[CANCER
RESEARCH
Relationship
Transplanted
27 Part 1, 1041-1052, June 1967
between
Tumors
Oxygen and Glucose
in Vivo
Consumption
by
PIETRO M. CULLINO, FLORA H. GRANTHAM, ANITA H. COURTNEY, AND ILONA LOSONCZY
Laboralorn of Biochemistry, National
Cancer Institute,
KIH,
USPHS,
SUMMARY
The in vivo consumption of oxygen and glucose was studied in
relation to growth in Walker Carcinoma 256, Hepatoma 5123,
and Fibrosarcoma 4956 transplanted in rats. Glucose not elim
inated as lactate or carbon dioxide in the efferent blood was pre
sumed to be retained by the tumor. The retention was so high
that growth alone could not account for it. As an alternative, the
elimination of glucose by the tumors in some unknown manner
is suggested. The weight-doubling times of the tumors were
independent of oxygen consumption and lactate production. The
tumors needed oxygen to survive; there was no indication, how
ever, that in vivo tumor metabolism shifted from respiration to
glycolysis when the supply of oxygen was deficient. The opposite
was found to be true: glucose consumption and lactate elimina
tion were in direct proportion to the oxygen utilized and a lack
of oxygen blocked both of them. The fraction of glucose trans
formed into lactate was maximal during insulin-induced hypoglycemia. These glucose-starved tumors did not produce lactate
during glucose refeeding despite a large glucose utilization.
Neither lack of oxygen nor large glucose consumption appeared
to be the dominant causes of in vivolactate production by tumors.
Experimental increases in the lactate content of subcutaneous
tissue could be obtained in the absence of any tumor. The
possibility that glycolysis is related to changes of cellular com
ponents not necessarily involved in the neoplastic process is
suggested.
INTRODUCTION
The measurement of the in vivo consumption of oxygen and
glucose by transplanted rat tumors was reported in 2 accompany
ing papers (8, 9). The ex(>eriments described here concern the
interdependence of respiration and glycolysis in these tumors.
In vivo glycolysis was discovered when the lactate content of
blood draining an organ bearing a tumor was found to be higher
than in the absence of the tumor (3, 17). The finding was con
firmed by the observation of increased blood lactate (12) and
decreased pH (15) following administration of glucose. Glycolysis
was considered essential for the existence of the tumor in vivo
when sections of Jensen sarcoma deprived of oxygen for 72 hours
but supplied with large amounts of glucose were still able to
develop into a tumor upon transplantation. The sarcoma did
not grow when glucose was not available (13). The finding was
interpreted to demonstrate that tumor cells could derive from
Received September 30, 1900; accepted January 11, 1907.
JUNE 1967
HEW, Hethesda, Maryland 20014
glycolysis all the metabolic energy required to survive for a
considerable period of time.
Cells grown in vivo are usually cultivated in media rich in
glucose, and large amounts of lactate are produced. However,
Graff et al. (4) observed that when two cell lines were grown in
a medium with a constant concentration of glucose of only 5 mg
per 100 ml, both lines grew well and would consume rather than
produce lactate. They considered the high glycolysis found in
vitro using media rich in glucose as a "detoxicating mechanism"
substituting for the hormonal barrier which, in the in vivo situa
tion, limited the flux of glucose across the cell membranes. Indeed
it was later found in vivo that the neoplastic cells grew in a
medium containing only a few mg percent of free glucose (16),
and the vascular wall served an important function in regulating
glucose transfer from the vascular into the interstitial compart
ment (8).
In the work presented here, an assessment was made of the
relative proportions of oxygen and glucose available and utilized
by tumors in vivo. Expriment« were performed in vivo to test
whether: (a) a relationship between tumor growth and glucose
or oxygen consumption could be found, (¿>)
the availability of
oxygen or glucose could switch tumor metabolism from respira
tion to glycolysis or vice versa, (c) the loss of lactate by the tumor
through the efferent blood could be interpreted in vivoby Graff's
hypothesis of a "detoxication mechanism," and (d) it was
possible to increase the concentration of lactate in the interstitial
fluid of the subcutaneous tissues prior to neoplastic transforma
tion.
MATERIALS AND METHODS
The procedures followed in the experiments reported here were
described in 2 accompanying papers (8, 9). Walker Carcinoma
256, transplanted in Sprague-Dawley rats and Hepatoma 5123
and Fibrosarcoma 4956 transplanted in Buffalo/N rats were used.
Each tumor was grown as a tissue-isolated preparation with one
afferent artery and one efferent vein (7). The blood flow and the
arterial-venous difference (A-V) in oxygen, glucose, lactate, and
carbon dioxide were measured and from these values the con
sumption of glucose and oxygen and the production of lactate
and carbon dioxide were evaluated in vivo as previously described
(8, 9). The detailed treatment of the tumor-bearing host (anes
thesia, hyper- and hypoglycemia, diabetes, anemia, etc.) was also
previously described (8, 9).
RESULTS
Tumor Growth
sumption
as Related
to Oxygon
and
Glucose
Con
1041
Downloaded from cancerres.aacrjournals.org on June 14, 2017. © 1967 American Association for Cancer Research.
[rP. M. Gullino, F. H. Grantham, A. H. Courtney, and I. Losonczy
Table 1 .summarizes, for convenient reference, the consumption
of oxygen and glucose as rejjorted in the two previous pa|>ers
(8, 9). A striking feature common to the three tumors studied was
the large in vivo glucose consumption as compared to the rela
tively small oxygen utilization. The extent of this diffeience
could be characterized as follows: If the fraction of glucose
eliminated as lactate in the efferent blood was subtracted from
the total glucose consumed, the quantity of glucose available to
the tumor for oxidation was obtained. Of this amount of glucose,
only 13%, 46%, and 8%, res|)ectively, could have been oxidized
by the oxygen consumed by Walker carcinomas, 5123 hepatomas,
and 4956 fibrosarcomas in vivo. The brain, which is the normal
tissue closest to the tumors as far as glucose consumption is
concerned (1.3 to 3.2 mmoles/hr/100 gm), normally utilizes an
amount of oxygen which could oxidize about 85% of the glucose
consumed (11, 14).
In the tumors, lactate production accounted for about 35 %
of the glucose consumed (Table 1) (8) if the assumption
glucose —>
2 lactate is accepted. Thus the fraction of glucose that
was neither glycolyzed nor oxidized was very large. The quantita
tive relationship between the size of this fraction and the rate of
tumor growth was studied.
The growth rate was determined from subcutaneous trans
plants rather than tissue-isolated preparations because surgical
manipulation necessary for measurement might have impaired
growth. About 300 ing of tumor fragments were implanted by
trocar in each animal. Forty rats were used for each tumor,
subdivided into 4 groups of 10 animals. Each group received
fragments from a single donor tumor and was sacrificed half at
the first and half at the second removal time (see below). The
donor tumors were grown intraperitoneally because in this
manner small nodules with a minimum of necrosis were obtained.
The subcutaneous transplants were removed, res]>ectively, at
TABLE 1
the 6th and 8th day for Walker carcinomas, 15th and 20th day
for 5123 hepatoma, and 10th and 15th day for Fibrosarcoma
Tumor Growth as Compared with in Vivo Consumption of
Oxygen or Glucose and Production of Lactate
4956, since preliminary ex]>eriments had shown that during these
periods growth was fastest. From the semilogarithmic plot of
weight increment over the chosen interval, the number of hours
TumorsWalker
needed by the tumor to double its weight was calculated (weighttime
con(hours)"3319661OxygenTotal
sup
con
elimin glucose
sup
doubling
time). It should be emphasized that the growth rate of
con
sumed62.84.20.8As%
ated1'3.82.01.7As%
plied455746GlucoseTotal
sumed5.43.12.5As%
plied282332LactateTotal
sumed353234
tumors older than the age group selected is slower, probably
because of necrosis. The values of oxygen or glucose utilization
Carci
and lactate production were measured in tumors within the same
256Hepatoma
noma
5123Fibrosarcoma4956Weight-doublins
age group as those utilized for the growth rate determination.
Previous work had shown (7) that the growth rates of subcutane
ous and ovarian transplants of the same tumor were equal.
Under our experimental conditions tumor growth was related
" Calculated from the semilog. Plot of time: dry weight increase
neither to consumption of oxygen or glucose, nor to production of
over the period of maximal growth of subcutaneous transplants.
' mmoles/hr/100 gm wet weight. The relationship did not lactate (Table 1).
On the assumption that all carbon dioxide eliminated by the
change when the experimental data were referred to dry weights.
Water content in gm/100 gm tumor: Walker carcinoma = 82.2, tumor was derived from glucose, the weight-doubling time was
Hepatoma 5123 = 80.0, Fibrosarcoma 4956 = 84.1.
compared with the amount of glucose which did not leave the
TABLE 2
Tumor Growth and Glucose Consumption
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)335123
hepatoma13.4(21.6-5.6)4.3(5.6-2.6)3.6(4.7-2.2)5.5(11.3-0.8)196(1
fibrosarcoma10.8(17.3-4.3)3.7(6
bloodGlucose
Glucose removed from
gmgm/24
hr/100
lactateCQjeliminated as
gmgm/24
hr/100
eliminated in efferent blood (all COa elimi
glucose)Glucose
nated was considered as derived from
lactateor
which did not leave the tumor as
tumorgrowth
COa and should have contributed to
3)]Weight-doubling
[1 - (2 +
foundWeight-doubling
time
gmgm/24
hr/100
gmhourshoursWalker
hr/100
time expected from glucose re
tention11Unitsgm/24
" In parentheses, 95% confidence limits. Number of determinations concerning the values given in Lines 1-4 = 65 for Walker carci
noma, 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 Walker
carcinoma = 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 should
double the tumor weight.
1042
CANCER RESEARCH VOL. 27
Downloaded from cancerres.aacrjournals.org on June 14, 2017. © 1967 American Association for Cancer Research.
Oxygen and Glucose Consumption by Tumors in Vivo
10 r
6r
SEibi-0.21
SEla)-0.30
cc
o
E
01
o
o
4
ce
O
E
en
O
O
E
E
cT
ÃœJ
tvl
O
o
E
E
(M
J
O
4
I
8
I
12
O
O
O
co
GLUCOSE UTILIZED
mmoles/hour/IOOgm
TUMOR
CHART1. Relation between glucose and oxygen utilization.
Each point (•)represents one animal. Glucose utilization was
changed by the increase of glycemia to various levels. Tumors
which utilized more glucose also utilized more oxygen. Curve for
Walker carcinoma; hepatoma and fibrosarcoma behaved in the
same way.
tumor via lactate or carbon dioxide (Table 2). If all this glucose
contributed to growth, one would expect a doubling time for
carcinomas, 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 and
fibrosarcomas was roughly equal to the amount of glucose re
tained while the hepatomas grew less than glucose retention
required. When compared with the growth rate, the large reten
tion of glucose suggests that the neoplastia tissue is able to trans
form and eliminate glucose in some unknown manner. (Prelimi
nary determinations did not show any appreciable difference in
the pyruvate content of afferent and efferent blood of Walker
carcinomas.)
Option
between
Respiration
and Glycolysis
The possibility that in vivo glycolysis increased during oxygen
shortage was studied in two types of experiments. In the first,
a group of animals bearing tumors were kept under anesthesia
for various periods of time, breathing room air or air enriched
with 10' ; oxygen. Some of these rats were diabetic; others were
hyper- or hypoglycémie.Thus the consumption of glucose and
oxygen and the production of lactate and carbon dioxide was
JUNE 1967
O
4
8
12
GLUCOSE CONSUMED
mmoles/hour/IOOgm
TUMOR
CHART2. Relation between glucose and oxygen consumption
by Walker carcinomas in diabetic rats. Each point (•)represents
one animal. The different amount of glucose consumed depended
upon the different degree of hyperglycemia produced by the
alloxan treatment. Tumors with large utilization of glucose con
sumed also more oxygen. Hepatomas and fibrosarcomas showed a
similar relationship.
compared in tumors under a relatively large range of glucose and
oxygen availability compatible with the host's survival.
When the tumors consumed small amounts of oxygen, only
small amounts of glucose were utilized (Chart 1). Hyperglycémie
rats, which steadily increased their consumption of glucose (8),
also utilized a larger amount of oxygen, and the consumption
was directly related to the level of glycemia (260-390 mg
glucose/100 ml plasma) (Chart 2). The lactate removed by the
efferent blood was also directly related to the oxygen utilized by
the tumor (Chart 3). There was no indication that tumors with
small amounts of oxygen available discharged a larger quantity
of lactate in the efferent blood (Chart 4). The carbon dioxide
clearing through the efferent blood was found to be directly
related to the lactate eliminated (Chart 5) and to the oxygen
utilized.
1043
Downloaded from cancerres.aacrjournals.org on June 14, 2017. © 1967 American Association for Cancer Research.
P. M. Cullino, F. H. Grantham, A. H. Courtney, and I. Losonczy
14
CC
O
12
E
CP
O
o
Õ--2.24+I.I4X
SE!b/'--0.19
SE(a)=0.66
IO
o
E
E
0123456789
02UTILIZED,mmoles/hour/IOOgm
TUMOR
CHAUT3. Relation between oxygen utilized and lactate eliminated. Each point (0) represents one tumor. The difference in oxygen
utilized per unit weight was obtained when the oxygen available was reduced to various degrees in animals anesthetized for different
periods and breathing air or pure oxygen. Tumors utilizing more oxygen eliminated more lactate. Curve fur Walker carcinoma; hepatoma and fibrosarcoma behaved in the same way.
In the second type of experiment, oxygen and glucose consump
tion and lactate and carbon dioxide production were measured
for the same tumor, first under "standard" conditions and then
there was no indication that hypoxemia increased glucose con
sumption despite a large availability.
after a severe anemia. The reduction of oxygen available to the
tumor was obtained by .substituting 5-6 ml of blood with 4-5
ml of plasma. The "standard" conditions were: (a) normoglycemia and air breathing in one group of animals, and (o) hyperglycemia (270-700 ing glucose/100 ml plasma) and oxygen
breathing in a second group. In the latter group, the plasma which
was substituted for blood to produce anemia contained an amount
of glucose proportionally larger than normal. The exprimen tal
conditions were thus arranged to measure glucose consumption
and lactate discharge when the oxygen available was sharply
reduced, but the glucose content of plasma was either kept at
physiologic levels or increased to levels which are known (8) to
augment glucose consumption.
The abrupt withdrawal of oxygen in normoglycemic animals
(Table 3, upper part) invariably induced a marked increase of
CÛ4production. However, glucose and lactate determinations
did not offer any clear indication that hypoxemia increased
glucose consumption or lactate elimination. In hyperglycémie
animals (Table 3, lower part), where the amount of glucose con
sumed was larger than normal, the withdrawal of oxygen pro
duced, most of the times, a sharp decrease of glucose consumption
and lactate elimination. Indeed, in several of these animals the
efferent blood of the tumor contained slightly more glucose than
the afferent blood while the reverse was true for lactate. Again,
Gljoolysis
1044
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 the
cellular level was experimentally tested under in vivo conditions.
It was shown previously (8) that the 3 tumors studied discharged
as lactate in the efferent blood about 35% of the glucose removed
from the afferent blood and that insulin had no appreciable effect
on their glucose uptake. In this expriment 2 sets of determina
tions were performed on the same tumor, the first when a severe
insulin-induced hypoglycemia was present in the host and the
second, immediately after normo- or hyperglycemia was restored.
The lactate eliminated per mole of glucose consumed was com
pared for each tumor during starvation and subsequent
abundance of glucose.
Hypoglycemia was produced by the intravenous injection of
20-25 units of insulin. Decrease of glucose concentration in
plasma started within 60 mintues and the first samples were
taken at 90 minutes, at a time when the hyixiglycemia had been
severe for 15-30 minutes. After sampling, 60 or 150 mg of glucose
in 1.0 ml of saline were injected i.v. into each rat. Within a few
minutes glycemia was usually at normal levels after the lower
dose and 2- to 3-fold the physiologic level after the high dose. The
second sample was obtained 5 to 15 minutes after the onset of
normo- or hyperglycemia.
CANCER RESEARCH VOL. 27
Downloaded from cancerres.aacrjournals.org on June 14, 2017. © 1967 American Association for Cancer Research.
Oxygen and Glucose Consumption by Tumors in Vivo
ce
O
14
Y--2A2+OA7X
SEib)--0.08
'12
E
en
O
SE(a); 0.69
—
~o
E
E
UJ
Lu
^
,
<I
J
i
i
J
IO
12
I I
14
I
I
16
i i i
18
20
i
i
22
02 AVAILABLE,mmoles/hour/IOOgm TUMOR
CHART4. Relation between oxygen available and lactate eliminated. Each point (•)represents one tumor. The difference in oxj-gen
available per unit weight was obtained when animals were anesthetized for various periods of time and were breathing air or pure oxy
gen. Tumors with more oxygen available produced more lactate. Curve for Walker carcinoma; hepatoma and fibrosarcoma behaved in
the same way.
Under .severe glucose starvation, the tumors discharged in the
efferent blood a proportionately larger amount of lactate than
during the subsequent hyperglycemia (Table 4). Moreover,
during the large consumption of glucose which followed the
period of glucose starvation, the efferent blood of several tumors
contained less lactate than the afferent blood. The tumors were
in 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 normal
conditions, was found to be 0.7 (8). However, during glucose
starvation the values were constantly higher, and during the
subsequent hyperglycemia they were below 0.7 and some of them
even below zero (Chart 6).
This finding contradicts the hypothesis of glycolysis as a
detoxication mechanism, which predicts a lower lactate elimina
tion during hypoglycemia. It is possible that under a severely
imbalauced glucose supply, the effect of a detoxication mech
anism might not be detected, or that insulin per se alters the
lactate-to-glucose ratio. In our experiments, however, there was
no indication of this effect.
In Vivo Induction of High Lactate Levels in the Interstitial
Fluid
The accumulation of lactate in the interstitial fluid was found
to be a consistent result whenever neoplastia cells were trans
planted subcutaneously (6). liefere grafting, the subcutaneous
interstitial fluid contained glucose and lactate at a concentration
JUNE 1967
only slightly less than that of plasma. As soon as the neoplastic
cells were injected, free glucose disappeared and lactate concen
tration started to rise. Experiments were designated to see
whether the same sequence of events could be obtained in the
absence of neoplastic cells. The following findings show that this
can be accomplished.
Two microi>ore chambers (6) were placed into the pouch of the
subcutaneous tissue of the same rat, one in the interscapular and
the other in the sacral region. One chamber had walls formed by
T\V millipore filters (0.45 n pore diameter) and the other cham
ber had walls of SS Very Dense filters (0.1 n pore diameter). Both
filters were able to exclude the subcutaneous tissue cells yet let
the interstitial fluid surrounding these cells fill the chamber. This
fluid was analyzed for glucose and lactate at various times after
the implantation.
For the first 4 days the glucose and lactate content of both
chambers did not change from the original level and, as compared
with plasma, the glucose concentration was 20-30% lower while
the lactate level was about equal in most cases. After 15 days,
however, the glucose concentration in all chambers was lower
than at the 4th day and the lactate concentration was several fold
higher than in plasma. In the chambers with T\V millipore filter,
the difference was generally more pronounced than in chambers
with SS Very Dense filters; in fact, in some of the millipore
chambers, free glucose was undetected while the lactate level was
equal to that found in malignant tumors (6) (Table 5). At that
time the connective tissue surrounding the two chambers showed
no morphologic indication of neoplastic transformation, and,
1045
Downloaded from cancerres.aacrjournals.org on June 14, 2017. © 1967 American Association for Cancer Research.
P. M. Cullino, F. H. Grantham, A. H. Courtney, and I. Losonczy
7
Õ--0.54+0.4IX
SE(b)--0.09
SE (a)--0.60
CE
O
e
J
en
O
O
5
o
4
o
E
E
(SJ
O
8
IO
12
14
LACTATE ELIMINATED.mmoles/hour/lOOgmTUMOR
CHART5. Relation between láclateand carbon dioxide elimination. Each point (•)represents one tumor. The tumor-bearing animals
were anesthetized for different periods and were breathing either air or pure oxygen. Each tumor used oxygen in proportion to the
amount available and as the lac tat«eliminated augmented the total carbon dioxide also increased. The same type of curve was observed
when oxygen utilized was substituted for lactate eliminated. Curve for Walker carcinoma; hepatoma and fibrosarcoma behaved in the
same way.
TABLE 3
Effect 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 = for
Rats 10, 11, 12 indicates that the amount of O»
consumed was indistinguishable from the amount avail
able because the values were so low as to defie accurate determination. Hyperglycemia was produced by
4-5 subcutaneous injections of 200mg dextrose each at about a 1-hour interval.
availableBefore12.67.16.715.07.226.213.312.117.96.823.044.8After3.20.50.54.02.16.53.82.25.10.21.01.5Oj
consumedBefore3.52.62.83.05.38.35.23.44.59.521.8After1.00.20.31.02.20.60.71.4^0.2Sl.O££1.5Glucose
consumedBefore3.91.65.05.97.32.69.010.714.122.834.769.2After3.71.54.82.19.64.0-1.62.21.80.8-1.8-4.2Lactate
producedBefore3.44.58.83.89.56.95.84.01.86.826.55.6After1.05.99.61.43.05.5-8.09.0-8.9producedBefore4.92.44.32.50.95.713.09.521.010.331.134.4A
RatNo.123456789101112Oi
104«
CANCER RESEARCH VOL. 27
Downloaded from cancerres.aacrjournals.org on June 14, 2017. © 1967 American Association for Cancer Research.
Oxygen and Glucose Consumption by Tumors in Vivo
TABLE 4
Comparison between Láclate Eliminated and Glucose Utilized by the Same Tumor during a
Severe Hypoglycemia Followed by Hyperglycemia
Hypoglycemia was produced by the i.V. injection of 20 units of insulin in animals bearing fibrosarcomas or Walker carcinomas and of 25 units in the animals with hepatomas. Hyperglycemia was pro
duced 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 glucose
utilized was lower during hyperglycemia with the exception only of Rat 4 of the Walker carcinoma
group and Rat 1 of the hepatoma group. Negative sign (—) = subtraction of lactate from blood instead
of elimination.
Tumor and rat
numberFibrosarcoma123456789Walker
utilized
eliminated
utilized
plasma
plasma
(mmole/
(mmole/
(mmole/
(mmole/
glucose
glucose
hr/100
gm)
hr/100
gm)
hr/100
gm)
hr/100 gm)
'0.030.0200.060.050.730.170.340.240.590.420.401.650.370.230.7400.140.140.22000.75Lactate
(A)
(mg/100
ml)113227233567452063352346625585214862123635Glucose
(mg/100
ml)143160257272557613748860173777791494224535477084758133344489500533Glucose
(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.
(C)0.060.560.300.340.250.891.344.307.036.056.432.444.242.903.8920.90.310.600.672.164.823
(D)-0.01-0.05-0.220.180.110.150.340.330.182.0803.643.042.402.831.800.770.6
carcinoma1234567Hepatoma1234567HypoglycemiaArterial
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 the
subcutaneous tissue of the lumbar region and each animal of the
second group received an SS filter chamber in the same area.
Both chambers ultimately produced sarcomas in about the
same number (Table 6). Morphologically,
the predominant cell
type was a large fibroblast and the cells were mostly collected in
bundles of irregular orientation (Fig. 2). The intercellular sub
stance was abundant with occasional deposition of calcium salts.
There was no morphologic difference between the fibrosarcomas
produced with either filter. All tumors started around the chamber
and grew to surround it. The host was killed within one month
JUNE 1967
after the tumor was first observed and all the sarcomas grew
after transplantation.
Fibrosarcomas arose earlier in the SS filter
group 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 were
first 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 tissues
around the chambers.
DISCUSSION
The comparison between glucose and oxygen consumed in vivo
by transplanted tumors has confirmed what had already been
shown in vitro, i.e., a large consumption of glucose contrasted to
a small oxygen utilization. In the past glycolysis was usually the
main objective of studies using in vitro technics to compare
1047
Downloaded from cancerres.aacrjournals.org on June 14, 2017. © 1967 American Association for Cancer Research.
P. M. Gullino, F. H. Grantham, A. H. Courtney, and I. Losonczy
3
Q
LU
LU
N
LU
LU
•Hypoglycemìa
A Hyperglycémie
98
Lu
O
O
O
O
EO
A°
vx-
A/.
A-?
I
2345678
GLUCOSE UPTAKE,mmoles/hour/IOOgm
TUMOR
CHART6. Relation between glucose uptake and ratio láclateeliminated/glucose utilized (Fibrosarcoma 4956). Each tumor is indi
cated by a number and represented by 2 values: (a) during hypoglycemia (•)and (6) 5-15 minutes after dextrose was given intrave
nously (A). During glucose starvation the values of the ratio were all higher than after hyperglycemia. Since two moles of láclate can
be produced by each mole of glucose the values found for No. 2 during glucose starvalion suggest lhal glucose was utilized from the
neoplastic lissue. Note thai Tumors 1, 2, and 3 utilized instead of producing lactate during hyperglycemia which followed glucose
starvation. (Tumor 3 showed no appreciable utilization of glucose or elimination of láclale during glucose starvation, therefore no
value appears on the graph.)
TABLE 5
Giocose 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 by
TW millipore filters (TW-MF), and interstitial fluid from the chamber formed by SS Very Dense filters
(SS-VD). At 15 days from the implant, glucose content was lower and lactate contenÃ-was higher lhan at
the 4th day in both chambers. The differences were often more evident in TW millipore filter chambers
where the lactate level was in some specimens similar to that found in tumors.
daysGlucosePlasma144167167181155160TW-MF144141109115140140SS-VD124113127177151152LactatePlasma463550413140TW-MF624852486048SS-VD60605548465015
daysGlucosePlasma144188174170160158TW-MF38075008SS-VD506179919690LactatePlasma3429412
Rat
No.1234564
1Sprague-Dawley cf and 9 .
1048
CANCER RESEARCH VOL. 27
Downloaded from cancerres.aacrjournals.org on June 14, 2017. © 1967 American Association for Cancer Research.
Oxygen and Glucose Consumption by Tumors in Vivo
TABLE 6
questions of survival capacity and susceptibility of tumor cells
to chemotherapeutic attack through their energy metabolism.
SS Very Dense Filters"
It was therefore crucial to ascertain whether in vivo neoplastic
Each animal received one chamber in a pouch of the subcu
cell ]>opulations could shift from respiration to fermentation or
taneous tissue of the lumbar region. The skin was sutured with vice versa as a function of the oxygen supply.
cotton and the stitches were removed during the second week.
We were unable to find any indication that the neoplastia
tissues
could supplement an in vivo oxygen deficiency by an
of
of
of rats
of
FilterSS
rats at
at first
tumors
observation appearance
increase in glycolysis. Tumors with low oxygen consumption
start127No. tumor95Totalproduced74Time (mo.)2020Time(mo.)1011131415161812151718
utilized proportionally smaller amounts of glucose. When gly
colysis was measured in the same tumor before and after an
DenseTW
Very
acute shortage of oxygen, there was no clear indication that
glucose utilization or lactate production was enhanced during
hypoxemia. Actually, any increase of glucose consumption re
quired an increase in oxygen utilization. Moreover, when an
abrupt deprivation of oxygen was inflicted to the tumor during
active glucose consumption, both utilization of glucose and
milliporeNo.
elimination of lactate ceased.
Despite the fact that glycolysis is not an exclusive pro|>erty of
the neoplastic tissues, it is, however, a remarkable metabolic
feature of most tumors and its significance has traditionally been
bound to the nature of neoplasia. The hy]X)thesis that neoplastia
°
Sprague-Dawleyrats rf1and 9 .
cells have an excessive influx of glucose and that lactate elimina
tion is a "detoxication" mechanism (4) could not be confirmed
and oxygen consumption of tumors. In vivo,it was found
in vivo. Neoplastic cells grew in a fluid practically devoid of
that glycolysis accounted for about 35% of the glucose utilized;
glucose and the vascular wall maintained a gradient between
however, a much larger fraction was not eliminated by the tumors
plasma and interstitial fluid (8). During normoglycemia, the
either as lactate or as carbon dioxide. The growth rate of many
amount
of glucose reaching the cells was less than they were able
of these tumors could not account for the large uptake of glucose.
to
handle
and, during moderate hyperglyeemia, the glucose
An alternative is that the neopla-stic tissue eliminated ]>art of
consumption increased while the lactate elimination remained
the glucose consumed in an unknown manner.
around 35% of the glucose intake.
The in vivo respiration of tumor tissue was small when com
The in vivo observations on lactate elimination during severe
pared with the glucose consumption. Quantitatively, however,
glucose
starvation suggest that glycolysis should not be studied
the absolute Q02 values were not excessively low. For instance,
solely from the point of view of energy metabolism, as it is
the oxygen consumption of rat heart muscle at rest is reported
usually done. When hypoglyeemia imposed a severe restriction on
(20) to be Qo2 = 3.0 to 8.0, and the Q02 of the tumors studied
glucose
consumption, the tumors eliminated practically all the
ranged from 1.0 to 4.2 (9). Our in vivo observations suggest the
utilized
glucose as lactate. However, when a large supply of glu
same conclusion provided by in vitro studies: The respiration of
the neoplastic tissues was much smaller than that of some tissues cose followed glucose starvation, not only did lactate elimination
with high oxidative rates, like the liver (1), however not so low cease despite a huge amount of glucose utilized, but even the
as to be considered deficient in comparison with other normal lactate of the afferent blood was consumed. The oxygen supply
was evidently adequate to co]>e with the increased metabolic
tissues with lower oxidative rates (2, 16, 19).
The relationship between respiration and growth is poorly needs, and energy requirement was probably not the major
understood. The seemingly logical e.\i>ectation that faster growth factor involved in lactate elimination by the neoplastic cells.
A large production of lactate was obtained in the subcutaneous
requires larger oxygen consumption is probably not true in these
tissues
long before any neoplastic transformation was observed.
simple terms. This is illustrated by the fact that in vivo the
fibrosarcomas used about J of the Oa of hepatomas but grew Actually, the lactate content of the subcutaneous interstitial fluid
twice as fast; both tumors converted about 35% of glucose into increased in all animals bearing microix>re chambers, despite the
lactate and fibrosarcomas consumed about 20% less glucose than fact that not all of them develojjed a sarcoma during 20 months
hepatomas. [The transformation of glucose into glycogen is of observation. The relationship between neoplastic transforma
negligible in 5123 hepatoma growing in normoglycemic host tion and excessive production of lactate in vivois difficult to assess
in the light of the results reported here. In beef erythrocytes, as
(18).]
The in vivo comparison between respiration and glycolytic an example, the glycolyzing enzymes seem to be concentrated on
caj>acity is in accord with the in vitro results. The amount of the cell membrane (5). A change in the ectobiologic characters
of the plasma membrane (10) could result, for instance, in an
glucose glycolyzed was higher than the quantity of glucose that
excessive production of lactate by these cells. The possibility
the tumors could oxidize. With both glycolytic and oxidative
energy pathways available, the neoplastic cell has often been that neoplastic transformation is only one of several conditions
considered to be in an advantageous position, vis à vis the non- able to produce an "unspeeific" increase of lactate production
Production of Fibrosarcomas by T\\' Millipore and
neoplastic cell. This assumption is relevant to the ini|X)rtant
JUNE 1967
should be kept in mind.
1049
Downloaded from cancerres.aacrjournals.org on June 14, 2017. © 1967 American Association for Cancer Research.
P. M. Gullino, F. H. Grantham, A. H. Courtney, and I. Losonczy
ACKNOWLEDGMENTS
We gratefully acknowledge the suggestions of Ur. Dean Burk
and Dr. Mark Woods and the help of Dr. Edmund A. Gehan, Na
tional Cancer Institute, Biometry Branch, and Mr. Donald Hill.
REFEKKÃŒVCES
1. Brauer, R. W., Leong, G. F., and Holloway, li. J. Oxygen
Consumption 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 in
Cancer Cells. Science, Ig4: 270-271, 1956.
3. Cori, C. F., and Cori, G. T. The Carbohydrate Metabolism of
Tumors. II. Changes in the Sugar, Lactic Acid, and COzcombining 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 Tannenbaum, M. The Significance of Glycolysis. J. Nati. Cancer
Inst., 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 of
Integrated Metabolie Pathways with Membranes. I. Glycolytic 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 Inter
stitial Fluid of Solid Tumors. Cancer Res., Õ4:780-798, 1964.
7. Gullino, P. M., and Grantham, F. H. Studies on the Exchange
of Fluids Between Host and Tumor. I. A Method for Growing
Tissue-isolated
Tumors in Laboratory
Animals. J. Nati.
Cancer Inst., %T:679-693, 1961.
8. Gullino, P. M., Grantham, F. IL, and Courtney, A. H. Glucose
Consumption by Transplanted Tumors in Vivo. Cancer Res.,
27: 1031-1040, 1967.
9. Gullino, P. M., Grantham, F. H., and Courtney, A. H. Utili
zation of Oxygen by Transplanted
Tumors in Vivo. Cancer
Res., 27: 1020-1030, 1967.
1050
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 Pas
sive Hyperventilation
on Cerebral Blood Flow, Cerebral
Oxygen Consumption, Cardiac Output and Blood Pressure of
Normal Young Men. J. Clin. Invest., go: 107-119, 1945.
12. Norman, T. I)., and Smith, A. B. The Blood Lactic Acid of
Tumor-bearing and Tumor-free Mice. Cancer Res., 16: 10271031, 1956.
13. Okamoto, Y. ÃœberAnaerobiose von Tumorgewebe. Biochem.
Z., 160: 52-65, 1925.
14. Scheinberg, P., and Stead, E. A., Jr. The Cerebral Blood Flow
in Male Subjects as Measured by the Nitrous Oxide Tech
nique. Normal Values for Blood Flow, Oxygen Utilization,
Glucose Utilization and Peripheral Resistance with Observa
tions 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 der
Wasserstoffionen Konzentration
der Gewebe Bei Lebenden
Tieren Mit Hilfe der Kapillar-glas-electrode.
Handbuch der
Biologischen Arbeitsmethoden.
Abderhalden, Abt. V., Teil
10: 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 Metabolism
of 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. Com
parative Biochemistry of Hepatomas II. Isotope Studies of
Carbohydrate Metabolism in Morris Heputoma 5123. Cancer
Res., 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. Federation
Proc., 21: 994-998, 1962.
CANCER
RESEARCH
Downloaded from cancerres.aacrjournals.org on June 14, 2017. © 1967 American Association for Cancer Research.
VOL. 27
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, subcutaneous
connective tissue; F, filter (In this case SS Very Dense was used. However, no appreciable difference was observed when TW millipore
filters formed the walls of the chamber.); SIF, subcutaneous interstitial fluid (protein precipitates). The morphology of the cells was
not appreciably different from that of cells of normal subcutaneous connective tissue without chamber despite the fact that the fluid
collected contained about twice the amount of lactate than plasma. The number of cells and of dilated capillaries was slightly higher
than normal around the filters. II & K, X 150.
JUNE 1967
IO.')1
Downloaded from cancerres.aacrjournals.org on June 14, 2017. © 1967 American Association for Cancer Research.
P. M. Gullino, F. H. Grantham, A. H. Courtney, and I. Losonczy
v r H'v
FIG. 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
Downloaded from cancerres.aacrjournals.org on June 14, 2017. © 1967 American Association for Cancer Research.
VOL. 27
Relationship between Oxygen and Glucose Consumption by
Transplanted Tumors in Vivo
Pietro M. Gullino, Flora H. Grantham, Anita H. Courtney, et al.
Cancer Res 1967;27:1041-1052.
Updated version
E-mail alerts
Reprints and
Subscriptions
Permissions
Access the most recent version of this article at:
http://cancerres.aacrjournals.org/content/27/6_Part_1/1041
Sign up to receive free email-alerts related to this article or journal.
To order reprints of this article or to subscribe to the journal, contact the AACR Publications
Department at [email protected].
To request permission to re-use all or part of this article, contact the AACR Publications
Department at [email protected].
Downloaded from cancerres.aacrjournals.org on June 14, 2017. © 1967 American Association for Cancer Research.