new the role of leakage of tubular fluid in anuria due to mercury...

Journal of Clinical InvestigationVol. 46, No. 5, 1967

The Role of "Leakage" of Tubular Fluid in Anuria Dueto Mercury Poisoning *

NORMANBANK,t BERTRANDF. MUTZ, ANDHAGOPS. AYNEDJIAN(From the Department of Medicine, New York University School of Medicine,

New York, N. Y.)

Summary. The role of "leakage" of tubular fluid in anuria produced bymercury poisoning was studied in rats by micropuncture techniques. Afteran initial brisk diuresis, almost all animals were completely anuric 24 hoursafter HgCl2 injection. Lissamine green injected intravenously in the earlystage of anuria appeared in the beginning of the proximal tubule, but the colorbecame progressively lighter as the dye traversed the proximal convolutions.The dye was barely visible in the terminal segments of the proximal tubule;it did not appear at all in the distal tubules. These observations suggestthat the proximal epithelium had become abnormally permeable to Lissaminegreen.

Tubular fluid to plasma inulin (TF/PIn) ratios and inulin clearance weremeasured in individual nephrons at three sites: early proximal tubule, lateproximal tubule, and distal tubule. It was found that TF/PN ratios wereabnormally low in the late proximal and distal tubules. Inulin clearance wasnormal at the beginning of the proximal tubule but fell by more than 60% bythe late proximal convolutions. Thus, the proximal tubule had also becomepermeable to inulin.

We conclude from these observations that anuria in mercury poisoningcan occur in the presence of a normal glomerular filtration rate. The ab-sence of urine flow appears to be due to complete absorption of the filtratethrough an excessively permeable tubular epithelium. The driving force af-fecting this fluid absorption is probably the colloid oncotic pressure of theperitubular capillary blood.

IntroductionThe pathologic physiology of anuria in acute

renal failure remains obscure in spite of extensivestudy. The three mechanisms that have been pos-tulated are: a severe reduction in renal blood flowand glomerular filtration rate (GFR), obstructionof the tubules by casts or edema, and excessive

*Submitted for publication July 12, 1966; acceptedJanuary 4, 1967.

This work was supported by grants from the NationalHeart Institute (HE-05770), the American Heart Asso-ciation, and the Life Insurance Medical Research Fund.

tCareer Scientist of the Health Research Council ofNew York City. Address requests for reprints to Dr.Norman Bank, Dept. of Medicine, New York Uni-versity School of Medicine, 550 First Ave., New York,N. Y.

back diffusion of the filtrate through a damaged tu-bular epithelium. To our knowledge, only threestudies have been published in which these variouspossibilities were examined by direct observationof the kidney in zrvo in animals with experimentalacute renal failure (1-3). Richards reportedthat, in the anuric stage of cyanide or mercurypoisoning in frogs, flow of fluid from the glomeru-lus into the beginning of the proximal tubule wasnormal or even increased (1). The failure of anyurine to appear in the ureter was attributed togross leakage of tubular fluid from more distalsegments of the proximal tubule. More recently,Oken and his co-workers, studying acute renalfailure due to mercury poisoning (2) and hemo-globinuria (3) in rats, found GFR per nephron

695

BANK, MUTZ, AND AYNEDJIAN

to be significantly reduced. They concluded thatanuria in these two experimental models resultedfrom intrarenal vasoconstriction and a subsequentfall in filtration rate.

In the present study, we have reinvestigatedthe role of leakage of tubular fluid in anuriacaused by mercury poisoning. The techniquesused were rapid sequence photography of thesurface tubules after intravenous Lissamine greeninjection, and quantitative measurements of inu-lin clearance at various points along the nephron.The findings support the view that the permeabilityof the tubules is markedly increased and non-selective during the anuric stage of mercurypoisoning. Glomerular filtration rate, on the otherhand, as evaluated by inulin clearance measuredin early segments of the proximal tubule, waswithin the normal range. Weconclude, therefore,that the mechanism of anuria in these animals wasexcessive back diffusion of the glomerular filtratethrough abnormally permeable tubules.

MethodsMale white rats weighing between 225 and 300 g were

fed a sodium-free diet for 3 or 4 days before the induc-tion of acute renal failure. The reason for eliminatingsodium from the diet was to determine whether a reab-sorptive defect for sodium occurred during the courseof the illness. After 3 or 4 days on the diet, a controlurine specimen was collected in a metabolic cage. Tapwater was allowed ad libitum during the collection ofurine, but food was withheld. Water intake as well asurine output was measured in each animal. Acute renalfailure was then induced by subcutaneous injection ofHgCls solution in a dosage of 3.0 to 3.5 mg per kg bodyweight of the mercuric ion. At this dosage, all animalssurvived at least 48 hours, and 60% underwent a diure-sis and recovered after a period of anuria. During thefirst 16 hours after HgCln injection, a marked diuresisoccurred, but beginning at about 20 hours after injectionand lasting for an additional 24 to 30 hours, the ratswere either totally anuric or severely oliguric. All Lis-samine green and micropuncture studies during the anuricstage of the disease were carried out between 24 and 48hours after Hg++ injection. In 22 animals that survivedthe anuric period, daily urine collections were continuedfor up to 2 weeks, and in 8 rats of this group, observa-tions on the function of surface tubules were made withthe aid of Lissamine green at various intervals duringthis time. The measurements made on all urine speci-mens were volume, sodium and potassium concentrations(internal standard flame photometery), and osmolality(Fiske osmometer).

The surgical preparation of the rats was as describedpreviously (4). The left ureter was catheterized with

mOsm/Kg H20 600d

mEq/L ' t ltoo~ ~~~a

mEq/L 50>

250-.

350

URINE 25

HgCL2 DAYSinjection

FIG. 1. URINE VOLUME, SODIUM, POTASSIUM, AND OS-MOLALITY AFTER HgCls INJECTION. The rats were fed asodium-free diet for 3 or 4 days before HgCl2 injectionand continued on the diet for the 8 days afterward. Thevertical lines indicate the observed range.

polyethylene Lubing and the bladder emptied by syringeat the beginning and end of each experiment in order todetermine whether there was any urine flow from eitherkidney. Mean arterial blood pressure was measuredin all experiments by a U-tube mercury manometer con-nected through its sidearm to a catheter secured in thecarotid artery. Only those experiments were acceptedin which mean blood pressure remained between 100 and130 mmHg. Approximately 4 to 5 ml of isotonic NaClwas given intravenously during surgical preparation ofthe rat to replace body fluid losses and to preclude thepossibility of salt depletion. A constant infusion of iso-tonic NaCl was administered at 0.025 ml per minutethroughout the experiment.

Lissamine green experiments. In 15 rats, 24 to 48hours after HgCl2 administration, the surface of the kid-ney was photographed sequentially after rapid intravenousinjection of 0.04 to 0.07 ml of 10%o Lissamine green, asdescribed by Steinhausen (5). In these animals, therenal capsule was stripped away to improve the clarityof the photographs, and the kidney was bathed in con-tinuously flowing saline warmed to 370 C. The photo-graphic equipment consisted of a Leitz Ultropak ob-jective and illuminating system, a Zeiss Ukatron elec-tronic flash unit (operated at 60 watt-seconds), and aLeica camera back with a battery-operated motor forautomatic film advance. High speed Ektachrome film(ASA 160) was used throughout. Pictures were takenat various intervals after Lissamine green injection, de-pending upon the appearance of the dye in the struc-tures seen on the surface of the kidney.

696

TUBULARLEAKAGEIN MERCURYPOISONING

2

3

4

NORMAL MERCURYFIG. 2.

697


Measurement of TF/PJn ratios and inulin clearanceper nephron. In eight rats, 24 to 30 hours after Hg++injection, TF/Pln ratios and inulin clearance were mea-sured at various points along the proximal and distaltubules. A priming dose of 25 Ac "4C-labeled inulincarboxylic acid' was given intravenously, and a sus-taining dose of 0.4 ,c per minute in isotonic saline wasstarted 30 minutes before tubular fluid collections werebegun. Early proximal segments were identified by in-jecting Lissamine green intravenously and observing thefirst intratubular appearance of the dye. Late proximalsegments were identified as those clustered about promi-nent branching capillaries (6), and distal tubules by thethin, glistening appearance of their epithelium. Tubularfluid collections were made from each of these three lociin each animal. A column of heavy castor oil stainedwith Sudan black was maintained just distal to the col-lection site to assure that all fluid coming from the glo-merulus was aspirated into the pipette. The period ofcollection was carefully timed with a stopwatch and thevolume of the collected fluid measured as described byWindhager and Giebisch (7). The inulin concentrationin tubular fluid and plasma was measured by liquid scin-tillation counting as previously described (8). Plasmainulin concentration at the midpoint of each tubular fluidcollection was determined from the plasma inulin curve,blood samples being collected from the cut tail at 30-minute intervals during the experiment. The plasmainulin concentrations were corrected for a plasma watercontent of 94%. Inulin clearance per nephron was cal-culated from the expression TF flow X TF/PIn. Atthe end of each tubular fluid collection, the nephron wasinjected with colored latex and the site of the collectiondetermined more precisely by microdissection.

In five normal rats also prepared on a sodium-free dietfor 3 or 4 days, the potent diuretic furosemide 2 was ad-ministered intramuscularly in a dose of 40 mg per kgbody weight. Urine output and water intake were mea-sured for the next 24 hours after which time TF/Plnratios and inulin clearance along the nephron were mea-sured as described above.

Results

Urine volume, sodium, potassium, and osmo-lality (Figure 1). The data obtained from a totalof 35 rats followed for various lengths of timeup to 8 days after HgCl2 injection are shown inFigure 1. All animals were maintained on a so-dium-free diet for 3 or 4 days before Hg++ ad-ministration and were continued on the diet dur-ing the 8 days afterward. During the first 16hours after Hg++ injection, a marked diuresis oc-

1 New England Nuclear Corp., Boston, Mass.2 The furosemide was kindly supplied by Hoechst Phar-

maceuticals, Inc., Cincinnati, Ohio.

curred, the average urine volume being 18 ml.Potassium constituted the major urinary cationduring the diuresis, as it did in the control period.Water ingestion paralleled or exceeded urine out-put during the 16-hour period, the average intakebeing 25 ml. Starting at about 20 to 24 hours,urine flow fell to zero in almost all animals andwas greatly reduced in the others. Between 24and 48 hours after Hg++ injection, the animalswere virtually anuric, even though water ingestionremained normal in many of them. Urine outputresumed in 22 of the 35 animals after this periodand exceeded control volumes on the third andfourth days after Hg++. Sodium excretion wasmarkedly increased during the early recovery pe-riod in spite of the absence of sodium in the diet.Urinary potassium excretion, on the other hand,was lower than in the control period, as was totalurinary osmolality. Between the fourth andseventh days, urinary sodium, osmolality, potas-sium, and volume all returned to control levels.Although the entire anuric and recovery periodswere telescoped as compared with the usual courseof acute renal failure in man (9), the pattern of re-ciprocal changes in urinary sodium and potassiumconcentrations and the reduction of urinary osmo-lality are typical of the findings in man (10).

Appearance of the kidney and observations withLissamine green (Figure 2). At 24 to 28 hoursafter 3.0 to 3.5 mg per kg of Hg++ was injectedsubcutaneously, the epithelial lining of the proxi-mal tubules appeared poorly defined, pale, andgranular in those segments closest to prominentbranching surface blood vessels. These blood ves-sels, one of which can be seen clearly in the upperleft photograph in Figure 2, have been identifiedas efferent arterioles and the adjacent tubularsegments as the most distal portions of the proxi-mal convolutions present on the surface of thekidney (6). Other tubular segments, presum-ably proximal to the abnormal ones, appearedmore normal. When they were examined between30 and 48 hours after Hg++ injection, a more dif-fuse involvement of the proximal tubules was seen,with almost all segments showing a loss of theusual sharp definition characteristic of the normalproximal epithelium. Proximal tubule luminacould be seen throughout the anuric stage of thedisease, but their diameters often could not bejudged accurately because of the granularity and

698


opacity of the overlying epithelium. In contrast

to the normal kidney, only a few patent distal tu-

bules could be recognized in kidneys examinedbetween 24 and 48 hours, but their thinner more

glistening epithelium appeared normal.The appearance of Lissamine green after rapid

intravenous injection is illustrated in Figure 2.A sequence of pictures of a normal kidney isshown at the left from top to bottom, and one of a

Hg++-poisoned kidney is shown on the right. Thelatter experiment was carried out 24 hours afterHg++ injection. At the time of study, the animalhad been totally anuric for 8 hours. Moreover, no

urine appeared in either the left ureteral catheteror bladder during the experiment. In both ani-mals, Lissamine green appeared in the surfaceperitubular capillaries 4 seconds after its intrave-nous injection. The dye next appeared in the firstsegment of the proximal tubule 7 seconds afterinjection. After this point, however, the color in-tensity decreased progressively in the Hg+-poisoned animal as the dye traversed the proximalconvolutions. Toward the end of the proximaltubule, in those segments nearest to the prominentbranching arterioles, only a diffuse light green

color could be seen. In contrast, in the normalrat, the dye intensity remained essentially un-

changed from the beginning to the end of the prox-

imal tubule. The marked difference between thenormal and diseased kidneys can be best appreci-ated by comparing adjacent photographs in Fig-ure 2, which were taken at comparable times afterLissamine green injection. In addition to theseobservations, we noted that Lissamine green never

reappeared in the distal tubules of the anuric ani-mals, as it does in normal rats (5). This was thecase even with the few patent distal tubules thatwere recognized by their characteristic thin epi-thelium. Similar observations were made in 14other anuric rats.

TF/PI, ratios and inulin clearance per nephron(Figure 3, Tables I and II). In eight rats stud-ied at 24 to 30 hours after Hgr+ injection, TF/Pi.ratios were measured at various points along thenephron, and inulin clearance per nephron was de-

FIG. 2. APPEARANCEOF KIDNEY SURFACEIN VIVO DUR-

ING PASSAGEOF LISSAMINE GREEN. In both the normaland anuric animals, the pictures were taken at 4, 7, 15,and 25 seconds, respectively, after intravenous injectionof 10% Lissamine green. Magnification 110 X.

termined. In five of the eight, no urine appearedin either the ureteral catheter draining the experi-mental kidney or in the bladder during the 3 to 4hours of the experiment. In the other three ani-mals, a volume of urine too small to fill the ure-

teral catheter was produced by the experimentalkidney. The urinary to plasma inulin (U/P11)ratios in these three animals were 9.56, 3.95, and1.91, respectively.

The control group consisted of five normal ratsfed a sodium-free diet for 3 or 4 days and givenfurosemide intramuscularly 24 hours before mi-cropuncture study. Urine output on the day pre-

ceding the experiment averaged 26.4 ml and wateringestion, 26.0 ml. In contrast to the Hg++-in-jected animals, urinary sodium concentration rose

markedly during the diuretic period to 43 mEqper L. During micropuncture on the followingday, each of the five control rats excreted urine at

nondiuretic rates, the average flow from the ex-

perimental kidney being 8.1 ul per minute per kg.The U/P1,, ratio in this group averaged 484.

The TF/P10 ratios are shown in Figure 3plotted on the ordinate against sites of collectionon the abscissa. The ratios were strikingly low in

both the proximal and distal tubules of the anuricanimals, as compared with the control animals andwith normal values reported by numerous investi-gators (7, 8, 11-13). The distal TF/Pin ratiosfound in the furosemide-treated rats are somewhatlower than those previously reported for non-

diuretic normal rats, perhaps due in part to the

500 S Hgq- POWNEDX4.RC RATS 0

0 NON-DI~ElC CONTROLRATS0

4.00- 0 00

TF/P,, 0 0

3.00~~~~~~~~~~0 00

2DO 0 *

00 0*

1.00 W,,^l, ,* .,10 aD 30 40 s0 670 0 20 40 60 80 l00PER CENT PROXMALTUBLE PER CENT DISTAL TUW..E

FIG. 3. TUBULARFLUID TO PLASMA INULIN (TF/Pi.)RATIOS IN ANURIC Hg++-POISONED RATS, AND NONDIURETICCONTROLS. A striking inability to concentrate inulin isevident in the anuric animals.

699


TABLE I

Inulin clearance measured at various points along the nephronin a mercury-poisoned anuric rat*

InulinTF TF flow clearance

Time sample TF/Pin rate per nephron Location

minutes nl/ nl/ %lengthmin/kg min/kg

0 Infusion: NaCl, 150 mEqper L at 0.125 ml per minute40 Inulin-14C, 25 uc intravenously41 Infusion: NaCl, 150 mEq per L; inulin-14C, 17 puc per

ml at 0.025 ml per minute92- 99 1 1.10 92 101 15 P

115-124 2 1.54 20 31 78 D144-151 3 1.29 40 52 61 P170-177 4 1.08 35 38 65 P192-202 5 1.39 14 19 80 D222-237 6 0.98 10 10 56 D

* Abbreviations: TF/Pi. = tubular fluid to plasma inulin ratio;nl = nanoliters; %length = per cent of total length of proximal ordistal tubule.

large sodium diuresis on the preceding day and asubsequent decrease in medullary hypertonicity.

In addition to TF/PIn ratios, tubular fluid flowrates were measured in these experiments, andinulin clearance per nephron was calculated for var-

TABLE II

Inulin clearance measured at various pointsalong the nephron*

Early Lateproximal proximal Distal

nt/min/kg nt/min/kg nl/min/kgHg++ 110 426t 40.5 426.2 28.6 +20.6

rats p <0.001$Control

rats 103±421 88.9±26.0 91.2420.80.1 <p <0.2 0.2 <p <0.3

* Inulin clearance per nephron was calculated from the expressionTF flow X TF/Pin.

t Data expressed as mean 4 standard deviation.* p values compare late proximal or distal tubular inulin clearance

with early proximal tubular inulin clearance.

ious collection sites. A typical experiment on ananuric rat is shown in detail in Table I, and thedata from all eight anuric and five control animalsare summarized in Table II. Inulin clearancemeasured in early proximal segments was notsignificantly different from that measured in thecontrol animals or that previously reported fromthis laboratory in which random collections from

FIG. 4. APPEARANCEOF KIDNEY SURFACE IN VIVO AFTER CLINICAL RECOVERYFROM HgC1S-INDUCED ACUTE RENAL

FAILURE. Intraluminal cellular debris obstructs many tubule segments. The internal diameter of the functioningtubules is 25%o larger than normal. Magnification 65 X.

700


the proximal tubule had been made (11). Whenmeasured in the late proximal segments, however,inulin clearance averaged only 37% of the earlyproximal values and in the distal tubule fell to26%. In the control rats, the site of tubular fluidcollection had no apparent influence upon calcu-lated inulin clearance per nephron.

Recovery stage (Figure 4). In eight rats thathad recovered from acute renal failure, Lissaminegreen photographic studies were carried out atperiods ranging from 5 to 17 days after Hg++ in-jection. Urinary sodium, potassium, and osmo-lality had all returned to control levels by thistime. In spite of the apparently normal urine,however, examination of the surface of the kidneyrevealed numerous tubules that contained whatappeared to be cellular debris. Lissamine greendid not pass through these segments, and we as-sumed that such nephrons were partially or com-pletely blocked. Adjacent tubules through whichLissamine green did flow were significantly di-lated. A typical example of the appearance of thesurface of the kidney 17 days after Hg++ injectionis shown in Figure 4. The heterogeneity of thenephron population is clearly evident. The aver-age luminal diameter of the functioning tubules inthis field was 29.5 pu, as compared with an averageof 23.5 in the normal rats. These observationssuggest that tubular obstruction, which can occurduring the anuric stage of mercury poisoning (2),may persist for some time after the diuresis andcould possibly account for certain functional ab-normalities found long after clinical recovery froman episode of acute renal failure (14, 15).

DiscussionSeveral observations indicate that the Hg++-in-

jected rats were in a true state of acute renal fail-ure when Lissamine green and micropuncturestudies were carried out. First, after an initialdiuresis induced by the HgCl2 injection, no urinewas excreted for periods ranging from 4 to 24hours before study, in spite of the fact that waterintake remained normal in most animals. Fur-thermore, during the acute studies where oneureter was catheterized and the bladder emptied bysyringe, little or no urine appeared from eitherkidney, even though 4 to 5 ml of saline had beenadministered during the surgical preparation ofthe animal. It seems unlikely that the absence of

urine flow was due to dehydration secondary tothe large diuresis on the preceding day, since waterintake during the diuretic period exceeded urineoutput. To exclude this possibility more rigor-ously, however, we injected normal sodium-de-prived control rats with furosemide, which induceda significantly greater diuresis than did the HgCl2and, moreover, resulted in considerable loss ofsodium in the urine.3 In contrast to the Hg++-poisoned animals, all furosemide-treated rats ex-creted quantities of urine during the acute ex-periments that were easily measurable. Thus, theabsence of urine flow in the Hg++-injected ratscannot be attributed to dehydration. Finally, inthree of the Hg++-poisoned animals in which asmall amount of urine did appear in the ureteralcatheter, U/P1n ratios averaged only 5.14. Inthe control rats with larger urine flows, U/PI,ratios averaged 484. These observations are com-patible with a diagnosis of acute renal failure inthe HgCl2-injected rats.

In the experiments with Lissamine green in theanuric animals, the dye appeared in the capillarieson the surface of the kidney and in the beginningof the proximal tubule in what was judged to be anormal amount and time after its intravenous in-jection. However, as the dye progressed alongthe proximal convolutions, its concentration ap-peared to decrease markedly so that in most casesonly a pale diffuse coloration was visible in theterminal segments on the surface of the kidney.Moreover, no dye was ever seen to reappear inthe distal tubules as it does in normal rats, where,in the nondiuretic state, it is seen in its most con-centrated form (5). It seems likely that theseobservations were due to an increase in the perme-ability of the tubules to Lissamine green ratherthan to some alteration in the hydrodynamics oftubular fluid flow. Although it is theoreticallypossible that an abnormally slow moving column oftubular fluid might allow longitudinal diffusion ofLissamine green and therefore dilution of the dye,

3 The fact that the HgCl-induced diuresis was accom-panied by a large increase in K+ excretion but not in Na+excretion could be explained by a proximal site of actionof the compound. Inhibition of proximal Nat reabsorp-tion could result in a large distal Na-K exchange, es-pecially since the rats were salt-deprived. Furosemide,on the other hand, has been shown to act principally onthe ascending limb of the loop of Henle (16), and thusincreased Nat excretion in spite of salt deprivation.

701


this has not been observed during acute reductionsin GFRproduced by controlled hemorrhage (17,18), constriction of the abdominal aorta (13), or

increased ureteral pressure (13). With very se-

vere reductions in filtration rate and slowing oftubular fluid flow produced by hematin injection,color intensity is increased in the proximal tubule,and the dye is almost black when it reaches thedistal tubule (19). Finally, the rate of tubularfluid flow in early segments of the proximal tu-

bule was within normal limits in the Hg++-poisonedrats, as indicated by the inulin clearance measure-

ments in Table II. Thus, the loss of Lissaminegreen color observed in these experiments cannot

be accounted for by a reduction in filtration rate

or in tubular fluid flow. Although retention ofwater in the tubular lumen as occurs, for ex-

ample, in partially or completely obstructed neph-rons might lead to dilution of the dye, this expla-nation seems unlikely, since the proximal tubuleswere not obviously dilated and it has been demon-strated that intraluminal pressure is not elevatedin Hg++-produced anuria (2). The observationsare most compatible, therefore, with the view thatthe tubules had become excessively permeable to

Lissamine green. Since there was a concentra-

tion gradient of the dye between tubular lumen andperitubular capillary blood (the vascular phase ofdye passage is over before the tubular phase be-gins, as seen in Figure 2), dye might have beenlost by diffusion down a concentration gradient.The site of increased permeability appeared to bethe more distal segments of the proximal convo-

luted tubule in the earliest stage of the diseasestudied, but deeper segments such as the pars

recta and loops of Henle may also have been in-volved. Our observations are in agreement withthose of Richards (1), who described leakage ofphenolsulphonphthalein from the proximal tubulesof both excised and in situ frog kidneys that hadbeen poisoned with HgCl2.

Further evidence of loss of selective permeabilityof the tubules was found in the micropuncture ex-

periments. Measurement of TF/P1. ratios alongthe proximal and distal convoluted tubules showeda striking inability to concentrate inulin. Thehighest ratio found in the proximal tubule was

only 1.40 and that in the distal tubule, 2.08, andmany measurements were not significantly higherthan 1.0. These values differ markedly from

those measured in the proximal and distal tubulesof the nondiuretic control animals. That theselow TF/P1n ratios were not due to retention ofwater in the lumen but rather to leakage of inulinout of the lumen is indicated by the inulin clear-ance per nephron measurements summarized inTable II. In the control animals, inulin clearancewas essentially uniform when measured at vari-ous sites along the nephron. In the experimentalanimals, however, an average fall in inulin clear-ance of 63%o was found when samples were col-lected from the late proximal tubule, and a fur-ther fall was observed in the distal tubule. Theaverage value for distal tubular inulin clearancemay be misleadingly high, however, as it repre-sents collections from what appeared to be a re-duced number of patent distal tubules. An ap-parent reduction in the number of recognizabledistal tubules could be the result of complete ab-sorption of the glomerular filtrate more proximallyin many of the nephrons, with subsequent collapseof the distal lumina. In any case, the data indi-cate that much of the filtered inulin had left the lu-men along the length of the proximal tubule, andthat even more was lost before fluid reached thedistal convolution. These observations thus pro-vide a physiological counterpart for histologicaland histochemical studies that have shown the ma-jor site of injury in mercury poisoning to be themore distal segments of the proximal convolutedtubule and the pars recta (20-22).

From the data in Table II, it should be notedthat inulin clearance measured in the early proxi-mal tubule was essentially the same in the anuricanimals as in the control rats; both values agreeclosely with those previously reported for normalrats (11). Thus, glomerular filtration rate wasapparently within the normal range in the totallyanuric animals. This observation also agreeswith that of Richards, who noted a brisk flow offluid from the glomerulus into the beginning ofthe proximal tubule in anuric mercury-poisonedfrogs (1). Since our data demonstrated a majordisruption of tubular permeability, we assumedthat anuria in these animals was due to completeabsorption of a normally formed glomerular fil-trate along the length of the nephron. The siteof final absorption of the filtrate and collapse ofthe lumen was presumably distal to the surfaceproximal convolutions, since in the early stages of

702


total anuria fluid could still be aspirated from themost distal portions of the proximal tubule on thesurface of the kidney.

The mechanism of complete absorption of theglomerular filtrate through an abnormally perme-able epithelium is not entirely clear. Since mer-cury poisoning probably inhibits active sodiumtransport (23), some other driving force must havebeen responsible for passive movement of wateracross the epithelium. The only other force ofsufficient magnitude to accomplish this is the col-loidal oncotic pressure of the peritubular capillaryblood, which is equivalent to approximately 35 cmof water (2). Normally, colloidal oncotic pressureplays little or no role in influencing water ab-sorption (24), presumably because of the perme-ability characteristics of the interposed cell mem-branes. In mercury nephrotoxicity, however,membrane permeability might increase to thepoint where this considerable force can be exertedmore directly on the fluid in the tubular lumen.

The observations made in this study differ inseveral respects from those reported by Flaniganand Oken (2), who also studied mercury poisoningin rats by micropuncture techniques. These au-thors found a reduction in individual nephron fil-tration rates averaging 55% when measured at 6hours after a dose of HgCl2 approximately threetimes larger than we used. TF/P1, ratios in theiranimals rose progressively along the length of theproximal tubule and were indistinguishable fromthose measured in normal rats. The reason forthese differences between their study and ours isnot entirely clear, but they could be explained bya dual effect of Hgr on the kidney at different doselevels and by the time after Hg++ injection whenmicropuncture experiments were carried out. Ithas been found that a variety of mercury com-pounds cause a sharp reduction in renal blood flowand glomerular filtration rate shortly after theirintravenous injection (25, 26). This effect, whichis thought to be due to renal vasoconstriction, istransitory unless the dose is high enough to causea fall in blood pressure. In this case, the reduc-tion in renal blood flow and GFRlasts for severalhours and may prevent the expected diuresis (25).That such a response may have occurred shortlyafter Hg++ injection in the rats studied by Flaniganand Oken is suggested by the fact that their animalshad little or no initial diuresis. Instead, urine

flow fell to oliguric levels soon after drug adminis-tration. Furthermore, since they injected the Hgintramuscularly, it seems reasonable that renalvasoconstriction might have persisted for longerperiods than after intravenous injection. The re-duction in single nephron GFRthat they found at6 hours might have been due, therefore, to a pro-longed effect on the renal blood vessels. The toxiceffect on the tubules that we observed at 24 to 48hours after Hgj injection had apparently not yetbecome manifest at 6 hours.4 Because our animalsresponded to a smaller dose of Hgw with a largediuresis, we assumed that renal blood flow andGFR were not severely reduced during the firsthours after the injection. From this interpreta-tion of the two studies, we conclude that, in highdosages, HgCl2 may produce oliguria or evenanuria by causing a reduction in blood flow andGFR. This effect should be transient (25, 26),however, and should not in itself produce the ab-normalities in urinary electrolyte composition seenin patients with acute renal failure (10) and inour rats. The more prolonged anuria producedby Hg++ poisoning, which can occur in the pres-ence of a normal filtration rate, appears to be dueto a toxic effect on the tubules leading to increasedpermeability and complete absorption of the glo-merular filtrate.

References1. Richards, A. N. Direct observations of change in

function of the renal tubule caused by certainpoisons. Trans. Ass. Amer. Phycns 1929, 44, 64.

2. Flanigan, W. J., and D. E. Oken. Renal micro-puncture study of the development of anuria in therat with mercury-induced acute renal failure. J.clin. Invest. 1965, 44, 449.

3. Oken, D. E., M. L. Arce, and D. R. Wilson. Gly-cerol-induced hemoglobinuric acute renal failurein the rat. I. Micropuncture study of the devel-opment of oliguria. J. clin. Invest. 1966, 45, 724.

4. Bank, N. Relationship between electrical and hy-drogen ion gradients across rat proximal tubule.Amer. J. Physiol. 1962, 203, 577.

5. Steinhausen, M. Eine Methode zur Differenzierungproximaler und distaler Tubuli der Nierenrindevon Ratten in vivo und ihre Anwendung zur

'The possibility that a reduction in renal blood flowproduced by large doses of Hg++ may actually afford someprotection to the tubules is suggested by the fact that only33% of Flanigan and Oken's rats became totally anuricin 24 hours. With smaller doses of Hg++, almost all ofthe rats in this study were anuric at 24 hours.

703


Bestimmung tubularer Str6mungsgeschwindig-keiten. Pflugers Arch. ges. Physiol. 1963, 277, 23.

6. Gertz, K. H., J. A. Mangos, C. Braun, and H. D.Pagel. On the glomerular tubular balance in therat kidney. Pfluigers Arch. ges. Physiol. 1965,285, 360.

7. Windhager, E. E., and G. Giebisch. Micropuncturestudy of renal tubular transfer of sodium chloridein the rat. Amer. J. Physiol. 1961, 200, 581.

8. Bank, N., and H. S. Aynedjian. A micropuncturestudy of renal bicarbonate and chloride reabsorp-tion in hypokalaemic alkalosis. Clin. Sci. 1965, 29,159.

9. Swann, R. C., and J. P. Merrill. The clinical course

of acute renal failure. Medicine (Baltimore)1953, 32, 215.

10. Meroney, W. H., and M. E. Rubini. Kidney functionduring acute tubular necrosis: clinical studies anda theory. Metabolism 1959, 8, 1.

11. Glabman, S., H. S. Aynedjian, and N. Bank. Mi-cropuncture study of the effect of acute reductionsin glomerular filtration rate on sodium and waterreabsorption by the proximal tubules of the rat.J. clin. Invest. 1965, 44, 1410.

12. Gottschalk, C. W. Renal tubular function: lessonsfrom micropuncture. Harvey Lect. 1963, series58, 99.

13. Rector, F. C., Jr., F. P. Brunner, and D. W. Seldin.Mechanism of glomerulotubular balance. I. Ef-fect of aortic constriction and elevated uretero-pelvic pressure on glomerular filtration rate, frac-tional reabsorption, transit time, and tubular sizein the proximal tubule of the rat. J. clin. Invest.1966, 45, 590.

14. Finkenstaedt, J. T., and J. P. Merrill. Renal func-tion after recovery from acute renal failure.New Engl. J. Med. 1956, 254, 1023.

15. Price, J. D. E., and R. A. Palmer. A functionaland morphological follow-up study of acute renalfailure. Arch. intern. Med. 1960, 105, 90.

16. Suki, W., F. C. Rector, Jr., and D. W. Seldin. Thesite of action of furosemide and other sulfona-mide diuretics in the dog. J. clin. Invest. 1965,44, 1458.

17. Steinhausen, M., A. Loreth, and S. Olson. Mes-sungen des tubularen Harnstromes, seine Bezie-hungen zum Blutdruck und zur Inulin-Clearance.Pfluigers Arch. ges. Physiol. 1965, 286, 118.

18. Steinhausen, M. Personal communication.19. Gessler, U., and K. Schroder. Experimental studies

on the GFR during acute anuria. Presented atthe Third International Congress of Nephrology,Washington, 1966.

20. Rodin, A. E., and C. N. Crowson. Mercury nephro-toxicity in the rat. I. Factors influencing the lo-calization of the tubular lesions. Amer. J. Path.1962, 41, 297.

21. Rodin, A. E., and C. N. Crowson. Mercury nephro-toxicity in the rat. II. Investigation of the in-tracellular site of mercury nephrotoxicity by cor-related serial time histologic and histoenzymaticstudies. Amer. J. Path. 1962, 41, 485.

22. Taylor, N. S. Histochemical studies of nephrotox-icity with sublethal doses of mercury in rats.Amer. J. Path. 1965, 46, 1.

23. Kleinzeller, A., and J. H. Cort. The mechanism ofaction of mercurial preparations on transportprocesses and the role of thiol groups in the cellmembrane of renal tubuler cells. Biochem. J. 1957,67, 15.

24. Giebisch, G., R. M. Klose, G. Malnic, W. J. Sullivan,and E. E. Windhager. Sodium movement acrosssingle perfused proximal tubules of rat kidneys.J. gen. Physiol. 1964, 47, 1175.

25. Farah, A. Renal vascular changes produced by themercurial diuretic salyrgan. Naunyn-Schmiede-berg's Arch. exp. Path. Pharmak. 1952, 215, 29.

26. Vargas, R., and E. J. Cafruny. Effects of mercurialcompounds on renal perfusion pressure. J. Phar-macol. exp. Ther. 1962, 135, 112.

704

new the role of leakage of tubular fluid in anuria due to mercury...

Documents