[CANCER RESEARCH 27 Part 1, 1041-1052, June 1967

Relationship between Oxygen and Glucose Consumption by

Transplanted Tumors in Vivo

PIETRO M. CULLINO, FLORA H. GRANTHAM, ANITA H. COURTNEY, AND ILONA LOSONCZY

Laboralorn of Biochemistry, National Cancer Institute, KIH, USPHS, HEW, Hethesda, Maryland 20014

SUMMARY

The in vivo consumption of oxygen and glucose was studied inrelation to growth in Walker Carcinoma 256, Hepatoma 5123,and Fibrosarcoma 4956 transplanted in rats. Glucose not eliminated as lactate or carbon dioxide in the efferent blood was presumed to be retained by the tumor. The retention was so highthat growth alone could not account for it. As an alternative, theelimination of glucose by the tumors in some unknown manneris suggested. The weight-doubling times of the tumors wereindependent of oxygen consumption and lactate production. Thetumors needed oxygen to survive; there was no indication, however, that in vivo tumor metabolism shifted from respiration toglycolysis when the supply of oxygen was deficient. The oppositewas found to be true: glucose consumption and lactate elimination were in direct proportion to the oxygen utilized and a lackof oxygen blocked both of them. The fraction of glucose transformed into lactate was maximal during insulin-induced hypo-glycemia. These glucose-starved tumors did not produce lactateduring glucose refeeding despite a large glucose utilization.Neither lack of oxygen nor large glucose consumption appearedto be the dominant causes of in vivolactate production by tumors.Experimental increases in the lactate content of subcutaneoustissue could be obtained in the absence of any tumor. Thepossibility that glycolysis is related to changes of cellular components not necessarily involved in the neoplastic process issuggested.

INTRODUCTION

The measurement of the in vivo consumption of oxygen andglucose by transplanted rat tumors was reported in 2 accompanying papers (8, 9). The ex(>eriments described here concern theinterdependence of respiration and glycolysis in these tumors.

In vivo glycolysis was discovered when the lactate content ofblood draining an organ bearing a tumor was found to be higherthan in the absence of the tumor (3, 17). The finding was confirmed by the observation of increased blood lactate (12) anddecreased pH (15) following administration of glucose. Glycolysiswas considered essential for the existence of the tumor in vivowhen sections of Jensen sarcoma deprived of oxygen for 72 hoursbut supplied with large amounts of glucose were still able todevelop into a tumor upon transplantation. The sarcoma didnot grow when glucose was not available (13). The finding wasinterpreted to demonstrate that tumor cells could derive from

Received September 30, 1900; accepted January 11, 1907.

glycolysis all the metabolic energy required to survive for aconsiderable period of time.

Cells grown in vivo are usually cultivated in media rich inglucose, and large amounts of lactate are produced. However,Graff et al. (4) observed that when two cell lines were grown ina medium with a constant concentration of glucose of only 5 mgper 100 ml, both lines grew well and would consume rather thanproduce lactate. They considered the high glycolysis found invitro using media rich in glucose as a "detoxicating mechanism"

substituting for the hormonal barrier which, in the in vivosituation, limited the flux of glucose across the cell membranes. Indeedit was later found in vivo that the neoplastic cells grew in amedium containing only a few mg percent of free glucose (16),and the vascular wall served an important function in regulatingglucose transfer from the vascular into the interstitial compartment (8).

In the work presented here, an assessment was made of therelative proportions of oxygen and glucose available and utilizedby tumors in vivo. Expriment« were performed in vivo to testwhether: (a) a relationship between tumor growth and glucoseor oxygen consumption could be found, (¿>)the availability ofoxygen or glucose could switch tumor metabolism from respiration to glycolysis or viceversa, (c) the loss of lactate by the tumorthrough the efferent blood could be interpreted in vivoby Graff'shypothesis of a "detoxication mechanism," and (d) it was

possible to increase the concentration of lactate in the interstitialfluid of the subcutaneous tissues prior to neoplastic transformation.

MATERIALS AND METHODS

The procedures followed in the experiments reported here weredescribed in 2 accompanying papers (8, 9). Walker Carcinoma256, transplanted in Sprague-Dawley rats and Hepatoma 5123and Fibrosarcoma 4956 transplanted in Buffalo/N rats were used.Each tumor was grown as a tissue-isolated preparation with oneafferent artery and one efferent vein (7). The blood flow and thearterial-venous difference (A-V) in oxygen, glucose, lactate, andcarbon dioxide were measured and from these values the consumption of glucose and oxygen and the production of lactateand carbon dioxide were evaluated in vivoas previously described(8, 9). The detailed treatment of the tumor-bearing host (anesthesia, hyper- and hypoglycemia, diabetes, anemia, etc.) was alsopreviously described (8, 9).

RESULTS

Tumor Growth as Related to Oxygon and Glucose Consumption

JUNE 1967 1041

Research. on November 25, 2020. © 1967 American Association for Cancercancerres.aacrjournals.org Downloaded from

[r-

P. M. Gullino, F. H. Grantham, A. H. Courtney, and I. Losonczy

Table 1 .summarizes, for convenient reference, the consumptionof oxygen and glucose as rejjorted in the two previous pa|>ers(8, 9). A striking feature common to the three tumors studied wasthe large in vivo glucose consumption as compared to the relatively small oxygen utilization. The extent of this diffeiencecould be characterized as follows: If the fraction of glucoseeliminated as lactate in the efferent blood was subtracted fromthe total glucose consumed, the quantity of glucose available tothe tumor for oxidation was obtained. Of this amount of glucose,only 13%, 46%, and 8%, res|)ectively, could have been oxidizedby the oxygen consumed by Walker carcinomas, 5123 hepatomas,and 4956 fibrosarcomas in vivo. The brain, which is the normaltissue closest to the tumors as far as glucose consumption isconcerned (1.3 to 3.2 mmoles/hr/100 gm), normally utilizes anamount of oxygen which could oxidize about 85% of the glucoseconsumed (11, 14).

TABLE 1Tumor Growth as Compared with in Vivo Consumption of

Oxygen or Glucose and Production of Lactate

TumorsWalker

Carcinoma256Hepatoma

5123Fibrosarcoma4956Weight-doublins

time(hours)"3319661OxygenTotalcon-sumed62.84.20.8As%sup

plied455746GlucoseTotalcon

sumed5.43.12.5As%supplied282332LactateTotal

eliminated1'3.82.01.7As%

glucosecon

sumed353234

" Calculated from the semilog. Plot of time: dry weight increaseover the period of maximal growth of subcutaneous transplants.

' mmoles/hr/100 gm wet weight. The relationship did notchange when the experimental data were referred to dry weights.Water content in gm/100 gm tumor: Walker carcinoma = 82.2,Hepatoma 5123 = 80.0, Fibrosarcoma 4956 = 84.1.

In the tumors, lactate production accounted for about 35 %of the glucose consumed (Table 1) (8) if the assumptionglucose —>2 lactate is accepted. Thus the fraction of glucose that

was neither glycolyzed nor oxidized was very large. The quantitative relationship between the size of this fraction and the rate oftumor growth was studied.

The growth rate was determined from subcutaneous transplants rather than tissue-isolated preparations because surgicalmanipulation necessary for measurement might have impairedgrowth. About 300 ing of tumor fragments were implanted bytrocar in each animal. Forty rats were used for each tumor,subdivided into 4 groups of 10 animals. Each group receivedfragments from a single donor tumor and was sacrificed half atthe first and half at the second removal time (see below). Thedonor tumors were grown intraperitoneally because in thismanner small nodules with a minimum of necrosis were obtained.The subcutaneous transplants were removed, res]>ectively, atthe 6th and 8th day for Walker carcinomas, 15th and 20th dayfor 5123 hepatoma, and 10th and 15th day for Fibrosarcoma4956, since preliminary ex]>eriments had shown that during theseperiods growth was fastest. From the semilogarithmic plot ofweight increment over the chosen interval, the number of hoursneeded by the tumor to double its weight was calculated (weight-doubling time). It should be emphasized that the growth rate oftumors older than the age group selected is slower, probablybecause of necrosis. The values of oxygen or glucose utilizationand lactate production were measured in tumors within the sameage group as those utilized for the growth rate determination.Previous work had shown (7) that the growth rates of subcutaneous and ovarian transplants of the same tumor were equal.

Under our experimental conditions tumor growth was relatedneither to consumption of oxygen or glucose, nor to production oflactate (Table 1).

On the assumption that all carbon dioxide eliminated by thetumor was derived from glucose, the weight-doubling time wascompared with the amount of glucose which did not leave the

TABLE 2Tumor Growth and Glucose Consumption

Glucose removed frombloodGlucose

eliminated aslactateCQj

eliminated in efferent blood (all COaeliminated was considered as derived fromglucose)Glucose

which did not leave the tumor aslactateorCOa and should have contributed totumorgrowth

[1 - (2 +3)]Weight-doublingtimefoundWeight-doubling

time expected from glucose retention11Unitsgm/24

hr/100gmgm/24

hr/100gmgm/24

hr/100gmgm/24

hr/100gmhourshoursWalker

carcinoma23.3(27.6-18.6)"8.2(9.5-6.9)2.2(2.6-1.7)12.9(15.5-12.0)33(26-40)335123hepatoma13.4(21.6-5.6)4.3(5.6-2.6)3.6(4.7-2.2)5.5(11.3-0.8)196(176-210)874956fibrosarcoma10.8(17.3-4.3)3.7(6.0-1.3)0.4(2.7-0.3)6.7(10.6-2.7)61(56-65)57

" In parentheses, 95% confidence limits. Number of determinations concerning the values given in Lines 1-4 = 65 for Walker carcinoma, 23 for 5123hepatoma, and 19for 4956fibrosarcoma, and in Line 5 = 40 animals each tumor.

*The doubling time expected from simple accumulation of glucose into the tumor was calculated as follows: 100 gm of wet Walkercarcinoma = 17.8 gm dry weight. In 24 hours 12.9gm glucose were added to the tumor and in 33 hours 17.8 gm of added glucose shoulddouble the tumor weight.

1042 CANCER RESEARCH VOL. 27

Research. on November 25, 2020. © 1967 American Association for Cancercancerres.aacrjournals.org Downloaded from

6 r

cco

E01o 4o

EE

cTÃœJtvl

(MO

SEibi-0.21SEla)-0.30

J I I

O 4 8 12

GLUCOSE UTILIZEDmmoles/hour/IOOgm TUMOR

CHART1. Relation between glucose and oxygen utilization.Each point (•)represents one animal. Glucose utilization waschanged by the increase of glycemia to various levels. Tumorswhich utilized more glucose also utilized more oxygen. Curve forWalker carcinoma; hepatoma and fibrosarcoma behaved in thesame way.

tumor via lactate or carbon dioxide (Table 2). If all this glucosecontributed to growth, one would expect a doubling time forcarcinomas, hei»tomas, and fibrosarcomas, respectively, of 33,87, and 57 hours. The doubling time actually found was 33, 196,and 61 hours (Table 2). The weight increase of carcinomas andfibrosarcomas was roughly equal to the amount of glucose retained while the hepatomas grew less than glucose retentionrequired. When compared with the growth rate, the large retention of glucose suggests that the neoplastia tissue is able to transform and eliminate glucose in some unknown manner. (Preliminary determinations did not show any appreciable difference inthe pyruvate content of afferent and efferent blood of Walkercarcinomas.)

Option between Respiration and Glycolysis

The possibility that in vivo glycolysis increased during oxygenshortage was studied in two types of experiments. In the first,a group of animals bearing tumors were kept under anesthesiafor various periods of time, breathing room air or air enrichedwith 10' ; oxygen. Some of these rats were diabetic; others werehyper- or hypoglycémie.Thus the consumption of glucose andoxygen and the production of lactate and carbon dioxide was

Oxygen and Glucose Consumption by Tumors in Vivo

10 r

ceO

Een

OO

oEE

OO

coO

O 4 8

GLUCOSE CONSUMEDmmoles/hour/IOOgm TUMOR

12

CHART2. Relation between glucose and oxygen consumptionby Walker carcinomas in diabetic rats. Each point (•)representsone animal. The different amount of glucose consumed dependedupon the different degree of hyperglycemia produced by thealloxan treatment. Tumors with large utilization of glucose consumed also more oxygen. Hepatomas and fibrosarcomas showed asimilar relationship.

compared in tumors under a relatively large range of glucose andoxygen availability compatible with the host's survival.

When the tumors consumed small amounts of oxygen, onlysmall amounts of glucose were utilized (Chart 1). Hyperglycémierats, which steadily increased their consumption of glucose (8),also utilized a larger amount of oxygen, and the consumptionwas directly related to the level of glycemia (260-390 mgglucose/100 ml plasma) (Chart 2). The lactate removed by theefferent blood was also directly related to the oxygen utilized bythe tumor (Chart 3). There was no indication that tumors withsmall amounts of oxygen available discharged a larger quantityof lactate in the efferent blood (Chart 4). The carbon dioxideclearing through the efferent blood was found to be directlyrelated to the lactate eliminated (Chart 5) and to the oxygenutilized.

JUNE 1967 1043

Research. on November 25, 2020. © 1967 American Association for Cancercancerres.aacrjournals.org Downloaded from

P. M. Cullino, F. H. Grantham, A. H. Courtney, and I. Losonczy

CCO

14

12

ECP

Oo IO

oEE

Õ--2.24+I.I4XSE!b/'--0.19SE(a)=0.66

0123456789

02UTILIZED,mmoles/hour/IOOgm TUMOR

CHAUT3. Relation between oxygen utilized and lactate eliminated. Each point (0) represents one tumor. The difference in oxygenutilized per unit weight was obtained when the oxygen available was reduced to various degrees in animals anesthetized for differentperiods and breathing air or pure oxygen. Tumors utilizing more oxygen eliminated more lactate. Curve fur Walker carcinoma; hepa-toma and fibrosarcoma behaved in the same way.

In the second type of experiment, oxygen and glucose consumption and lactate and carbon dioxide production were measuredfor the same tumor, first under "standard" conditions and then

after a severe anemia. The reduction of oxygen available to thetumor was obtained by .substituting 5-6 ml of blood with 4-5ml of plasma. The "standard" conditions were: (a) normogly-cemia and air breathing in one group of animals, and (o) hyper-glycemia (270-700 ing glucose/100 ml plasma) and oxygenbreathing in a second group. In the latter group, the plasma whichwas substituted for blood to produce anemia contained an amountof glucose proportionally larger than normal. The exprimen talconditions were thus arranged to measure glucose consumptionand lactate discharge when the oxygen available was sharplyreduced, but the glucose content of plasma was either kept atphysiologic levels or increased to levels which are known (8) toaugment glucose consumption.

The abrupt withdrawal of oxygen in normoglycemic animals(Table 3, upper part) invariably induced a marked increase ofCÛ4production. However, glucose and lactate determinationsdid not offer any clear indication that hypoxemia increasedglucose consumption or lactate elimination. In hyperglycémieanimals (Table 3, lower part), where the amount of glucose consumed was larger than normal, the withdrawal of oxygen produced, most of the times, a sharp decrease of glucose consumptionand lactate elimination. Indeed, in several of these animals theefferent blood of the tumor contained slightly more glucose thanthe afferent blood while the reverse was true for lactate. Again,

there was no indication that hypoxemia increased glucose consumption despite a large availability.

Gljoolysis as a "Detoxicatioii Mechanism"

Graff's concept of enhanced lactate production as a conse

quence of excessive glucose input due to lack of control at thecellular level was experimentally tested under in vivo conditions.It was shown previously (8) that the 3 tumors studied dischargedas lactate in the efferent blood about 35% of the glucose removedfrom the afferent blood and that insulin had no appreciable effecton their glucose uptake. In this expriment 2 sets of determinations were performed on the same tumor, the first when a severeinsulin-induced hypoglycemia was present in the host and thesecond, immediately after normo- or hyperglycemia was restored.The lactate eliminated per mole of glucose consumed was compared for each tumor during starvation and subsequentabundance of glucose.

Hypoglycemia was produced by the intravenous injection of20-25 units of insulin. Decrease of glucose concentration in

plasma started within 60 mintues and the first samples weretaken at 90 minutes, at a time when the hyixiglycemia had beensevere for 15-30 minutes. After sampling, 60 or 150 mg of glucosein 1.0 ml of saline were injected i.v. into each rat. Within a fewminutes glycemia was usually at normal levels after the lowerdose and 2- to 3-fold the physiologic level after the high dose. Thesecond sample was obtained 5 to 15 minutes after the onset ofnormo- or hyperglycemia.

1044 CANCER RESEARCH VOL. 27

Research. on November 25, 2020. © 1967 American Association for Cancercancerres.aacrjournals.org Downloaded from

Oxygen and Glucose Consumption by Tumors in Vivo

14ceO

'12

Een

O

—~o

EE

UJ

Lu^ ,

<I

Y--2A2+OA7XSEib)--0.08

SE(a); 0.69

J i i J I I I I i i i i iIO 12 14 16 18 20 22

02 AVAILABLE,mmoles/hour/IOOgm TUMOR

CHART4. Relation between oxygen available and lactate eliminated. Each point (•)represents one tumor. The difference in oxj-genavailable per unit weight was obtained when animals were anesthetized for various periods of time and were breathing air or pure oxygen. Tumors with more oxygen available produced more lactate. Curve for Walker carcinoma; hepatoma and fibrosarcoma behaved inthe same way.

Under .severe glucose starvation, the tumors discharged in theefferent blood a proportionately larger amount of lactate thanduring the subsequent hyperglycemia (Table 4). Moreover,during the large consumption of glucose which followed theperiod of glucose starvation, the efferent blood of several tumorscontained less lactate than the afferent blood. The tumors werein effect using lactate from the arterial blood, not producing it,despite the large quantity of glucose which they were consuming.The ratio of lactate eliminated to glucose utilized, under normalconditions, was found to be 0.7 (8). However, during glucosestarvation the values were constantly higher, and during thesubsequent hyperglycemia they were below 0.7 and some of themeven below zero (Chart 6).

This finding contradicts the hypothesis of glycolysis as adetoxication mechanism, which predicts a lower lactate elimination during hypoglycemia. It is possible that under a severelyimbalauced glucose supply, the effect of a detoxication mechanism might not be detected, or that insulin per se alters thelactate-to-glucose ratio. In our experiments, however, there wasno indication of this effect.

In Vivo Induction of High Lactate Levels in the InterstitialFluid

The accumulation of lactate in the interstitial fluid was foundto be a consistent result whenever neoplastia cells were transplanted subcutaneously (6). liefere grafting, the subcutaneousinterstitial fluid contained glucose and lactate at a concentration

only slightly less than that of plasma. As soon as the neoplasticcells were injected, free glucose disappeared and lactate concentration started to rise. Experiments were designated to seewhether the same sequence of events could be obtained in theabsence of neoplastic cells. The following findings show that thiscan be accomplished.

Two microi>ore chambers (6) were placed into the pouch of thesubcutaneous tissue of the same rat, one in the interscapular andthe other in the sacral region. One chamber had walls formed byT\V millipore filters (0.45 n pore diameter) and the other chamber had walls of SS Very Dense filters (0.1 n pore diameter). Bothfilters were able to exclude the subcutaneous tissue cells yet letthe interstitial fluid surrounding these cells fill the chamber. Thisfluid was analyzed for glucose and lactate at various times afterthe implantation.

For the first 4 days the glucose and lactate content of bothchambers did not change from the original level and, as comparedwith plasma, the glucose concentration was 20-30% lower whilethe lactate level was about equal in most cases. After 15 days,however, the glucose concentration in all chambers was lowerthan at the 4th day and the lactate concentration was several foldhigher than in plasma. In the chambers with T\V millipore filter,the difference was generally more pronounced than in chamberswith SS Very Dense filters; in fact, in some of the milliporechambers, free glucose was undetected while the lactate level wasequal to that found in malignant tumors (6) (Table 5). At thattime the connective tissue surrounding the two chambers showedno morphologic indication of neoplastic transformation, and,

JUNE 1967 1045

Research. on November 25, 2020. © 1967 American Association for Cancercancerres.aacrjournals.org Downloaded from

P. M. Cullino, F. H. Grantham, A. H. Courtney, and I. Losonczy

7 Õ--0.54+0.4IXSE(b)--0.09

CEO

e Jen

OO

5 4o

oEE

(SJO

SE(a)--0.60

8 IO 12 14

LACTATE ELIMINATED.mmoles/hour/lOOgmTUMOR

CHART5. Relation between láclateand carbon dioxide elimination. Each point (•)represents one tumor. The tumor-bearing animalswere anesthetized for different periods and were breathing either air or pure oxygen. Each tumor used oxygen in proportion to theamount available and as the lac tat«eliminated augmented the total carbon dioxide also increased. The same type of curve was observedwhen oxygen utilized was substituted for lactate eliminated. Curve for Walker carcinoma; hepatoma and fibrosarcoma behaved in thesame way.

TABLE 3Effect of Acute Anemia on Glucose Consumption and Láclateand CO«Production of Walker

Carcinomas (inmoles/hr/100 gm wet tumor)The first 6 animals were normoglycemic, the last 6 were hyperglycémie(270-700mg/100 ml plasma).

Before and 10-15 min after, 5-6 ml of blood were substituted with 4-5 ml of plasma. The negative sign(—)indicates the opposite of consumption or production depending on the column. The sign = forRats 10, 11, 12 indicates that the amount of O»consumed was indistinguishable from the amount available because the values were so low as to defie accurate determination. Hyperglycemia was produced by4-5 subcutaneous injections of 200mg dextrose each at about a 1-hour interval.

RatNo.123456789101112OiavailableBefore12.67.16.715.07.226.213.312.117.96.823.044.8After3.20.50.54.02.16.53.82.25.10.21.01.5OjconsumedBefore3.52.62.83.05.38.35.23.44.59.521.8After1.00.20.31.02.20.60.71.4^0.2Sl.O££1.5GlucoseconsumedBefore3.91.65.05.97.32.69.010.714.122.834.769.2After3.71.54.82.19.64.0-1.62.21.80.8-1.8-4.2LactateproducedBefore3.44.58.83.89.56.95.84.01.86.826.55.6After1.05.99.61.43.05.5-8.09.0-8.9-1.00-12.0COiproducedBefore4.92.44.32.50.95.713.09.521.010.331.134.4After9.911.012.34.64.310.010.110.610.71.75.63.5

104« CANCER RESEARCH VOL. 27

Research. on November 25, 2020. © 1967 American Association for Cancercancerres.aacrjournals.org Downloaded from

Oxygen and Glucose Consumption by Tumors in Vivo

TABLE 4Comparison between Láclate Eliminated and Glucose Utilized by the Same Tumor during a

Severe Hypoglycemia Followed by HyperglycemiaHypoglycemia was produced by the i.V. injection of 20 units of insulin in animals bearing fibrosar-

comas or Walker carcinomas and of 25 units in the animals with hepatomas. Hyperglycemia was produced byi.v. injection of 60-150 mg dextrose to each animal. Two samples from each tumor: (a) 90 minutes

after insulin; (6) 5 to 15 minutes after i.v. injection of dextrose. The ratio of lactate eliminated to glucoseutilized was lower during hyperglycemia with the exception only of Rat 4 of the Walker carcinomagroup and Rat 1 of the hepatoma group. Negative sign (—) = subtraction of lactate from blood insteadof elimination.

Tumor and ratnumberFibrosarcoma123456789Walker

carcinoma1234567Hepatoma1234567HypoglycemiaArterial

plasmaglucose

(mg/100ml)113227233567452063352346625585214862123635Glucose

utilized(mmole/

hr/100 gm)(A)'0.030.0200.060.050.730.170.340.240.590.420.401.650.370.230.7400.140.140.22000.75Lactate

eliminated(mmole/

hr/100 gm)(B)0.050.0600.110.040.640.100.360.251.302.921.220.281.391.960.650.190.380.210.410.5900.70(B)(A)1.63.001.80.800.880.591.061.042.26.93.00.23.78.50.20.192.711.501.860.5900.93HyperglycemiaArterial

plasmaglucose

(mg/100ml)143160257272557613748860173777791494224535477084758133344489500533Glucose

utilized(mmole/

hr/100 gm)(C)0.060.560.300.340.250.891.344.307.036.056.432.444.242.903.8920.90.310.600.672.164.8231.101.70Lactateeliminated

(mmole/hr/100 gm)

(D)-0.01-0.05-0.220.180.110.150.340.330.182.0803.643.042.402.831.800.770.650.280.290.04-0.570.53(D)(C)-0.17-0.10-0.730.540.440.170.250.080.030.3401.50.720.830.730.092.481.00.410.130.008-0.020.31

indeed, a very "mild reaction" to the presence of the chamber

normally observed (Fig. 1).

On the assumption that an increase of glycolysis might indicate

a rapid malignant transformation of the cells, both chambers

were tested for their ability to produce tumors. Two groups of

rats were prepared and observed for 20 months. Each animal of

the first group received one TW filter chamber in a ixmch of thesubcutaneous tissue of the lumbar region and each animal of thesecond group received an SS filter chamber in the same area.

Both chambers ultimately produced sarcomas in about thesame number (Table 6). Morphologically, the predominant celltype was a large fibroblast and the cells were mostly collected inbundles of irregular orientation (Fig. 2). The intercellular substance was abundant with occasional deposition of calcium salts.There was no morphologic difference between the fibrosarcomasproduced with either filter. All tumors started around the chamberand grew to surround it. The host was killed within one month

after the tumor was first observed and all the sarcomas grew

after transplantation. Fibrosarcomas arose earlier in the SS filtergroup and 5 liad already dcvelo|x>d when the second sarcoma

appeared in the TW millipore group (Table 5). The fibrosarcomas

arose about one year after the increased glycolytic rates werefirst observed at the site of the chamber implant. Moreover, not

all animals developed tumors within 20 months of observation

even though all had shown an increase of lactate in the tissuesaround the chambers.

DISCUSSION

The comparison between glucose and oxygen consumed in vivoby transplanted tumors has confirmed what had already beenshown in vitro, i.e., a large consumption of glucose contrasted toa small oxygen utilization. In the past glycolysis was usually themain objective of studies using in vitro technics to compare

JUNE 1967 1047

Research. on November 25, 2020. © 1967 American Association for Cancercancerres.aacrjournals.org Downloaded from

P. M. Gullino, F. H. Grantham, A. H. Courtney, and I. Losonczy

3

QLU

LU

LU

O

LUN

Lu

OO

O

EO

•HypoglycemìaA Hyperglycémie

98

vx-A°

A/.

A-?

I 2345678GLUCOSE UPTAKE,mmoles/hour/IOOgm TUMOR

CHART6. Relation between glucose uptake and ratio láclateeliminated/glucose utilized (Fibrosarcoma 4956). Each tumor is indicated by a number and represented by 2 values: (a) during hypoglycemia (•)and (6) 5-15 minutes after dextrose was given intravenously (A). During glucose starvation the values of the ratio were all higher than after hyperglycemia. Since two moles of láclate canbe produced by each mole of glucose the values found for No. 2 during glucose starvalion suggest lhal glucose was utilized from theneoplastic lissue. Note thai Tumors 1, 2, and 3 utilized instead of producing lactate during hyperglycemia which followed glucosestarvation. (Tumor 3 showed no appreciable utilization of glucose or elimination of láclale during glucose starvation, therefore novalue appears on the graph.)

TABLE 5Giocose and Láclale Present in the Interstitial Fluid Collected with Micropore Chambers

from the Subcutaneous Tissue of \ormal Rats" (»ig/100ml)

Three samples were taken from each animal: plasma, interstitial fluid from the chamber formed byTW millipore filters (TW-MF), and interstitial fluid from the chamber formed by SS Very Dense filters(SS-VD). At 15days from the implant, glucose content was lower and lactate conteníwas higher lhan atthe 4th day in both chambers. The differences were often more evident in TW millipore filter chamberswhere the lactate level was in some specimens similar to that found in tumors.

RatNo.1234564

daysGlucosePlasma144167167181155160TW-MF144141109115140140SS-VD124113127177151152LactatePlasma463550413140TW-MF624852486048SS-VD60605548465015daysGlucosePlasma144188174170160158TW-MF38075008SS-VD506179919690LactatePlasma342941284755TW-MF841236454114134SS-VD896756376473

1Sprague-Dawley cf and 9 .

1048 CANCER RESEARCH VOL. 27

Research. on November 25, 2020. © 1967 American Association for Cancercancerres.aacrjournals.org Downloaded from

Oxygen and Glucose Consumption by Tumors in Vivo

TABLE 6Production of Fibrosarcomas by T\\' Millipore and

SS Very Dense Filters"

Each animal received one chamber in a pouch of the subcutaneous tissue of the lumbar region. The skin was sutured withcotton and the stitches were removed during the second week.

FilterSS

VeryDenseTW

milliporeNo.

ofrats at

start127No.

of ratsat firsttumor95Total

tumorsproduced74Time

ofobservation

(mo.)2020Time

ofappearance

(mo.)1011131415161812151718

°Sprague-Dawleyrats rf1and 9 .

and oxygen consumption of tumors. In vivo,it was foundthat glycolysis accounted for about 35% of the glucose utilized;however, a much larger fraction was not eliminated by the tumorseither as lactate or as carbon dioxide. The growth rate of manyof these tumors could not account for the large uptake of glucose.An alternative is that the neopla-stic tissue eliminated ]>art ofthe glucose consumed in an unknown manner.

The in vivo respiration of tumor tissue was small when compared with the glucose consumption. Quantitatively, however,the absolute Q02 values were not excessively low. For instance,the oxygen consumption of rat heart muscle at rest is reported(20) to be Qo2 = 3.0 to 8.0, and the Q02 of the tumors studiedranged from 1.0 to 4.2 (9). Our in vivo observations suggest thesame conclusion provided by in vitro studies: The respiration ofthe neoplastic tissues was much smaller than that of some tissueswith high oxidative rates, like the liver (1), however not so lowas to be considered deficient in comparison with other normaltissues with lower oxidative rates (2, 16, 19).

The relationship between respiration and growth is poorlyunderstood. The seemingly logical e.\i>ectation that faster growthrequires larger oxygen consumption is probably not true in thesesimple terms. This is illustrated by the fact that in vivo thefibrosarcomas used about J of the Oa of hepatomas but grewtwice as fast; both tumors converted about 35% of glucose intolactate and fibrosarcomas consumed about 20% less glucose thanhepatomas. [The transformation of glucose into glycogen isnegligible in 5123 hepatoma growing in normoglycemic host(18).]

The in vivo comparison between respiration and glycolyticcaj>acity is in accord with the in vitro results. The amount ofglucose glycolyzed was higher than the quantity of glucose thatthe tumors could oxidize. With both glycolytic and oxidativeenergy pathways available, the neoplastic cell has often beenconsidered to be in an advantageous position, vis àvis the non-neoplastic cell. This assumption is relevant to the ini|X)rtant

questions of survival capacity and susceptibility of tumor cellsto chemotherapeutic attack through their energy metabolism.It was therefore crucial to ascertain whether in vivo neoplasticcell ]>opulations could shift from respiration to fermentation orvice versa as a function of the oxygen supply.

We were unable to find any indication that the neoplastiatissues could supplement an in vivo oxygen deficiency by anincrease in glycolysis. Tumors with low oxygen consumptionutilized proportionally smaller amounts of glucose. When glycolysis was measured in the same tumor before and after anacute shortage of oxygen, there was no clear indication thatglucose utilization or lactate production was enhanced duringhypoxemia. Actually, any increase of glucose consumption required an increase in oxygen utilization. Moreover, when anabrupt deprivation of oxygen was inflicted to the tumor duringactive glucose consumption, both utilization of glucose andelimination of lactate ceased.

Despite the fact that glycolysis is not an exclusive pro|>erty ofthe neoplastic tissues, it is, however, a remarkable metabolicfeature of most tumors and its significance has traditionally beenbound to the nature of neoplasia. The hy]X)thesis that neoplastiacells have an excessive influx of glucose and that lactate elimination is a "detoxication" mechanism (4) could not be confirmed

in vivo. Neoplastic cells grew in a fluid practically devoid ofglucose and the vascular wall maintained a gradient betweenplasma and interstitial fluid (8). During normoglycemia, theamount of glucose reaching the cells was less than they were ableto handle and, during moderate hyperglyeemia, the glucoseconsumption increased while the lactate elimination remainedaround 35% of the glucose intake.

The in vivo observations on lactate elimination during severeglucose starvation suggest that glycolysis should not be studiedsolely from the point of view of energy metabolism, as it isusually done. When hypoglyeemia imposed a severe restriction onglucose consumption, the tumors eliminated practically all theutilized glucose as lactate. However, when a large supply of glucose followed glucose starvation, not only did lactate eliminationcease despite a huge amount of glucose utilized, but even thelactate of the afferent blood was consumed. The oxygen supplywas evidently adequate to co]>e with the increased metabolicneeds, and energy requirement was probably not the majorfactor involved in lactate elimination by the neoplastic cells.

A large production of lactate was obtained in the subcutaneoustissues long before any neoplastic transformation was observed.Actually, the lactate content of the subcutaneous interstitial fluidincreased in all animals bearing microix>re chambers, despite thefact that not all of them develojjed a sarcoma during 20 monthsof observation. The relationship between neoplastic transformation and excessive production of lactate in vivois difficult to assessin the light of the results reported here. In beef erythrocytes, asan example, the glycolyzing enzymes seem to be concentrated onthe cell membrane (5). A change in the ectobiologic charactersof the plasma membrane (10) could result, for instance, in anexcessive production of lactate by these cells. The possibilitythat neoplastic transformation is only one of several conditionsable to produce an "unspeeific" increase of lactate production

should be kept in mind.

JUNE 1967 1049

Research. on November 25, 2020. © 1967 American Association for Cancercancerres.aacrjournals.org Downloaded from

P. M. Gullino, F. H. Grantham, A. H. Courtney, and I. Losonczy

ACKNOWLEDGMENTS

We gratefully acknowledge the suggestions of Ur. Dean Burkand Dr. Mark Woods and the help of Dr. Edmund A. Gehan, National Cancer Institute, Biometry Branch, and Mr. Donald Hill.

REFEKKÃŒVCES

1. Brauer, R. W., Leong, G. F., and Holloway, li. J. OxygenConsumption by the Isolated Rat Liver. San Francisco, Calif. :U. S. Naval Radiological Defense Laboratory, TH-573. 1963.

2. Burk, 1)., and Schade, A. L. On Respiratory Impairment inCancer Cells. Science, Ig4: 270-271, 1956.

3. Cori, C. F., and Cori, G. T. The Carbohydrate Metabolism ofTumors. II. Changes in the Sugar, Lactic Acid, and COz-combining Power of Blood Passing Through a Tumor. J.Biol. Chem., 65: 397-405, 1925.

4. Graff, S., Moser, H., Kastner, D., Graff, A. M., and Tannen-baum, M. The Significance of Glycolysis. J. Nati. CancerInst., 34: 511-519, 1965.

5. Green, D. E., Murer, E., Hultin, H. O., Richardson, S. H.,Salmon, B., Brieley, G. P., and Baum, H. Association ofIntegrated Metabolie Pathways with Membranes. I. Glyco-lytic Enzymes of the Red Blood Corpuscle and Yeast. Arch.Biochem. Biophys., 11$: 635-647, 1965.

6. Gullino, P. M., Clark, S. H., and Grantham, F. H. The Interstitial Fluid of Solid Tumors. Cancer Res., Õ4:780-798, 1964.

7. Gullino, P. M., and Grantham, F. H. Studies on the Exchangeof Fluids Between Host and Tumor. I. A Method for GrowingTissue-isolated Tumors in Laboratory Animals. J. Nati.Cancer Inst., %T:679-693, 1961.

8. Gullino, P. M., Grantham, F. IL, and Courtney, A. H. GlucoseConsumption by Transplanted Tumors in Vivo. Cancer Res.,27: 1031-1040, 1967.

9. Gullino, P. M., Grantham, F. H., and Courtney, A. H. Utilization of Oxygen by Transplanted Tumors in Vivo. CancerRes., 27: 1020-1030, 1967.

10. Kalckar, H. M. Galactose Metabolism and Cell "Sociology."

Science, 150: 305-313, 1965.

11. Kety, S. S., and Schmidt, G. F. The Effects of Active and Passive Hyperventilation on Cerebral Blood Flow, CerebralOxygen Consumption, Cardiac Output and Blood Pressure ofNormal Young Men. J. Clin. Invest., go: 107-119, 1945.

12. Norman, T. I)., and Smith, A. B. The Blood Lactic Acid ofTumor-bearing and Tumor-free Mice. Cancer Res., 16: 1027-1031, 1956.

13. Okamoto, Y. ÃœberAnaerobiose von Tumorgewebe. Biochem.Z., 160: 52-65, 1925.

14. Scheinberg, P., and Stead, E. A., Jr. The Cerebral Blood Flowin Male Subjects as Measured by the Nitrous Oxide Technique. Normal Values for Blood Flow, Oxygen Utilization,Glucose Utilization and Peripheral Resistance with Observations on the Effect of Tilting and Anxiety. J. Clin. Invest.,28: 1163-1171, 1949.

15. Voegtlin, C., Kahler, II., and Fitch, R. H. Die Bestimmung derWasserstoffionen Konzentration der Gewebe Bei LebendenTieren Mit Hilfe der Kapillar-glas-electrode. Handbuch derBiologischen Arbeitsmethoden. Abderhalden, Abt. V., Teil10: a. 667-684, 1935.

lü.Warburg, O. On Respiratory Impairment in Cancer Cells.Science, 1Õ4:269-270, 1956.

17. Warburg, O., Wind, F., and Negelein, E. On the Metabolismof Tumors in the Body. In: The Metabolism of Tumors, Ch.XV, p. 254. London: Constable and Co., 1930.

18. Weber, G., Morris, H. P., Love, W. C., and Ashmore, J. Comparative Biochemistry of Hepatomas II. Isotope Studies ofCarbohydrate Metabolism in Morris Heputoma 5123. CancerRes., a/: 1406-1411, 1961.

19. Weinhouse, S. On Respiratory Impairment in Cancer Cells.Science, 124: 2G7-268, 1956.

20. Whalen, W. J. Energetics of Isolated Muscle. FederationProc., 21: 994-998, 1962.

1050 CANCER RESEARCH VOL. 27

Research. on November 25, 2020. © 1967 American Association for Cancercancerres.aacrjournals.org Downloaded from

Oxygen and Glucose Consumption by Tumors in Vivo

F •

SIF

1

Flo. 1. Cross-section through a Doioroporechamber sampling subcutaneous interstitial fluid. C, blood capillaries; SC, subcutaneousconnective tissue; F, filter (In this case SS Very Dense was used. However, no appreciable difference was observed when TW milliporefilters formed the walls of the chamber.); SIF, subcutaneous interstitial fluid (protein precipitates). The morphology of the cells wasnot appreciably different from that of cells of normal subcutaneous connective tissue without chamber despite the fact that the fluidcollected contained about twice the amount of lactate than plasma. The number of cells and of dilated capillaries was slightly higherthan normal around the filters. II & K, X 150.

JUNE 1967IO.')1

Research. on November 25, 2020. © 1967 American Association for Cancercancerres.aacrjournals.org Downloaded from

P. M. Gullino, F. H. Grantham, A. H. Courtney, and I. Losonczy

vr H'vFIG. 2. Histologie aspect of a sarcoma produced by a TW Millipore chamber. Fibroblast-like cells grouped in bundles with different

orientation. Nuclei highly irregular, hyperchromatic, and with some mitotic figures. II & E, X 190.

1052 CANCER RESEARCH VOL. 27

Research. on November 25, 2020. © 1967 American Association for Cancercancerres.aacrjournals.org Downloaded from

1967;27:1041-1052. Cancer Res   Pietro M. Gullino, Flora H. Grantham, Anita H. Courtney, et al.  

in VivoTransplanted Tumors Relationship between Oxygen and Glucose Consumption by

  Updated version

  http://cancerres.aacrjournals.org/content/27/6_Part_1/1041

Access the most recent version of this article at:

   

   

   

  E-mail alerts related to this article or journal.Sign up to receive free email-alerts

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

  Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/27/6_Part_1/1041To request permission to re-use all or part of this article, use this link

Research. on November 25, 2020. © 1967 American Association for Cancercancerres.aacrjournals.org Downloaded from