[CANCERRESEARCH42, 1078-1081, March19821
Nuclear Catalyzed Antibiotic Free Radical Formation
Nicholas R. Bachur, Malcolm V. Gee,and RosalindD. Friedman
Laboratory of Clinical Biochemistry, Baltimore Cancer Research Program, Division of Cancer Treatment, National Cancer Institute, Balitmore Maryland 21201
Nucleiandmicrosomeswerepreparedfromtissuesof maleSprague
Dawleyrats (150 to 200 g) fed ad ilbitum.Liver nucleiwere isolated
ABSTRACT
Nuclei isolated from rat liver, heart, and kidney catalyze
oxygen consumption in the presence of reduced pyridine nu
cleotide (NADPH) and quinane or quinane-imine antibiotics
such as Adriamycin, daunarubicin, actinomycin 0, mitomycin
C, and streptonigrin. The Kmand Vmaxvalues for NADPH were
2.4 x 10-@ M and 3 x 10@ mol 02 per mm per mg protein and
Km values
for the antibiotics
ranged
from
1 .4 x 1 O@ M to 5.9
x 10-6 M. Metabolism of the anthracycline antibiotics, i.e.,
reductive glycasidase reaction, occurs in reaction mixtures
after all oxygen is consumed. During the reaction, free-radical
species of Adriamycin and daunarubicin are detectable by
electron paramagnetic reasonance spectrometry. These ob
servations indicate that some cytotoxic antibiotics can be ac
tivated to a free-radical state at the site where damage to
nuclear DNA may result.
It is well established that quinone and quinone-imine contain
ing antibiotics damage DNA in living cells and produce such
effects as DNA strand breaks and strand cross-linking, cis
chromatid exchange, and ultimately mutagenic action (21 , 22).
Damage to cellular DNA appears to require the biological
activation of these antibiotics presumably through a reduction
process (1, 2, 17, 23). In our previous work, we have shown
that quinone and quinone-imine containing anticancer antibiot
ics produce free radicals and are catalytically converted to
free-radical intermediates by microsomes. Subsequently, we
demonstrated that the purified enzymes NADPH cytochrome
reductase
weremincedfinelyandhomogenizedina VirTis45 homogenizer(VirTis
Co., Gardiner, N. V.) for 2 30-sec intervals at medium speed; the
homogenate
was filtered
through
4 layers of cheesecloth
(4) and xanthine
oxidase
(1 4) catalyze
the
drug free-radical formation through a single-electron reduction
process.
We have postulated the ‘
‘site-specific
free-radical' ‘
concept
as a mechanism for the nuclear DNA damage caused by these
free radical-producing antibiotics (3). However, we have been
concerned over the physical distance from the microsomal
enzymes to the target DNA and whether highly reactive free
radical forms can traverse this physical distance. Since nuclei
themselves contain NADPH cytochrome P.450 reductase (5)
and xanthine oxidase (6), as well as glutathione (19), catalase
Verhoeven and DeMoor (20). Liver microsomes were prepared as
described previously by Omura and Takesue (13).
P388cellsweremaintainedin maleBALB/c x DBA/2 F1(hereafter
called CD2F1) mice. For nuclei preparations,
cells were harvested
once in phosphate-bufferedsaline,suspendedin a hypotonicbuffer
(1 0 [email protected],
1 0 mM Tris,
1 .5 mM MgCI2, pH 7.4), and stirred
Elvehjem homogenizer (50 strokes). Cell breakage was checked by
light microscopy. The homogenate was made 0.25 M in sucrose and 5
for
For electron microscopy,rat liver nuclei, preparedas described
above, were centrifuged and the pellet was prepared according to the
method of Sanel (1 6) except that Karnofsky's
fixative was used and
the washeswere with 1% sucrosein 0.1 M sodiumcacodylateand
0.01 % potassium chloride. By this method, the nuclei we isolated were
freeof microsomalcontamination(Fig. 1).
Oxygenconsumptionwasdeterminedwitha Clark-typeelectrodein
a YellowSpringsInstrumentCo.Model53 BiologicalOxygenMonitor.
All measurementswere made at 37°.The reaction mixtures (final
volume,1 ml)contained0.2 Mpotassiumphosphatebuffer(pH 7.0; in
somecases,pH 7.5), 5 mMNADPH(or NADHwhereindicated),1o@
M to 1 o-@ M drug,
and
0.3
to 1 .0 mg nuclear
protein.
The
buffer
was
aeratedin thechamberfor 3 mm,nucleiwereadded,andtheelectrode
was placedin contactwith the solution.After 1 mmfor equilibration,
NADPHwasinjectedintothe reactionmixture.AfterI additionalmmof
equilibration, the drug was introduced into the system, and drug
induced oxygen consumption was recorded for 1 to 3 mm during the
linearphaseof the reaction.
EPR' spectra were obtainedwith a Varian E-i 09 CenturySeries
spectrometer
in a rectangular
dual cavity operating in TE-i 04 mode.
Strongpitch at g 2.0028 in the referencecavitywasusedto evaluate
the g values. Nuclei (containingapproximately1 mg protein) were
preincubatedon icefor I 5 mmwith0.2 Mpotassiumphosphatebuffer,
pH 7.0, containing 0.1% Triton N-i 01 . NADPH (final concentration, 5
mM)and drug (final concentration, 1 mM)were added, the mixture was
deoxygenated
an EPRtube.Thetemperaturewasmaintainedat 37°.
by bubbling with N2, and an aliquot was transferred
to
anticancer agents.
AND METHODS
Adriamycin HCI,daunorubicin HCI,actinomycin D, streptonigrin, and
mitomycin C were supplied by the Drug Development Branch, National
CancerInstitute.NADPHand NADHwereobtainedfrom P-LBiochem
icals, Milwaukee, Wis. Triton N-i 01 was purchased from Sigma Chem
ical Co., St. Louis,Mo.
Received September 16, 1980; accepted December 9, 1981.
1078
gently
for 15 mmat 4°.The suspensionwasthen homogenizedin a Potter
ability of isolated, purified nuclei to catalyze free radical for
mation with quinone and quinone-imine
7
days after i.p. implantation of 1o@cells. The tumor cells were washed
(1 8), and superoxide dismutase (1 0), we have investigated the
MATERIALS
and the
unfilterablematerialwas rehomogenized;and the secondhigh-speed
centrifugationwasomitted.Kidneynucleiwerepreparedaccordingto
mM in CaCI2, and the method of Ohly et a!. (1 2) was followed
isolation of nuclei.
INTRODUCTION
P-450
according to the method of Kasper (7). Heart nuclei were prepared as
described by Nair eta!. (11) with the following modifications: the hearts
RESULTS
Unlike microsomes, which consume oxygen endogenously
in the presence of NADPH, rat liver nuclei have little if any
measurable endogenous oxygen consumption (Table 1). How
ever, when the quinone antibiotics Adriamycin, daunorubicin,
streptonigrin, or mitomycin C or the quinone-imine drug acti
,The
abbreviation
used
is:EPR,
electron
paramagnetic
resonance.
CANCERRESEARCHVOL. 42
Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1982 American Association for Cancer Research.
@
@
0
t@
Nuc!ear Cata!yzed Antibiotic Free Radica!s
murine leukemia cells did not show a preference.
During the oxygen consumption reaction with daunorubicin
and rat liver nuclei, we saw no evidence of daunorubicin
metabolism (Chart 1). However, once anaerobiosis occurs (ap
proximately 20 mm), the metabolite deoxyaglycone appears as
a reaction product while the daunorubicin disappears (data not
shown).
We examined the nuclear reactions for free radical formation
by EPA spectroscopy. When liver nuclei and NADPH are in
cubated with daunorubicin or Adriamycin, characteristic free
radical signals of these antibiotic free radicals (with respective
g values of 2.0037 and 2.0035) are observed (Chart 2). Free
,,
@
;‘.@‘
.
,. a
@
@:,@. ,,.;@‘
radical signals are absolutely dependent on the presence of
active nuclei, reduced pyridine nucleotide, and antibiotic. Con
trol reactions lacking any component or containing heat-mac
tivated nuclei failed to yield the free radical signals.
Heart and kidney nuclei also generate
.,
@
p j.'—
@
drug free-radical
signals of characteristic g values with daunorubicin as sub
•@
@@:;P'
•
.‘
strate. Whereas we have obtained
nuclear generated
free
radical signals for streptonigrin, we have not observed signals
for mitomycin C or actinomycin D. This is probably related to
;@X;@b@
I'1'@
,..,@,.
the free radical half-life or other free radical characteristics
of
these agents.
@
‘V.'
•:
@
@:
..@‘
-,
?ft@#@a―.-%@l@...
.,@,
@4v,,4@i.@t,:::i;@
..
.
‘
-S@@•‘
•@
-4'DISCUSSION
4'
‘S.
.,,d6...
‘
@@i:@i@,.-:.':
The mechanism or mechanisms by which antibiotics cause
.;.i;'::'.-.
Fig. 1. Electron micrograph of isolated ral liver nuclei Nuclei were prepared
for electron microscopy as described in Materials and Melhods
x 7800
nomycin D are added to nuclei, rapid oxygen utilization occurs.
Oxygen consumption is totally dependent on 3 factors: me
tabolizing nuclei, the electron donor NADPH, and the drug.
Heat treatment of nuclei at 100°for 10 mm or freezing at —
20°
overnight inactivates them, but activity is stable for 24 to 36 hr
if the nuclei are kept at 4°.This contrasts to microsomal activity
which is stable for indefinite periods at —20°.
Since oxygen consumption is a convenie@@
measurement of
superoxide
mM NADPH, rat liver nuclei, and 5 x I 0@ M antibiotic.
With daunorubicin
as substrate, the rat liver nuclei have a
distinct preference for NADPH as cofactor although they will
utilize NADH (Table 2).
Characteristics of the reactivities of both quinone and qui
none-imine antibiotics with liver nuclei indicate that the system
is saturable and follows Michaelis-Menten
protein)Nuclei
0Nuclei(control)
0Nuclei+ NADPH (control)
0Nuclei+ daunorubicin
11.2Nuclei
+ NADPH + daunorubicin
9.33Nuclei
+ NADPH + Adriamycin
10.6Nuclei
+ NADPH + actinomycin D
9.06Nuclei
+ NADPH + mitomycmnC
+ NADPH + streptonigrin
1.5Table
from
5.9
x
I O_6 M to 1 .4 x
NADPH.Oxygen
daunorubicin,
3).
In
the case of streptonigrin, the Kmis 23-fold lower than the other
drugs, but the V@, is comparable.
Nuclei from other rat tissues and from P388 murine leukemia
cells were isolated and studied for their abilities to catalyze
drug-mediated oxygen consumption. We examined the nuclei
for pH optima and cofactor preferences. All nuclei show optimal
oxygen consumption at pH 7.0 to 7.5 in potassium phosphate
buffer. When tested with NADPH and NADH, nuclei from normal
concentrations
of NADH
(X
or
Pyridine
protein)NADPH
nucleotide
Km (x 1o@ M)
3.0NADH
2.4
42.0
1 0-8 mol 02 uti
lized/min/mg
5.2
Table 3
antibioticsReaction
Apparent nuclear kinetic constants for
mixtures contained 0.2 M potassium phosphate buffer (pH 7.0), 5
mM NADPH,
rat liver nuclei, and varying
concentrations
drugs.V,,@
of the respective
(x 1o@ mol 02 uti
protein)Daunorubicin
Drug
Adriamycin
Actinomycin D
Mitomycin C
0.88a
Streptonigrin
tissues
prefer
NADPH
ascofactor,
butnuclei
from
theP388
MARCH
rat liver nuclei, and varying
consumption was determined as described in ‘Materials
Methods.―Vm@@x
and
kinetics. Saturation
1 O@ M (Table
1
2Apparent
cofactorsReaction
nuclear kinetic constants for pyridine nucleotide
0'@'M
mixtures contained 0.2 Mpotassium phosphate buffer (pH 7.0), 1
is achieved at drug concentrations approximating 10@ M ex
cept for streptonigrin which saturates about one-tenth of that
concentration. These data fit Lineweaver-Burk plots to yield
Km5 ranging
02
o@mol
consumption (x 1
02
utilized/mm/mgSample
formation through free radical formation (2), we
have used oxygen consumption to quantify the reactivity of the
nuclei for several characteristics.
@
Table 1
Antibiotic stimulation of nuclear oxygen consumption
Reaction mixtures contained 0.2 M potassium phosphate buffer (pH 7.0), 5
Mean
Km(M)
1.4
1.3
1.2
1.4
5.9
±0.67
±0.81
±1.2
±0.07
±1.9
x
x
x
x
x
Iized/min/mg
10@4a
10@
iO@
iO@
10°
1.6
1.5
1.8
1.1
1.1
±0.61
±0.93
±0.93
±1.0
±
± S.D. of 3 experiments.
1982
Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1982 American Association for Cancer Research.
1079
@1@
N. R. Bachur et a!.
r
C
0
m
U,
C,
9@
8m
z
7C
w
0
z
x
6-i
5U)
z
4
3
0
I-
U
a.
LI
MINUTES
Chart 1. Relationship of daunorubicin-stimulated oxygen consumption and
daunorubicin biotransformation by rat liver nuclei. A reaction mixture with a final
volume of 2.0 ml was set up in the oxygen electrode vessel containing: 0.2 Ii
potassium phosphate buffer, pH 7.0; 5 m@,iNADPH; 1o@ M daunorubicin; and
rat liver nuclei (2 mg protein). At various times after injection of the daunorublcin,
1O-@d
aliquots were removed and placed In tubes containing 40 @sI
of ethanol. Ten
@l
from each tube was spotted on a thin-layer chromatographic plate, and the
plate was developed once in CHCI3:methanol:H2O(80:20:3). The fluorescent
spotsmigratingwiththesolventfront(daunorubicin
aglycone)werescrapedfrom
the plate, placed in 2.0 ml of 0.3 M HCI:50% ethanol, and the fluorescence of the
extracted aglycones were determined on an Aminco fluorimeter (excitation, 470
nm; emission, 585 nm).
w@—\r'--'--―@@---.-------'-CONTROL
LIVER NUCLEI + D
/@
@
\\ ,,@/
@/\\
LIVER
NUCLEI
+A
@@@@HEART
NUCLEI+D
@
KIDNEY
NUCLEI+D
_‘__\__@—
STANDARD
PITCH
Chart 2. EPR spectra of anthracyclines in the presence of nuclei and NADPH.
Reaction mixtures contained the same components as those used for oxygen
consumption with the following modifications. Nuclei were preincubated on ice
for 15 mm with 0. 1% Triton N-i 01 , and the entire reaction mixtures were gassed
with N2 prior to placement in the EPR tube. Conditions for EPR are given in
“Materials
and Methods.' ‘
Modulation amplitude for liver nuclei, 1OG;modulation
amplitude for heart and kidney nuclei, 2G. A, Adriamycin; D, daunorubicin.
damage to DNA and chromosomes has been an enigma. Cer
tainly it is clear that the quinone and quinone-imine containing
antibiotics cannot produce DNA damage without bioactivation.
Investigators have shown that mitomycin C and streptonigrin
require reductive activation to produce DNA damage or cyto
toxicity (17, 23). More recently, investigations have shown that
those agents, as well as the anthracyclines, will damage DNA
if reduced
chemically
(8, 9). Our studies
show that all of these
agents are capable of being reduced to free radicals through
microsomal enzymes and specifically
P-450
reductase,
xanthine
oxidase,
by NADPH-cytochrome
and other single-electron
reducing flavoproteins (15). We have questioned whether the
life time of the antibiotic free radicals would be sufficiently long
to allow these very reactive compounds to traverse the distance
between a microsomal site of origin and the target nuclear
Chart3. Possiblereactionpathwaysof quinone-typedrugsin thecell.FPREO,
reduced flavoprotein; FP0@,oxidized flavoprotein.
The ‘
‘site-specific
free-radical' ‘
process is based on 2 struc
tural characteristics of these cytatoxic antibiotics. First, each
of these antibiotics is structurally adapted to bind specifically
at some macramolecular receptor site such as DNA where it
can inflict damage to the host cell. Secondly, each of these
molecules is inherently capable of single-electron reduction to
a free radical state to activate the antibiotic molecule. Since
many of the quinone and quinone-imine containing antibiotics
are known to bind to DNA specifically and also are known to
inflict their damage at the DNA level, we presume that DNA
binding sites are more or less common to these antibiotics. It
is also possible, however, that other specific sites exist within
the host cell where the antibiotics can inflict damage at a
primary level. If the antibiotic free radicals generate secondary
free radicals, damage can be inflicted at secondary levels by
such activated molecules as superoxide, hydroxyl radical, hy
drogen peroxide, etc.
We have shown that nuclei from normal rat tissues and from
murine leukemia cells possess the enzymatic activity to activate
the quinone and quinone-imine antibiotics to free-radical
states. Since the enzymatic activity is located immediately at
the site of DNA residency, it is quite feasible that antibiotic
enters the nucleus, is activated at that site, and then reacts
with the resident DNA and inflicts cytotoxic damage.
We can postulate possible reaction mechanisms of free
radical-generating
antibiotics
within the cell (Chart 3). The
quinone or quinone-imine drug crosses the cell membrane, and
once inside, it may take 2 paths. One path is to travel directly
to the nucleus and be reduced by flavoproteins located in the
nuclear stroma. The result of this enzymatic reduction is the
formation of semiquinane free radical intermediates which may
react with DNA directly, or the drug free radical can react with
oxygen to farm superoxide free radical and subsequently hy
drogen peroxide and hydroxyl radicals. These secondary-level
free radicals may also damage DNA or other nuclear structures.
The regenerated quinane group is then free to repeat this
single-electron reduction as a cyclic process. The antibiotic
may also bind to DNA or other nuclear structures as an adduct,
and this reacted baund quinone or quinone-imine system may
be available for continued single-electron reduction to free
radical and further secondary free-radical generation in situ. In
the cytoplasm, a second pathway for quinone or quinone-imine
DNA.
activation is available. Endoplasmic reticulum flavoproteins or
mitochondrial enzymes also catalyze the formation of drug free
radicals. These radicals may react primarily with cellular cam
ponents such as cell membranes, or they may generate sec
1080
CANCERRESEARCHVOL. 42
Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1982 American Association for Cancer Research.
Nuc!ear Cata!yzed Antibiotic Free Radicals
ondary free radicals which may also damage cellular compa
agents. Biochem. Biophys. Res. Commun., 76: 705—7i0, 1977.
nents.
Itmay
bepossible
that
some
semiquinone
free
radicalsMichelsonet al. (eds),SuperoxideandSuperoxideDismutases.NewYork:
i 0. Michelson, A. M. Toxicity of superoxide radical anion, p. 200. In: A. M.
produced at the endoplasmic reticulum may travel into the
nucleus and react as noted above.
1 1. Nair, K. G., Rabinowitz, M., and Tu, M. C. Characterization of the ribonucleic
ACKNOWLEDGMENTS
12. OhIy, K. W., Mehta, N. G., Mourkides, G. A., and Alivisatos, S. G. A. Isolation
We wouldlike to thankDr. WilliamCasparyfor assistancein obtainingthe
EPR spectra and Dr. Frances Sanel for assistance with electron microscopy of
isolated nuclei.
REFERENCES
1. Bachur, N. R., Gee, M. V., and Gordon, S. L. Enzymatic activation of
actlnomycln D to free radical state. Proc. Am. Assoc. Cancer Res., 19: 75,
1978.
2. Bachur,N. R., Gordon,S. L, and Gee, M. V. Anthracyclineantibiotic
augmentation
of microsomal
electrontransportandfreeradicalformation.
Mol.Pharmacol.,13:901-910, 1977.
3. Bachur, N. R., Gordon, 5. L., and Gee, M. V. A general mechanism for
microsemal activatIon of quinone anticancer agents to free radicals. Cancer
Re.., 38: 1745—1750, 1978.
4. Bachur,N.R.,Gordon,S.L, Gee,M.V.,andKon,H.NADPH
cytochrome
P.450 reductue activation of quinone anticancer agents to free radicals.
Proc.Nati.Aced.Sd.U.5. A.,76:954-957,1979.
5. Bartok, 0. M., Galeotti, T., and Azzi, A. Production of superoxide anions
and hydrogen peroxide in Ehrlich ascites tumor cell nuclei. Biochim. Bio
pI@e. Acts, 497: 622—626, 1977.
Academic Press, Inc., 1977.
acid synthesized in an isolated nuclear system from rat heart muscle.
Biochemistry, 6: i 898—1
902, 1967.
and characterizationof histonesfrom Ehrlichascitestumorafter in vitro
labeling with lysine-'4C. Arch. Biochem. Biophys., 118: 631 —636,1967.
13. Omura, T., and Takesue, S. A new method for simultaneous purification of
cytochrome b5 and NADPH-cytochrome c reductase from rat liver micro
somes.J. Biochem.(Tokyo),67: 249—257,
i 970.
i 4. Pan, S., and Bachur, N. R. Xanthine oxidase catalyzed reductive cleavage
of anthracycline antibiotics and free radical formation. Mol. Pharmacol., 17:
95—99,1980.
15. Pan, S.. Gutierrez, P., and Bachur, N. R. Flavoprotein catalysis of anthra
cycline antibiotic reductive cleavage, free radical formation and free radical
conversions. Fed. Proc., 39: 310, 1980.
16. Sanel, F. Studies of neoplastic
myelomonocytic
cells in BALB/c
mice
producing infectious C-type viruses. Cancer Res., 33: 671 —678.1973.
17. Schwartz, H. S., Sodergren, S. E., and Philips, F. S. Mitomycin C: chemical
and biological studies on alkylation. Science (wash. D. C.), 142: 1181—
1183, 1963.
18. Siebert, G.. Humphrey, G. B., Themann, H., and Kersten, W. Neue Kriterien
zur Beurteilung isolierter Zellkerne. Hoppe-Seyler's Z. Physiol. Chem., 340:
51—72,
1965.
19. Srivastava, S., and Beutler, E. The transport of oxidized glutathione from the
erythrocytesof variousspeciesin the presenceof chromate.Biochem.J.,
6. Hashimoto, S. Developmental changes in rat liver xanthine oxidase(s) and
Its Intracellular distrIbution. Agric. Biol. Chem., 38: 1697—1
702, 1974.
114:833—837,1969.
20. Verhoeven, G., and DeMoor, P. Kinetic data on the formation of 17$hydroxy-5a-androstan-3-one in renal nuclei from male and female rats.
Endocrinology, 91: 54—64,1972.
7. Keeper, C. B. IsolatIon and properties of the nuclear envelope. Methods
21 . Vig, B. K. Genetic toxicology of mitomycin C, actinomycins, and Adriamycin.
Enzymol., 31A: 279—282,1974.
Mutat.Res.,49: 189—238,
1977.
8. Lown,J. W.,andSim,S. K. Studiesrelatedto antitumorantibioticsVIII.
Cleavage
ofDNAbystreptonigrin
analogues
andtherelationship
toantineo
plasticactivity.Can.J.Blochem.,
54:446—452,
1976.
22. Whang-Peng, J., Leventhal, B. G., Adamson, J. W., and Perry, S. The effect
9. Lown,J. W.,Sirn,S. K., Majumdar,K. C.,andChang,R.Y.Strandscission
23. White, H. L, and White, J. R. Lethal action and metabolic effects of
streptonigrin on Escherichia coli. Mol. Pharmacol., 4: 549—565,1968.
ofDNAbyboundAdriamycln
anddaunorubicin
inthepresence
ofreducing
MARCH 1982
of daunomycin on human cells in vivo and in vitro. Cancer Res., 23: 113—
121, 1969.
1081
Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1982 American Association for Cancer Research.
Nuclear Catalyzed Antibiotic Free Radical Formation
Nicholas R. Bachur, Malcolm V. Gee and Rosalind D. Friedman
Cancer Res 1982;42:1078-1081.
Updated version
E-mail alerts
Reprints and
Subscriptions
Permissions
Access the most recent version of this article at:
http://cancerres.aacrjournals.org/content/42/3/1078
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 16, 2017. © 1982 American Association for Cancer Research.