ON THE MECHANISMOF IMPAIRMENT OF RENALCONCENTRATINGABILITY IN POTASSIUMDEFICIENCY *

By ANDRZEJMANITIUS,t HOWARDLEVITIN,t DAVID BECK§ ANDFRANKLIN H. EPSTEIN 11

(From the Department of Internal Medicine, Yale University School of Medicine,New Haven, Conn.)

(Submitted for publication October 14, 1959; accepted December 17, 1959)

A decrease in the ability of the kidneys to ex-crete a highly concentrated urine is one of theearliest hallmarks of potassium depletion in ani-mals and in humans (1). According to presentviews of the mechanism of the renal concentratingprocess (2), this might be the result of a) im-pairment of the countercurrent multiplier sys-tem by which a high concentration of sodium (andurea) is created and maintained in the inter-stitial space of the renal papilla, and/or b) de-creased permeability of the walls of the collectingducts and distal tubules to the back-diffusion ofwater. If the first alternative were correct, analy-sis of the renal papilla from potassium-deficientanimals excreting a maximally concentrated urineshould reveal a lower concentration of sodiumthan that present in the papilla of normal animals.Should the second mechanism be operative, thefall in maximum urinary osmolality observed inpotassium deficiency might be entirely unassoci-ated with a decrease in papillary sodium, or elsewould be out of proportion to it.

In the present experiments samples of maxi-mally concentrated urine and renal papilla, me-dulla and cortex from potassium-depleted ratsand dogs were analyzed. By placing all animalson a sodium-free diet prior to sacrifice, difficultiesin the interpretation of tissue analyses consequentto high concentrations of sodium in the urinewere avoided. The results indicate that the

* Aided by grants from the American Heart Associ-ation, the National Heart Institute, the Lawrence M.Gelb Foundation, and a contract (MD-116) with theOffice of the Surgeon General, Department of the Army.

t Special Fellow of the Rockefeller Foundation.Present address: Second Medical Service, MedicalAcademy, Gdansk, Poland.

t Advanced Fellow of the American Heart Associa-tion.

§ Fellow of the Dazian Foundation.II Established Investigator, American Heart Associa-

tion.

concentration of sodium and total solutes in therenal papilla is indeed decreased by potassium de-pletion, though the decrease is not as great as thefall in maximum urinary osmolality.

METHODSI. Rats

White male Sprague-Dawley rats initially weighing150 to 200 g were divided into the following groups.The composition of the diets is given in Table I.

Group IE. Twenty-four animals were fed a lowpotassium diet for 29 days, followed by 5 days of alow sodium, low potassium diet before sacrifice.

Group IC. Twenty-two rats received a normal dietfor 23 days, followed by 5 days of a low sodium diet.

Group IIE. Twenty-eight rats received a high so-dium, low potassium diet for 6 days. Desoxycorticoster-one acetate (DCA), 0.2 mg per rat, was injected sub-cutaneously daily for 5 days. Following this, a low so-dium, low potassium diet was given for 9 days.

Group IIC. Twenty-four animals received a highsodium diet for 6 days and a low sodium diet for 5 days.

In order to obtain enough papillary tissue for ac-curate analysis, 3 or 4 animals were kept in a cage andanalyses of urine, plasma and tissue were performed on

pooled material.Twenty-four hours before sacrifice food and water

were withheld. Twelve hours later 50 mUof vasopres-sin in oil was injected subcutaneously and micturition wasinduced. The rats were then placed in metabolism cageswith screens so placed as to deflect feces, and urine wascollected under oil for 12 hours. Although precautionswere taken to prevent evaporation, some (estimatedfrom separate experiments as under 15 per cent) un-doubtedly occurred as the urine passed over the screenand into the funnels. Rats were then removed fromtheir cages, anesthetized with sodium pentobarbital,and the abdominal cavity was opened. Five ml of bloodwas aspirated anaerobically from the abdominal aortaand immediately thereafter both kidneys and about 4 gof thigh muscle were removed. Separate experimentsestablished that the concentration of sodium in the re-nal papilla was not appreciably affected by preliminarybleeding in the manner described. The kidneys were sec-tioned longitudinally and slightly off center so that freeaccess to the entire papilla was obtained (Figure 1A).The papilla was removed approximately at the point in-

684

HYPOSTHENURIAIN POTASSIUMDEPLETION

TABLE I

Composition of diets

High sodium, Low sodium, Low sodium,Normal Low potassium High sodium low potassium normal potassium low potassium

g g g g g gNaCl 6.04 6.04 19.52 19.52NaH2PO4*H20 4.09KCl 7.4K2HPO4 11.60 11.60 2.4Mineral mixture 30.37 30.37 30.37 30.37 30.37 30.37Basic diet ad 1,000.0 ad 1,000.0 ad 1,000.0 ad 1,000.0 ad 1,000.0 ad 1,000.0

Diet content mEq/kg mEq/kg mEq/kg mEq/kg mEq/kgNa 103 133 334 334K 133 133 114Cl 103 103 334 334 100

Mineral mixture Basic dietg g

CaCO3 2,400 Sucrose 458CaHPO4 496 Lard 220MgSO4 1,186 Vitaminized casein 250MmSO4 30 Vitaminized corn oil 50ZnCl2 2.0 Choline 4CuSO4 2.4KI 6.4CoCI2 2.4Fe-citrate 220

dicated in Figure IB. Portionsand cortex were then obtained.

of the inner medulla

II. Dogs

Five female mongrel dogs weighing 7 to 15 kg, were

depleted of potassium by a high sodium, low potassiumdiet (Table I) plus 1.0 mg of DCAper kg of body weightinjected subcutaneously each day for 7 to 9 days, fol-lowed by a low sodium, low potassium diet for 8 more

days. A sixth animal received 45 g of potassium-ex-change resin (Kayexalate, Winthrop) in its low potas-sium diet each day for 13 days, and was then placed on a

low sodium, low potassium diet for 8 more days. Sixother dogs served as controls and were fed for 4 to 8days on the high sodium diet and then for 8 or 9 dayswith the low sodium diet. Water was withheld for 36hours before sacrifice. Twelve hours and 2 hours be-fore sacrifice, 5 units of vasopressin in oil was injectedsubcutaneously. A sample of venous blood was collectedanaerobically and the dog was lightly anesthetized withsodium pentobarbital. (The anesthetic dose of pento-barbital is considerably reduced in potassium-depletedanimals.) The bladder was emptied, washed with air,and urine was collected via an inlying catheter for thenext 15 to 30 minutes. The abdomen was then openedand both kidneys were removed and sectioned trans-versely. Samples of papilla tip (20 to 60 mg), innermedulla (300 to 400 mg) and cortex (about 1,500 mg)were immediately removed and placed in weighing bottles.

III. Analytical procedures

In the rat experiments, tissue from one kidney was

analyzed for urea and ammonia; tissue from the other

kidney, for water and electrolytes. In the experimentson dogs, samples of each renal papilla, medulla and cor-tex were divided into two approximately equal portions.The portion for urea and ammonia analysis was placedin a weighing bottle containing clean dry sand. Afterweighing, the tissue and sand were transferred to amortar and thoroughly ground. The mixture was sus-pended in a measured amount of water, transferred to aclosed test tube and immersed in boiling water for 5 min-utes. After cooling, the fluid was filtered and analyzedfor urea and ammonia (3). The second portion of renaltissue was weighed, then desiccated for 5 days at 950 Cand reweighed. Water content was determined aftergrinding, drying for another 24 hours at 950 C andweighing again. The ground dry tissue was then di-gested for 24 hours in concentrated nitric acid at 70° C.Sodium and potassium were determined by indirect flamephotometry on the filtered, diluted digest.

Muscle tissue was dried, extracted three times withanhydrous ether, ground, and digested with nitric acidbefore determination of sodium and potassium with aflame photometer.

Osmolality of urine and plasma was measured withthe Fiske osmometer. Urine was diluted to an osmo-lality less than 1,000 mOsmper kg prior to the determina-tion of freezing point. CO2 content of plasma wasmeasured with the van Slyke apparatus or the Natelsonmicrogasometer. "Total solute" concentration in tis-sue was calculated as the sum of 2 X (Na + K + NH4)+ urea, where Na, K, NH, and urea are expressed asmillimoles per kilogram of tissue water. It must be em-

phasized that for many reasons (4) this value is at best

685

A. MANITIUS, H. LEVITIN, D. BECKAND F. H. EPSTEIN

A

FIG. 1. A. RAT KIDNEY SECTIONED LONGITUDINALLY.

The arrow indicates the approximate point at which thepapilla tip was excised. B. RAT KIDNEY AFTER EXCISION

OF PAPILLA.

only a rough approximation of the osmotic activity ofwater in tissue.

RESULTS

0

04)

04)

0.

0

.0

0

.<0

0

0

0

4).0

4)

0

0

I. Rats (Tables II and III). As reported manytimes by others (5, 6), potassium depletion pro-duced by dietary means alone (Group IE) and byDCA (Group IIE) was associated with decreasedweight gain, polyuria, renal hypertrophy, hypo-kalemia, alkalosis, a decrease in muscle potassiumand an increase in muscle sodium (Table II).The potassium content of the papilla and cortexof depleted rats was lower than that of controls(Table III) but this was not true for the renalmedulla.

On a sodium-free diet, potassium-depleted rats

4)

Iz

co

04

Cdz

.)C:o)

4- - ')o - e-4 1 -

4- UC CS 0ef). 0C')

wE460 eli i 46 4 6E -4 4. V "t H V

+- M 0%

$- tu2 W.)

E -H

f4-

e, -.

ro-H

00V'C Xo °oRtN,- -H

.''110U') -#)-H

46

- W) No t. O' -4

o 0lo eU)6i (64 46 6

V -H '-H V

of) oU) o00 oa46 46d 6

A 4- -f r- -H A

0

-HlV

oo

H HV> -It 00 \ C 1)int- 00 C toa t , 14 C4 M f) M Ln \0 00 0

ce iC c-i 'i 6 c-,6 65J~N -A + -HV

CU

N -.1C

66tr C1)46._ t

cE 00

z

N

z

beC

%41-

O 0

000 \O teOI- 6O0%00 00 W) C'10 - mV') _- _- -

e

-^ -+ V '-n-H -^ V

U') -4r -It 0 (N in 0t 0

rz~ ~6ci161 cli -Z04-- 0%(N 00%" V -- H V

Ch, o O - t O-00H 0 "-H --

-HV~ -H+V

-0 ON (N, - - e -A'

e 0 %6N6 6 e_ 4~ C

+ -+ A + V14

too-C

0

IS

4 C-4

00ON

C C;66o

ON %0d4 CNr- m

00

\0 -4

-

V

0

V;

U') m 00

OfiN r1IrXO0%~-H 4--- 4--l

0 V 4--\0\0(N (NIC' U)o~ -4 tt 6; C

-H -H V

Nm to VI)0 enIt 00 0

HV V

00oo o m0-or4 oo40 elci \4

-H O H V CNH 4-H A4-C.- eq r- -4 0 4t- -~ -.

tn .4 _. CNI o oo m _- eq oC-4 6C5 e C eli 65C

-H V -H V

in) -4

0 eq -t00 0 I-d'0 4- 0O 0

c-o o o6 cj

-H H A -H -Ht V

00 o c C4 c)en C so

e" v4 -H V ` H -l

6

V

(-4 --t Rt 00

C14 (N d-4 (No%0 4-- %0I- I--,

\- o - \ O -

0 0_4.) w.

0

.co i-o

-.

l-4)

S._

-o4)

CtCZ

EU)14-~

s:

100

CCd*

CZW

41)E)n

*b- 11

* --

686

HYPOSTHENURIAIN POTASSIUMDEPLETION

C-4

U); - CoR oH '-i A

0-H -' V

.4)- o

Bx 0Co.R~ .Ci CCCC O

S '0-H°t A

.4)~~~~u'00 F-F 0U) 0F-- F- Vo

o r-~ 0-4 0_ -)I _

vq I- -! In

_ o-' O

'0 -.

'00 O 00 0_A v-H CH

o to _ ) 0

A,- -+ V

C01- CIF-U 0I" F-p It;~ 0l eq - O - 0 C - 000 0

0 l c0 - Vco°° 4 VOt- 0

t O4

V

eb 0C 0ut_ -_ 0O

-4 f H V

t 0 0t0 0 0o%0C -4 C-40C -4 It4 F

V .-+- _. V

t- 0 0C 000 00 '00 0o _

-HV -H - A

-H.)

A - -H - A

0' ' 0 0o 0 0

G o 0 - 0U-H+ -H -H V

C4 C4 t- C4 Nt o0 00 C-Cm "O

00oe CUo ) 00'oO

'- b-H Ao H VV

.in . . . .4

~C0 0 C~- 000 000 -H 00 -H V 00 FH- -H v

C 0o 0 No0 0- 04-H V -H A

2(A4oiuC:w~

I-_

~o Xo_

687

excreted more sodium in the urine than did nor-mal controls deprived of sodium for a similar orshorter period (Table II).

The concentration of sodium and urea in renaltissue of normal rats increased progressively fromcortex to papilla (Table III), as was reported byUllrich and Jarausch in dogs (7). No such gradi-ent was observed for potassium. The concentra-tion of sodium in water of papillary tissue wasconsiderably higher than that in peripheral plasma,even though the sodium concentration of urine wasextremely low. "Total solute" concentration ofpapillary tissue was about one-half the osmolalityof urine, but tended to increase as the sample ex-cised for analysis became smaller and smaller asthe tip of the papilla was approached, (Figure 2).The concentration of potassium in the urine ofnormal rats on a sodium-free diet was so high(400 to 500 mEq per L) when compared withthe concentration of potassium in the renal papilla(70 to 90 mEq per kg H2O) that urine couldnot possibly have made up more than 15 to 20per cent of the water of the papilla, and probablywas a good deal less than that.

Maximum urinary osmolality was approxi-mately halved by potassium depletion. This wasassociated in both groups of depleted rats witha significant reduction in the amount of sodiumper 100 g of dry solids of renal papilla. In groupIE, the decrease in concentration of sodium, ex-pressed as milliequivalents per kilogram of tissue

2000-

% _i

4)0

+Cd

w0+0

r. +CU'-

0

00

*1 114

Tissue Solutes

00.

mM/KgH20

000O

4

* *@

*-:-

5 6 7 8Wedgt of Papeil mg.

9 10 11

FIG. 2. CONCENTRATION OF "TISSUE SOLUTES" [2 X(NA + K + NH4) + UREA] IS PLOTTED VERSUS WEIGHTOF ANALYZED SAMPLE OF PAPILLA. As the tip of thepapilla is approached (i.e., as the sample analyzed be-comes smaller) the concentration of solutes in tissuewater rises.

E 0'0 U)cU).

C- m -H

~" 00 0'tot

C4 - -H

144)

0u

Cd

cd.a

Cd

CU4)

z

Cdz

Mz

CUz

CUz

- t:0 45

-oa)

z

CU

CU:cd 4,

.0

bO._p

3

*

t.

H

lie

el

- 4)

C-

Q

0

0

4I

0

Q

rlQ

KA

cO)

w" -

--o _ o

--, 0*

$04 F- t

SoE - -H

-s, m.fZ ,

A. MANITIUS, H. LEVITIN, D. BECKAND F. H. EPSTEIN

pq $:?% 6 6J~' -Ho4 r U) 0CN U

00 00 0 0%-H `4-H

0

V

6A

/\4

0dt U) Ce1 e1 6U) d14 ) encq1 -4 U) -

-H -H V

o %0

C-I -

-H

U)0

A

0

V

U) CN

0 t0- U)

- V

0~o 00 o Voo'IO 00 oE 00o on CKi C%

_ -H -H V

-o 00 -' -U) -4 0% 0% 06 4 C-i o

+ -HV

0% 00

-H

o

-I

-HCA

A

in 00 U) %o 1'

cN- q-H A

00 NtUl)

41

U)4 00 C% C1'I

-< 44 dc i 6

t~ 2-H L.2 A

00

4 4

ge I I-

C- J-.r: r.

CdE

0

z

x

0

0.4

*

IZ40

t.

0

0

0

0V

0;

0

0

U)-o

._

CdI._

C0Cd(n

la

-o

r-*

Cd

Cd

0

U1)I-

Cd

K

0u

Cd

Cd

z

cdz

z

cdz

CUz

z

b4C -H -°

-A! -H+N

(9o0-Q +

$ *bk

Ui

A

0A

'In0

V

0A

- N oo U) oo o

o~0%

U)

{R - - V

. .

'tW-

g N +-00 00X A

U) 00

'. e

Q~

00

i-0 44 - -H

0-40

v0

V

;z tn ~ t- Coo t- C4 o1 0o

V

_;. em e

-'o 0

;4t

-H A

~o r.U)

0

V

4:40 .; w - 0)

-k M + -1 + V

.

w oow +oOA

S e e

E a0

z 3

688

zD

I">IC U)

1

N4 r- V4 (i o6wo 00U)E-H

E - C._-H

I. 00 V-4

sI * -H

*

0

C..

$)0q

*V2

$..O2

m *

C.)o

0s

0

-CU0-

z

C'z

Cdz

0

E4-4

0

114 00 Ul)

N1o -H

._--I;

U1Cd

r.

Si Cd01!045

A0+10

-0+CdU4i

U)001

* +

X11

4)_ W*) V+

689HYPOSTHENURIAIN POTASSIUMDEPLETION

TABLE VI

3*0No

meq/KgTissue Wot2

I)

.-.N-orntl Boas*---LoL@w Kdo@V

Y lWine owseI

Cortex MedutLL. ?apitta

FIG. 3. GRADIENT OF SODIUM CONCENTRATION FROMCORTEXTO MEDULLAIS PLOTTED INDIVIDUALLY FOR NORMALDOGS (SOLID LINES) AND POTASSIUM-DEPLETED DOGS(BROKEN LINES). The concentration of sodium in thepapilla is reduced in potassium-deficient animals, al-though not so much as is the maximal urinary osmolality(in parentheses).

water, did not reach statistical significance, but ingroup HIE a significant decrease in the concentra-tion of sodium in tissue water was apparent in bothmedulla and papilla. An unexpected decrease inthe water content of papilla and medulla was ap-parent in both groups of potassium-deficient rats."Total solute" concentration of papilla and me-dulla was significantly diminished by potassiumdeficiency, as a result of decreases in the concen-tration of both sodium and urea.

II. Dogs (Tables IV and V). Potassium-de-pleted dogs became weak, and some (not in-cluded here) died with symptoms of paralysis ofperipheral and respiratory muscles. Urinary ex-cretion of potassium was extremely low and, asin potassium-depleted rats, difficulty in conservingsodium was noted on a sodium-free diet.

Potassium deficiency was associated with a

Comparison of urinary osmolality, papillary solute concen-tration and papillary sodium concentration in normal

and potassium-depleted dogs

"Tissue PapillaryAnimal Uosm solutes"* Na

no. mOsm/ mmole/ mEq/kg H20 kg H20 kg H20

Normal dogs 75 1,628 1,448 27974 1,664 1,232 19869 1,728 1,518 30473 1,924 1,339 26267 2,244 2,006 42857 2,608 2,181 382

Potassium-depleted dogs 65 501 772 172

63 670 672 14658 812 916 12953 1,036 1,076 22270 1,184 1,174 25061 1,256 i1336 269

* "Tissue solutes" = 2 X (Na + K + NH4) + urea.

marked decrease in maximal urinary osmolalityand with a significant fall in the concentration ofsodium and urea in the renal papilla. (The de-crease in water content of medulla and papillaobserved in potassium-deficient rats was notnoted in dogs.) Figure 3 illustrates the de-creased gradient of sodium concentration betweencortex, medulla and papilla produced in dogs bypotassium depletion.

The calculated solute concentration of the pa-pilla of normal dogs was about 80 per cent of the

R ATS

12.0

U/tpuo6.0

[Ttdd Sokieslppi

20

1.0

NORMAL LOW-K

FIG. 4. POTASSIUM DEPLETION PRODUCEDA DECREASE

IN RATIO OF MAXIMAL URINARY OSMOLALITY (UOSM) TO

CONCENTRATION (IN MILLIMOLES PER KILOGRAM TISSUE

WATER) OF SODIUM AND OF "TOTAL SOLUTES" IN RENAL

PAPILLA.

A. MANITIUS, H. LEVITIN, D. BECKAND F. H. EPSTEIN

osmolality of urine. In potassium-deficient dogs,on the other hand, total solute concentration ofthe papilla was equal to or higher than the os-molality of urine (Table VI). These data sug-gest but do not prove the possibility that the back-diffusion of water from collecting ducts into theinterstitium of the papilla may have been im-paired by potassium deficiency. Similar altera-tions were observed in the ratio of urinary osmo-lality to the concentration of total solutes or so-dium in the papilla of potassium-deficient rats(Figure 4), although in the rat experimentsurine from several rats was pooled and collectedin funnels for 12 hours prior to sacrifice, and theabsolute values of such ratios are therefore lessmeaningful.

DISCUSSION

These experiments illustrate clearly the effectof potassium depletion in impairing the abilityof the kidneys to produce a concentrated urine.Although desoxycorticosterone was used in somerats and dogs to induce potassium deficiency, theresults cannot be ascribed to the influence of thishormone, per se, on the renal tubules, since itsadministration was discontinued many days be-fore the animals were sacrificed. It is unlikelythat the low sodium diet to which all animals weresubjected substantially affected maximum urinaryconcentration or the composition of the renal pa-pilla (8).

A priori, the decrease in concentrating abilityobserved in potassium deficiency might arise fromseveral sources. First, the distal convoluted tu-bules and collecting ducts might be less permeableto the outward diffusion of water under the in-fluence of vasopressin, so that water would beretained in the urine despite the maintenance of ahypertonic interstitial fluid in medulla and papilla.Giebisch and Lozano concluded that this was animportant cause of hyposthenuria in potassium-depleted dogs, because these animals excreted ahypotonic urine during mild mannitol diuresisdespite infusions of vasopressin (9). In the pres-ent experiments, the decrease in urinary soluteconcentration induced by potassium depletion al-ways exceeded the fall in the concentration of so-dium or total solutes in the water of the papilla,and presumably, in the papillary interstitium.

These findings are compatible with some im-pairment of the back-diffusion of water from thecollecting ducts, but they do not suffice to es-tablish it.

An increased delivery of hypotonic urine to thedistal convolutions and collecting ducts might alsoplay a role in the hyposthenuria of potassium de-ficiency. It is conceivable, for example, that re-absorption of salt and water in the proximal tu-bule might be diminished, and that consequentlymore fluid would be delivered to the ascendinglimb of Henle's loop, where sodium would be re-absorbed without water. The present data donot bear directly on this possibility, but one mightexpect that under such circumstances the concen-tration of sodium in the tissue of the renal me-dulla would be increased over normal or un-changed, not diminished. It is noteworthy inthis connection that the total solute excretion perminute (corrected for body weight) was similarfor potassium-depleted animals and for normalcontrols.

A decrease in the concentration of sodium at-tained in the interstitial fluids of medulla and pa-pilla would be expected to result in hyposthenuria.The present experiments suggest that this is animportant cause of the decrease in renal concen-trating ability associated with potassium defi-ciency. The diminished concentration of sodiumin papillary tissue cannot be ascribed to a lowerconcentration of sodium in the urine of the po-tassium-depleted animals, since urinary sodiumconcentration in potassium deficiency was actu-ally higher than that of control animals. Althoughan increased volume of sodium-poor urine in thecollecting ducts of potassium-deficient, polyuricanimals might have reduced the papillary con-centration of sodium expressed as milliequivalentsper kilogram of tissue water, it would not haveproduced a decrease in the amount of sodium per100 g of dry solids.

The decrease in sodium concentration in renalmedulla and papilla observed in potassium-defi-cient animals might result from 1) alterationsin countercurrent flow through capillary loops ofthe vasa recta, 2) impairment of sodium reab-sorption from the loops of Henle, or 3) impair-ment of active reabsorption of sodium withoutwater from the collecting ducts. The data do not

690

HYPOSTHENURIAIN POTASSIUMDEPLETION

bear critically on the first possibility, which isdifficult to test with present methods. With re-gard to the second and third alternatives, thetendency of potassium-depleted rats and dogs ona sodium-free diet to lose sodium in the urinesuggests impairment of tubular reabsorption ofsodium at some site in the nephron. Some so-dium transport mechanisms are known to be sen-sitive to lack of potassium (10, 11). The highconcentration of sodium normally present in therenal medulla is thought to be a result of activetransport of sodium by the loops of Henle (12).Nevertheless, several considerations suggest thatin potassium deficiency, impaired reabsorptionof sodium from the collecting ducts may be inpart responsible for the diminished accumulationof sodium in the papilla. The osmolality of fluidobtained by micropuncture from the early distalconvolution of hydropenic potassium-depletedrats is hypotonic to plasma and similar to that ofcontrol animals (13). Potassium-deficient ratswith impaired concentrating ability are able todilute the urine normally in response to a waterload (14). These facts suggest that in potas-sium deficiency reabsorption of sodium withoutwater proceeds normally in the loop of Henle.Finally, the alterations in histological structureof the rat kidney produced by potassium deple-tion are most marked in the collecting ducts (15).

Active reabsorption of sodium from the collect-ing ducts has been demonstrated by Hilger,Klumper and Ullrich (16). As pointed out bythese authors and by Gottschalk and Mylle (12),if during hydropenia the collecting ducts arefreely permeable to water, the mere removal ofsodium from the urine by collecting duct epi-thelium might serve to maintain the hypertonicityof medullary interstitial fluid but could not in-crease it. If, however, sodium reabsorption inthis region were linked primarily or partly withthe creation of osmotically active particles (forexample, the addition to the urine of NH,+) (17),sodium reabsorbed by the collecting ducts couldbe trapped in the medulla by countercurrentcapillary flow (18) and could contribute im-portantly to the creation of an osmotic gradientbetween cortex and medulla. It seems reasonableto suppose that the collecting ducts play a rolein normal kidneys in establishing this gradient

and that this process is disrupted in potassiumdeficiency.

SUMMARY

1. The effect of potassium depletion on thecomposition of renal papilla, medulla and cortexwas studied in dogs and rats in an effort to clarifythe mechanism of the hyposthenuria produced bypotassium deficiency.

2. Potassium deficiency was associated with afall in the concentration of sodium and urea inthe papilla and medulla of hydropenic animals.

3. The ratio of urinary solute concentration topapillary solute concentration was uniformlylower in potassium-depleted than in normal ani-mals. This finding is compatible with but doesnot prove the hypothesis that potassium deficiencyimpairs the permeability of the collecting ductsto water.

4. It is suggested that the defect in renal con-centrating ability induced by potassium deficiencymay be at least in part a result of impaired reab-sorption of sodium by the collecting ducts.

ACKNOWLEDGMENT

The authors acknowledge gratefully the technicalassistance of Mrs. Nadia Myketey and Mrs. EvaTaborsky.

REFERENCES

1. Relman, A. S., and Schwartz, W. B. The kidney inpotassium depletion. Amer. J. Med. 1958, 24, 764.

2. Smith, H. W. The fate of sodium and water in therenal tubules. Bull. N. Y. Acad. Med. 1959, 35,293.

3. Conway, E. J. Microdiffusion Analysis and Volu-metric Error, 4th ed. London, Crosby Lockwoodand Son, Ltd., 1957.

4. Maffly, R. H., and Leaf, A. The potential of waterin mammalian tissues. J. gen. Physiol. 1959, 42,1257.

5. Hollander, W., Jr., Winters, R. W., Williams, T. F.,Bradley, J., Oliver, J., and Welt, L. G. Defect inthe renal tubular reabsorption of water associatedwith potassium depletion in rats. Amer. J.Physiol. 1957, 189, 557.

6. Muntwyler, E., and Griffin, G. E. Tissue electrolytecontent of potassium and protein-deficient rats-Proc. Soc. exp. Biol. (N. Y.) 1955, 89, 349.

7. Ullrich, K. J., and Jarausch, K. H. Untersuchungenzum Problem der Harnkonzentrierung und Harn-verdunnung. Pfluig. Arch. ges. Physiol. 1956, 262,.537.

691

A. MANITIUS, H. LEVITIN, D. BECKAND F. H. EPSTEIN

8. Manitius, A., Levitin, H., Beck, D., and Epstein, F. H.Unpublished observations.

9. Giebisch, G., and Lozano, R. The effects of adrenalsteroids and potassium depletion on the elaborationof an osmotically concentrated urine. J. clin. In-vest. 1959, 38, 843.

10. Huf, E. G., and Wills, J. Influence of some inorganiccations on active salt and water uptake by isolatedfrog skin. Amer. J. Physiol. 1951, 167, 255.

11. Ussing, H. H. Ion transport across biologicalmembranes in Ion Transport across Membranes.New York, Academic Press Inc., 1959.

12. Gottschalk, C. W., and Mylle, M. Micropuncturestudy of the mammalian urinary concentratingmechanism: Evidence for the countercurrent hy-pothesis. Amer. J. Physiol. 1959, 196, 927.

13. Gottschalk, C. W., Mylle, M., Winters, R. W., andWelt, L. G. Micropuncture study of the osmolalityof renal tubular fluid in potassium-depleted rats(abstract). J. clin. Invest. 1958, 37, 898.

14. Levitin, H., Manitius, A., Beck, D., and Epstein, F. H.Urinary dilution in potassium deficiency. To bepublished.

15. Oliver, J., MacDowell, M., Welt, L. G., Holliday,M. A., Hollander, W., Jr., Winters, R. W., Wil-liams, T. F., and Segar, W. E. The renal le-sions of electrolyte imbalance. I. The structuralalterations in potassium-depleted rats. J. exp.

Med. 1957, 106, 563.16. Hilger, H. H., Klimper, J. D., and Ullrich, K. J.

Wasserruckresorption und Ionentransport durchdie Sammelrohrzellen der Saugetierniere. Pflug.Arch. ges. Physiol. 1958, 267, 218.

17. Ullrich, K. J., Hilger, H. H., and Klfimper, J. D.Sekretion von Ammoniumionen in den Sammel-rohren der Saugetierniere. Pflug. Arch. ges.

Physiol. 1958, 267, 244.18. Berliner, R. W., Levinsky, N. G., Davidson, D. G.,

and Eden, M. Dilution and concentration of theurine and the action of antidiuretic hormone.Amer. J. Med. 1958, 24, 730.

692