The Handling of Immunoreactive Vasopressin by theIsolated Perfused Rat Kidney

RALPHRABKIN, LEONARDSHARE, PAULA. PAYNE, JUDYYOUNG,and JOANCROFTON,Departments of Medicine and Physiology and Biophysics, University of TennesseeCenter for the Health Sciences, Memphis, Tennessee 38163

A B S T RA C T Using the isolated rat kidney perfusedwith an artificial medium containing glucose as the solefuel, we studied the renal handling of immunoreactivearginine vasopressin (AVP) and determined the effectof various factors on the ability of the kidney toremove AVP.

In control kidneys perfused with AVP at concentra-tions below 116 uU/ml, the organ clearance of AVP(OCAVP) was 1,145±47 (SE) p1l/min, whereas glomeru-lar filtration rate (GFR) averaged 515±37 ,ul/min. Filtra-tion could thus account for up to 45% of the OCAVP,the balance presumably being cleared from the peri-tubular circulation. Of the AVPfiltered, only 38%couldbe recovered in the urine (urinary clearance AVP av-eraged 205±12 ,p/min) suggesting that the balance wastaken up by the tubular epithelium and degraded. Frac-tional excretion of filtered AVP rose significantly inthe presence of anoxia and cold (10°C) to 49 and 59%,respectively, but was not affected by ouabain or highlevels of AVP (458±58 ,uU/ml).

The OCAVPwas not significantly altered by the ab-sence of glucose in the perfusate, anoxia, or ureteralligation, maneuvers that were associated with signifi-cant reductions in GFR. In these and the control ex-periments, there was a significant inverse correlationbetween GFRand peritubular clearance emphasizingthe importance of the latter (r = -0.749; P < 0.001).Cold, ouabain, and high concentrations of AVPreducedthe clearance of AVP by the kidneys.

On the basis of these studies we conclude that thekidney clears AVP from the circulation via two path-

Portions of this work were presented at the InternationalConference on the Neurohypophysis, Key Biscayne, Fla., No-vember 1976; the Southern Society of Clinical Investigation,New Orleans, La., January 1977; and the Annual Meeting ofthe Endocrine Society, Chicago, Ill., June 1977.

Address reprint requests to Dr. R. Rabkin, NephrologyDivision, Stanford University School of Medicine, Stanford,Calif. 94305.

Receivedfor publication 22 September 1977 and in revisedform 18 September 1978.

6

ways, glomerular clearance and peritubular clearance.This exposes both the luminal and contraluminal sur-faces of the tubular cells to the hormone, allowing thesecells to remove AVP from the filtrate and the peritubu-lar compartment. Noteworthy is the observation thatunder several conditions when GFRfalls reducing theglomerular clearance of AVP, peritubular clearance in-creases and the total clearance of AVP by the kidneyremains constant.

INTRODUCTION

The kidney plays an important role in regulating theplasma levels of several peptide hormones includingvasopressin. This is achieved in part by degradationand in part by excretion of the hormone in the urine.In addition to being the target organ for the action ofvasopressin, the kidney avidly removes this peptidefrom the circulation and accounts for between one-third and one-half of its metabolic clearance rate (1).Despite several studies (1), confusion exists as to theexact manner in which the kidney handles vasopressin.This is in part the consequence of insensitive bio-assay procedures, the rapid half disappearance timeof the hormone, and the variable design of earlierstudies.

The purpose of the present study, carried out withisolated perfused rat kidneys, was to assess the mannerin which the kidney handles immunoreactive argininevasopressin (AVP)1 and to determine which of severalmetabolic factors might affect this process. The isolatedkidney perfused with a continuously recirculated elec-trolyte-albumin solution was chosen as the experi-mental model for several reasons. First, it has a per-fusion flow about six times that of the normal in vivorenal blood flow, which in the presence of a glomeru-lar filtration rate roughly one-half that in vivo may tend

1 Abbreviations used in this paper: AVP, arginine vasopres-sin; GFR, glomerular filtration rate; OCAVP, organ clearanceof AVP.

J. Clitn. Invest. © The American Society for Clinical Investigation, Inc., 0021-9738/79/011/006/08 $1.00Volume 63 January 1979 6-13

to exaggerate processes occurring on the contraluminalside of the tubular cell, thus facilitating their study.Second, with a bloodless perfusion medium variousextraction procedures necessary for the assay of vaso-pressin are avoided, and the precision of the assay isthus increased. Third, various maneuvers may be per-formed with the isolated kidney that are impossible tocarry out in vivo.

METHODS

The study was carried out with male Sprague-Dawley ratsweighing 250-350 g and allowed free access to food andwater. Anesthesia was produced with intraperitoneal pento-barbital (60 mg/kg). The kidneys were then isolated and per-fused according to the method of Nishiitsutsuji-Uwo et al.(2) as modified by Ross et al. (3). After an intravenous injec-tion of mannitol, 50 mg/100 g, the abdomen was opened by amidline incision and the right ureter was catheterized withpolyethylene-10 tubing. Heparin (200 U) was administeredintravenously and a thin glass cannula was inserted into thesuperior mesenteric artery and then into the aorta. The per-fusion flow was started and the glass cannula was then in-serted into the right renal artery after which the kidney wasrapidly excised and placed in the perfusion apparatus. Therenal vein was not cannulated. The kidneys were perfusedwith a Watson Marlow MHRE3pulsatile flow pump (Watson-Marlow Ltd., Cornwall, England). After a period of equilibra-tion, perfusion pressure was maintained at an effective maxi-mumpressure of 100 mmHg unless otherwise stated. Theperfusion apparatus was housed in -a plexiglass cabinet andthe temperature was maintained at a75C. The perfusion me-dium (3) consisted of a modified Krebs-THenseleit bicarbonatesolution containing 6.7 g of albumin per 100 ml, gassed con-tinuously with 95% oxygen and 5%-CO2. Bovine serum al-bumin (Fraction V, Miles Laboratories, Inc., Elkhart, Ind.)was dissolved in 860 ml of Krebs-Henseleit bicarbonate towhich 40 ml of 1 Msodium bicarbonate was added; this wasdialysed at 4°C against six to eight changes of Krebs-Henseleitbicarbonate for 72 h. The final medium contained in milli-moles per liter, sodium 143, potassium 5, calcium 2.25, mag-nesium 1.18, bicarbonate 18, chloride 11:6, phosphate 1.2, sul-fate 0.8, and glucose 5 mM at pH 7.4 when gassedwith 5%CO2.

4 min after the start of the perfusion [14C]carboxyl-inulin(Amersham/Searle Corp., Arlington Heights, Ill.) was added tothe medium and a control perfusate sample was obtained. At10 min synthetic arginine vasopressin (Bachem, Torrance,Calif.; 380 U/mg) was added to achieve concentrations rang-ing from 35 to 116 ,uU/ml in the initial perfusate sample ob-tained at 20 min. After allowing 15 min from the start of theperfusion for the stabilization of renal function, seven 10-minurine collections with midpoint perfusate samples (1.2 ml)were obtained. In some experiments the kidneys were thenremoved from the perfusion apparatus and further perfusatesamples were obtained. The volume of perfusion medium atthe start of midpoint sampling was =63 ml; sample volumesremoved were not replaced. To avoid loss of AVP by ad-herence to glass, urine samples were collected in tubes thathad been lightly coated with albumin and allowed to dry.

Experimental groups. The kidneys were perfused underseveral different conditions: (a) Six kidneys were perfusedunder standard conditions referred to as controls. (b) Six kid-neys were perfused with a glucose-free medium. (c) Fourkidneys were studied under anoxic conditions. This wasachieved by gassing the medium with 95% nitrogen and 5%

CO2. (d) Four kidneys were perfused with ouabain (4 mM)added to the medium before starting the perfusion. (e) Fourkidneys were perfused in the presence of high levels of AVP(458+58 1xU/ml). (f) Four kidneys were perfused with cold(10°C) medium by circulating cold water through a glass jacketsurrounding the oxygenator. (g) Obstruction of the ureter offour kidneys was achieved by clamping the ureteral catheter5 min before collecting the first perfusate sample.

Binding of AVP to albumin was assessed by ultrafiltrationstudies employing a Diaflo XM50 membrane in an AmiconModel TCF1O ultrafiltration system (Amicon Corp., Lexing-ton, Mass.). Ultrafiltration of AVP added to freshly preparedmedium was compared with that of AVP added to Krebs-Hen-seleit bicarbonate. The two solutions were gassed with 95%02 and 5% CO2 before ultrafiltration, and the ultrafiltrationsystem was pressurized with the same gas mixture. The tem-perature was maintained at 370C by placing the apparatusin the kidney perfusion cabinet. At the end of 18 min of ultra-filtration, samples of retentate and ultrafiltrate were obtainedand assayed for AVP.

Analytical methods. Arginine vasopressin was deter-mined in unextracted perfusate and urine samples by meansof a radioimmunoassay as previously described (4). The anti-body used in the immunoassay did not cross-react with lysinevasopressin, arginine vasotocin, oxytocin or angiotensins I andII. Between-assay coefficient of variation (n = 21) was 8.9%and the within-assay coefficient of variation based on fiveassays was 7.3±1.4% (SE). Extraction procedures were notused, as recovery of AVP added to medium averaged 94+1.81%, whereas recovery of AVP added to urine averaged94.7+2.1%. Concentrations were not corrected for incompleterecovery. Sodium was measured with an IL 143 flame photom-eter (Instrumentation Laboratory, Inc., Lexington, Mass.).[14C]carboxyl-inulin was counted in a Nuclear Chicago Isocap300 liquid scintillating counter (Nuclear Chicago, Chicago,Ill.).

Calculations. In each experiment the organ clearance ofAVP (OCAVP), which represents the volume of perfusate fromwhich AVPis completely and irreversibly removed each min-ute, was calculated between successive 10-min samples ac-cording to Mortimore et al. (5) as OCAVP= 2.3 V/At log ([Cl- CQ/[C2 - C<]), where V = volume of perfusion fluid cor-rected for perfusate sampling; C, and C2 = initial and finalconcentrations over time interval t; and C, = asymptote ap-proached by C after prolonged perfusion. C,, was taken as zerobecause prolonged perfusion at low concentrations of AVPindicated the eventual disappearance of the peptide. Use ofthis formula was validated by the finding that (a) the disap-pearance of AVP from the medium followed approximatelyfirst order kinetics (Fig. 1A), and (b) the clearance of [14C]inu-lin determined by this formula was similar to the clearanceas determined by the standard urine clearance formula de-scribed below (in the control experiments these clearancevalues were 515+37 and 545+37 ,ul/min, respectively; P> 0.05, Student's paired t test). Based on the assumption thatAVP is freely filtered at the glomerulus and that none of thefiltered hormone is reabsorbed intact into the peritubular cir-culation, the glomerular clearance of AVPwas taken to equalthe glomerular filtration rate, and the peritubular clearancewas taken to equal the difference between the organ clear-ance and the glomerular clearance of AVP.

The urinary clearance of [14C]carboxyl-inulin and of AVPwere calculated from the urine:plasma concentration ratiosand the urine flow. The clearance of ['4C]inulin was used asa measure of glomerular filtration rate (GFR). Fractional ex-cretion of filtered AVP was calculated as the urinary AVPclearance/GFR x 100 and fractional excretion of filtered sodiumas sodium clearance/GFR x 100. Perfusion flow was determined

Vasopressin Handling by the Isolated Kidney 7

100 -

90 -

80-

70

E 60

A 50

$ 40w

(n

DL 30w

a.

A37°C

I-

1-

201P

10 20

B

AVP

30 40 50 60 70 80 0 10 20 30 40 50 60 7

10 c

70 80

TIME IN MINUTES

FIGURE 1 (A) Disappearance of AVP from the perfusion me-dium of control kidneys perfused at 37°C. (B) Disappearanceof AVP from the perfusion medium of kidneys perfused at10°C.

from the rate at which renal venous effluent filled a funnel-shaped siphon held continuously in place below the kidney.

The average of all the measured values in each experi-ment was taken as the value for that experiment. Results areexpressed as the mean±SEM. The data were statistically ana-lyzed (6) with a one-way analysis of variance, and if the Fratio was significant, Duncan's new multiple range test wasused to make comparisons among all the means. A level ofP < 0.05 was regarded as significant. Student's t test for non-paired data was used to evaluate the ultrafiltration data. Re-gression analysis was performed by the method of leastsquares.

RESULTS

Albumin binding studies

Ultrafiltration of medium containing AVP through aDiaflo XM50 membrane yielded an AVP concentra-tion ratio of ultrafiltrate to retentate of 0.78+0.07 (n= 4). This did not differ significantly from the AVPconcentration ratio of 0.75+0.01 (n = 4) achieved withultrafiltration of Krebs-Henseleit bicarbonate contain-ing AVP, indicating that significant binding of AVPto bovine serum albumin does not occur.

Perfusion studiesThe organ clearance of AVP (Table I). In control

kidneys the organ clearance of AVP, determined fromits rate of disappearance from the perfusate, averaged1,145+47 ,ul/min. As shown in Fig. 1A the disappear-ance of AVP followed approximately first order ki-netics. In several experiments, to exclude the possibil-ity that the disappearance of AVPwas a result of spon-

taneous degradation or to degradation by enzymesreleased from the kidney, several samples of perfusatewere obtained after removing the kidneys from thecircuit. In the absence of the kidney, no decrease inAVP concentration occurred (Fig. 2).

In kidneys perfused without glucose in the medium,the OCAVPwas 983+ 53 ,ul/min, a value not significantlydifferent from the control group. Similarly, the OCAVPwas not significantly altered when kidneys were per-fused under anoxic conditions (1,037±66 ,ul/min), norwhen the ureters were obstructed (992±50 ,l/min).

Addition of ouabain to the medium was associatedwith a significant drop in the OCAVP(866±60 IAI/min)as compared to the control value. Similarly, high levelsof AVP (initial concentration 315-592 ,uU/ml) signifi-cantly depressed the OCAVP(877±69 p1A/min). Coolingthe kidney to 10°C greatly reduced the ability of thekidney to remove AVPfrom the perfusate (Fig. 1B). Inthese experiments the organ clearance of AVPwas cal-culated on the assumption that Ca equals zero in thecold. Thus the value obtained, 204±63,1A/min, shouldbe regarded as an approximation, particularly as the fallin perfusate AVP concentration was small, introducingsome error into the calculation. GFRdropped signifi-cantly to 199±33 ,ul/min. From these data it appearsthat peritubular clearance was essentially abolished inthe cold and that AVPwas being removed by glomeru-lar filtration, a process dependent on mechanical en-ergy.

Contribution of glomerular clearance (assumed to

E

-

:

ILQIato

60[

KIDNEY

0IT50-

40F

30[

201

lo1

85 95 105 115 125 135 145

TIME IN MINUTES

FIGURE 2 Effect of removing the kidney from the perfusioncircuit on the disappearance of AVP from the medium of acontrol (0) and an anoxic (x) kidney. Kidneys were removedafter 85 min of perfusion.

8 R. Rabkin, L. Share, P. A. Payne, J. Young, and J. Crofton

TABLE IFunction* itl Isolated Rat Kidneys Perfuised utncder Different Coniditions twith AVP (74.8±4.5 ,UImnl)

High levelAVP

Glucose-free Ouabain (458+58Control medium Anoxia (4 mM) AU/ml) Cold (10°C) Obstruction

No. of experiments .6 6 4 4 4 4 4Effective maximum perfusion

pressure. mmlg .100 100 100 100 100 190 100

OCAVP, pllnritn 1,145±47 983±53 1,037±66 866±604 877±691 (-204)(±63)§ 992+50GFR= GCAVP,,AllInin 515±37 262±404 87±144 195±94 485±7 199±334PCAVP, ,ull/min 631±66 721±33 995±1154 671±59 392±74-UCAVP, ,l/minl 205±12 130±194 48±74 75±54 212± 12 113±131PC/OCAVP, % 54.6±3.8 73.7±2.84 91.7±1.14 77.2±1.94 43.5±5.14GC/OCAVP, % 45.4±3.8 26.3±2.84 8.3±1.14 22.8±1.94 56.5±5.1-UC/GCAVP=FEAVP, % 38.1±2.6 44.9±1.9 49.1±2.64 34.1±1.8 45.3±2.1 59.2±5.34FENA, % 8.2±0.8 26.8± 1.64 62.8±3.34 35.6± 1.84 12.2±2.0 78.5±4.74RPF, mlli/in 37.9±2.7 37.1±1.5 42.1±2.6 33.6±0.7 36.1± 1.2 31.0±0.54 35.0±2.0

Abbreviations used in this table: AVP, arginine vasopressin; OCAVP, organ clearance of AVP; GFR, glomerular filtrationrate; GCAVP= gloinerular clearance of AVP; PCAXP, peritubular clearance of AVP; UCAVP, urinary clearance AVP; FEAVP,fractional excretion of filtered AVP; FENA, fractional excretion of filtered sodium; RPF, renal perfusion flow.* Values are means±SEM.4 Significantly different when compared to controls using Duncan's new multiple range test vith P < 0.05.§ This value is an approximation acnd thus has not been compared statistically with the controls.

equal GFR) and peritubular clearanice to the organclearance of AVP(Table I, Fig. 3). In control kidneys,GFR(515+37 ,ul/min) was 45.4+3.8% of the OCAVP. Inthe absence of glucose, GFR fell to 262+40 ,ul/min,which was 26.3+2.8% of the OCAVP, and in the presenceof anoxia GFR fell to 87±+14 pul/min, being 8.3±1.1%of the OCAVP. Under both these conditions, despite asignificant fall in GFR, the OCAVPremained essentiallyunchanged from the control values because of an in-crease in peritubular clearance (Table I, Fig. 3). In thecontrol, glucose-free, an(d anoxic kidneys, there was asignificant inverse correlation (P < 0.001; r = -0.749)between the glomerular and peritubular clearance ofAVP (Fig. 3). Although the GFRwas not measured inthe obstructed kidneys, it is reasonable to assume thatit was severely reduced. Nevertheless, the OCA'P wasnot found to differ from the control values. Presum-ably, under these circumstances AVP was being re-moved almost entirely by peritubular clearance.

In kidneys perfused with high levels of AVP, therewas a significant fall in OCAVPdespite maintenanceof GFRat a level (485±7 ,ul/min) similar to the controlvalues. This would suggest that the rate of filtration ofAVP was maintained and that the fall in OCAVPwas aconsie(quence of a(l depressed rate of clearance fromii theperitubular circulation. In fact, peritubular clearance inthese kidneys was significantly lower than in the con-trol group (392±74 ul/min vs. 631±66 ul/min). In theouabain perfusedl kidneys, the decrease in OCAVPcouldbe accouinted for by the significant fall in GFR (195±+9 pU1/m1iI1).

Fraction(al excretio)n offiltered AVPand sodium (Ta-ble I). In control kidneys, the fractional excretion offiltered AVP was 38.1+2.6%, and the fractional excre-tion of sodium was 8.2+0.8%. Expressing the resultsin the control kidneys in a different manner, it can becalculated (urinary clearance AVP/OCAVP) that urinaryexcretion of AVPaccounted for 17.9+±0.9% of all the im-munoreactive AVPremoved by the kidney. In glucose-free perfusions, the glomerular clearance and urinarvclearance of AVP fell proportionately, hence fractionalexcretion of filtered AVP (44.9+1.9%) did not differsignificantly from that of the controls; fractional excre-tion of filtered sodium (26.8±1.6%) did, however, in-crease siginificantly. Similarly, in ouabain-treated kid-neys fractional excretion of filtered AVP (34.1±1.8%)was not increased, whereas there was a significacnt in-crease in the fractional excretion of sodium (35.6± 1.8%). Anoxia was associated with a proportionatelygreater fall in GFR than in urinary clearance of AVPor sodium resulting in a significant increase in the cal-culated fractional excretion of filtered AVP(49.1±2.6%)and filtered sodium (62.8±3.3%). Cold perfusions werealso associated with a significant increase in the frac-tional excretion of AVP (59.2±5.3%) and of sodium(78.5±4.7%). In kidneys perfused with high levels ofAVP there was an insignificant rise in the fractionalexcretion of AVP (45.3±2.1%) and of sodium (12.2±2.0%) as coompared to the controls.

Renial perfusion flow (Table I). Except for the coldlgroup, perfuision flow in the different treatment groupsdid not differ significantly from control values (37.9

Vaso pressis IIHandlinig by the Isolated Kidntey 9

1100

1000 lx x

4E3

c0

p

a

.0

0

0

0

.

0

r. -0.749

P ( 0.001

100 200 300 400 500 600 700

Glomerular Clearance A V P(ul /min)

FIGuRE 3 Relationship between peritubular clearance andglomerular clearance of AVP. The latter has been assumed tobe equal to the GFRand the former has been calculated asbeing equal to the difference between the organ clearanceof vasopressin and GFR. The calculated regression line forvalues from control (0), glucose-free (0), and anoxic (x) kid-neys has been drawn (y = 960.74 - 0.68x). In these treatmentgroups the organ clearance of vasopressin did not differ sig-nificantly from the controls. Each kidney is depicted and thedifferent groups are represented by specific symbols.

+2.7 ml/min). Cold was associated with marked vaso-constriction, and despite perfusing the kidneys at pres-sures -90 mmHg higher than the controls, the averageflow (31±0.5 ml/min) remained significantly lower thanthat of the controls.

DISCUSSION

The earlier work on the renal handling of vasopressinyielded conflicting reports that have resulted in con-siderable confusion. (See review of Lauson [1]). Theseconflicting reports have been a consequence of severalfactors including the use of relatively unreliable bio-assay procedures. Recently, with the advent of sensi-tive radioimmunoassays for vasopressin (7-10), muchof the difficulty in studying vasopressin metabolismhas been resolved. In the present study, the antibodyused in the radioimmunoassay was specific for AVPto the extent that it has been tested; possible metabo-lites of AVP have not been tested. Furthermore, assynthetic AVP was the sole peptide added to the per-fusion medium, we were assured that only immuno-

reactive AVP and possibly also immunoreactive frag-ments thereof were measured. It is possible that thekidney may release immunoreactive fragments into theperfusate or the urine which would result in some error,probably small, in the calculated AVP clearances. Ifpresent in the perfusate these fragments would resultin an underestimation of the organ clearance and peri-tubular clearance of AVP; if present in the urine only,an overestimation of urinary clearance would occur.Nevertheless, even if this did occur it would not alterthe validity of our overall description of the mannerin which the isolated kidney handles AVP. It is alsopossible that all the intact AVP removed is degradedwithin the kidney to small fragments without the re-lease of hypothetical immunoreactive fragments intoperfusate or urine. Our calculations would then ac-curately reflect the true clearance of immunoreactiveintact AVP. A further consideration is the possibilitythat the kidney may modify the structure of AVP withloss of immunoreactivity, but not biological activity.There is however no documented evidence that sucha process occurs.

In the past there has been considerable controversyregarding the question of whether or not AVP bindsto plasma proteins (1, 11). By showing that AVP doesnot bind appreciably to the albumin perfusing the iso-lated kidney, interpretation of our data has been simpli-fied. Thus it is reasonable to assume that AVP (molwt 1,084) passes through the glomerular filtration bar-rier with little hindrance, and that the glomerularclearance of AVP approximates the GFR. As the GFRin the control kidneys was less than half the OCAVP,it is apparent that filtration cannot account for all theAVP cleared by the kidney and that peritubular clear-ance must play a major role in the clearance of thehormone. AVP leaving the peritubular circulationcould theoretically be removed from the renal inter-stitium by tubular cells or by lymphatic drainage. How-ever, as lymph from the isolated kidney drains intothe perfusion medium, lymphatic drainage cannot ac-count for the AVP leaving the peritubular circulation.This suggests that uptake by the contraluminal aspectof the tubular cells occurs. It is likely that at leastpart of the AVP cleared from the peritubular compart-ment binds to specific receptors on the contraluminalplasma membranes of the distal tubule and collectingducts with subsequent activation of vasopressin-sensi-tive adenylate cyclase at this site (12-15). Apart fromserving to stimulate the formation of cyclic-AMP, theAVP removed from the peritubular compartment isprobably degraded by the rich supply of enzymes pres-ent in the plasma membrane (16) and in the cell (17-19). Although there is no evidence in support thereof,it is conceivable that the hormone may be returnedto the circulation after undergoing a minor structuralmodification without loss of biological activity, but with

10 R. Rabkin, L. Share, P. A. Payne, J. Young, and J. Crofton

loss of immunoreactivity, and hence not measured inthis study. It is also possible that some of the AVP istransported intact across the tubular cells and secretedinto the tubular lumen (20, 21).

Further evidence suggesting that peritubular clear-ance plays a major role in the removal of immuno-reactive AVP by the isolated rat kidney is providedby the observation that under certain conditions, (le-spite a marked fall in GFR, the perfused kidney main-tains a constant rate of AVP clearance. In this respect,the kidney handles AVPin the same manner as it doesinsulin, a polypeptide of mol wt 6,000 (22). Althoughdifficult to explain, there are several possible mecha-nisms for this observation. Based on histochemicalstudies, Venkatachalamn and Karnovsky (23) have sug-gested that the intrarenal distribution of protein leavingthe peritubular vessels may depend on regional ratesof tubular fluid reabsorption, bulk outward flow ofwater tending to wash the proteins toward the capil-laries. It is therefore conceivable that the fall in bulkflow associated with a drop in GFRmay enhance theperitubular clearance of AVP, accounting for the con-stancy of OCAVP. Another possible explanation is thatcontraluminal uptake may be regulated in part by theamount of AVP presented to and taken up by the lu-minal surface of the tubular cell. Thus when the filteredload drops, more AVPis extracted from the peritubularcompartment. Constancy of OCAVPat varying glomeru-lar filtration rates would also occur if AVPwere poorlyfiltered and removal were solely by peritubular clear-ance. However, as discussed earlier, AVP probablypasses through the glomerular barrier with little restric-tion. Finally, if all the filtered AVP were reabsorbedand returned unaltered to the peritubular circulation,then a drop in GFRwould not affect the OCAVP. Thisis unlikely in view of the rich supply of AVPdegradingenzymes in the kidney (17, 18).

The possibility of peritubular clearance of AVP haspreviously been raised by Ginsburg (20) and by Harveyet al. (21). However, the findings of Lauson et al. (24)were not in agreement, for they reported that the totalrenal AVP clearance in dogs was slightly less than theGFR. In these earlier studies AVP was measured bybioassay. With a sensitive radioimmunoassay, whichcorrelated closely with the bioassay, Shade and Share(25) found that at physiological plasma antidiuretic hor-mone levels and at normal renal plasma flows the totalrenal clearance of antidiuretic hormone in dogs wvasequal to the GFRand that the urinary AVP clearanceaveraged 59% of the GFR. However, when renalplasma flow was reduced to 25% of control levels, de-spite the cessation of glomerular filtration, renal AVPclearance only fell to 20% of the control values, sug-gesting the presence of a postglomerular mechanismfor clearing AVP.

In the present study, only 18% of the total amount

of AVP removed by the control kidneys was excretedin the urine, the balance presumably being degradedwithin the kidney (19). Fractional excretion of AVPwas 38% of the filtered load, indicating that AVP isremnoved from the tubular lumen. Theoretically, lumi-nal rem--oval could be achieved in several ways incelud-ing transport into the peritubular fluid of intact AVP,absoiption across the luminal membrane with intracel-lular degradation, degradation by the luminal plasmnamembrane or degradation by luminal fluid with reab-sorption, or excretion of the degradation products. Aswvith other peptides (26, 27), absorption of intact AVPanid any products of degradation would be expectedto take place in the proximal tubule and niot in thedistal nephron where, as discussed earlier, contralumi-nal uptake of AVP may occur. Absorbed AVP is pre-sumably degraded wvithin the tubular cells; transtubu-lar transport of intact hormone being unlikely in viewof the active AVP degrading system in the kidney(18). Although the exact fate of filtered AVP is notknowni, the report by Walters and Bowman(19) and thereview of Lauson (1) suggest that degradation occursbeyond the proximal tubule, possibly by the action of adistal luminal membrane-bound enzyme. Clearlyfurther studies are indicated to determine how thettubules han(dle filtered AVP.

Having established that both the luminal and thecontraluminal aspect of the tubular cells are exposedto the AVP cleared from the renal circulation, we ex-amine(l various factors that might affect the removalof the peptide by these two surfaces of the cells. Ininterpreting the results obtained, it should be recog-nized that we have made two assumptions: (a) thatthe various maneuvers employed do not alter the imi-permeability of the luminal membrane to inulin and(b) that in the presence of a normally functioning lumi-nal mechanism for removing AVP, urinary clearanceof AVP will fall in proportion to a fall in GFR, andfractional excretion of filtered AVP will remain con-stant, as occurred in our glucose-free and ouabain ex-periments. As to luminal impermeability to inulin, mi-croperfusion studies carried out in vivo (28) and in vitro(29, 30) indicate that luminal impermeability to inulinis not altered by inhibition of oxidative metabolism,cold (210C), ouabain (0.1 mM), or the absence of glu-cose. Hence it is likely that luminal impermeabilityto inulin was maintained under the various conditionseminployed in our study. Regarding the luminal handlingof AVP, anoxia and cold appeared to increase the frac-tional excretion of filtered AVP, which was nlot signifi-cantly affected by the absence of exogenous glucose,high levels of AVP or ouabain. As it has been showwnthat the degradation of AVP by tissue homogenatesis independent of oxygen (17), it appears that the in-creased fractional excretion of filtered AVP by theanoxic kidneys was a consequence of a depression of

Vasopressin Handlitng by the Isolated Kidney 11

a reabsorptive rather than a degradation process. Cold,which may inhibit enzyme activity and also transportprocesses, may have increased fractional excretion offiltered AVPby depressing degradation and reabsorp-tion. These findings are similar to those observed instudies on the handling of low molecular weight pro-teins (31, 22) (lysozyme and insulin) by the isolatedperfused rat kidney. Similarly, Miller et al. (32) foundthat the luminal uptake of horseradish peroxidase (molwt, 40,000) was abolished by cold and by interfer-ence with oxidative metabolism, but not by the absenceof exogenous fuels. These workers also found that oua-bain inhibited the luminal uptake of horseradish perox-idase, whereas in our study ouabain did not interferewith the fractional excretion of AVP, suggesting thatsmall peptides are handled in a manner different to thatof larger proteins. The possibility that inhibition ofluminal removal of AVP in the presence of cold oranoxia might be secondary to the inhibition of sodiumtransport and not a result of a direct effect on a meta-bolic energy-dependent transport process is worthy ofconsideration. However, in the kidneys perfused with-out glucose or in those perfused with ouabain, frac-tional excretion of filtered sodium increased signifi-cantly in the absence of a significant increase in thefractional excretion of filtered AVP, suggesting that atleast in part luminal AVP reabsorption is not linked tosodium reabsorption.

Extending our observations to the contraluminal as-pect of the tubular cell, we found that contraluminaluptake is also sensitive to cold, being virtually abol-ished at 10°C, and that this process does not requireexogenous glucose as a fuel. In contrast to its effecton the luminal uptake of AVP, anoxia did not inhibitcontraluminal uptake. On the contrary, although GFRfell to 8%of the OCAVPin the absence of oxygen, peri-tubular clearance increased and the OCAVPwas main-tained at control levels. This suggests that contralumi-nal uptake is not dependent on oxidative metabolismand, furthermore, that peritubular clearance of AVP isable to increase in proportion to the fall in glomerularclearance of AVP and so maintain a relatively con-stant OCAVP.

The ability of the kidney to clear AVP from the peri-tubular circulation and maintain the OCAVPat controllevels was inmpaired by high levels of AVPand by cold.High levels of AVP, although not interfering with lumi-nal uptake, may have impaired contraluminal uptakeby saturating contraluminal AVP degradation, trans-port, or receptor binding. Cold may also have inhibitedcontraluminal uptake by inhibiting AVP degradation,transport, or receptor binding (14). In addition, as bothAVP and cold are vasoactive agents, it is possible thata redistribution of intrarenal perfusate flow (33) mayhave impaired AVP clearance from the peritubular cir-culation.

The data presented in this study indicate that AVPis removed by the tubular cells from the lumen of thetubules and from the peritubular compartment. In addi-tion the study suggests that removal of AVP is a tem-perature-dependent process consisting of two differentsystems. The one system that is associated with theluminal aspect of the cell is not depressed by highconcentrations of AVP, but may possibly depend on oxi-dative metabolism. The other system associated withthe contraluminal aspect of the tubular cell does notappear to require oxidative metabolism but is de-pressed by high concentrations of AVP. It is noteworthythat under several conditions when GFRfalls, decreas-ing the amount of AVP cleared by filtration and thusthe load of AVP handled by the luminal aspect of thecell, peritubular clearance increases and the rate atwhich the kidney clears AVP remains constant.

In interpreting the data presented in this study, rec-ognition must be made of the differences between re-nal function in the isolated organ and in the intactanimal. In the isolated kidney GFRis reduced, whereasperfusion flow is several fold greater than in vivo, asituation that might tend to exaggerate events on thecontraluminal aspect of the tubules and also reducethe role of filtration. Thus, the role of peritubular AVPclearance would probably be exaggerated and that ofglomerular AVP clearance reduced. In addition to aspecies difference this may accouint for the quantitativedifference noted earlier, between the renal handling ofAVP in the intact dog and the isolated rat kidney.Nevertheless, the isolated perfused rat kidney has al-lowed an approach to studying the renal handling ofAVP under a variety of conditions and in particularhas permitted the study of peritubular events. This hasresulted in the unequivocal demonstration of a peri-tubular mechanism for AVP removal that differs fromthat of the luminal mechanism.

ACKNOWLEDGMENTS

We gratefully acknowledge the technical assistance of DonMiles and the secretarial assistance of Martha Perser andJoanne Brown.

This study was supported by grants in aid from the Ameri-can Heart Association (76-876), with funds contributed in partby the Tennessee Heart Association, and grants AM19774-01and HL12990 from the U. S. Public Health Service. Com-puter assistance was supported by U. S. Public Health Servicegrant HL-09495.

REFERENCES

1. Lauson, H. D. 1974. Metabolism of the neurohypophysialhormones. Halatdb. Physiol. 4(Sect. 7. Endocrinology):287-393.

2. Nishiitsutsuji-Uwo, J. M., B. D. Ross, and H. A. Krebs.1967. Metabolic activities of the isolated perfused rat kid-ney. Biocheml. J. 103: 852-862.

3. Ross, B. D., F. H. Epstein, and A. Leaf. 1973. Sodium

12 R. Rabkin, L. Share, P. A. Paytne, J. Young, and J. Croftoni

reabsorption in the perfused rat kidney. Am. J. Physiol.225: 1165-1171.

4. Shade, R., and L. Share. 1976. Metabolic clearance ofimmunoreactive vasopressin and immunoreactive [131II-iodo vasopressin in the hypophysectomized dog. Endo-crinology. 99: 1199-1206.

5. Mortimore, G. E., F. Tietze, and D. Stetten, Jr. 1959.Metabolism of Insulin-I 31. Studies in isolated, perfusedrat liver and hindlimb preparations. Diabetes. 8: 307-314.

6. Kirk, R. E. 1968. Experimerital Design: Procedures forthe behavioral sciences. Roger E. Kirk, editor. WadsworthPublishing Co., Belmont, Calif. 93-94.

7. Robertson, G. L., L. A. Klein, J. Roth, and P. Gorden.1970. Immunoassay of plasma vasopressin in man. Proc.Natl. Acad. Sci. U. S. A. 66: 1298-1305.

8. Beardwell, C. G. 1971. Radioimmunoassay of argininevasopressin in human plasma.J. Clin. Endocrinol. Metab.33: 254-260.

9. Robertson, G. L., E. A. Mahr, S. Athar, and T. Sinha.1973. Development and clinical application of a newmethod for the radioimmunoassay of arginine vasopres-sin in human plasma. J. Clin. Invest. 52: 2340-2352.

10. Husain, M. K., N. Fernando, M. Shapiro, A. Kagan, andS. M. Glick. 1973. Radioimmunoassay of arginine vaso-pressin in human plasma.J. Clin. Endocrinol. Metab. 37:616-625.

11. Baumann, G., and J. F. Dingman. 1976. Distribution,blood transport, and degradation of antidiuretic hormonein man. J. Clin. Invest. 57: 1109-1116.

12. Dousa, T., 0. Hechter, I. L. Schwartz, and R. Walter.1971. Neurohypophyseal hormone-responsive adenylatecyclase from mammalian kidney. Proc. Natl. Acad. Sci.U. S. A. 68: 1693-1697.

13. Campbell, B. J., G. Woodward, and V. Borberg. 1972.Calcium-mediated interactions between the antidiuretichormone and renal plasma membranes. J. Biol. Chem.247: 6167-6175.

14. 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.

15. Rajerison, R., J. Marchetti, C. Roy, J. Bockaert, and S.Jard. 1974. The vasopressin-sensitive adenylate cyclaseof the rat kidney. J. Biol. Chem. 249: 6390-6400.

16. Musolan, C., and B. J. Campbell. 1975. Inactivation ofantidiuretic hormone by plasma membranes isolated fromporcine kidney medulla. Rev. Roum. Biochim. 12: 47-52.

17. Dicker, S. E., and A. L. Greenbaum. 1956. Inactivationof the antidiuretic activity of vasopressin by tissue ho-mogenates. J. Physiol. (Lond.). 132: 199-212.

18. Walter, R., and H, Shlank. 1975. Differences in the en-

zymatic inactivation of arginine vasopressin and oxytocinby rat kidney homogenate. Endocrinology. 96: 811-814.

19. Walter, R., and R. H. Bowman. 1973. Mechanism of in-activation of vasopressin and oxytocin by the isolated per-fused rat kidney. Endocrinology. 92: 189-193.

20. Ginsburg, M. 1957. The clearance of vasopressin fromthe splanchnic vascular area and the kidneys. J. Endo-crinol. 16: 217-226.

21. Harvey, N., J. J. Jones, and J. Lee. 1967. The renal clear-ance and plasma binding of vasopressin in the dog.J. En-docrinol. 38: 163-171.

22. Rabkin, R., and A. E. Kitabchi. 1978. Factors influencingthe handling of insulin by the isolated rat kidney.J. Clin.Invest. 61: 169-175.

23. Venkatachalam, M. A., and M. J. Kamovsky. 1972. Extra-vascular protein in the kidney. An ultrastructural study ofits relation to renal peritubular capillary permeability us-ing protein tracers. Lab. Invest. 27: 435-444.

24. Lauson, H. D., M. Bocanegra, and C. F. Beuzeville. 1965.Hepatic and renal clearance of vasopressin from plasmaof dogs. Am. J. Physiol. 209: 199-214.

25. Shade, R. E., and L. Share. 1977. Renal vasopressin clear-ance with reductions in renal blood flow in the dog. Am.

J. Physiol. 1: F341-F347.26. Pullman, T. N., S. Oparil, and F. A. Carone. 1973. Fate

of labeled angiotensin II microinfused into individualnephrons in the rat. Am. J. Physiol. 228: 747-751.

27. Carone, F. A., T. N. Pullman, S. Oparil, and S. Nakamura.1976. Micropuncture evidence of rapid hydrolysis ofbradykinin by rat proximal tubule. Am. J. Physiol. 230:1420-1424.

28. Green, R., and G. Giebisch. 1975. Ionic requirements ofproximal tubular sodium transport. I. Bicarbonate andchloride. Am. J. Physiol. 229: 1205-1215.

29. Schafer, J. A., S. L. Troutman, M. L. Watkins, and T. E.Andreoli. 1978. Volume absorption in the pars recta. I."Simple" active Na+ transport. Am. J. Physiol. 3: F332-F339.

30. Schafer, J. A., C. S. Patlak, and T. E. Andreoli. 1975. Acomponent of fluid absorption linked to passive ion flowsin the superficial pars recta. J. Gen. Physiol. 66: 445-471.

31. Maack, T. 1975. Renal handling of low molecular weightproteins. Am. J. Med. 58: 57-64.

32. Miller, A. T., D. M. Hale, and K. D. Alexander. 1965.Histochemical studies on the uptake of horseradish by ratkidney slices. J. Cell Biol. 27: 305-312.

33. Johnson, M. D., C. S. Park, and R. L. Malvin. 1977. Anti-diuretic hormone and the distribution of renal corticalblood flow. Am. J. Physiol. 232: Fl1-F116.

Vasopressin Handling by the Isolated Kidney 13