[CANCER RESEARCH 44,4458-4464,
October 1984]
Efflux of Adenine Nucleotides in Mitochondria from Rat Tumor Cells of
Varying Growth Rates1
Brad W. C. Lau and Samuel H. P. Chan2
Department of Biology, Biological Research Laboratories, Syracuse University, Syracuse, New York 13210
ABSTRACT
The efflux of adenine nucleotides was studied in mitochondria
isolated from normal rat liver, host livers, and the tumors from
four Morris hepatoma lines of varying growth rates. [3H]Adenosine diphosphate (ADP) or [3H]adenosine triphosphate (ATP) was
preloaded to the energized mitochondria, and the initial rates of
exchange with unlabeled extramitochondrial nucleotides were
measured with the carboxyatractyloside stop method. Results
indicate that the V™«
values of ATP efflux in mitochondria from
fast and intermediately growing tumors (hepatoma cell lines
7777, 7800, and 5123D) are significantly smaller than that of
host or normal liver mitochondria, while in slow growing tumor
(line 16) the Vâ„¢, is not different. On the other hand, for ADP
efflux, the opposite (namely, higher in tumor than in host) is
observed in the mitochondria of fast growing tumors. Preincubation with the divalent cation ionophore A23187 and calcium
chelator ethyleneglycolbisOS-aminoethyl ether)-/V,/V'-tetraacetic
acid increases the efflux of both ATP and ADP (to a lesser
extent) in these tumor mitochondria, indicating that the extraor
dinarily high concentrations of calcium form complexes with
adenine nucleotides (particularly ATP) and thus lower the effec
tive concentrations of free nucleotides for translocation. Together
with previously published results (R. L. Barbour and S. H. P.
Chan, Cancer Res., 43: 1511-1517, 1983) on lower nucleotide
uptake rates in these tumor mitochondria, we propose that the
lower ATP efflux and higher ADP efflux rates may cause a futile
cycle of ADP transport across the mitochondria! membrane
which may contribute to high rates of aerobic glycolysis (by
stimulating Key glycolytic enzymes such as hexokinase and
phosphofructokinase) observed in these fast and intermediately
growing tumors.
INTRODUCTION
High aerobic glycolytic activity in tumor cells was first discov
ered by Warburg (38) more than 50 years ago. A number of
investigators have explained this phenomenon in terms of cytcsolic or mitochondrial alterations in the regulation of high-energycontaining intermediates (for example, see Refs. 4 and 31). We
think it possible that factors regulating the metabolism of acienine
nucleotides in the mitochondrial matrix could become altered in
the tumor cells and elicit a lack of integration in the cytosol
resulting in an increased lactic acid production.
The transport of adenine nucleotides through the mitochondrial
membrane is carried out by a carrier protein on a one-for-one
basis (30). Adenine nucleotide in one side of the mitochondrial
membrane can exchange with the same (homologous) or differ
ent (heterologous) nucleotides on the other side of the mem
brane. Under a normal, energized state, a heterologous ex1This work was supported by NIH Grant CA-20454.
2To whom requests for reprints should be addressed.
Received January 20, 1984; accepted July 13,1984.
4458
change of extramitochondrial ADP with matrix ATP is preferred
to other types of exchange reactions (20, 23, 36).
This transport mechanism is dependent on several factors,
such as the energy state of mitochondria (28, 37), the pool size
of internal free adenine nucleotides (28), and the levels of certain
endogenous inhibitors (6, 25). In a series of Morris hepatoma
lines, the initial velocity measurements of the uptake of ADP and
ATP in mitochondria have shown a Vm«x
significantly lower than
that of normal rat liver mitochondria (3). An inverse correlation
was found between the Vmaxof ADP uptake and the growth
rates of these tumors. Decreased rates of nucleotide uptake in
tumor cells can be partly due to (a) diminished pool size of both
total and exchangeable adenine nucleotides in the matrix and (b)
elevated levels of Ca2+ in the mitochondria from these tumor
lines, especially Morris Hepatoma 7777 (7), which shows a 70fold increase in mitochondrial Ca2+. The excessive Ca2+ levels
may reduce effective free nucleotides for translocation. This
paper reports results on the efflux of adenine nucleotides in
mitochondria isolated from tumors of different growth rates
(Morris Hepatoma 7777, rapidly growing, poorly differentiated;
7800 and 5123D, intermediately growing, well differentiated; and
16, slowly growing, highly differentiated). Kinetic studies on
tumor mitochondria preincubated with the divalent cation ion
ophore A23187 (33) provide information on the effect of the high
Ca2+ levels on translocase activity in these tumor mitochondria.
A preliminary report of this study was presented in abstract form
(8,22).
MATERIALS AND METHODS
Materials. Carboxyatractyloside, oligomycin, ruthenium red, m-chlorocarboxylcyanide phenylhydrazone, A23187, ATP, ADP, AMP, a-oxoglutarate, BSA3 (Fraction V), and EGTA were purchased from Sigma
Chemical Co., St. Louis, MO. Tetrabutylammonium
hydroxide was a
product of Aldrich Chemical Co., Milwaukee, Wl. [2,8-3H]ADP (28 Ci/
mmol) and [2,8-3H]ATP (25 C¡/mmc4)were purchased from New England
Nuclear, Boston,
MA. Male Sprague-Dawley
rats (250 g) were from
laconic Farm, Germantown, NY. Male Buffalo rats weighing 140 to 180
g were obtained from Simonsen Laboratories, Gilroy, CA, and shipped
to Howard University, Washington, DC, where the Morris 7777, 7800,
5123D, and 16 cell lines were transplanted to the hind legs. After the
growth of tumor was confirmed in these animals (usually within a few
days), they were transferred to our laboratory. All other reagents used
were of highest purity commercially available.
Adenine Nucleotide Transport Assay. The study of efflux of adenine
nucleotides reported in this paper refers to a backward transport of one
kind of adenine nucleotide from the matrix of mitochondria in exchange
for another nucleotide species in the extramitochondrial compartment,
i.e., ADP»,for ATPou, or ATP«for ADPcu,, with the subscripts "in" and
"out" indicating adenine nucleotide inside or outside mitochondria, re-
9 The abbreviations used are: BSA, bovine serum albumin; HPLC, high-pressure
liquid chromatography; EGTA, ethyleneglycolbis(/3-aminoethyl ether)-«, W-tetraacetic acid.
CANCER
RESEARCH
VOL. 44
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1984 American Association for Cancer Research.
Adenine Nucleotide Efflux in Tumor Mitochondria
spectively. Exchange for homologous adenine nucleotide was also ex
amined as indicated in the individual legends. The efflux rate of mito
chondria! adenine nucleotides was measured by the carboxyatractyloside
stop method (2, 37). Mitochondria were first energized by incubating in
50 ml KCI medium containing 115 mM KCI, 21 mM Tris-HCI (pH 7.4),
1.05 mm EDTA, 5.25 mM a-oxoglutarate, and 1% BSA at room temper
ature for 5 min before the addition of oligomycin (2 ^g/mg protein). Since
tumor mitochondria exhibit slower uptake rates and contain less amounts
of endogenous ATP and ADP than do the normal liver mitochondria (3),
the loading of radioactive ATP or ADP into tumor mitochondria required
higher levels of radioactive isotopes than that of normal liver. Therefore,
the energized mitochondria were loaded with higher levels of radioactive
ATP or ADP (10 nC\ for normal or host liver mitochondria and 20 nC\ per
50 mg tumor mitochondria! protein) at 0°for 30 and 5 min, respectively,
following the procedure of Pfaff and Klingenberg (29). The preloaded
mitochondria! pellets were washed twice immediately with 30 ml of 0.25
M sucrose medium. The transport reaction of adenine nucleotides from
mitochondria was initiated by adding graded amounts of counterexchanging unlabeled adenine nucleotides from a syringe apparatus into
microfuge tubes that contain 1 ml KCI medium and 1 mg mitochondria!
protein. A vortex evaporator was used to provide adequate mixing during
the reaction. After 10 sec of mixing at 0°,the reaction was terminated
by the injection of 50 /¿Iof 200 ¿<M
carboxyatractyloside
which was
delivered by a second series of Hamilton syringes. The reaction mixture
was centrifugea by a Beckman microcentrifuge for 4 min. A fraction of
the supernatant was transferred to another tube and counted for radio
active adenine nucleotide that was transported out of the mitochondria;
the pellet was suspended in 200 n\ of 2% sodium dodecyl sulfate and
counted for radioactivity that remained in the loaded mitochondria. The
rate of adenine nucleotide efflux was calculated from the percentage of
total radioactive adenine nucleotide in the supernatant medium. Any
small increases of radioactivity in the extramitochondrial compartment
were corrected by the subtraction of new sets of zero time controls.
Apparent kinetic constants were calculated from data on the first half
[higher substrate concentrations because of the biphasic nature of kinetic
plots (2) as will be shown under "Results"] of the curve by the direct
linear plot method of Eisenthal and Comish-Bowden
(12). When mito
chondria were incubated with ionophore, 20 ¿»M
A23187, 1.5 mM EGTA
was added in the presence of 0.25 M sucrose and 1% serum albumin
and incubated for 5 min at room temperature before nucleotide efflux
assays were performed.
Determination of Mitochondria! Adenine Nucleotides. Intramitochondrial adenine nucleotides were extracted from mitochondria (10 to
15 mg mitochondrial protein) by 5% perchloric acid at 4°for 30 min and
neutralized by KOH. Determination of adenine nucleotides was per
formed by a Beckman HPLC system (Model 112 solvent delivery system
and Model 153 optical unit). A HPLC column [üchrosorb RP-18 (5 /mi)]
was used for the isocratic elution of adenine nucleotides in a mobile
phase of 0.03 M KH2PO4, 0.01 M tetrabutylammonium hydroxide, 19%
acetonitrile, and 0.1% triethanolamine, pH 3.5, and detected at 254 nm.
The resolved nucleotide peaks were quantitatively compared with stand
ards by using a Hewlett-Packard Model 3390A integrator.
Isolation of Mitochondria. Mitochondria of the host and normal rat
liver were isolated according to the method of Bustamante et al. (5) in
220 mM mannitol, 70 mM sucrose, 2 mw 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid, 1 mM EDTA (pH 7.4), and 1% BSA. The hepatoma mitochondria were isolated as before (7) using the procedure of
Kaschnitz ef al. (19). Since the tumor size is an important factor for
obtaining mitochondrial preparations with good respiratory control ratios
(3), the size ranges of tumors used were (in g): 2.1 ± 0.9 (S. D.),
Hepatoma 7777; 4.05 ±1.32, Hepatoma 5123D; 7.65 ±2.15, Hepatoma
7800; and 12.2 ±3.1, Hepatoma 16. It appeared that it is more critical
to use smaller tumors of the fast-growing line than slow-growing cell
lines. Using these sizes, the resultant mitochondrial preparations had
respiratory control ratios of 2.5 or above (using succinate as substrate)
and appeared to have intact double membranes as examined by electron
microscope (7). Mitochondria were suspended in 0.25 M sucrose and
1% BSA to give a 100- to 150-mg protein/ml concentration
immediately
prior to their use.
RESULTS
The time course for the efflux of adenine nucleotides across
the mitochondrial membrane varies somewhat among the normal
rat liver, different hepatomas, and their host livers (Chart 1). For
example, the effluxes of ATP and ADP in normal liver mitochon
dria remain linear for at least 30 and 40 sec, respectively. The
efflux behavior of Hepatoma 16 (a slow-growing, well-differen
tiated tumor) is similar to that of the normal rat liver mitochondria.
The time course of both ADP and ATP effluxes remained linear
until 30 sec of incubation (except in the exchange of ATP for
intramitochondrial ADP in tumor mitochondria). The homologous
efflux rates of ADP in mitochondria of both Tumor 16 and its
host liver were higher than was the heterologous exchange,
whereas the heterologous ATP exchange was higher in these
mitochondria than was the homologous exchange. For all other
tumors and their host livers, the effluxes of both ATP and ADP
in mitochondria in exchange with the extramitochondrial nucleo
tides started to drop or level off after 15 sec. It is of particular
interest that the hepatomas affect the transport activity of the
host liver mitochondria. In any case, all the subsequent kinetic
studies of efflux were performed for 10 sec.
In order to ascertain that carboxyatractyloside
is similarly
effective on tumor mitochondrial membrane, 3 identical sets of
nucleotide exchange reaction were carried out in the mitochon
dria of Hepatoma 7800 and its host liver (Table 1). Centrifugation
of the samples was performed at 0, 5, and 10 min after the
termination of reaction. There were essentially no difference in
radioactive counts in the mitochondrial pellets regardless of
whether there is a postponement in separating them from the
media, thus indicating a complete stop of efflux of nucleotides
from the tumor mitochondria by carboxyatractyloside. This is in
contrast to the results of Woldegioris ef al. (43), who reported
that tumor mitochondria from the same cell line were somewhat
less susceptible to the inhibition of carboxyatractyloside than
were the host liver mitochondria. This difference could be due to
the detergent effect (42) of carboxyatractyloside at the higher
concentration (100 /IM) used in their study. In fact, many other
laboratories used only 5 to 10 MM carboxyatractyloside for the
complete inhibition of the adenine nucleotide transport system
in rat liver mitochondria (10, 21, 37).
The transport of adenine nucleotide in normal mitochondria is
dependent on (a) the levels of endogenous inhibitors (6, 25), (b)
the mitochondrial energy states (28, 37), and (c) the pool size of
exchangeable adenine nucleotides within the mitochondrial ma
trix (28). We reported that the rates of uptake of adenine nucleo
tides in tumor mitochondria were lower than that of normal rat
liver mitochondria, but the lower transport activity was not due
to the effects of endogenous inhibitor (e.g., long-chain acyl-CoA
esters) or the mitochondrial energy states (3). Compared to the
results previously obtained by enzymatic methods (7), the ade
nine nucleotide levels determined by the HPLC procedure were
slightly higher (probably reflects higher accuracy by HPLC), but
the overall pattern was similar (Table 2). The total adenine
nucleotide levels of both tumor and host liver mitochondria were
diminished as a function of the growth rate of the tumor; i.e., the
faster the growth rate of a tumor, the smaller is the pool size of
its endogenous adenine nucleotides in both tumor and host liver
OCTOBER 1984
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1984 American Association for Cancer Research.
4459
8. W.C. Lau and S. H. P. Chan
.15
MH5123D
.12
.09
.06
.03
120
O
10
15
20
25
10
15
20
25
30
0.5
0.4
o>
0.3
o
0.2
0.1
O
1.0
MH16
0.8
0.6
0.4
0.2
120
O
O
10 20 30
120
60
Time (see)
Chart 1. Time course of adenine nudeotkJe efflux transport activity in mitochondria from normal rat liver, various hepatomas, and their host livers. One mg of energized
mitochondria preloaded with either [2.8-3H]ATP (O, tumor; A, host liver; O, normal liver) or [2,8-3H]ADP (•,tumor; A, host liver; •,normal liver) were suspended in 0.25
M sucrose solution and exchanged with either ADP or ATP, respectively (heterologous exchange) in KCI medium for from 5 to 120 sec. The exchange reaction was
terminated by adding 50 nl of 200 MM carboxyatractyloside at th indicated times. Isolation of mitochondria and determination of efflux rate are described in 'Materials
and Methods". Homologous exchanges of ADP p. tumor; D, host liver) and ATP (C, tumor; A, host liver) in Hepatoma 16 are also presented.
Table 1
Effectiveness of carboxyatractyloside on the inhibition of adenine nucleotide efflux
in tumor (Hepatoma 7800) and host liver mitochondria
The ADP transport was assayed as described in 'Materials and Methods,"
except that the reaction mixture was centrifugea at 0, 5, and 10 min after the
termination of the efflux reaction by adding 10 MMcarboxyatractyloside. After the
removal of the reaction mixture supernatant, the radioactive ADP retained in the
mitochondrial pellet was counted by liquid scintillation.
timesMitochondrial
sourceHepatoma
Table 2
Adenine nucleotide levels In mitochondria of normal rat liver,
Morris tumors and host livers
Mitochondrial AMP, ADP, and ATP levels were measured by HPLC-neutralized
HCIO. extracts of 10 to 15 mg mitochondria.
Mitochondrial
No. of measurementsNormalMH16"tumor
source
cpm at following delay
min14,160
Host
liverMH5123D
8534
proteinAMP8.0
±1.2"7.5
±1.22.6
±0.71.6
±1.5
±1.26.9
8.0
±0.5
0.72.8
2.0 ±
±0.4
±0.20.6
1.6
±2.5
±1.03.2
7.9
±0.2
0.43.9
2.3 ±
±0.2
±0.41.0
1.3
±0.3
0.63.1
5.9 ±
±0.5
0.63.6
3.4 ±
±0.1
±0.20.9
1.3
7800
Host liver0
18,7505min13,540 19,53010min14,420 18,205
mitochondria. It was somewhat unexpected that the AMP levels
in host liver mitochondria of both tumor lines, 7800 and 7777,
were much lower than that in other tumor host livers and the
normal rat liver. On the other hand, the AMP levels of these 2
tumor lines were more than 50% lower than the control value. It
was generally observed that ATP levels in the tumor mitochon
dria (except Hepatoma 16) were lower than in the host liver
mitochondria which were very similar to each other. On the
contrary, the mitochondrial ADP levels in all the tumor lines
exhibited higher values than did their corresponding host liver
mitochondria. Reduction of ATP levels and accumulation of ADP
in tumor mitochondria could be partly due to intrinsic ATPase
activities in these mitochondria (1). What is more important is
that this change of adenine nucleotide pool size may attribute to
altered efflux rates as can be seen later in this report.
Kinetic analysis on the efflux of adenine nucleotides in normal
rat liver mitochondria showed biphasic double reciprocal plots
4460
tumor
Host
liverMH7800
tumor
Host
liverMH7777
tumor
Host liver98
±0.2*
44
±0.4
±0.3
±0.0
4nmol/mg 4.9 ±0.2ADP4.0
2.4 ±0.3ATP1.5
1.6
Mean ±S.E.
" MH, Morris hepatoma.
(Chart 2). Similar kinetic behavior was also observed in various
tumors and host liver mitochondria. As interpreted in our report
on the uptake of adenine nucleotide in mitochondria (2), this
pattern is also a possible indication of negative cooperativity
between adjacent carriers in the membrane or the presence of
endogenous, tight-binding competitive inhibitors. Since the sig
nificance of kinetic constants obtained at low substrate concen
trations would depend on the inconclusive interpretation of this
plot, the apparent kinetic constants reported in the present study
were determined from data at the high substrate concentration
range.
CANCER RESEARCH
VOL. 44
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1984 American Association for Cancer Research.
Adenine Nucleotide Efflux in Tumor Mitochondria
Chart 2. Lineweaver-Burk ptot of adenine
nucleotide efflux in normal rat liver mitochon
dria. Initial velocities of [3H]adenine nucleotide
(AdN) efflux were measured as described in
"Materials and Methods." Energized mitochon
dria were first loaded with labeled ADP or ATP
(3 MCi/10 mg protein) for 5 or 10 min and
washed twice in the homogenization medium.
The washed mitochondria were then sus
pended in 0.25 M sucrose solution with 1% BSA
(100 mg/6 ml protein). The final efflux assay
mixture was composed of 110 mu KCI, 20 m M
Tris-HCI (pH 7.4), 1.05 mM EDTA, 5 mw oxoglutarate, 1% BSA, and about 5 mg mitochondria
per set of assay. The reaction was started by
a simultaneous injection of 5 different concen
trations of either ADP or ATP and terminated
by the addition of 50 pi of 200 ,<M carboxyatractyloside to the mixture. Three different types of
effluxes were included: ATP«-^ ADPM (O),
ADPh —ATP«,(A), and ADP,, -- AD?«,,(D).
0.4
1.2
0.8
1.6
I AdN I"1,
Results in Table 3 show the apparent kinetic constants from
the measurement of the initial rates of exchange between matrixlabeled ATP and extramitochondrial ADP in normal rat liver, host
liver, and tumor mitochondria. Both K„,
and Vâ„¢*for ATP efflux
are significantly lower in the hepatoma mitochondria than that of
their host livers and normal rat liver. The rapidly growing tumor
(Hepatoma 7777) particularly showed a substantially lower rate
of efflux of ATP than did its host liver and other tumors. It seems
that an inverse correlation exists between the Vâ„¢xvalues of
efflux rates and the growth rates of the tumor. In fact, the efflux
rate of ATP in Hepatoma 16 is not lower than that of its host
liver and normal liver mitochondria. It is also of interest to note
that the ATP efflux rates in the host livers of the faster-growing
tumors were higher than that of the slow-growing tumor, indi
cating a possible interaction between that tumor and its host
liver (40).
In normal respiring mitochondria, since ATP bears one more
negative charge than does ADP at physiological pH, it is prefer
entially transported into the extramitochondrial space (20). In the
presence of uncouplers, which negate the membrane potential,
this preference of ATP efflux is cancelled. In the present study,
the tumor mitochondria of hepatoma 7777 and 5123D in the
presence of the uncoupler /D-chlorocarboxylcyanide phenylhydrazone exhibited increases in Vmax,while the Kmvalues did not
vary to any significant extent. Therefore, membrane potential
seems to play an important role, but it does not constitute the
total cause of decreased ATP efflux rates as observed in these
tumor mitochondria.
Endogenous divalent cations (Ca2+ and Mg2+) are known
chelators for oxyanions (41), such as ADP3" and ATP4", to form
cation-adenine nucleotide complexes. Complexation of adenine
nucleotides by Ca2+ greatly reduces the pool size of free ex
changeable nucleotides in mitochondria and thus the exchange
rates. The rapidly growing tumor mitochondria were found to
contain large amounts of Ca2+ that impede the transport rates
Table3
Apparent kinetic constants of ATP efflux in control rat liver and hepatoma
mitochondria: effects of uncoupler and ionophore
Measurement of the initial rates of efflux of ATP was described in 'Materials
and Methods." The mitochondria preloaded with ß.S-'HlATP (specific activity,
>40,000 cpm/mg protein) were suspended in 0.25 M sucrose solution; a fraction
with 1 mg mitochondria! protein was first mixed with the KCI medium before the
reaction was started by the injection of exogenous ADP (final concentration, 1 to
50 MM).The exchange reaction was terminated after 10 sec by adding 50 rf of 200
»IM.Uncoupled mitochondria was prepared in KCI medium containing 4 ¡Mmchtorocarboxylcyanide phenylhydrazone. In the ionophore experiments, unlabeted
mitochondria were incubated with ionophore A231 87 in the presence of EGTA for
5 min at room temperature. The V™,and K„
are expressed in terms of nmol/mg/
min and #IM,respectively.
Mitochondrial
sourceNormal
uncouplerVIT«1.20.61.210.41.850.871.261.81.05K*.9.766.5
ionophorey1.551.064.30.62.20.854.02.0
liverHepatoma
23DHost 51
liverHepatoma
7800Host
liverHepatoma
7777Host
liverHepatoma
+ruthenium
7777
redHepatoma
16Host
liverHepatoma
(TT)aHost 16
liver (TT)ControlVâ„¢,1.10.551.20.521.580.231.70.481.61.30.970.51Kâ„¢19.73.17.92.0710.43.99.56.442
" TT, efflux of ATP in exchange with extramitochondrial ATP.
of both ADP and ATP (3). Since we knew from the early experi
ment that efflux rate of ATP was reduced, we wanted to examine
whether it is due to the increased Ca2+ levels in tumor mitochon
dria. When incubated with A23187, mitochondria from hepatoma
7777, which contains up to 70 times as much Ca2+ as does the
normal rat liver mitochondria (3), improved their ATP efflux rates
up to 4-fold, while those with lower Ca2+ levels improved their
OCTOBER 1984
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1984 American Association for Cancer Research.
4461
B. W. C. Lau and S. H. P. Chan
ATP efflux rates only 1.5- to 2-fold. A23187 also enhances the
ATP efflux in host liver mitochondria, although to a smaller
magnitude. Therefore, ionophore exerts its action on the efflux
of ATP probably by modifying the membrane structure of mito
chondria in addition to its disposal effect on Ca2+.
Eboli ef al. (11) found an essentially unchanged adenine nucleotide transport activity (ATPh ^ ADP««)
in Tumor 3924A
mitochondria which was treated by ionophore A23187 to reduce
its endogenous Ca2+ pool from 300 to 46 nmol/ml. This discrep
ancy has prompted us to further examine the Ca2+ levels in the
tumor mitochondria treated with the ionophore. As shown in
Table 4, the tumor (Hepatoma 7800) mitochondria contain 137
nmol Ca2+/mg protein, which is about 30-fold higher than that in
the normal rat liver mitochondria (3.98 nmol/mg). It is of interest
that, as the growth rate of hepatoma 7800 accelerated and the
tumor size became larger, the mitochondria! Ca2+ content also
concomitantly increased drastically (Ref. 7; see also "Discus
sion"). Host liver of this tumor line contains slightly less Ca2+
than does the normal control. After treatment with A23187, the
tumor mitochondrial Ca2+ level dropped to 6.33 nmol/mg,
whereas the host liver mitochondria showed a minimal decrease
in its endogenous Ca2+ pool. The tremendous reduction of Ca2+
Tables
Apparent kinetic constants of ADP efflux in control liver and hepatoma
mitochondria: effects ofuncoupler and ionophore
Experimental conditions were similar to those described in Tabte 3, except that
the mitochondria were preloaded with [2,8-3H]ADP (specific activity, >30,000 cpm/
mg protein) in exchange with exogenous ATP (4 to 40 pM final concentration) The
. and K„
are expressed in terms of nmol/mg/min and pu. respectively.
Mitochondrial
sourceNormal
uncoupterv™,1.822.341.443.672.34.21.72.91.1K„21
ionophorev„»1.293.951.445.5
liverHepatoma
23DHost 51
liverHepatoma
7800Host
liverHepatoma
7777Host
liverHepatoma
16Host
liverHepatoma
(DD)aHost 16
liver (DD)perseV™.0.872.221.052.751.753.31.60.640.362.130.63K„60.66.634.85.252.015.160.
DD, efflux of ADP in exchange with extramitochondrial
ADP.
increase of V^ of ATP efflux in Tumor 7777 mitochondria may
be caused by a higher membrane potential and thus reflect the
electrogenic nature of the adenine nucleotide transport system
in tumor mitochondria. This slight increase of efflux activity
7800 mitochondria is much lower (44 nmol/mg) when the mito
chondria were isolated in the presence of 5 UM ruthenium red. It represents only a small percentage of recovery. As mentioned
earlier, it is unlikely that membrane potential is totally responsible
is probable that the tumor mitochondria not only accumulate
excessive amounts of Ca2+ in vivo but also tend to take up a in causing the lower ATP efflux rate in tumor mitochondria.
large amount of Ca2+during the course of isolation. Nevertheless,
A comparative study was made on the efflux of ATP in Tumor
even a 44-nmol/mg level of Ca2+ represents a 7-fold higher
16 mitochondria between the homologous and heterologous
content than the Ca2+ content when the tumor mitochondria
exchange of adenine nucleotides (Table 3). Both Tumor 16 and
its host liver mitochondria have smaller Vâ„¢ but much greater
were treated with A23187. Studies on the efflux rate of ATP and
Kmvalues in their homologous exchange than their correspond
ADP in Hepatoma 7800 mitochondria isolated from tissue ho
ing kinetic values in the heterologous exchange. In comparing
mogenized in medium containing 5 pu ruthenium red also
the Vmaxfor Tumor 16 to its own host liver and host livers in
showed higher Vm«than did the control tumor mitochondria
other tumor lines, the experimental results show that this slowly
(data not shown). Moderately increased V™«
and «„,
on the efflux
growing tumor does not show much deviation in the efflux of
of ATP were also obtained in Tumor 7777 mitochondria (Table
ATP from the normal tissues and host livers.
3) isolated in the presence of ruthenium red.
Table 5 shows the apparent kinetic constants for the ADP
The unsuccessful trial by Eboli ef al. (11) to deplete the tumor
mitochondria of excessive endogenous Ca2+ to normal range (3 efflux in mitochondria of normal rat liver, hepatomas, and their
host livers. In contrast to ATP efflux, the V™«
of labeled ADP in
to 5 nmol/mg) could possibly be due to a relatively low pH in the
the matrix of tumor were twice as fast as the normal control,
EGTA solution. For our successful partial depletion, we used
whereas the Km values were smaller. There also appears to be
isotonic sucrose solution in the presence of 1% BSA and 1.5
a direct correlation between the growth rates of the tumor and
rnw EGTA, pH 7.4, for suspending the final mitochondrial pellet
the ADP efflux rates in tumor mitochondria; i.e., the more rapidly
before the treatment of A23187.
growing tumor exhibits a higher ADP efflux activity, in all tumor
Since ruthenium red also inhibits State 4 respiration and may
mitochondria, the efflux of ADP showed uniformally higher V™«
create a larger membrane potential (35), it is probable that the
and lower Kmthan that of their host liver mitochondria. A homol
Table 4
Reduction of Ca2* levels in Morris Hepatoma 7800 mitochondria
ogous exchange of ADP in Tumor 16 mitochondria further illus
by ionophore A23187
trates that tumor mitochondria have indeed higher Vmaxvalues
Mitochondria were incubated with 20 ¡IMA23187 in 5 ml of EGTA buffer (1 5
than do the host liver mitochondria. This comparative study of
(DM EGTA, 0.25 sucrose, and 1% BSA, pH 7.4) at room temperature for 5 min.
homologous versus heterologous ADP efflux has revealed that
After being washed twice in EGTA buffer, the mitochondria were suspended in 2%
sodium dodecyl sulfate and measured for Ca2+ by atomic absorption spectrophohomologous exchange in tumor mitochondria shows an even
tometry. Ruthenium red-treated mitochondria were isolated from tumor tissue
more drastic difference between tumor mitochondria and their
which was homogenized in the presence of 5 MMruthenium red.
host liver counterparts.
nmol Ca*+/mg protein
The uncoupled tumor mitochondria showed a faster rate of
efflux of ADP as observed in host liver and normal control.
liver2.85
Control
Apparently, the efflux of ADP in tumor mitochondria is also
+ A23187
6.33
2.46
dependent on the membrane potential as indicated by a progres
•
Ruthenium red during isolation780013744.2Host
sive increase of efflux rates in their uncoupled state. Again
in the tumor mitochondrial matrix thus may implicate the pres
ence of higher levels of available exchangeable ADP and ATP.
Another interesting finding shows that the Ca2+ levels in Tumor
4462
CANCER
RESEARCH
VOL. 44
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1984 American Association for Cancer Research.
Adenine Nucleotide Efflux in Tumor Mitochondria
High Ca2+ as the Main Culprit for Changes in Efflux Rates.
ionophore A23187 preincubation enhanced the efflux of ADP in
tumor mitochondria, manifesting the release from divalent cation
complexes of free adenine nucleotides available for the transport
by the carrier across the mitochondrial membrane.
We had previously reported enormous elevation of intramitochondrial Ca2+ levels in various tumor lines which may account
DISCUSSION
inhibition of adenine nucleotide exchange activity (16). Similar
results were observed by Eboli ef al. (11). When the Ca2+ levels
Most malignant tumors exhibit increased aerobic glycolytic
activities, but the reason(s) and the mechanism are largely un
clear. This study and our previous results indicate that an altered
adenine nucleotide transport in tumor mitochondria may be
responsible. Apparently, the adenine nucleotides are being trans
ported across the tumor mitochondrial membrane through a futile
cycle. As reported earlier, the apparent V™«
values for the uptake
of ATP and ADP in tumor mitochondria are smaller than that in
normal tissue, and under the condition of active respiration ATP
uptake is higher than ADP uptake. At the same time, as indicated
in this report, the apparent V™«
values for the efflux of ATP in
tumor mitochondria are much smaller than that of their host liver
and normal rat liver mitochondria, whereas the opposite is ob
served for the efflux of ADP. Consequently, comparing to normal
tissue, the tumor mitochondria are transporting more ATP from
the cytosol and in turn supply the cytosol with excess ADP, thus
generating an energy-deficient mitochondrial state and a higher
ADP/ATP ratio in the cytosol. Not only does the tumor cell
deprive itself of efficient oxidative phosphorylation by mitochon
dria but also its glycolytic activity in cytosol is also greatly
stimulated by the elevated ADP/ATP ratio (18, 34).
Lowered Efflux Rates due to Smaller ATP and ADP Pool
Size. The pool size of the endogenous adenine nucleotides
deserves a closer look for its effect on the mitochondrial efflux.
Pfaff and Klingenberg (29) have established that the transport
rates of adenine nucleotide depend on the levels of the free
exchangeable ATP or ADP in the mitochondria. The higher ADP
levels and lower ATP levels in the tumor mitochondria, especially
those in the more rapidly growing tumors, seem to explain at
least partially why ADP efflux rates are faster and ATP efflux
rates are slower in the tumor mitochondria than that in their host
liver mitochondria.
Another pronounced difference is the great reduction of AMP
levels in these tumor mitochondria, especially in Tumor 7777. A
similar observation was also reported by Eboli ef al. (11) in
another rapidly growing tumor, 3924A, in which the AMP levels
were 20% less than the normal control while the total exchange
able pool size (ADP + ATP) remained similar to that in the liver.
Lower levels of AMP in tumor mitochondria could drastically
reduce the potential reserve capacity to replenish the exchange
able adenine nucleotides, further diminishing the availability of
substrates for the transport system. Under certain metabolic
stress when energy is quickly depleted, the mitochondrion can
utilize its potential reserve capacity to generate more ATP by
phosphorylating AMP through the GTP-AMP-P, transferase (17)
and/or adenylate kinase reactions (9, 39). The adenylate kinase
may exert its action on the inner mitochondrial membrane to
phosphorylate AMP into ADP. Although this issue remains con
troversial, levels of AMP remain important to maintain the pool
size of exchangeable adenine nucleotides. Therefore, altered
nucleotide transport kinetics in tumor mitochondria is not simply
determined by the availability of free ATP and ADP; any decrease
in AMP levels may further affect the low activity of nucleotide
transport in these mitochondria.
in normal rat liver mitochondria were raised to that close to tumor
mitochondria, depressed ATP efflux rates similar to that of tumor
mitochondria were observed. Therefore, artificial reduction of
Ca2+ in the mitochondrial matrix can hypothetically release more
nucleotides from the Ca2+-nucleotide chelated complexes. In
for the observed lower free ATP concentration in the matrix (7).
Loading of Ca2+ in normal rat liver mitochondria resulted in the
deed, the presence of divalent cation ionophore A23187 greatly
enhances the efflux rates of ATP in tumor and to a lesser extent
host liver mitochondria, demonstrating that the inhibitory action
by Ca2+ on the carrier was reversed. This result was compatible
with the work of Duszynski ef al. (10) that A23187 stimulates
translocase activity. Similarly, the efflux of ADP was also in
creased by the action of ionophore. This indicates that some
ADP was also freed from the Ca2+-ADP complex and becomes
available for the translocase activity.
Slower uptake rates of both ADP and ATP in tumor mitochon
dria could primarily lead to smaller ATP pool size, and less ATP
is phosphorylated from the incoming ADP. In addition, the pres
ence of excessive Ca2+ makes the readily free adenine nucleo
tides unavailable for the transport protein. Therefore, it is con
cluded that these 2 factors both contribute synergically to a
smaller V,™«
of ATP efflux in the tumor mitochondria. Williams
(41) reported that the binding affinity of Ca2+ to anions depends
on the number of anionic groups available. Since ADP3" bears
one less negative charge than does ATP4" at physiological pH,
it has lower binding affinity towards Ca2+ (26). On one hand, the
extent of ADP bound by Ca2+ is less in the matrix of tumor
mitochondria; on the other hand, the total pool size of ADP is
larger than that of ATP. It is conceivable that ADP stands a
better chance to compete for the binding to the carrier and thus
results in a faster rate of efflux into the extramitochondrial
compartment.
It is worth noting that recently the growth rate of Hepatoma
7800 has increased quite drastically. It was only within 2 to 3
weeks (instead of 3 to 5 weeks previously) that the tumor had
grown almost full size to 2 to 3 cm in diameter. Concomitant
with faster growth rate, the Ca2+ level in this tumor line is also
higher than previously observed (7). If this elevated Ca2+ is a
result of increased growth rate of the tumor, the lowered ATP
efflux rates in the more rapidly growing tumor mitochondria
caused by the excessive Ca2+ levels are inversely correlated to
the size of the tumor. This finding is consistent with the postu
lation by Emmelot and Bos (13) that tumor size may influence
mitochondrial functions. It is of interest to note that fast-growing
tumors influence the mitochondrial nucleotide transport of their
host liver, indicating possible hormonal or other interactions
between the tumor and its host (40).
The question remains as to why tumor mitochondria accu
mulate such elevated amounts of Ca2+ and how. The transport
of Ca2+ across the mitochondrial membrane is facilitated by
another carrier protein which requires energy for its function (24).
It appears that many tumor cells tend to retain more Ca2+ than
Mg2+ in their mitochondria (27, 32). Since the Ca2+ camer is
membrane bound, its activity is greatly dependent on the physical
state of the mitochondrial membrane, including the quantity and
OCTOBER 1984
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1984 American Association for Cancer Research.
4463
a W. C. Lau and S. H. P. Chan
species of lipid moiety in the membrane. Changes in these lipid
moieties (including phospholipids and cholesterol) result in struc
tural alteration in tumor membranes (14) and associated changes
in mitochondrial membrane-bound enzymes (26). Hepatoma mi
tochondria contain much higher level of cholesterol than do the
normal rat liver mitochondria (15). We also found that the inner
mitochondrial membranes of various Morris hepatoma lines con
tain significantly higher amounts of cholesterol.4 Since both
adenine nucleotide transport protein and the Ca2+ carrier are
located in the inner mitochondrial membrane, the increased
cholesterol levels might affect the membrane fluidity and there
fore change the normal function of these carriers.
To summarize, from the studies of uptake and efflux of adenine
nucleotide in tumor mitochondria, it is clear that a transient higher
level of ADP is retained in the cytosolic compartment. In addition
to higher ADP efflux rates, lower ATP efflux rates in tumor
mitochondria result in a smaller ATP pool size and a higher ADP/
ATP ratio in the extramitochondrial space. As a consequence,
several key enzymes in the glycolytic pathway, such as hexokinase and phosphofructokinase, are stimulated to higher capacity
in the production of lactic acid, thus contributing to the elevated
rates of aerobic glycolysis as observed in many malignant tumor
lines.
REFERENCES
1. Barbour, R. L, and Chan, S. H. P. Mitochondrial ATPase activities of hepatoma
BW7756 and ascites tumor cells. Influence of Mg~ ions, free fatty acids and
couplers. J. Bid. Chem., 253: 367-376, 1978.
2. Barbour, R. L., and Chan, S. H. P. Characterization of the kinetics and
mechanism of the mitochondrial ADP-ATP carrier. J. Blot. Chem 256- 19401948, 1981.
3. Barbour, R. L., and Chan, S. H. P. Adenine nucleotide transport in hepatoma
mitochondria and its con-elation with hepatoma growth rates and tumor size
Cancer Res., 43: 1511-1517, 1983.
4. Bustamante, E., and Pedersen, P. L. High aerobic glycolysis of rat hepatoma
cells in culture: rote of mitochondrial hexokinase. Proc. Nati. Acad. Sci USA
74: 3735-3739, 1977.
5. Bustamante, E., Soper, J. W., and Pedersen, P. L. A high-yield preparative
method of isolation of rat liver mitochondria. Anal. Biochem., 80: 401-408
1977.
6. Chan, S. H. P., and Barbour, R. L. Regulation of mitochondrial adenine
nucleotide transport by cytosolic fatty acid binding protein. In: C. P. Lee, G.
Schatz, and L. Ernster (eds.), Membrane Bioenergetics, pp. 521 -532. Reading,
MA: Addison-Wesley, 1979.
7. Chan, S. H. P., and Barbour, R. L. Adenine nucteotide transport in hepatoma
mitochondria. Characterization of factors influencing the kinetics of ADP and
ATP uptake. Biochim. Btophys. Acta, 723: 104-113, 1983.
8. Chan, S. H. P., Lau, B. W. C., and Weber, L. Adenine nucleotide transport in
Monis hepatoma mitochondria. J. Nutr. Growth Cancer, 1: 117, 1983.
9. Das, S. K. The submitochondrial localization adenylate kinase: an enzymatic
marker for the inner surface of the outer membrane of lung mitochondria in
guinea pig. Biochem. Btophys. Res. Commun., 703: 1145-1148, 1981.
10. Duszynski, J., Savina, M. K., and Wojtczak, L. Effect of the divalent ionophore
A231 87 on the translocation of adenine cation nucteotides in liver mitochondria
FEBSLett., 86:8-13, 1978.
11. Eboli, M. L., Malmstrom, K., Galeotti, T., Lopez-Alarcon, L., and Carafoli, E.
Calcium transport and transkxatton of adenine nucteotides in mitochondria
from Morris hepatoma 3924A. Cancer Res., 39: 2737-2742, 1979.
12. Eisenthal, R., and Comish-Bowden, A. The direct linear ptot. A new graphical
procedure for estimating enzyme kinetic parameter. Biochem. J., 739: 715720, 1974.
13. Emmetot, P., and Bos, C. J. Studies on isolated tumor mitochondria. Oxidative
phosphorylation and fatty acid oxidation by rat hepatoma mitochondria. Br J
Cancer, 75: 373-385, 1961.
14. Feo, F., Canuto, R. A., Garcea, R.. Avogadroa Villa, M., and Celasco, M. Lipid
phase transition and breaks in the Arrhenius plots of membrane-bound en
zymes in mitochondria from normal rat liver and hepatoma AH-130. FEBS
4 B. W. C. Lau, L. Weber, M. B. Magio, and S. H. P. Chan, manuscript in
preparation.
4464
Lett., 72: 262-266, 1976.
15. Feo, F., Canuto, R. A., Garcea, R., and Gabriel, L. Effect of cholesterol content
on some physical and functional properties of mitochondria isolated from adult
rat liver, fetal liver, cholesterol-enriched liver and hepatomas AH-130 3924A
and 5123. Btochim. Biophys. Acta, 473:116-134,1975.
16. Gomez-Puyou, A., Tuena De Gomez-Puyou, M., Klapp, M., and Carafoli, E.
The effect of calcium on the translocation of adenine nucteotides in rat liver
mitochondria. Arch. Biochem. Btophys., 794: 399-404,1979.
17. Hektt, H. W., and Schwalbach, K. The participation of GTP-AMP-P, transferase
in substrate tevel phosphate transfer of rat liver mitochondria. Eur J Biochem
7:199-206,1967.
18. Hess, B., and Chance, B. Metabolic control mechanisms. VI. Chemical events
after glucose addition to ascites tumor cells. J. Biol. Chem 236- 239-246
1962.
19. Kaschnitz, R. M., Hatefi, Y., and Morris, H. P., Oxidative phosphorylation
properties of mitochondria isolated from transplantaWe hepatoma. Biochim
Biophys. Acta, 449: 224-235,1976.
20. Klingenberg, M. Metabolite transport in mitochondria. An example for mtraceilular membrane function. Essays Biochem., 6:117-159,1970.
21. Klingenberg, M., and Appel, M. A possible mechanism for respiration-depend
ent efflux of Mg tons from liver mitochondria. FEBSLett., 779:195-199
1980
22. Lau, B. W. C., and Chan, S. H. P. Efflux of adenine nudeottoes in hepatoma
mitochondria. Fed. Proc., 42:2012,1983.
23. Lauquin. G. J. M., Devaux, P. F., Bienvenue, A., Villiers, C., and Vignais, P. V.
Spin-labeled acyl atractytoskte as a probe of the mitochondrial adetiosine
diphosphate carrier. Asymmetry of the carrier and direct lipid environment
Biochemistry, 76:1202-1208,1977.
24. Lehninger, A. L., Carafoli, E., and Rossi, C. S. Enzyme-linked movements in
mitochondrial systems. Adv. Enzymol., 29: 259-320,1967.
25. Lemer, E., Shug, A. L., Elson, C., and Shrago, E. Reversible inhibition of
adenine nucleotide transtocatton by long chain fatty acyl coenzyme A esters
in liver mitochondria of diabetic and hibernating animals. J. Biol Chem 2471513-1519,1972.
26. Liepkalns, B. A., and Spector, A. A. Alteration of the fatty acid composition of
Ehrlich ascites tumor cell lipids. Biochem. Biophys. Res. Commun., 63-10431047,1975.
27. Meli, J., and Bygrave, F. L. The rote of mitochondria in modifying calciumsensitive cytoplasmic metabolic activities. Biochem. J., 728:415-420,1972.
28. Pfaff, E., Heldt, H. W., and Klingenberg, M. Adenine nucteotide transtocatton
of mitochondria. Kinetics of the adenine exchange. Eur. J. Chem 70- 484493,1969.
29. Pfaff, E., and Klingenberg, M. Adenine nucleotide translocatton of mitochondria.
1. Specificity and control. Eur. J. Biochem., 6: 66-79,1968.
30. Pfaff, E., Klingenberg, M., and Heldt, H. W. Unspecific permeation and specific
exchange of adenine nucleotides in liver mitochondria. Biochim. Biophys Acta
704:312-315,1965.
31. Racker, E. Bioenergetics and the problem of tumor growth. Am Sci 60-5663,1972.
32. Rasmussen, H. Cell communication, calcium ion, and cyclic adenosine monophosphate. Science (Wash. DC), 770: 404-412,1970.
33. Reed, P. W., and Lardy, H. A. A23187: a divalent cation ionophore. J. Btol
Chem., 247: 6970-6977,1972.
34. Scholnick, P., Lang, D., and Racker, E. Regulatory mechanisms in carbohy
drate metabolism. IX. Stimulation of aerobic glycolysis by energy-linked ion
transport and inhibition by dextran sulfate. J. Biol. Chem., 248: 5175-5182
1973.
35. Siliprandi, D., Toninelto, A., Zoccarato, F., and Siliprandi, N. Is there a common
binding center in the ADP-ATP carrier for substrate and inhibitors. Biochem
Btophys. Res. Commun., 78: 23-27, 1977.
36. Vignais, P. V., and Lauquin, G. J. M. Mitochondrial adenine nucleotide transport
and its rote in the economy of the cell. Trends Biochem. Sci., 4: 90-92,1979
37. Vignais, P. V., Lauquin, G. J. M., and Vignais, P. M. Kinetics and binding
properties of ADP/ATP carrier as a function of the carrier environment. In: L.
Packer and A. Gomez-Puyou (eds.), Bioenergetics, Biogenesis and Membrane
Structure, pp. 109-125. New York: Academic Press, Inc., 1976.
38. Warburg, O. The Metabolism of Tumors. Translated by F Dicker LondonConstable, 1930.
39. Watanabe, K., and Kubo, S. Mitochondrial adenylate kinase from chicken liver
Purification, characterization and its cell-free synthesis. Eur. J. Biochem 723587-592,1982.
40. Weber, G., Stubbs, M., and Morris, H. P. Metabolism of hepatomas of different
growth rates in situ and during ischemia. Cancer Res., 37: 2177-2183,1971
41. Williams, R. J. P. Calcium tons: their ligands and their functions. Brachem. Soc
Symp., 39:133-138.1974.
42. Wojtczak, L. Effects of tang-chain fatty acids and acyl-CoA on mitochondrial
permeability, transport, and energy-coupling processes. J. Btoenerq. Btomemb., 8:293-311,1976.
43. Woktegtoris, G., Voss, S., and Shrago, E. Restoration of normal function of
the adenine nucteotide translocase from hepatoma mitochondria after its
reconstitution into a liposome system. Fed. Proc., 42:1947,1983.
CANCER
RESEARCH
VOL. 44
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1984 American Association for Cancer Research.
Efflux of Adenine Nucleotides in Mitochondria from Rat Tumor
Cells of Varying Growth Rates
Brad W. C. Lau and Samuel H. P. Chan
Cancer Res 1984;44:4458-4464.
Updated version
E-mail alerts
Reprints and
Subscriptions
Permissions
Access the most recent version of this article at:
http://cancerres.aacrjournals.org/content/44/10/4458
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 18, 2017. © 1984 American Association for Cancer Research.