STUDIES ON THE EXCHANGE OF POTASSIUM BETWEENTUMOUR CELL AND MEDIUM

. ARTHUR LASNITZKI

(From the Biochemical Laboratory of the Cancer Research Institute, University of Berlin, Germany,and the Cancer Research Department, University of Manchester, England)

The present paper deals with the manner in which the exchange of potas­sium between tumour cell and medium is effected, and the means by which itmay be controlled, a subject which appears to be of importance in view ofthe rOle of potassium in the life processes within the tumour cell. Theinvestigation was suggested by the results of research on the influence ofpotassium and calcium upon the energy-metabolism, especially the fermenta­tion capacity, of tumours (2-8).

METHODS

The experiments were performed in vitro, and the general conditions were,as far as possible, the same as those employed in the metabolism experimentsmentioned.

Tumour Material: Jensen rat sarcoma was used, being particularly suit­able as it contains little stroma. The tumours were produced in the usualway, by subcutaneous implantation of minced tumour tissue in adult rats.

Treatment of the Tissue: In each experiment a tumour, removed imme­diately after the animal had been killed, was divided into several portions, theperipheral areas of which were cut, by means of a razor, into a fairly largenumber of thin slices. These were put, at room temperature, in groups, intoRinger's solution modified by varying the amounts of potassium and calcium.In the earlier experiments the portions of tumour were immersed in the solu­tion and then sliced, the razor being moistened only with the particularsolution concerned. In the later experiments the tumour, while being protectedagainst drying, was cut into slices with a dry razor, and slices were put al­ternately into the various media. Both methods gave similar results. Asfar as possible the thickness of the slices was limited to that advocated byWarburg (1), but because of the necessity of employing all the availablematerial the use of somewhat thicker slices could not be avoided. For ourinvestigations, however, this should be of no consequence.

After the removal of any necrotic tissue, the tumour slices were washedtwice by agitating them carefully in 40 or 50 c.c. of the modified Ringer'ssolution for ten minutes. Then, after each slice had been rinsed rapidly indistilled water, they were transferred to weighing glasses. In the earlier ex­periments individual groups of slices were transferred to the weighing glassesone after the other, for the most part in the order in which they had beenprepared; in the later experiments the slices belonging to the different groupswere, as in their preparation, transferred alternately. In both instances it

513

514 ARTHUR LASNITZKI

was the intention to equalise, as far as possible, the average time during whichthe slices in the different groups were kept in contact with the mediurn. Thelength of this time, always amounting to several hours, depended on the totalnumber of slices to be handled, the amount of necrotic tissue to be removed,the assistance available, and other incidental factors.

Frequently, small pieces of tumour, after being freed from any necrotictissue and rinsed rapidly in distilled water, were transferred to weighingglasses without further treatment. These pieces served for estimating theoriginal potassium content of the tissue. In some of these experiments theinvestigation of the potassium exchange was restricted to one medium only.

Media: The solutions in which the tumour slices were suspended weresimilar to those used in the experiments on the influence of potassium andcalcium upon the fermentation of tumours. Their composition is shown inTable I.

TABLE I: Composition of Modified Ringer's Solution, Used as Media

WithoutWithout Without

Without Withcalcium, calcium, withpotassium with normal increased potassium, potassium

and content of content of with andcalcium potassium potassium calcium calcium

NaCI. ........ 0.124 M 0.1215 M 0.084 M 0.1215 M 0.119 MKCI .......... - 0.0025 M 0.04 M - 0.0025 MCaCI2 ••••••••. - - - 0.0018 M 0.0018 MNaHCO•...... ~ 0.025 M ~

Glucose ....... ~ 0.2% ~

As will be seen, potassium and calcium chlorides are added to the mediumin place of an isotonic amount of sodium chloride. Attention may further bedrawn to the fact that the media contain a physiological amount of sodiumbicarbonate, but no definite amount of free carbonic acid. This conditionexisted, also, during the preparation period of the fermentation experiments,but not during the manometric measurements, which were carried out afterthe media had been saturated with about 5 per cent CO2 in nitrogen. In thefermentation experiments the media attained, therefore, a rather constantphysiological PH, but this was not so in the experiments now being considered.Colorimetric estimations showed that the pH was within physiological limitsonly in the beginning (- 7.3 to 7.4), since the bicarbonate stock solutionsused for the preparation of the media had been saturated with CO2 • Laterthe pa gradually increased, finally reaching values of 8.4 to 8.6. It is un­likely, however, that this difference should interfere with a comparison be­tween the two series of investigations. The same may be said of the differ­ence in temperature.

Estimation of Potassium: The tissue, placed in weighing glasses, wasdried at about 100°, and the dry weight then estimated. Afterwards thetissue was ashed by dissolving it in fuming nitric acid, pouring the solutioninto a micro-Kjeldahl flask, and heating with the addition of more fumingnitric acid and of hydrogen peroxide. Finally the ash residue was dissolved

EXCHANGE OF POTASSIUM BETWEEN TUMOUR CELL AND MEDIUM 515

in distilled water, and the solution brought to a volume of 25 c.c, Of thissolution 10 c.c. were pipetted into each of two centrifuge tubes. (In somecases the solution was, as a whole, transferred into a centrifuge tube, andbrought to a volume of 10 c.c.) The estimation of potassium was carriedout according to the method of Kramer and Tisdall. After removal of am­monia, the potassium was precipitated by addition of 20 c.c. of the cobalti­nitrite reagent to each 10 C.c. of ash solution. The precipitate was then sep­arated by centrifuging, and washed five times with distilled water. Theoxidation was carried out in two stages: (1) preliminary oxidation by meas­ured volumes of 0.02 M KMn04 solution, and (2) final titration with 0.01M KMn04 solution, both being standardised against 0.01 M oxalic acid. Theagreement between the two single estimations was, on the whole, satisfactory.

The total amount of potassium was calculated according to the linearequation:

K = 2.5·(m·V -a) (or, respectively, with 1.0 as factor)

in which K is the quantity of potassium in mg., V the (average) volume ofused KMnO. solution in c.c. (considered as 0.01 M), while m and a are con­stants, the value of which had been found by similar estimations carried out,under the same conditions as above, with a series of KCI solutions of knownconcentrations. (The value of m was 0.069, that of a was 0.042.) Forsmall quantities of potassium, however, this equation could not be utilized,the volume of used KMn04 solution in this range being less than would beexpected. In such cases the determination was performed graphically. Fromthe quantity of potassium thus determined the amount of potassium per 100mg. dry weight of tissue was finally calculated.

RESULTS

The individual results obtained in 40 experiments are shown in Table II,the potassium content of the tissue being calculated to 0.05 mg. of potassiumper 100 mg. dry weight of tissue. The figures in parentheses indicate the ac­tual amount of tissue (mg. dry weight) employed for the corresponding estima­tion of potassium. The age of the tumour and the approximate average timeduring which the slices had been kept in contact with the media are also given.It is to be noted that neither age of tumour nor time of contact, within thegiven limits, appears to bear any relation to the results.

Original Potassium Content of the Tissue: In 27 different experimentsthe original potassium content of the tissue was estimated. The average valuewas found to be 1.78 mg. potassium per 100 mg. dry weight. This agrees inorder of magnitude with figures obtained by a number of authors workingwith other tumours, both spontaneous and experimental. As to the variationof individual values, there appears to be a slight tendency for higher valuesto be more frequent in the later than in the earlier experiments. The stand­ard deviation is + 0.32.

In 4 additional experiments, which are not included in the Table, theoriginal potassium content in four or five different parts of the tumour wasestimated, without any further test. The difference between the single values

516 ARTHUR LASNITZKI

and the corresponding mean values averaged about 7 per cent. Similarly inexperiments Nos. 18 and 22 the estimation was carried out twice, and hereagain the agreement was satisfactory.

Diffusibility of Tissue Potassium: In 15 experiments the original potas­sium content of tissue was compared with the potassium content of tissuefrom the same tumour after suspension in the solution containing neither

TARLE II: Potassium Content of Jensen Rat Sarcoma Originally andAfter Suspension in Different Media

(Mg. potassium per 100 mg. dry weight)

Tis8ue suspended In Ringer's solution

Withpotassium

andcalcium

(0.0025 Mxci,

0.0018 MCaCl.)

Withoutpotassium,

withcalcium

(0.0018 MCaCI.)

Withoutcalcium.

withpotassium

(0.04 MKCI)

Withoutcalcium,

withpotasaium(0.0025 M

KCI)

Withoutpotassium

andcalcium

1----:-----:-----,-----:--,------1 ~i~~~fcontactwith the

media(Approx­imate)

Hours

Ttssuenon ...

suspended

Ageof thetumor

indaysafter

inocu­lation

DateEx·peri­

mentNo.

0.15 (127)0.30 (124)0.30 (86)

0.25 (156)0.20 (118)

0.35 (67)0.30 (131)

0.20 (171)

0.45 (43)

- 0.55 (183) -2.10 (157) -

- 0.50 (102) -- 0.40 (74) -

1.40* (21) - -

- - 1.20 (99)- 0.65 (103) 1.50 (71)- - 1.55 (44)

1.05 (154) -- - 0.65 (III)- - 1.15 (191)- - 0.55 (III)

I Dec. 10, '302 Dec. 18, '303 Jan. 13, '314 Jan. 27, '315 Nov. 7, '326 Nov. 8,'327 Nov. 29, '328 Dec. 2, '329 Dec. 16, '32

10 Jan. 6, '33II Mar.24, '3312 Apr. 7, '3313 Jan. 12, '3414 Jan. 19, '3415 Jan. 31. '3416 Feb. 2, '3417 Feb. 21, '3418 Feb. 28. '3419 Mar. 2. '3420 Mar. 12, '3421 Mar. 14, '3422 Mar.20,'3423 Mar. 26. '3424 Apr. 9, '3425 Apr. 12, '3426 Apr. 25, '3427 May 3, '3428 May 9, '3429 Nov. 29, '3430 Nov. 30, '3431 Dec. 5, '3432 Dec. 19, '3433 Feb. 4, '3534 Feb. 20, '3535 Feb. 21, '3536 Mar. I, '3537 Mar. 12, '3538 Mar. 14, '3539 Mar. 19, '3540 Apr. 8, '35

? - 0.25 (114)? - 0.35 (l05~

28 - 0.50 (7121 - 0.40 (21617 -18 - 0,00 (203)21 1.00 (91) 0.10 (178)24 0.15 (275)20 1.25 (t96)20 1.80 (156) 0.25 (361)21 1.55 (121)21 1.80 (t42) 0.55 (172)15 1.70 (81) 0.35 (212)22 0.40 (107)16 2.00 (96)18 1.40 (40) 0.25 (106)15 1.95 (41) 0.30 (133)22 2.00" (70) 0.25 (141)24 1.10 (55) -17 2.15 (59) -19 1.75 (92) 0.25 (200)25 2.10* (67). -31 - 0.30 (81)24 - 0.25 (91)27 1.45 (37)14 2.00 (t9)22 1.65 (34)28 1.50 (43)2223 2.00 (131)28 2.05 (80)19 2.15 (73)24 1.95 (57)1920 1.95 (62)28 ~

18 12.15 (43)20 2.10 (53)25 1.85 (16)24 1.75 (60)

0.40 (104)

0,50 (211)0.30 (162)0.20 (174)

0.35 (237)

0.75 (77)

0.45 (104)

0.40 (92)0.40* (108)

o.sor (146)0.45 (93)

0.60 (53)

0.45 (64)

0.30 (107)

0.40 (51)

0.40 (185)0.60 (46)0.40 (150)

1.15 (210)

0.90 (279)0.35 (196)0.70 (96)

0.45 (131)

1.30 (89)

0.35 (140)

1.60 (52)1.85t (116)

1.45 (84)1.60 (66)

0.55 (73)

- 0.40 (96)

1.55 (39) -

1.45 (114)0.95 (204)0.80 (199)

1.15 (161)

0.90 (61)1.35 (84)

1.15 (33)

0.90 (66)

1.30 (47)

1.35 (127)

1.45 (115)

?32312133313213?5314315144131441421441434335134

~I31453

• Mean values of two parallel estimations {In parentheses corresponding mean values of the amounts nft Mean values of three parallel estimations tissue employed.

potassium nor calcium. The average potassium content of the non-suspendedtissue was here again 1.78 mg. per 100 mg. dry weight, while the correspondingvalue for the tissue which had been suspended was 0.28 mg. (Fig. 1). Theaverage decrease in potassium content of the tissue was, therefore, about 85per cent, the loss in potassium being in the majority of experiments between80 and 90 per cent.

From this result it is concluded that at least most of the potassium con­tained in Jensen rat sarcoma cells is freely diffusible, under the given condi-

EXCHANGE OF POTASSIUM BETWEEN TUMOUR CELL AND MEDIUM 517

tions. In order to determine whether the residual potassium 1 could be stillfurther reduced by a prolongation of contact with the medium, in 2 of theexperiments (Nos. 31 and 35) part of the tissue was left in the solution fortwo hours longer than is indicated in the Table. A further decrease ofabout 0.10 mg. potassium per 100 mg. dry weight was obtained in both cases.Thus it appears that the diffusion of potassium had not been completed withinthe given time periods, and that the average amount of potassium which canbe liberated from the tissue probably exceeds the limit of 85 per cent. Theinfluence of the time factor cannot be deduced, however, by comparison of theresults obtained in different experiments.

Potassium Content of the Tissue in Relation to that of the Medium: Thequestion now arises as to what extent the potassium content of the tissue canbe increased by the addition of potassium to the medium. In 9 experimentsthe potassium content of tissue after suspension in the solution without potas­sium and calcium was compared with the potassium content of tissue from the

FIG. 1. POTASSIUM CONTENT OF JENSEN SARCOMA AFTER SUSPENSION IN A MEDIUM WITHOUT

POTASSIUM AND CALCIUM, COMPARED WITH THE ORIGINAL POTASSIUM CONTENT

(AVERAGE OF IS TESTS)

In this and the following figures the columns indicating the original potassium content oftissue, and the potassium content of tissue after suspension in media (1) without potassium andcalcium, (2) with potassium only, (3) with calcium only, and (4) with both potassium and cal­cium, are marked respectively: 0, Na, Na + K, Na +Ca, Na + K + Ca.

same tumour after suspension in a similar solution to which a normal amountof potassium (0.0025 M KCI) was added. The average potassium contentof the tissue per 100 mg. dry weight after suspension in the former mediumwas 0.27 mg.; after suspension in the latter it was 0.41 mg. (Fig. 2a). Thusthe average increase was about 50 per cent. Although the absolute differenceappears to be small, its correctness is supported by the fact that, so far asthe direction of the result is concerned, all the experiments were in accord.

As the potassium content of the medium is increased, the increase in thepotassium content of the suspended tissue also becomes more pronounced. In5 experiments the potassium content of tissue after suspension in the solutionwithout potassium and calcium was compared with the potassium content oftissue from the same tumour after suspension (a) in a similar solution con­taining potassium in the concentration given above (0.0025 M KCI) and (b)in a similar solution containing potassium in a higher concentration (0.04 M

1 In this connection see also other experiments in which the potassium content of tissue aftersuspension in a potassium-free and calcium-free medium was estimated.

518 ARTHUR LASNITZKI

KCI). The average potassium content of the tissue per 100 mg. dry weightwas 0.30 mg. after suspension in the potassium-free medium, 0.44 mg. aftersuspension in the medium with a normal potassium content, and 1.36 mg. aftersuspension in the medium with a high potassium content (Fig. 2b ). Thus,while the potassium content of the medium was increased 16 times, the increasein the potassium content of the tissue was, on an average, only about 7.6 timesgreater. Some variation of this value was observed in individual experiments.

Since it appeared possible that the values found for the potassium contentof tissue after suspension in the potassium-containing media might be toosmall, because of washing away of part of the potassium within the tissuewhile it was being rinsed in distilled water, tests were performed to find outwhether the time during which the tissue was in contact with distilled waterinfluenced the result. This was done with the tissue which had been sus­pended in solution containing 0.0025 M KCI in experiments Nos. 20 and 22,and with the tissue which had been suspended in solution containing 0.04 M

aFIG. 2. EFFECT OF THE ADDITION OF POTASSIUM TO THE MEDIUM ON THE POTASSIUM CONTENT

OF JENSEN SARCOMA

(a) Addition of 0.0025 M KCI (average of 9 tests).(b) Addition of 0.04 M KCI (average of 5 tests).

KCI in experiments Nos. 25 and 30. As usual, every tissue slice was rapidlymoved to and fro in distilled water, but in experiments Nos. 22 and 30 aboutone third of the slices were moved only once, a second third were moved 4times, which corresponds roughly to the ordinary manner of rinsing, andthe last third 7 times; while in experiments Nos. 20 and 25 the slices weremoved once and 4 times only. It was found that the differences in the potas­sium content of the tissue, separated in this way, were only slight, if noticeableat all, and were of such a nature that no influence of the kind mentionedcould be detected." (It is for this reason that only the mean values obtainedfrom the two or three parallel estimations are given in the Table.) Thus wemay assume that the potassium content as estimated represents in each casethe true potassium content of the tissue under the given conditions.

An approximate calculation was also made of the surplus of tissue potas­sium, which is due to, and apparently in equilibrium with, the potassium ofthe medium, per unit of tissue water. The original water content of the tissue

2 In experiment No. 33, however, part of the tissue which had been suspended in the solutioncontaining 0.04 M KCI was washed thoroughly for about one minute in distilled water, andthereby the potassium content of the tissue was reduced to nearly one half.

EXCHANGE OF POTASSIUM BETWEEN TUMOUR CELL AND MEDIUM 519

is about 80 per cent, but that of the suspended tissue, especially in the absenceof calcium in the medium, is probably somewhat greater. Assuming that thetissue suspended in the solution with 0.0025 M KCI contains 85 + 2 percent water, we shall find that 1 gm. of tissue water contains, on an average,a surplus of 0.25 + 0.04 mg. potassium. As the medium contains about0.10 mg. potassium per 1 gm. of water, the average ratio:

Increase of potassium concentration in tissue waterPotassium concentration in medium water

is, therefore, --- 2.5 + 004. The accuracy of this ratio depends on the validityof our assumption that the water content of the suspended tissue is 85 + 2per cent. With a water content lower than 83 per cent the ratio will increaseabove 0.29, while if the water content exceeds 87 per cent it will decreasebelow 0.21. It must be remembered, also, that only the free water, not thatbound to the cell colloids, can be a suitable solvent for potassium. The ratiothereby will naturally be increased, but to what extent we do not know. Inany case it may be concluded, with some reservation, that although the averagesurplus of tissue potassium, due to the presence of potassium in the medium,is rather small, it is at least twice as great as it would be if its distributionthroughout the tissue were a matter of solubility alone. The most likelyexplanation appears to be that part of the potassium concerned is adsorbedby certain structural elements of the tumour cell. Needless to say, the re­maining part, which is in true solution, will not be contained in the cells alone,but also, probably to a slight extent, in the spaces between the cells.

Our conception is further supported by the fact that the surplus of tissuepotassium did not increase 16 times, but, on an average, only about one halfthat figure, when the medium contained 0.04 M instead of 0.0025 M KCI.While, under this condition, that part of the surplus which is in true solutionwill probably become 16 times greater, the amount adsorbed must increaseless than that. The extent of this latter increase it is difficult to determine,but a simple numerical consideration shows that it cannot be great.

Finally, it is quite possible that it is not the surplus but the total amountof tissue potassium which, in reality, is in equilibrium with the potassiumof the medium, seeing that a true equilibrium in the potassium-free mediummay possibly correspond with a complete absence of potassium in the tissue.Special investigations will be necessary to ascertain this possibility, whichwould, obviously, favour our conception even more than the result hererecorded.

Potassium Loss of Tissue in Potassium-containing Media: The rathersmall increase in the potassium content of the tissue on the addition of0.0025 M KCI to the medium indicates that, under this condition, the libera­tion of potassium from the tissue was still considerable. In 10 experimentsthe potassium content of tissue after suspension in the solution containing0.0025 M KCI was directly compared with the original potassium content oftissue from the same tumour. An average of 0048 mg. potassium per 100mg. dry weight was here obtained for the suspended tissue, and an averageof 1.92 mg. for the non-suspended tissue, both values being somewhat higherthan before (Fig. 3a). Thus the average decrease of the potassium content

520 ARTHUR LASNITZKI

was 75 per cent, i.e., only a little less than the average loss of potassiumobserved in tissue after suspension in the potassium-free medium. This factappears to be of particular interest because a KCl concentration of about0.0025 M is approximately that contained in media which, normally, are usedfor investigations on cell metabolism, both respiration and fermentation.(The same KCl concentration was to a large extent utilized in the metabolismexperiments already mentioned.) It will be seen that, on an average, threequarters of the amount of potassium present in tumour tissue as taken fromthe body is not in equilibrium with the amount of potassium contained insuch media. The potassium content has, therefore, to be increased if equi­librium is to be obtained.

Even if the potassium content of the medium was 16 times higher thannormal, however, complete equilibrium could not be attained. In 7 experi­ments the potassium content of tissue after suspension in the solution contain­ing 0.04 M KCl was compared with the original potassium content of tissuefrom the same tumour. An average of 1.61 mg. potassium per 100 mg. dry

a b

FIG. 3. POTASSIUM CONTENT OF JENSEN SARCOMA AFTER SUSPENSION (a) IN A MEDIUM CONTAIN­

ING 0.0025 M KCI (AVERAGE OF' 10 'Ii:sTS) AND (b) IN A MEDIUM CONTAINING 0.04 M KCl

(AVERAGE OF 7 TESTS), EACH COMPARED WITH ORIGINAL CONTENT

weight was here obtained for the tissue which had been suspended (a highervalue than before), and an average of 1.80 mg. for the non-suspended tissue(Fig. 3b). Thus we had still an average loss in potassium of 11 per cent,and in order to obtain complete equilibrium a further increase in the potas­sium content of the medium appears to be necessary.

With certain reservations the figures giving the average 'percentage decrease of thepotassium content of tissue which had been suspended in the potassium-free medium, andin the two potassium-containing media, can be utilized for checking our previous findingsregarding the increase in the potassium content of tissue by the addition of potassium tothe medium. The figures 85 per cent, 75 per cent and 11 per cent indicate that the potas­sium content increases, approximately, in the proportions 15:25 and 15:89 respectively.This would mean for the medium with 0.0025 M KCI an average increase from 0.27 to0.45 mg. potassium per 100 mg. dry weight, while for the medium with 0.04 M KCI theaverage increase would be 7.4 times higher. The agreement with the results of our directestimations is quite satisfactory.

The Effect of Calcium: As most of the potassium contained in the tumourcell diffuses freely into a medium in which potassium is absent, and the libera­tion of potassium is not much less if the medium contains potassium in a

EXCHANGE OF POTASSIUM BETWEEN TUMOUR CELL AND MEDIUM 521

physiological concentration, it was of interest to ascertain whether this dif­fusion could be inhibited by the addition of other cations to the medium.Among the cations in question the most important one appears to be calcium.

In 10 experiments the potassium content of tissue after suspension in thesolution without either potassium or calcium was compared with the potassiumcontent of tissue from the same tumour after suspension in a similar solutionto which a certain amount of calcium (0.0018 M CaCI2 ) was added. Onan average, the tissue which had been suspended in the control medium con­tained 0.31 mg. potassium per 100 mg. dry weight, while with the addition ofcalcium it contained 0.55 mg. (Fig. 4a). Thus it is evident that the presenceof calcium in the medium caused a definite increase in the potassium contentof the tissue, the average increase being about 80 per cent. The absolutedifference is not great, but its correctness is again supported by the fact thatthe result of each individual experiment was in the same direction.

The effect of calcium generally became more pronounced, both absolutelyand relatively, if potassium was present in the medium. In 8 experiments

a bFIG. 4. EFFECT OF THE ADDITION OF CALCIl;M (0.0018 M CaCl.) TO THE MEDIUM ON THE

POTASSIUM CONTENT OF JENSEN SARCOMA

(a) Medium without potassium (average of 10 tests).(b) Medium containing 0.0025 M KCl (average of 8 tests).

the potassium content of tissue after suspension in the solution containing0.0025 M KCI was compared with the potassium content of tissue from thesame tumour after suspension in a similar solution containing, in addition,0.0018 M CaCI2 • An average of 0.43 mg. potassium per 100 mg. dry weightwas found for the tissue which had been suspended in the control medium,and an average of 1.18 mg. for the tissue which had been suspended in thecalcium-containing medium (F:ig. 4b). Thus the average increase in thepotassium content of the tissue caused by the presence of calcium in themedium was about 175 per cent. It is to be noted that the average time ofcontact with the media was approximately the same in this and the formergroup of experiments (three and a half hours).

From these results it is concluded that calcium is capable of inhibitingthe diffusion of potassium from the tumour cell into the surrounding medium.There does not seem to be any other possible explanation." The influence

B In this connection it is interesting to note that the tissue suspended in the two calcium­containing media appeared to be more compact and less transparent than that suspended in thecalcium-free media, a difference which may well give indications regarding the explanation of thecalcium etlect.

522 ARTHUR LASNITZKI

of calcium appears to be more effective if the medium contains a small amountof potassium. To a certain extent this may be due to a lessening of thediffusion gradient, but, on the other hand, there is the possibility that thepresence of potassium in the medium may, as such, support the effect ofcalcium. Finally, it is probable that the effect would be similar when diffusiontakes place in an opposite direction.

Potassium Loss of Tissue in Calcium-containing Media: It is clear thatthe outward diffusion of potassium was, at the given (physiological) calciumconcentration, not completely inhibited, but only delayed. This may bequantitatively demonstrated by a comparison of the potassium content oftissue after suspension in the two calcium-containing media with the originalpotassium content of tissue from the same tumour. For this comparison thereare available 6 experiments in which potassium was absent from the medium,and 11 in which it was present. The tissue contained an average of 0.54 mg.potassium per 100 mg. dry weight after suspension in the potassium-free

a b

FIG. 5. POTASSIUM CONTENT OF JENSEN SARCOMA (a) IN A MEDIUM CONTAINING 0.0018 M CaCI,

(AVERAGE OF 6 TESTS) AND (b) IN A MEDIUM CONTAINING 0.0025 M KCI AND 0.0018 M CaCI,

.(AVERAGE OF 11 TESTS), EACH COMPARED WITH THE ORIGINAL CONTENT

medium, and 1.09 mg. after suspension in that containing potassium, while anaverage of 1.78 and 1.64 mg. respectively was obtained for the original potas­sium content of the tissue (Fig. 5a and b). The average decrease in potas­sium content of tissue was, therefore, 70 per cent in the case of the potassium­free medium, and 34 per cent in the case of the potassium-containing medium.Thus the average percentage loss of potassium was reduced to about one halfif potassium were present in the medium, the average time of contact beingapproximately the same for both media (three. and a half hours).

The figures giving the average percentage decrease in the potassium content of tissuewhich had been suspended in the two calcium-containing media, and in the two correspond­ing control media, can be utilized, again with reservation, for checking the results givenin the previous section. A decrease of 85 and 70 per cent respectively when potassium isabsent, and a decrease of 75 and 34 per cent respectively when potassium is present, areapproximately equivalent to an increase in the potassium content of the tissue by the actionof calcium in the proportions of 15:30 and 25:66 respectively. This would mean in theformer case an increase from 0.31 to 0.62 mg. potassium per 100 mg. dry weight, and inthe latter case an increase from 0.43 to 1.13 mg. The agreement is good in one case; inthe other it may be considered as satisfactory in view of the very unequal number of controland test experiments.

EXCHANGE OF POTASSIUM BETWEEN TUMOUR CELL AND MEDIUM 523

Combined Effect of Potassium and Calcium: Finally, in order to estimatethe combined effect of both cations, we have to compare the potassium contentof tissue after suspension in the solution containing 0.0025 M KCI and0.0018 M CaCI" with the potassium content of tissue from the same tumourafter suspension in a similar solution free of potassium and calcium. Forthis comparison 9 experiments are available. An average of 0.26 mg. potas­sium per 100 mg. dry weight was obtained for the tissue which had beensuspended in the control medium, and an average of 1.14 mg. for the tissuewhich had been suspended in the medium with both potassium and calcium(Fig. 6). The average increase was, therefore, nearly 3)/;1 times.

Thus the combined effect is clearly greater than the sum of the two singleeffects exercised by potassium and calcium separately. The relative increaseis approximately 2)/;1 times with regard to the absolute as well as the per­centage figures. This result again demonstrates that the influence of calciumis more effective if the medium contains a small amount of potassium.

FIG. 6. En-EeT OF ADDITION OF BOTH POTASSIUM AND CAl.CIUM (0.0025 M KCI, 0.0018 M CaCU

TO TilE MEDIUM ON THE POTASSIUM CONTENT OF JENSEN SARCOMA (AVERAGJe or 9 TJeSTS)

GENERAL DISCUSSION

Potassium Exchange and Energy-metabolism: As was mentioned in thebeginning of this paper, the investigations considered were suggested by theresults of researches on the influence of potassium and calcium upon theenergy-metabolism of the tumour cell, in particular its fermentation capacity(2-8). With Jensen rat sarcoma and Flexner-Jobling rat carcinoma, it hadbeen shown that the anaerobic fermentation depended to some extent upon thepresence of potassium and calcium in the medium, the increase caused by thetwo cations being greater than that caused by a single one. As regards theaction of potassium alone, the effect occurred not only if the cation was addedto the medium immediately, but also if it was added after the tissue had fer­mented for a while in the control medium. This, however, was not so in thecase of calcium.

In connection with our observation concerning the relation of the potas­sium content of Jensen sarcoma to the concentration of KCI in the medium,it is of interest to note that the effect of potassium upon the intensity offermentation was, on the whole, not much altered in the concentrations used,which varied in nearly all the relevant experiments from 0.0025 to 0.04 M.This result was obtained with both tumours. Thus it must be concluded

524 ARTHUR LASNITZKI

that, although the surplus of tissue potassium, due to the presence of potas­sium in the medium, increases greatly (probably in both tumours), this haslittle influence on the effect of potassium upon fermentation.' This be­haviour cannot, meanwhile, be fully explained, although an explanation mayperhaps be found by considering the action of potassium as consisting of twocomponents, one of which furthers and the other inhibits the process offermentation.

The fact that calcium in a potassium-free medium was capable of in­creasing the fermentation intensity only if it was added immediately, but notif it was added subsequently, may be explained by assuming that the actionof calcium is indirect, taking place by an inhibition of the diffusion of potas­sium from the tumour cell into the medium. In this way calcium acts onlyas long as an outward diffusion of potassium takes place, but is unable to actwhen diffusion has practically come to an end. We have seen that this ex­planation is well supported by the result of our investigations on the effectexercised by calcium upon the potassium content of Jensen sarcoma placedin a potassium-free medium. On the other hand, the fact that the fermenta­tion intensity of tissue was greater in a medium containing both potassium(0.0025 M KCI) and calcium (0.0018 M CaCl~) than in a medium containingonly potassium (as above) " might suggest that, even when potassium waspresent, calcium had acted in a similar manner as in the absence of potassium,on the supposition that the intracellular potassium was not in equilibriumwith that of the medium. In view of the fact that the potassium loss in tissueafter suspension in a medium with 0.0025 M KCI was not much less thanthat in tissue which had been suspended in a potassium-free medium, and thatthe addition of 0.0018 M CaCI~ to the former medium increased the potassiumcontent of the tissue considerably, our explanation is seemingly also correctin the case of the potassium-containing medium.

It appears, however, questionable whether calcium acts solely by inhibit­ing the outward diffusion of potassium, for it is difficult to understand why adefinite increase in the fermentation intensity of tissue, compared with thatobtained in a medium with 0.0025 KC1, apparently occurred only if the cor­responding increase in tissue potassium took place by means of the inhibitoryeffect of calcium, but not by means of an increase in the potassium concen­tration of the medium. In addition, it is possible that calcium acts by com­pensating the presumed inhibitory component of the action of potassium, butif we compare quantitatively the increase in tissue potassium with that infermentation intensity we must admit that this compensation will not becomplete.

Finally, it may be remarked that these considerations probably apply, also,to the action of potassium and calcium exercised upon the respiration of thetumour cell. It was found that both cations were capable of increasing theintensity of respiration, the combined effect of both being clearly greater thanthat of either. The differences observed between the effects upon respiration

4 It appears unlikely that this conclusion will be essentially altered by the differences in theconditions of the two series of investigations, including the difference resulting from the factthat, in the fermentation experiments concerned, potassium was subsequently added to the medium.

"The experiments concerned were carried out with Jensen sarcoma alone, but there is littledoubt that the result will apply also to the Flexner-J obling carcinoma.

EXCHANGE OF POTASSIUM BETWEEN TUMOUR CELL AND MEDIUM 525

and fermentation appeared to be of a quantitative nature only, except forthe fact that potassium was unable to act upon respiration if it were subse­quently added to the medium. Even this difference, however, does not influ­ence our conclusions.

Relation between Intracellular and Extracellular Potassium in the Body:The fact that the potassium originally present in the tumour cell was, underour conditions in vitro, not in equilibrium with the potassium of a mediumcontaining 0.0025 M KCI prompted an inquiry into the behaviour of thetumour cell under natural conditions. If one assumes that the intracellularpotassium is freely diffusible to about the same extent in vitro and in vivo, andthat the potassium content of the medium surrounding the tumour cell in vivois not much greater than that of our medium," it is of importance to find thereason for the high potassium content of tumour tissue as taken from thebody. In this connection it is possible that in vivo the conditions as to equi­librium are different from those in vitro, in so far as the tumour cell is capableof binding more potassium, for instance, by a greater adsorptive power.' Onthe other hand, it may well be that the high potassium content is due to anactive concentration of potassium by the cell by expenditure of energy, thusproducing and maintaining a non-equilibrium. This capacity might disappearoutside the body, so that a quick equilibration takes place if it is not inter­fered with by the presence of calcium. The question which of these twopossibilities will ultimately prove to be correct must be left for the future toanswer. Obviously the question is part of a general problem concerning thetumour cell as well as the normal cell.

SUMMARY

( 1) The exchange of potassium between tumour cell and medium wasstudied, at room temperature, with Jensen rat sarcoma as the tumour materialand modified Ringer's solution, similar to that utilized in fermentation ex­periments, as the medium.

(2) It was found that most of the potassium originally present in thetumour cell was freely diffusible, the potassium content of the tissue beingreduced, on an average, by about 85 per cent after it had been suspended ina medium without potassium and calcium.

(3) The average potassium content of the tissue became about 50 per centgreater if 0.0025 M KCI were added to the former medium. Although theabsolute increase was rather small, it indicated that part of the surplus oftissue potassium, due to the presence of potassium in the medium, did notexist in true solution but in a state of adsorption. Compared with the originalpotassium content of the tissue, there remained an average decrease of 75

6 According to the results of P. K. Smith and A. H. Smith (9) the potassium content of rat'sserum is about twice as great as that of our medium, while the value found by Heller and Paul(10) is about five times as great. But even if the potassium content of our medium were increasedfive times, a rough estimation shows that still about SO to 60 per cent of the intracellular potassiumcould not be in equilibrium with that of the medium.

7 As far as the difference in temperature is concerned, it appears unlikely that the highertemperature of the body will increase the amount of adsorbed potassium, in view of the fact thatthe temperature coefficient of adsorption is, as a rule, negative.

526 ARTHUR LASNITZKI

per cent, indicating that the greater part of the potassium originally presentin the tumour cell was not in equilibrium with that of the medium.

(4) If the KCI concentration of the medium was increased from 0.0025to 0.04 M, the surplus of tissue potassium increased, on an average, nearlyeight times, a finding which from the point of view of adsorption was to beexpected. In the medium with 0.04 M KCI the potassium content of thetissue was, on an average, still about 10 per cent less than its original potas­sium content.

(5) The potassium content of the tissue also increased, if 0.0018 M CaCLwere added to a medium containing no potassium. The average increasewas about 80 per cent, while the average decrease, compared with the originalpotassium content, was 70 per cent.

(6) If the same amount of CaCI~ were added to a medium containing0.0025 M KCI, a more pronounced increase in the potassium content of thetissue was generally obtained. The average increase was about 175 per cent.Compared with the original potassium content of the tissue, there was nowfound an average decrease of about 35 per cent.

(7) The effect of calcium on the potassium content of the tissue, in thepotassium-free as well as the potassium-containing medium, was probably dueto an inhibition of the diffusion of potassium from the tumour cell into themedium.

(8) The combined effect of potassium (0.0025 M KCI) and calcium onthe potassium content of the tissue was approximately 20 times greater thanthe sum of the effects exercised by potassium and calcium separately.

(9) The significance of these results with regard to the influence ofpotassium and calcium on the energy-metabolism of the tumour cell, in par­ticular its fermentation capacity, is discussed, and the relation in whichintracellular and extracellular potassium may find themselves in the body isalso considered.

Note. The author wishes to express his gratitude to Dr. W. Cramer of London, forsupplying strains of the Jensen rat sarcoma, and to Dr. M. Lasnitzki for her valuableassistance.

REFERENCES

1. WARBURG, 0.: The Metabolism of Tumours, English Translation by F. Dickens, London,Constable and Co., 1930.

2. LASNITZKI, A.: Ztschr. f. Krebsforsch. n . 116, 1928.3. LASNITZKI, A., AND ROSENTHAL, 0.: lliochem. Ztschr. 20i: 120. 1929.4. LASNITZKI, A., AND ROSENTHAL, 0.: Biochern, Ztschr. 262: 203, 1933.5. LASNITZKI, A.: Biochem. Ztschr. 264: 285, 1933.6. LASNITZKI, A.: Protoplasm a 22: 274, 1934.7. LASNITZKI, A., AND ROSENTHAL, 0.: Biochem. Ztschr. 281: 395, 1935.8. LASNITZKI, A.: Biochem. Ztschr. 285: 101, 1936.9. SMITH, P. K., AND SMITH, A. H.: J. BioI. Chern. 107: 673, 1934.

10. HELLER, V. G., AND PAUL, H.: ]. BioI. Chern. 105: 655, 1934.