THE DISTRIBUTION OF BODYWATERANDELECTROLYTESIN ADRENALINSUFFICIENCY'

BY HAROLDE. HARRISONAND DANIEL C. DARROW

(From the Department of Pediatrics, Yale University School of Medicine, New Haven)

(Received for publication September 17, 1937)

The mode of action of the hormone of theadrenal cortex necessary for life has been thesubject of much discussion. Among the promi-nent hypotheses for explaining the symptoms ofadrenal cortical insufficiency are loss of sodiumand chloride with accompanying changes in thedistribution of body water (1, 2, 3); toxic effectsof potassium (4, 5, 6, 7); disturbance in the dis-tribution of body water not clearly related to lossof sodium or retention of potassium (8); anddisturbances in metabolism or carbohydrate (9).Since it seems, at present, bootless to attempt torelate the anomalies in carbohydrate metabolismto the disturbances in metabolism of water andelectrolyte, this aspect of the function of theadrenal cortex will not be discussed. The presentinvestigation was undertaken to determine by tis-sue analyses the actual distribution of water andelectrolyte in the tissues of adrenalectomized ratsin relation to the toxic action of potassium andthe therapeutic effect of sodium salts.

EXPERIMENTALPROCEDURE

Healthy male albino rats weighing approximately 200grams were adrenalectomized under ether anesthesia viathe usual lumbar route. The majority of the animalswere fed a low potassium diet of the following composi-tion: lactalbumin 15 per cent, sucrose 25 per cent, dextrin32 per cent, crisco 22 per cent, cod liver oil 1 per cent,yeast powder 2 per cent, bone ash 2 per cent and NaCl 1per cent. One group of animals was fed a ration con-taining 1.5 grams of K,HPO, per 100 grams of the abovediet. Diet 1 on analysis, was found to contain 0.03 percent potassium whereas Diet 2 contained 0.7 per cent po-tassium. The animals were given distilled water, and thefood and water intakes were measured daily.

The adrenalectomized animals were divided into groupsand treated as follows. The animals of the first groupwere sacrificed when symptoms of severe adrenal insuffi-ciency were manifest. The time of onset of symptomswas exceedingly variable on Diet 1 so that the ratscould not be sacrificed at stated intervals after operationif comparable results were to be obtained. The symp-

1 This work was aided by a grant from the Fluid Re-search Fund of Yale University School of Medicine.

toms used for the diagnosis of adrenal insufficiency wereanorexia, loss of weight, bloody nasal discharge, diar-rhea, muscular weakness and finally, marked diminutionof the intake of water. It was found that the rats al-most invariably died within 24 hours after they ceaseddrinking, so that it was possible by waiting for thissymptom to obtain a series of animals for analysis in asimilar state of severe insufficiency.

The animals of the second group were allowed to reachthe same stage of adrenal insufficiency as those of Group1 and were then treated by intraperitoneal injections of apotent adrenal cortical extract.2 The amounts givenwere sufficient to cure the symptoms although the ani-mals were fasted and given no sodium except that con-tained in the extract used. These rats were sacrificed at24 and 48 hour intervals after therapy was instituted.

The animals of the third group when in a similar stateof adrenal insufficiency were treated by daily intraperito-neal injections of 10 cc. of a solution containing 250 mm.NaCl and 50 mm. NaHCO, per liter. This treatmentwas continued until symptoms of insufficiency disappearedand the animals were thought to be in normal conditionagain. The animals were then sacrificed. These ratsreceived no cortical extract.

The control animals were male rats from the samecolony and of the same weight as the adrenalectomizedrats. Since rats developing adrenal insufficiency refuseto eat, the control animals were likewise fasted beforebeing sacrificed. A number of animals were treated ascontrols of caloric intake by giving only as much foodas the adrenalectomized rats ate. This procedure wasfound unnecessary, however, and subsequent controlswere uniformly fasted for 7 days and then sacrificed foranalysis.

A number of intact animals were also fed both the lowand the high potassium diets and then analyzed. Theseshowed no differences and served as fed controls.

METHODSOF ANALYSIS

Since sufficient blood and tissue for a complete analysiscould not be obtained from a single rat, the animals werepooled for analysis into groups of 2 to 4 animals. Therats were etherized and as much blood as could be ob-tained was withdrawn from the abdominal aorta withoutexposure to air. A portion of the blood was collected ina bottle containing potassium oxalate to prevent clotting,

2 Weare indebted to Dr. G. F. Cartland and the Up-john Co. for a generous supply of cortical extract.

77

HAROLDE. HARRISONAND DANIEL C. DARROW

while the remainder was allowed to clot, centrifuged, andthe serum separated.

Immediately after sacrificing the animals, the liverswere removed and a sample of muscle was dissected offrapidly and freed of fat and connective tissue. The skinwas removed in toto and the remainder of the carcass

treated separately. All of the tissues were put intocovered receptacles, weighed and then placed in an oven

at 105° C. and dried until constant weight was attained.The dried tissues were ground to a powder in a small

food chopper. The dried powdered tissues were kept inglass stoppered flasks and aliquots taken for analysis.

One cubic centimeter samples of serum were accuratelyweighed in platinum dishes and the water content ob-tained by drying 24 hours in an oven at 1050 C. Thedried residue was ashed in a furnace at 5000 C. for 2hours.

Serum chloride was determined by the method of VanSlyke and Sendroy (10). The tissue chloride was deter-mined after ashing the material at 4500 C. in the pres-

ence of a large excess of sodium carbonate. The ashwas dissolved in water with the aid of a little dilutenitric acid and then the usual Volhard titration was car-

ried out. This method was found exceedingly satis-factory and accurate recoveries could be obtained.Sodium was determined by the Butler and Tuthill modi-fication of the Barber and Kolthoff method (10) afterseparation of potassium by precipitation as the chloro-platinate. Potassium in serum and tissues was de-termined by a modification of the chloroplatinate pre-

cipitation in which the chloride of the K2PtCl, was

titrated by the Volhard technique. This method will bedescribed in detail elsewhere. The nonprotein nitrogenof the whole blood was determined by Folin's Nessleriza-tion method (10). Tissue phosphorus and nitrogen were

determined by the Fiske and Subbarow (11) and Kjeldahlmethods respectively. The tissue fat was determined byextracting the tissues with alcohol-ether mixture, andafter evaporation to dryness, re-extracting the residuewith petroleum ether. The petroleum ether soluble resi-due was weighed.

EXPERIMENTALRESULTS

The effect of the potassium content of the diet on

the survival of adrendectomized rats

The results of the present experiments agree

with previous reports that a high intake of potas-sium accelerates the onset and aggravates thesymptoms of adrenal insufficiency. Of the 19adrenalectomized rats on Diet 2 (potassium con-

tent 0.7 per cent) all but one showed symptoms ofadrenal insufficiency within 5 days after opera-tion. On the other hand, of 64 animals on Diet1 (potassium content 0.03 per cent) 29 (45 per

cent) survived 10 days or more without showingsymptoms; 14 (22 per cent) lived 15 days or

longer without symptoms of adrenal insufficiency.When 8 of the 14 animals surviving more than2 weeks were changed to a stock diet (Purina DogChow) which contains 0.6 per cent potassium and1.5 per cent sodium chloride, all developed symp-toms of adrenal insufficiency within 6 days.

The concentration of serum electrolytes

As shown in Table I, rats in adrenal insuffi-ciency show a constant but moderate reduction in

TABLE I

Serum water and electrolyte

Non.Chlo- So- Po- pro-Water ide dium tas- teinrddimsium nitro-

gen

grams mM. mm. mM. mpgm.pe e epe r%per100 lite lier lier 100

Adrenal insufficiencyMaximum.94.0 101 140 8.3 168Minimum . 9............... 94 131 6.4 50Average . 93......... . 97 136 7.4 96

Cortical extract,* 24 hours 93.3 92 132 60Cortical extract,t 48 hours 94.2 99 134 11.9 64Cortical extract,X 48 hours 94.3 101 143 5.5 22Saline treated .94.5 105 143 6.1 40Saline treated ............ . 102 143 5.7 28Control

Maximum . 9............... 107 146 5.9 45Minimum .9............... 102 140 4.5 29Average. 93.9 105 144 5.2 35

* Single injection of 5 cc.t Two injections totalling 9 cc.t Three injections totalling 12 cc.

concentrations of sodium and chloride in serumbut no significant change in total water. The con-centration of potassium in serum is moderatelyincreased. The marked increase in nonproteinnitrogen is invariably found in rats exhibitingsymptoms of adrenal insufficiency. The decreasesin sodium and chloride are considerably less thanthose reported in adrenalectomized dogs (1, 12)and cats (13), and human patients with Addison'sdisease (14, 15). This explains the failure ofRubin and Krick (16) to find an increasedurinary loss of sodium and chloride in adrenalec-tomized rats.

It should be recalled that the animals treatedwith extract received only water by mouth andno sodium chloride except the small amount con-tained in the cortical extract. Within 24 hoursafter a single injection of cortical extract, symp-tomatic recovery is obvious although the concen-

78

BODY WATERAND ELECTROLYTESIN ADRENALINSUFFICIENCY

tration of sodium and chloride in serum remainslow. When cortical extract is given for 48 hourswith only water by mouth, the concentrations ofsodium and chloride approach the normal figure.This restoration of the concentration of sodiumand chloride is explained by the loss of water andtissue with retention of these ions. With pro-longed intensive treatment normal concentrationsof potassium and nonprotein nitrogen are found.

Intraperitoneal administration of hypertonicsolutions of sodium chloride and sodium bicar-bonate restores to normal the concentrations ofall the serum constituents studied.

The total electrolyte of muscle, liver and totalanimal

In order to make the various figures comparable,the data in Table II are expressed per kilogram

TABLE II

Water and electrolyte of muscle, liver and the total animal(The results are expressed per kilogram of fat-free tissue.)

Adrenal Cor- saline Fasted Fedienui- eticalt treated controls controls

Number of animals 22 8 9 20 1t

MUSCLE

Water, grams ....... 779 778 770 774 774Potassium, mM.... 117 106 109 109 106Chloride, mM...... 15.0 14 15.4 16.4 14.5Sodium, mM....... 18.8 21 23.6 23.6 23.3Phosphorus, mM... 80 77 77.7 79 78

LIVER

Water, grams ....... 755 753 757 746 739Potassium, mM..... 92 95 98 95 95Chloride, mM...... 34 34.2 35.6 31.4 31.3Sodium, mM....... 34.6 32.0 28 30.5 29.0Phosphorus, mM... 117 121 110 113 114

TOTAL ANIMAL

Water, grams ....... 731 717 723 729 732Potassium, mM..... 85 79 79 77 76Chloride, mM...... 37 36 40 39 37Sodium, mM....... 50.5 53 56 55 52Phosphorus, mM.... 235 256 256 249 233

of fat-free tissue. In each case the result givenis the average for the number of animals indi-cated at the top of the column.

When compared with the saline treated and thefasted, control animals, the total bodies of rats

in insufficiency show a slight decrease in chlorideand sodium but a definite increase in potassium.No consistent change in water is demonstrated.A decrease in chloride is not shown when com-pared with fed controls, but since animals in in-sufficiency refuse to eat, the fasted controls areprobably a better standard of comparison. Theanimals treated with cortical extract show a per-sistence of the low concentration of chloride andsodium. However, the striking feature is restora-tion of potassium to practically normal figuresin the groups treated with either cortical extractor saline. The rats treated with salt or extractare the only ones showing relative loss of water.

The findings in muscle are probably chiefly re-sponsible for the changes in the animals as awhole, although similar changes cannot be ex-cluded in any of the tissues not separately exam-ined. In the muscle, the concentration ofpotassium is strikingly elevated in the animalsshowing adrenal insufficiency. In the adrenalec-tomized rats treated with either sodium chlorideor cortical extract, the muscle potassium returnsto normal. In the animals with adrenal insuffi-ciency the concentrations of water and chloridein muscle are unaltered while that of sodium isdefinitely reduced. Moreover, in rats treatedwith cortical extract, the sodium and chloride areessentially the same as in the untreated ones.

The livers are quite different from the musclesin that no increase in potassium occurs in adrenalinsufficiency. Apparently, the irregular valuesfor sodium and chloride in the liver depend to alarge extent on the amount of blood in the organsso that comparisons are probably not valid.

The concentrations of protein and electrolyte inintracellular water

Since the data in Table II do not separate thechanges in the tissues as a whole from thosewithin the cells, the data were calculated forTable III so as to show the concentrations of cellsolutes per 1000 cc. of intracellular water. Themethod of calculation is that described by Har-rison, Darrow and Yannet (17) based on thepremise that the chloride, except for the red cells,is practically exclusively an extracellular ion andtherefore the volume of extracellular fluid inthe muscle can be estimated from the total amountof muscle chloride divided by the concentration of

79

HAROLDE. HARRISON AND DANIEL C. DARROW

TABLE IIIThe concentration of Potassium, magnesium, phosphorus, and

protein in the sntracellular water of muscle and liver(The figures are expressed in terms of 1000 cc.

of intracellular water)

Muscle Liver

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)Num- Ad- hber of Po- Phos-a-m Pro is Po- Po.Poani- Group tea- pho- ne- tea pho- teinmale sium ru sium sislum rue

mM. mM. mM. grams mM. mM. mM. grams4 Adrenal insufficiency 186 128 345 1862 Adrenalinsufficienoy 189 128 20 347 1893 Adrenalinsufficiency 185 128 18 346 185 183 237 4513 Adrenal insufficiency 180 122 20 316 197 173 221 4673 Adrenal insufficiency 177 121 21 312 196 189 244 433

Average 183 125 20 333 191 182 234 450

2 Cortical extract* 154 114 20 299 178 174 226 4073 Cortical extractt 163 123 17 340 166 178 242 4553 Cortical extract$ 167 118 350 165 183 228 434

Average 161 118 18.5 330 169 178 232 432

4 Saline 167 121 24 345 167 177 202 3435 Saline 173 123 346 173 169 190 306

Average 170 122 346 170 173 195 325

5 Fasted controls 173 128 348 1725 Fasted controls 164 119 24 346 164 175 210 4235 Fasted controls 171 124 343 1716 Fasted controls 168 119 22 349 167 182 239 474

Average 169 123 23 347 168 179 225 449

4 Fed controls 159 125 21 338 162 175 212 3992 Fed controls 161 118 20 341 164 180 220 3595 Fed controls 172 121 21 332 179 173 208 348

Average 164 121 21 337 168 176 213 369

* Single injection of 5 cc.t Two injections totalling 9 cc.t Three injections totalling 12 cc.

chloride in extracellular fluid. In the liver, theanalogous calculations are based on the sodiumcontent of the tissue since, in rat livers, perhapsowing to the large amount of red cells included,the chloride values are higher than those of thesodium, indicating that there is an excess ofchloride beyond that accounted for by the extra-cellular fluid. The concentration of protein ofthe cells is estimated by subtracting the non-protein nitrogen from the total nitrogen andmultiplying the resulting nitrogen by the conven-tional factor 6.25. This amount of protein maythen be assumed to be present in intracellularwater without gross error.

When the data are presented after these cal-culations (Table III), the concentrations ofpotassium, phosphorus and protein in the intra-cellular water of normal muscle are found to beremarkably constant. In the fasting animals theconcentration of protein in intracellular water isslightly higher than that of the fed controls.These findings suggest that, during starvation,the loss of water proceeds more rapidly from

muscle cells than the loss of nitrogen, producinga concentration of cellular protein.

The outstanding finding in the muscle of ratswith adrenal insufficiency is the invariably in-creased concentration of potassium in intracellu-lar water. This increased concentration of basein intracellular water is all the more strikingwhen it is recalled that the concentration of basein extracellular fluid is reduced in these animals.Furthermore, the increase of potassium withinthe cell is not associated with any change in theconcentration of magnesium which is the otherbase found within the cells in considerableamounts. If the concentration of proteinswithin the cells can be used as a measure ofcell dilution, it is seen that in 3 of the groupsof animals with adrenal insufficiency no dilutionof the muscle cells has occurred, whereas in 2groups a definite shift of water from the extra-cellular to the intracellular phase is indicated bythe decrease in concentration of protein.

All the animals treated either with cortical ex-tract or with hypertonic saline solution show areduction of the concentration of potassium inthe water of muscle cells to the values found inthe control animals. In the group sacrificedwithin 24 hours after a single injection of corticalextract, the concentration of protein in intracellu-lar water of muscle is low, indicating that recov-ery may occur although dilution of the cellpersists.

Column 7 of Table III shows the hypotheticalconcentrations of potassium which would obtainif the concentrations of intracellular protein wereconstant. For this purpose, the concentration ofprotein in the muscle cell water of the fastedcontrol animals was taken to be the standard.This calculation eliminates the effects of changesin the amount of intracellular water on the con-centration of potassium. Column 7 shows, there-fore, the actual change in content of cellularpotassium. The average " adjusted " concentra-tions of potassium in the intracellular water ofmuscle are practically identical in animals showingno symptoms of adrenal insufficiency, whereas inthe muscle of animals in severe adrenal insuffi-ciency, the concentration of potassium is in-creased about 14 per cent.

The concentrations of muscle phosphorus ofthe animals in adrenal insufficiency are slightly

80

BODY WATERAND ELECTROLYTESIN ADRENALINSUFFICIENCY

higher than those of the controls or the treatedanimals. However, the considerable scatteringof the values does not permit much emphasis tobe placed on these differences.

In the liver, the results are quite in contrast tothose of the muscles. The concentrations ofpotassium, phosphorus and protein in intracellularwater of liver are essentially the same in the ani-mals showing symptoms of adrenal insufficiencyand in the fasted controls. An incidental findingof some interest is the effect of starvation on theprotein concentration in the liver cell. The fastedanimals, whether adrenalectomized or fasted con-trols, show an increased concentration of proteinin the water of the cell indicating a relative lossof water from the cell. The concentrations ofprotein in the liver cell of adrenalectomized ani-mals treated with saline are comparable to thoseof the fed animals. The explanation is pre-sumably the fact that return of appetite was usedas one of the criteria of cure in the case ofadrenalectomized rats treated by daily injectionsof saline. Hence, the animals treated with salinewere not sacrificed until after they had resumedeating for at least one day. Whatever the effectof adrenal insufficiency on the livers, no change inconcentration of potassium in the liver cell isfound, although the normal concentration of potas-sium is slightly higher than that of muscle.

The effect of changes in concentration of sodiumin extracellular water on intracellular

hydrationHaving shown that in adrenal insufficiency the

concentration of base in intracellular water ofmuscle increases while the concentration of basein extracellular water decreases, it seemed impor-tant to determine what effect the changes in therelative concentrations of extracellular and intra-cellular base have on the water content of thecell. Since the movement of water into the cellsresults in a lowered concentration of solids in theintracellular water, the concentration of solids inthe intracellular water may be used as a measureof cellular hydration. The intracellular water ofmuscle tissue can be calculated as previouslydescribed, and the fat free solids of muscletissue are assumed to be entirely intracellu-lar. The previous experiments gave data show-ing a moderate variation in concentration of

extracellular electrolyte but in order to increasethe variations in adrenalectomized animals, deple-tion of extracellular electrolyte by the techniqueof Darrow and Yannet (18) was carried out onadrenalectomized rats. In order to producedeficit of extracellular electrolyte without adrenalinsufficiency, the same procedure was carried outon normal rats.

In the two groups of normal rats, 10 cc. of 5per cent glucose solution per 100 grams of bodyweight was injected into the peritoneal cavity.After an interval of 5 hours, a volume of asciticfluid equal to that injected was removed from theperitoneal cavity. One group of animals wassacrificed immediately while the other group wassacrificed after 48 hours during which only waterby mouth was given. These intact animalsshowed no obvious ill effects following the injec-tion of the glucose solution and its subsequentremoval.

Two groups of adrenalectomized rats showingearly evidences of adrenal insufficiency were in-jected with 7.5 cc. of 5 per cent glucose solutionper 100 grams of body weight. These animalsrapidly went into collapse and at the end of 3hours all were moribund. At this time, the peri-toneal fluid was removed and the animals sacri-ficed.

Two other groups of adrenalectomized ratswere similarly treated but at the time of the in-jection of the glucose solution, 5 cc. of corticalextract was also injected into each animal. Thistreatment protected the rats from prostration anddeath although they still manifested some symp-toms of adrenal insufficiency. After 3 to 4 hours,the peritoneal fluid was removed. One groupwas sacrificed at this time while each rat of theother group received 3 cc. of cortical extract andwas sacrificed 24 hours later. The blood andtissues were in all cases analyzed as in the pre-vious experiments.

The correlation between the concentration ofsolids in intracellular water of muscle and the con-centration of sodium in extracellular water isgraphically shown in Figure 1. Each pointrepresents the pooled analysis of a group of 2 to5 animals. Although there is considerable scat-tering, a definite relationship of the concentrationof intracellular solids to the concentration ofsodium in extracellular water is demonstrated.

81

HAROLDE. HARRISONAND DANIEL C. DARROW

w%__ h

J 350

J O

X-j 340. A

o 330

33M

M.320P

07

J' 310Ci0

A

110 120 130 140 150

MM. No, PER L. EXTRACELLULAR H20

FIG. 1. RELATION OF CONCENTRATIONOF SOLIDS IN

MUSCIE CELL TO CONCENTRATIONOF SODIUM IN EXmA-CELLULARWATERA Fasted control rats.O Rats in adrenal insufficiency.+ Adrenalectomized rats treated with saline.X Adrenalectomized rats treated with cortical extract.

Control rats depleted of sodium.* Adrenalectomized rats depleted of sodium (mori-

bund).Adrenalectomized rats depleted of sodium and pro-

tected with cortical extract.

In other words, as the concentration of sodiumin extracellular water decreases, water diffusesinto the cells in order to maintain osmotic equi-librium. The values for the intact, adrenalec-tomized and the treated adrenalectomized ratsall fall along the same line indicating that theconcentration of extracellular base is the deter-mining factor for concentration of intracellularsolids in both intact and adrenalectomized rats.Since the total body water of the rats representedin Figure 1 remained unchanged, the extent of theshift of water into the cells also measures thedegree of depletion of the volume of extracellu-lar fluid.

The injection of adrenal cortical hormone doesnot affect the distribution of body water exceptthrough changes in concentration of sodium inextracellular fluid, since the untreated and thetreated groups of adrenalectomized rats having

the same depletion of sodium by intraperitonealinjection of glucose show the same degree of dilu-tion of the cells, although the untreated groupwas moribund and the group treated with corticalhormone was protected from immediate death.Furthermore, the intact animals depleted of extra-cellular electrolyte show as marked dilution ofthe cells as the adrenalectomized rats, as is indi-cated by the similar reduction in concentrationof intracellular solids. Nevertheless, this samedegree of hydration of the cells is accompaniedby no obvious symptoms in intact rats while in theadrenalectomized rats a moribund state is induced.These experiments show that the concentration ofextracellular electrolyte is the chief factor con-trolling movement of water into and out of thecells and that changes in volume of either intra-cellular or extracellular fluid are not essentialaspects of adrenal insufficiency in rats nor neces-sary factors explaining the curative effect ofcortical extract. Furthermore, the increased con-centration of intracellular base in muscle of ratsin adrenal insufficiency plays no appreciable rolein the movement of water through the musclecell membrane.

DISCUSSION

In a series of papers Swingle and coworkers(19, 20, 21, 22) have offered evidence that, indogs, adrenal insufficiency is associated withhemoconcentration and lowered plasma volume.These investigators suggest that the chief physio-logical disturbance is depletion of the volume ofextracellular fluid owing to shift of water fromextracellular spaces into the cells (8, 23).Harrop (3) demonstrated that in adrenalec-tomized dogs in collapse the volume of extracellu-lar fluids is actually diminished. The mechanismof the loss of extracellular volume was thoughtto be a shift of water from the extracellular tothe intracellular compartments similar to thatdemonstrated by Darrow and Yannet (18) to beproduced as the result of depletion of extracellu-lar electrolyte by intraperitoneal glucose injec-tion.

In the present experiments on rats, the move-ment of water into and out of the cell is foundto depend chiefly on the concentration of base inthe extracellular water. Since rats in adrenalinsufficiency show only slight reduction of the

82

364

BODY WATERAND ELECTROLYTESIN ADRENALINSUFFICIENCY

concentration of sodium in the extracellular water,little or no changes in water distribution are found.The therapeutic action of cortical hormone inthe adrenalectomized rat may thus be entirely in-dependent of any changes in the volume of intra-or extracellular water. In contrast to rats, thegreater reduction of concentration of sodium inthe serum of dogs (1, 12), cats (13) and humans(14) with adrenal insufficiency may bring aboutan appreciable increase of intracellular water anda considerable reduction in the volume of extra-cellular fluid.

Nevertheless, although changes in distributionof body water may occur in adrenal insufficiencyin certain animals, evidence is accumulating thatreduction of the volume of extracellular waterdoes not play a deciding role in the genesis of thesymptoms following adrenalectomy. Swingle,Parkins, Taylor and Hays (23) observed thatadrenalectomized dogs improved rapidly whengiven large amounts of cortical extract eventhough no change in concentration of sodium inthe serum was found. Conversely, an adrenalec-tomized dog was found to develop severe symp-toms and signs of adrenal insufficiency followingwithdrawal of cortical extract, although becauseof anuria there was no appreciable reduction ofthe concentration of sodium in the serum (8).Their hypothesis that adrenal cortical hormonecontrols the movement of water from the cells tothe extracellular spaces independently of the con-centration of sodium in the extracellular water isuntenable. In the recovery experiments, if waterleft the cells to increase the volume of the extra-cellular fluids, an actual decrease of the sodiumand chloride concentrations in the serum shouldhave been found. Similarly, in the animal goinginto adrenal insufficiency while anuric, an inceaseof the concentration of serum sodium and chlorideshould have taken place if water entered the cellsand thus reduced the volume of extracellularwater. The assumption of Swingle, Parkins,Taylor and Hays (8) that sodium and chloridediffuse out of or into the cell together with water,thus maintaining the concentrations in the extra-cellular fluid unchanged, is not a likely one.Sodium and chloride have not been found withinthe tissue cells in sufficient quantities (24, 25, 26,16) to permit such an assumption. When dogsare treated with cortical extract without sodium

chloride, evidences of increase in plasma volumedo not necessarily indicate an increase in thevolume of extracellular fluid but rather a shift ofextracellular fluid from interstitial spaces into theblood vessels. These changes might be secondaryto improved circulation and vascular tone. Theavailable data indicate that in other animals as inthe rat, the action of adrenal cortical hormone isnot primarily a regulation of the distribution ofbody water.

The present data support and amplify previouswork indicating the importance of potassium inadrenal insufficiency. In adrenalectomized rats,the aggravation of symptoms and the accelerationof their onset by diets high in potassium confirmsprevious work in cats (6), dogs (5) and patientswith Addison's disease (7). As in dogs andhumans, a diet low in potassium protectsadrenalectomized rats from adrenal insufficiency.Furthermore, adrenalectomized rats frequentlyshow a high concentration of potassium in serum.The increased concentration of potassium inserum in adrenalectomized animals is apparentlythe result of changes in renal function sinceHarrop, Nicholson and Strauss (27) have shownthat excretion of potassium is diminished in ani-mals suffering from adrenal insufficiency and thatrecovery from adrenal insufficiency following theinjection of cortical hormone is accompanied bythe excretion of considerable quantities of potas-sium. Even in normal animals, injection ofcortical hormone produces a transitory loss ofpotassium in the urine (28). However, both theamount of potassium retained during developmentof adrenal insufficiency and the amount excretedduring recovery are too large to be contained inthe extracellular fluids. The suggestion ofZwemer and Truszkowski (29) that the increasedconcentration of potassium in the serum is broughtabout by increased diffusion of potassium fromthe cells is not supported by tissue analyses. Al-though these workers found that the concentra-tion of potassium in the muscle of adrenalec-tomized cats was reduced, this finding is explainedby the increase in muscle water and not by lossof potassium. If the potassium concentration isexpressed per unit of solids, an appreciable in-crease in muscle potassium is demonstrated.Hegnauer and Robinson (30) also showed thatthe concentration of potassium in muscle is cer-

83

HAROLDE. HARRISON AND DANIEL C. DARROW

tainly not reduced in adrenal insufficiency in catsand may be increased. In the rat and presumablyin other animals, the retention of potassium afteradrenalectomy can be accounted for by increasedconcentration in muscle cells as well as increasedconcentration in the extracellular fluid.

It is not possible to consider adrenal insuffi-ciency as a special case of potassium intoxicationowing entirely to increased concentration ofpotassium in extracellular water. In many exam-ples of severe adrenal insufficiency the concentra-tion of potassium in serum is only slightlyelevated. In occasional cases of chronic nephritisin man (31, 32) and in experimental roentgen-ray nephritis (33) in dogs, the concentration ofserum potassium may rise in the terminal stagesto much higher levels than that seen in Addison'sdisease in crisis or in adrenalectomized dogs.Ingle, Nilson and Kendall (34) have demon-strated that cortical hormone prolongs the lifeand capacity for work of rats from which boththe kidneys and adrenals have been removed. Insuch animals given cortical hormone, the con-centration of potassium in serum may be higherthan in animals dying of adrenal insufficiency.Furthermore, no regular correlation between theconcentration of potassium in serum and theseverity of the symptoms has been observed inadrenal insufficiency.

Since in the rats not subject to adrenal insuffi-ciency the concentration of potassium in musclecells is quite constant, the higher concentrationin adrenal insufficiency may be assumed to indi-cate important and profound changes in themuscle cell. Although at present one cannotcorrelate increased concentration of muscle potas-sium with definite physiological functions, cer-tain features of adrenal insufficiency point todisturbed function of the muscle. Rapid muscu-lar fatigue is an outstanding phenomenon in alladrenalectomized animals (35, 36) as well as inpatients with Addison's disease. Fenn and Cobb(37) have found that, normally, muscular activ-ity is associated with changes in the content ofpotassium in muscle cells since muscles stimu-lated repeatedly lose potassium and, during theprocess of recovery, potassium diffuses back intothe muscle cell. It will be important to determinewhether, in adrenal insufficiency, the concentra-tion of cell potassium is increased in cardiac and

smooth muscle and in the kidneys, as these tis-sues as well as skeletal muscle show marked dis-turbances of function.

The excess of potassium in the muscle cells inadrenal insufficiency must be present in an almostentirely unionized or osmotically inactive form.Otherwise, it would be impossible for increasedconcentration of base in intracellular water toexist concomitantly with decreased concentrationof base in extracellular water unless cell mem-branes were impermeable to water. That themembranes of muscle cells in adrenal insufficiencyare as freely permeable to water as those ofnormal animals is demonstrated by the compar-able dilution of cellular solids when the concen-tration of sodium in extracellular water isdecreased. Since so little is known of the actualcombinations of muscle potassium, the mechanismby which intracellular potassium becomes osmoti-cally inactive cannot be discussed.

It has been conclusively demonstrated that ad-ministration of relatively large amounts of sodiumsalts may protect adrenalectomized animals forlong periods (28, 39, 40). In spite of the rela-tively slight decrease in concentration of sodiumin serum in adrenalectomized rats, sodium saltshave been found just as effective in prolonginglife in adrenalectomized rats as in adrenalec-tomized dogs in which decrease in concentrationof sodium is much greater. In the experimentsreported above, hypertonic solutions of sodiumchloride and sodium bicarbonate were given for3 to 4 days. Improvement was never as rapid asfollowing the injection of cortical hormone, butafter 3 to 4 days apparent recovery occurred evenwhen food was withheld. Presumably, followingeach injection, the slight deficit of sodium andchloride was replaced, but apparent recovery onlyoccurred when large amounts of sodium chloridewere given over a considerable period. On analy-sis the animals cured with sodium salts showednormal concentrations of potassium in the intra-cellular water of muscle as well as normal con-centrations of sodium, chloride and potassium inextracellular fluids. Hence, one aspect of thecurative effect of sodium salts may be facilitationof the release of potassium from muscle cells andthe excretion of potassium in the urine.

The mechanism by which changes in the con-centration of sodium in extracellular fluid may

84

BODY WATERAND ELECTROLYTESIN ADRENALINSUFFICIENCY

influence the concentration of potassium withinthe muscle cell is unknown. However, it haslong been known that the effectiveness of adrenalcortical extract is enhanced by the simultaneousadministration of sodium chloride. Moreover,reduction of the concentration of sodium in extra-cellular fluids reduces the effectiveness of thehormone. This fact was demonstrated in the ex-periments in which adrenalectomized rats werefurther depleted of sodium by intraperitoneal in-jection of glucose solution. It had previouslybeen found that 5 cc. of cortical extract sufficedto produce a rapid cure of all symptoms in ratswhich had been allowed to progress to the terminalstages of insufficiency. However, in the adre-nalectomized rats further depleted of sodiumchloride, the same amount of extract merely pro-tected the animals from prostration and imme-diate death since symptoms such as diarrhea andmuscular weakness persisted and the blood non-protein nitrogen remained elevated. Further-more, muscle potassium was not restored tonormal as was always the case with adrenalec-tomized rats not further depleted of sodium andchloride. Future work may show an interrela-tionship between the regulation of the concentra-tion of potassium within the cell by the adrenalcortical hormone and the control of the concen-tration of sodium and other solutes in the extra-cellular water.

SUMMARY

By analysis of the tissues, the distribution ofbody water and electrolyte was studied in rats inthe terminal stages of adrenal insufficiency and inrats cured of adrenal insufficiency.

Following adrenalectomy, rats show diminishedconcentrations of sodium and chloride in theserum and increased concentrations of potassiumand nonprotein nitrogen. The changes in theconcentrations of sodium and chloride are lessmarked in rats than in dogs, cats and man inadrenal insufficiency.

The acceleration of the onset of symptoms ofadrenal insufficiency by a high intake of potas-sium previously found in dogs, cats and humansis confirmed in adrenalectomized rats.

The concentration of potassium in the musclecells is constantly increased in rats showingmarked symptoms of adrenal insufficiency,

whereas in animals cured of adrenal insufficiencyeither by adrenal cortical extract or by a hyper-tonic solution of sodium chloride and sodiumbicarbonate, the concentrations of potassium inthe muscle cells are identical with those foundin unoperated controls. No changes in the potas-sium content of the liver cells are found. Thesefindings may be associated with the functionalchanges of muscle in adrenal insufficiency.

In rats with adrenal insufficiency, as well ascontrol animals, the water content of the musclecell is inversely related to the concentration ofsodium in the extracellular water. The adrenalcortical hormone does not regulate the movementof water into and out of the cell except as it in-fluences the concentration of sodium in the extra-cellular water. In the rat, the symptoms ofadrenal insufficiency are not related to changes inthe distribution of body water.

BIBLIOGRAPHY1. Loeb, R. F., Atchley, D. W., Benedict, E. M., and

Leland, J., Electrolyte balance studies in adre-nalectomized dogs with particular reference to theexcretion of -sodium. J. Exper. Med., 1933, 57,775.

2. Loeb, R. F., Atchley, D. W., and Stahl, J., The r6leof sodium in adrenal insufficiency. J. A. M. A.,1935, 104, 2149.

3. Harrop, G. A., The influence of the adrenal cortexupon the distribution of body water. Bull. JohnsHopkins Hosp., 1936, 59, 11.

4. Hastings, A. B., and Compere, E. L., Effect of bi-lateral suprarenalectomy on certain constituents ofthe blood of dogs. Proc. Soc. Exper. Biol. andMed., 1931, 28, 376.

5. Allers, W. D., Nilson, H. W., and Kendall, E. C.,Studies on adrenalectomized dogs: The toxic ac-tion of potassium. Proc. Staff Meet., Mayo Clin.,1936, 11, 283.

6. Zwemer, R. L., and Truszkowski, R., The importanceof corticoadrenal regulation of potassium metab-olism. Endocrinology, 1937, 21, 40.

7. Wilder, R. M., Kendall, E. C., Snell, A. M., Kepler,E. J., Rynearson, E. H., and Adams, M., Intakeof potassium, an important consideration in Ad-dison's disease. A metabolic study. Arch. Int.Med., 1937, 59, 367.

8. Swingle, W. W., Parkins, W. M., Taylor, A. R., andHays, H. W., The influence of adrenal corticalhormones upon electrolyte and fluid distribution inadrenalectomized dogs maintained on a sodium andchloride free diet. Am. J. Physiol., 1937, 119, 684.

9. Britton, S. W., and Silvette, H., On the function ofthe adrenal cortex-general, carbohydrate and cir-culatory theories. Am. J. Physiol., 1934, 107, 190.

85

HAROLDE. HARRISONAND DANIEL C. DARROW

10. Peters, J. P., and Van Slyke, D. D., QuantitativeClinical Chemistry, Volume II. Methods. Wil-liams and Wilkins Co., Baltimore, 1932.

11. Fiske, C. H., and Subbarow, Y., The colorimetric de-termination of phosphorus. J. Biol. Chem., 1925,66, 375.

12. Harrop, G. A., Soffer, L. J., Ellsworth, R., andTrescher, J. H., Studies on the suprarenal cor-tex. III. Plasma electrolytes and electrolyte ex-cretion during suprarenal insufficiency in the dog.J. Exper. Med., 1933, 58, 17.

13. Baumann, E. J., and Kurland, S., Changes in theinorganic constituents of blood in suprarenalec-tomized cats and rabbits. J. Biol. Chem., 1926-27,71, 281.

14. Loeb, R. F., Chemical changes in the blood in Ad-dison's disease. Science, 1932, 76, 420.

15. Allot, E. N., Chemical changes in the blood in Ad-dison's disease and their alteration in response totreatment. Lancet, 1936, 1, 1406.

16. Rubin, M. I., and Krick, E. T., The salt and watermetabolism of adrenal insufficiency and partialstarvation in rats. J. Clin. Invest., 1936, 15, 685.

17. Harrison, H. E., Darrow, D. C., and Yannet, H., Thetotal electrolyte content of animals and its prob-able relation to the distribution of body water.J. Biol. Chem., 1936, 113, 515.

18. Darrow, D. C., and Yannet, H., The changes in thedistribution of body water accompanying increaseand decrease in extracellular electrolyte. J. Clin.Invest., 1935, 14, 266.

19. Swingle, W. W., Pfiffner, J. J., Vars, H. M., Bott,P. A., and Parkins, W. M., The function of theadrenal cortical hormone and the cause of deathfrom adrenal insufficiency. Science, 1933, 77, 58.

20. Swingle, W. W., Pfiffner, J. J., Vars, H. M., andParkins, W. M., The effect of hemorrhage on thenormal and adrenalectomized dog. Am. J.Physiol., 1934, 107, 259.

21. Swingle, W. W., Vars, H. M., and Parkins, W. M.,A study of the blood volume of adrenalectomizeddogs. Am. J. Physiol., 1934, 109, 488.

22. Swingle, W. W., Parkins, W. M., and Taylor, A. R.,Experiments on intact and adrenalectomized dogssubjected to sodium and chloride depletion by in-traperitoneal injections of glucose. Am. J.Physiol., 1936, 116, 430.

23. Swingle, W. W., Parkins, W. M., Taylor, A. R., andHays, H. W., Relation of serum sodium and chlo-ride levels to alterations of body water in theintact and adrenalectomized dog, and the influenceof adrenal cortical hormone upon fluid distribution.Am. J. Physiol., 1936, 116, 438.

24. Peters, J. P., Body Water, C. C. Thomas Co., Spring-field, 1935, Chapter 6.

25. Fenn, W. O., Electrolytes in muscle. Physiol. Rev.,1936, 16, 450.

26. Hastings, A. B., and Eichelberger, L., The exchange

of salt and water between muscle and blood. I.The effect of an increase in total body water pro-duced by the intravenous injection of isotonic saltsolution. J. Biol. Chem., 1937, 117, 73.

27. Harrop, G. A., Nicholson, W. M., and Strauss, M.,Studies on the suprarenal cortex. V. The influ-ence of the cortical hormone upon the excretion ofwater and electrolytes in the suprarenalectomizeddog. J. Exper. Med., 1936, 64, 233.

28. Harrop, G. A., and Thorn, G. W., Studies on thesuprarenal cortex. VI. The effect of suprarenalcortical hormone upon the electrolyte excretion ofthe intact normal dog. A proposed method ofcomparative assay. J. Exper. Med., 1937, 65, 757.

29. Zwemer, R. L., and Truszkowski, R., Potassium: Abasal factor in the syndrome of corticoadrenal in-sufficiency. Science, 1936, 83, 558.

30. Hegnauer, A. H., and Robinson, E. J., The waterand electrolyte distribution among plasma, redblood cells and muscle after adrenalectomy. J.Biol. Chem., 1936, 116, 769.

31. Rabinowitch, I. M., On the relative proportions ofsodium, potassium, calcium and magnesium inblood plasma in renal disease. J. Biol. Chem.,1924-25, 62, 667.

32. Nelken, L., and Steinitz, H., tUber den Gehalt desBlutserums an Calcium und Kalium bei Nieren-krankheiten. Ztschr. f. klin. Med., 1926, 103, 317.

33. Bolliger, A., and Breh, F., tSber die Mineralstoff-verinderungen des Blutes bei experimentellerNephritis, mit spezieller Berucksichtigung desKalium-Kalziumspiegels im Serum. Zentralbl. f.inn. Med., 1928, 49, 825.

34. Ingle, D. J., Nilson, H. W., and Kendall, E. C., Theeffect of cortin on the concentrations of some con-stituents of the blood of adrenalectomized rats.Am. J. Physiol., 1937, 118, 302.

35. Hales, W. M., Haslerud, G. M., and Ingle, D. J.,Time for development of incapacity to work inadrenalectomized rats. Am. J. Physiol., 1935, 112,65.

36. Everse, J. W. R., and DeFremery, P., On a methodof measuring fatigue in rats and its applicationfor testing the suprarenal cortical hormone (Cor-tin). Acta brev. Neerland., 1932, 2, 152.

37. Fenn, W. O., and Cobb, D. M., Electrolyte changesin muscle during activity. Am. J. Physiol., 1936,115, 345.

38. Harrop, G. A., Soffer, L. J., Nicholson, W. M., andStrauss, M., Studies on the suprarenal cortex. IV.The effect of sodium salts in sustaining the supra-renalectomized dog. J. Exper. Med., 1935, 61, 839.

39. Rubin, M. I., and Krick, E. T., Effect of adrenalec-tomy on salt metabolism in rats. Proc. Soc. Exper.Biol. and Med., 1933-34, 31, 228.

40. Allers, W. D., The influence of diet and mineral me-tabolism on dogs after suprarenalectomy. Proc.Staff Meet., Mayo Clin., 1935, 10, 406.

86