factors antagonistic to

ANP,etc.),limitingthenatriureticresponse to ANP.This is

suggested bythe observation thatevenin thepresenceofNEP

inhibition, thepeak UNaV after ANP administration or was

300-400

,ueq/min

incontrols and TIVC-Rdogs,butonly 100

,.eq/min

innonrespondingTIVCdogs (Table III).

The schemaillustratedinFig.8summarizeourideasand

explanationsofourexperimental findings.The leftpanel

indi-catesnormalevents.ANPandkininsarefreely filteredatthe

glomerularlevel(Fig. 8,site1).Atsite 2 inFig.8,NEP24:11

alongwithkinasesdestroyvirtuallyallofthefiltered kininsso

that none leaves the proximal tubule (27). In TIVC

nonre-sponders,excess NEPactivityatextrarenalsitescouldaccount

for thediminished baselinelevelsofplasma iANP (Table I).

Augmentedreabsorption ofthispeptideafterNEPinhibition

seemsunlikely giventhe extensivedegradationbyNEP24:11

withinthe brush borderoftheproximal convoluted tubule.As

well, previous studiesinthismodel( 1,2, 16)have shown that

differencesinplasmavolumecannotaccountfordifferencesin baselineplasmaiANP levels. Though plasma levelsofiANP didnotrise in TIVC-NR dogs afterNEPinhibition(asmight be

anticipated

if excessive nonrenal degradationwere

occur-ringin thisgroup),it should be notedthatfractional excretion ofANP(TableIII) rose morein thisgroup than theothers.

This rise in urinary excretion mighthave limited the rise in

plasmalevels.Inthislattergroup aswell, excessintrarenalNEP

activitycould leadtodiminished distal deliveryof ANP

com-paredwith TIVC responders. ThatNEPactivity is related to

kinin degradation seems clear. It is now believed that NEP 24:1 1isatleast asimportantaskininase 11 in degradingkinins withinthe nephron (12), and a recent reportindicates that NEP24:11 may accountfor 53-74% of intrarenal kinin

degra-dation (28). In rats, urinary kininase activity appears to be largely provided by NEP 24:1 1 (27 ).

NEP

24:11 INHIBITION

GLOI

MERULUS

T

P; Keg

PROXIMAL Neutral

PROXIALE

Endopeptidase

TUBULE _- 24:11

Iil

ANP Kinins (0.2-2% 0

tiftered 1 (a

bad) , \

(

Kinins

DISTAL ANP Knn

TUBULE

CORTICAL

LOOP _Inhibodby COLCINGDUCT

ANP, Kinins,... , + I

Thiazidos c Inhbited by .... Na

Amidan

Inheitedby f...1.A MC

ANP,Kinir

IMCD

' % t /_ C. GMP

MANP; Kinins

lrl

-SQ

#28603

XE

AXE

ANP Kinins v (5-17% (? amount) Kininase

fihered (_Inhibition

load)

Figure8.Asummaryofkinin-ANPinteractions as they occur within the nephron normally(left)andafter NEP 24:1 1 inhibition (right). See text for details.

kininactivity acting at moredistalsites.Apossible reductionin

distaldelivery of kininsin TIVCnonresponderssuggestthere

maybe reduced formation

_and/or

excessive

degradation

atthis site. Fig. 8, sites4and 5 indicatethe

major

pathways for

so-diumreabsorption by conductiveandelectroneutralpathways

in thecortical collectingduct andby conductivechannels in theIMCD. Others havealready

supplied

evidencethatkinins

mayplayanimportantpermissiveroleatthesetransportsites. Zeidel etal. (29) recently demonstratedthe

ability

of

brady-kinin to directly inhibit conductive Na' channels in rabbit

IMCD cells. Morerecently,Stoosetal.(30) reporteda

syner-gistic effectfor ANP andbradykininonshort circuitcurrent,a measure ofsodium transport in cultured cortical collecting

duct cells

(M-l line).

Eachagent

by

itselfwaswithout effecton

short circuitcurrentbut,when

given

together,reducedsodium

transportby15%whereas cGMPcontentofthe cellsincreased.

Wehavealso shown atsite5 inFig. 8 thatkininsmayplaya

permissive role for the actions ofcGMP, at least in TIVC nonresponders.

TherightpanelofFig. 8, summarizespossible effects after

NEP24:11inhibition. Thepresenceof SQ 28603atFig. 8, site

2would augment thedistaldeliveryofANPandkinins. Since, however, kinins are thought notto reach the distal nephron fromtheproximaltubulenormally,alackofkinins from this

source cannotbeinvokedasplayingaroleinthe tubular

insen-sitivitytoANPin TIVC nonresponders. Asmentioned previ-ouslythere may be asynthetic and/or degradativeproblem at the levelofthedistal nephron. This raisesthepossibilitythat

the NEPinhibitormaybeplayingabeneficialrole at Fig. 8, site

3,asillustrated in thediagram.Ithasbeen speculated that the

kallikrein-kininsystem mayclearANPfromprohormone at

thissite(23),andthismaybeadetermining factor.

In any event, after NEP inhibition, we postulate that

in-creaseddelivery ofkininsandANPdistallytoFig. 8, sites4and

5maypermitamagnified natriuresis inTIVC responders and

induceanatriuresisin nonrespondersasthedeliveryof kinins andANPreach beyond somecriticallevel. Since conductive and electroneutral pathways for sodium reabsorptionare

lu-minal, inhibition would presumably beatthis site. Certainly

bradykininreceptors( 12)andANP receptors(19)arethought

toexistatthis site. Our data alsosuggestthatkininsare

impor-tant for cGMP actions to occur, at least in TIVC

nonre-sponders. That both substances vizANPandkinins mustbe

present togetherforanANP-induced natriuresisto occur

re-ceivessupport fromtheobservationthat when antibodiesto ANP areadministeredNEPinhibitorslosetheirnatriuretic ef-fects (31). Although other investigators have certainly

sug-gested andprovidedevidencethatkininsmay beimportant for thenatriuretic effects ofANPafterNEPinhibition (20, 24, 26, 30),webelievethat oursisthefirst such evidencecollectedina

modelof sodium retentionandascites,wheredefects in kinin

availabilitymay serve as anexplanation forthe attenuated

na-triureticeffects ofANP.

In summary,we haveuncoveredtwo maneuversthatwill

convert TIVCdogs

unresponsive

tothe natriuretic effects of

ANPintoresponders.OneistoinhibitNEP24:11,the otheris

1434 L.Legault,P. Cernacek,M.Levy,E.Maher,andD.Farber

toprovideintrarenalkinins.Our observationsin TIVC

nonre-sponders thatpretreatment withbradykinin ispermissivefor

the natriureticeffects ofANP,urodilatin(anatriuretic peptide thought to escape intrarenal enzymatic cleavage), and cGMP,

impliesa finalcommonpathway. Becauseurodilatin escapes NEP cleavage but is still not natriuretic in TIVC

nonre-spondersuntilbradykininissupplied,itispresumablynot aug-menteddistal delivery of natriuretic peptidethat is critical after NEP inhibition, but rather theaugmented delivery of kinins.

Thisconcept receives support from the observations that both

aprotininand a BKA will attenuateanANP-induced

natriure-sis in TIVCresponders. How kinins interact with ANP isnot clear but they may modulate ligand (ANP) binding to IMCD

luminal receptors or in some way be critical for full signal

transduction.Whysome60% of our TIVCdogsin the present

studyappear to lack critical amounts of intrarenal kinins isnot

answeredbyourexperimentsbut may be duetodifferencesin

thelevel ofNEP (kininase) activitywithin therenal tubule.

Theobservation byusthatNEP-induced natriuresisoccurs in

TIVCnonresponders in the absence of raisedplasmalevels of

ANPbutseems tobeassociated with increased urinary

excre-tion ofANP supports theview that theANP-kinin interactions areprobablyoccurringattheluminalasopposedtothe basolat-eralsurface of the IMCD.

Acknowledgments

We aregratefulfor the competent technical skills of Mr.LuigiFranchi, Mrs. Christine Fechner, and Mrs. OlgaLawryk, and the secretarial expertise providedby Mrs. ChristinePamplin.The generous donation ofasupply ofSQ 28603 by theSquibbInstitute for Medical Researchis acknowledged.

Duringthesestudies Dr. Legault was a FellowoftheKidney Foun-dation ofCanada(KFC),and Mr. Farber was supported by the Sum-merStudentFellowshipprogramoftheKidney Foundation ofCanada. Dr.Maher was supported by aspecialstipend fromthe Dean'soffice, Faculty of Medicine, McGill University. Operating funds fromthe Medical ResearchCouncil ofCanada to M. Levy supportedthis re-search, as well as a grantfromtheKFC to M. Levy and P. Cernacek.

References

1.Maher, E.M., P.Cernacek, and M. Levy. 1989. Heterogeneous renal re-sponsestoatrial natriuretic factor I: chronic caval dogs.Am.J. Physiol.257

(Regu-latory Integrative Comp. Physiol.26):R1057-R1067.

2. Maher, E. M., P.Cernacek,and M. Levy. 1989. Heterogeneous renal re-sponsestoatrial natriuretic factorII:cirrhotic dogs.Am.J.Physiol.

257(Regula-toryIntegrativeComp.Physiol.26):R1068-R1074.

3. Lopez,C.,W.Jiminez,V. Arroyo, G.LaVilla,J.Goya, J. Claria, F.Rivera, and J. Rodes. 1989. Role of alteredsystemichemodynamicsin the blunted renal responsetoatrial natriureticpeptideinratswithcirrhosisandascites. J.Hepatol. (Amst.).9:217-226.

4.Salerno, J.,S.Badalamenti,P.Incerti,L.Capozza,and L.Mainardi.1988. Renal response toatrial natriuretic peptide inpatients with advanced liver cirrho-sis.Hepatology.8:21-26.

5.Maher, E.M., P. Cernacek, and M. Levy. 1990. Serial natriuretic response

toatrialpeptideinpreascitic bile duct ligated dogs. Can.J.Physiol.Pharmacol.

68:1396-1400.

6. Levy, M. 1990. Comparative effects of diuretics and atrial peptide in chronic caval dogs. Am. J. Physiol. 258(Renal Fluid Electrolyte Physiol. 27):F768-F774.

7.Maher,E.M., P. Cernacek, and M. Levy. 1990. Physiological features of edematous dogs unresponsive to atrial natriuretic peptide. Am. J. Physiol.

258(Renal Fluid ElectrolytePhysiol.27):F1490-F1496.

8.Legault, L.,P.Cernacek,and M.Levy. 1990.Attemptstomanipulatethe natriuretic responsetoANP in chronic cavaldogs. Clin.Invest. Med. 13:78. (Abstr.)

9. Brenner, B., B. J. Ballerman, M. E. Gunning, and M. L. Zeidel. 1990. Diversebiologicalactions ofatrial natriureticpeptide. Physiol.Rev. 70:665-699. 10.Margulies, K. B., P. G. Cavero, A. A. Seymour, N. G. Delaney, and J. C. Burnett, Jr. 1990. Neutral endopeptidase inhibition potentiates the renal actions of atrialnatriuretic factor. Kidney Int. 38:67-72.

11.Wilkins, M. R., S. L. Settle, P. T. Stockman, and P. Needleman. 1990. Maximizing the natriuretic effect of endogenous atriopeptin in a rat model of heart failure. Proc.Natl.Acad. Sci. USA.87:6465-6469.

12. Coyne, D.W., and A. R. Morrison. 1991. Kinins: biotransformation and cellular mechanisms of action. In Hormones, Antacoids, and the Kidney. S. Goldfarb andF. N. Ziyadeh, editors. Churchill Livingstone, New York. 264-280. 13.Levy, M. 1972. Effects of acute volume expansion and altered

hemody-namicsonrenaltubularfunctioninchroniccavaldogs.J.Clin. Invest. 51:922-938.

14.Mizelle,H. J., J. E.Hall, and D. A. Hildebrandt. 1989. Atrial natriuretic peptide and pressure natriuresis: interactions with renin-angiotensin system. Am. J.Physiol.257(Regulatory,IntegrativeComp. Physiol. 26):R1169-R I174.

15.Gerbes, A. L., M. C. Kollenda, A. M. Vollmar, J. Reichen, N. Vakil, and R. M.Scarborough. 1991. Altered density of binding sites for atrial natriuretic factor in bileduct-ligated rats with ascites. Hepatology. 13:562-566.

16. Maher, E. M. 1989. Atrial natriuretic factor in two canine models of ascites:cardiac release andheterogeneityof renal natriuretic response. Ph.D. thesis, McGillUniversity, Montreal, Canada.

17.Morali, G., S. Tobe, K. Skorecki, and L. Blendis. 1991. Modulation of ANF unresponsiveness by mannitol (M) in refractory ascites. J. Am. Soc.

Nephrol.2:409. (Abstr.)

18.Sonnenberg, H., U. Honrath, and D. R.Wilson.1990. In vivo microperfu-sionof inner medullary collecting ductinrats:effect of amiloride andANF. Am. J.Physiol.259(Renal Fluid Electrolyte Physiol. 28):F222-F226.

19.Figueroa,C.D.,H. M.Lewis,A.G. MacIver, J. C.Mackenzie,and K. C. Bhoola. 1991. Cellular localization of atrial natriuretic factor in the human kid-ney.Nephrol. Dial. Transplant.5:25-31.

20.Margulies,K.B., M. A.Perella, L. J. U. McKinley, and J. C. Burnett, Jr. 1991.Angiotensin inhibition potentiatestherenal responses to neutral endopep-tidase inhibitionindogs withcongestiveheartfailure.J.Clin.Invest. 88:1636-1642.

21.Sagnella,G. A., N. D. Markandu, M.G. Buckley,M. A.Miller,D.R. J.

Singer,F. P.Cappuccio,andG.A.MacGregor.1991.Atrial natriuretic peptides

inessentialhypertension:basalplasma levels andrelationshiptosodiumbalance. Can. J.Physiol. Pharmacol. 69:1592-1600.

22.Schulz-Knappe,P., K.Forssman,F.Herbst, D. Hock, R.Pipkorn,and W.G. Forssman. 1988.Isolationandstructuralanalysis of"urodilatin",a new

peptideof thecardiolatin(ANP)family,extractedfromhumanurine.Klin. Wo-chenschr. 66:752-759.

23.Goetz,K. L.1991.Renalnatriuretic peptide (urodilatin?)and

atriopep-tin: evolvingconcepts. Am. J. Physiol. 261(Renal Fluid Electrolyte Physiol. 30):F921-F932.

24.Smits, G.J., D. E.McGraw,and A. J.Trapani.1990.Interaction ofANP andbradykinin during endopeptidase 24:11inhibition:renaleffects.Am.J.

Phys-iol.258(Renal FluidElectrolyte Physiol. 27):F1417-F1424.

25.Beierwaltes,W. H.,0.A.Carretero,and A.G. Scicli. 1988. Renal hemody-namics in response to a kinin analogue antagonist. Am. J. Physiol. 255(Renal

Fluid Electrolyte Physiol. 24):F408-F414.

26.Sybertz,E.J., Jr.,P.J.S.Chui,R.W.Watkins,andS.Vemulapalli.1991. Neutralmetalloendopeptidase inhibitorsas ANFpotentiators: sites and

mecha-nismsof action. Can.J.Physiol. Pharmacol.69:1628-1635.

27.Carretero,0. A., and A.G. Scicli.1990. Kinins as regulators ofblood flow andblood pressure inhypertension.InPathophysiology, Diagnosis and Manage-ment.J. H.Laragh,and B. M. Brenner,editors. Raven Press, New York. 805-818.

28. Ura, N., 0.A.Carretero,and E.G. Erdos. 1987. Role of renal endopepti-dase 24:11in kininmetabolisminvitroand invivo.KidneyInt.32:507-513.

29.Zeidel,M.L.,K.JABS, D.Kikeri,and P.Silva. 1990. Kinins inhibit conductiveNa'uptake byrabbit innermedullary collecting duct cells. Am. J.

Physiol.257(RenalFluidElectrolytePhysio.27):Fl584-Fl591.

30.Stoos,B. A., 0.A.Carrectero, and J. L. Garuin. 1991. Potential roles of cAMP andproteinkinase C in thesynergisticaction of atrial natriuretic factor andbradykininin theM-1 corticalcollectingduct cell line.J.Am.Soc. Nephrol. 2:418. (Abstr.)

31.Samuels,G. F. R. 1990.Atriopeptidase inhibition: a new therapeutic modulator.ADecadeofANFResearch:International Hypertension