[CANCER RESEARCH 51.6066-6072. November 15.
1991]
Ascorbic Acid Enhances the Effects of 6-Hydroxydopamine
dependent DNA Strand Breaks and Related Processes in the Neuroblastoma
Cell Line SK-N-SH
Gernot Bruchelt,1 Ingrid U. Schraufstätter, Dietrich Niethammer, and Charles G. Cochrane
Department of Immunology, Research Institute ofScripps Clinic, La Jolla, California 92037 [G. B., I. U. S., C. G. C.J, and Department of Hematology and Oncology,
Children's Hospital, University of Tuebingen, Ruemelinstrasse 23, D-7400 Tuehingen, Federal Republic of Germany ¡I).N.j
ABSTRACT
Neuroblastoma cells accumulate ascorbic acid and iron. It was hypoth
esized that these features could be exploited for sensitizing neuroblastoma
cells for therapy in combination with reactive oxygen intermediates. In
the present study the effects of 6-hydroxydopamine (6-OIIDA) and IM),
on metabolic parameters critical for cell survival were investigated in
cells with low and high ferritin content in the presence and absence of
ascorbate. Human neuroblastoma SK-N-SH cells were pretrcated with
100 n\\ l i'SO, and 10 JÃŒM desferrioxamine, respectively, for 24 h yielding
cells with different ferritin contents. The effects of 6-OIIDA and H2O2
(25 nM-250 m\i ) in the absence and presence of I HIMascorbic acid on
DNA strand break formation, activation of poly(ADP-ribose) polymer-
ase, and finally decrease in NAD* and ATP concentration were investi
gated. All these parameters were influenced by 6-OHDA and IM), in a
concentration-dependent manner in a similar way. The effects were most
pronounced in ferritin-rich cells and in the presence of ascorbic acid.
Using isolated CCC PM2 DNA, 6-OHDA and ascorbic acid caused
strand breaks that were prevented in the presence of mannitol or desfer-
rithiocine. ! M >rimdialed strand breaks were observed only in the pres
ence of ascorbic acid. Based on these data and data published by others
a model explaining the deleterious effects of ascorbic acid on neuro
blastoma cells is presented. It is suggested that continuous application of
a high dosage of ascorbic acid might be a useful approach in neuro
blastoma therapy.
INTRODUCTION
Neuroblastoma stage IV has a poor prognosis (1). Therefore,
new therapeutic approaches are currently under clinical inves
tigation. Among them is autologous bone marrow transplanta
tion in combination with different purging procedures in order
to remove contaminating neuroblastoma cells (2-4). In one of
these purging systems bone marrow cells were incubated with
100 MM 6-OHDA2 in combination with 1 m\i AA. It was
assumed that neuroblastoma but not bone marrow cells would
take up the catecholamine-analogous compound 6-OHDA (5).
Inside the cells, H2O2 formed during the oxidation of 6-OHDA
(6) would destroy the neuroblastoma cells specifically (7, 8).
Furthermore, AA included in a 10-fold molar excess would
enhance these effects by acting as redox cycler (9. 10). However,
some of these theoretical considerations as arguments for purg
ing procedures proved to be incorrect. Since the oxygen con
centration present in the purging mixture (about 250 p\i) is
limited, most of the AA is not consumed by the redox cycling
Received 8/7/91; accepted 9/9/91.
The costs of publication of this article were defrayed in part by the payment
of page charges. This article must therefore be hereby marked advertisement in
accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1Recipient of a grant of Förderverein Leukämiekranker Kindcr.e.V.. Tuebin-
gen. Federal Republic of Germany. Supported by the Deutsche Krebshilfe. Grant
W54/89/NÌ4.To whom requests for reprints should be addressed. Permanent
address: Department of Hematology and Oncology. Children's Hospital. Univer
sity of Tuebingen. Ruemelinstrasse 23, D-7400 Tuebingen. Federal Republic of
Germany.
2The abbreviations used arc: 6-OHDA, 6-hydroxydopamine: AA. ascorbic
acid: DFO, desferrioxamine: CCC PM2 DNA, closed covalently circular PM2
DNA; PBS, phosphate-buffered saline; GSH, glutathione.
6066
and H2O2 on Iron-
process. Nevertheless, the cytotoxic effects of the combination
6-OHDA/AA were the more pronounced the more AA was
present (9). This means that AA must have caused some toxic
effects irrespective of its function as redox cycler. Furthermore,
oxidation of 6-OHDA occurs rapidly so that most of the reac
tive oxygen intermediates are already formed in the incubation
medium before a significant amount of 6-OHDA is taken up
selectively by neuroblastoma cells (11). Moreover, only a small
percentage of neuroblastoma cells can take up catecholamines
selectively (12); therefore both neuroblastoma and hemato-
poietic stem cells are equally exposed to the reactive oxygen
compounds. Neuroblastoma cells proved to be more vulnerable
to 6-OHDA/AA than hematopoietic cells (9, 11) although the
reason for this is unknown. We supposed that AA might sen
sitize neuroblastoma cells in concert with iron. Iron is found in
these cells in large amounts incorporated in ferritin (13, 14).
AA is concentrated in sympathetic nervous tissue (15, 16) and
should therefore accumulate in neuroblastoma cells if provided
in significant amounts. The potential cytotoxic effects of AA
alone or in combination with cytostatic drugs have been de
scribed for neuroblastoma and other cells (17-19). However,
some confusion exists since AA can act either as antioxidants
(e.g. Refs. 20 and 21) or as prooxidants, the latter especially in
combination with iron (22-24). Iron stored as ferritin is de
prived of its potential cytotoxic effect, but it can be released
from ferritin yielding ferrous iron (Fe2+) by many agents, among
them 6-OHDA and AA (25-28). Fe2+ catalyzes the formation
of the highly toxic OH radicals from H2O2 in the Fenton
reaction (29). In accordance with these findings, a mixture of
H2O: and ferritin was shown to produce OH radicals in the
presence but not in the absence of AA (30). AA and ferritin,
acting together, could therefore be an effective enhancer system
for therapeutic regimens that are operative via generation of
H2O2. The most sensitive cellular target structure for H2O2 is
obviously the DNA (31-34). H2O2 causes DNA strand breaks
that are followed by the activation of poly(ADP-ribose) polym-
erase; this enzyme consumes NAD"1",a process that is subse
quently associated with a loss of ATP (35, 36).
The aim of our study was to investigate the influence of AA
on the effects of 6-OHDA and H2O2 on these DNA-dependent
pathways using neuroblastoma cells with high and low ferritin
content in order to find a better rational basis for a potential
application of AA in the therapy of this tumor.
MATERIALS AND METHODS
Materials
Ethidium bromide, 6-hydroxydopamine, ascorbic acid, and 1,10-
phenanthroline were from Sigma, St. Louis, MO; CCC PM2 DNA was
from Boehringer Mannheim. Mannheim, Federal Republic of Ger
many; mannitol was from Calbiochem-Behring Corp., La Jolla, CA;
desferrioxamine and desferrithiocin were from Ciba-Geigy. Basle, Swit
zerland; |'H]NAD* (specific activity, 25 Ci/mmol) was from ICN,
 ASCORBATE. FERRITIN. AND REACTIVE OXYGEN COMPOUNDS
Irvine, CA; L-[carboxyl-"C] ascorbic acid (specific activity, 10-30 mCi/
mmol) was from Amersham-Buchler, Braunschweig, Federal Republic
of Germany.
Determination of DNA Strand Breaks (Cell-free System)
(a) Ethidium-binding Assay According to the Method of Lown (37).
Briefly, CCC PM2 DNA (10 p\\ 2 Mg)was incubated with 10 M!PBS or
with additives dissolved in PBS. After addition of 180 M' of ethidium
bromide-containing alkaline buffer the samples were heated for 4 min
at 96°C.Twenty M' of this solution were used for electrophoresis (see
below); the remaining 180 M' were mixed with an additional 1 ml
alkaline buffer before measurement of ethidium bromide fluorescence
on a Perkin-Elmer model 650-15 (Hitachi-Perkin Elmer Instruments,
Mountain View, CA).
(b) Electrophoresis. Twenty M'of DNA containing solution (200 ng,
see a) were mixed with 4 M!loading buffer. Electrophoresis was carried
out using 15 /¿Iof this solution on a 0.75% agarose gel [375 mg of
agarose were melted in 50 ml Tris-borie acid-EDTA buffer (8. l g Tris
base plus 4.125 g boric acid plus 0.7 g disodium EDTA-2 H2O/liter)
containing 2.5 n\ of ethidium bromide (10 mg/ml)]. Electrophoresis
was carried out using Tris-borie acid-EDTA buffer containing 0.5 mg
ethidium bromide/liter at 100 V for about 2 h.
Determination of Ferritin Content in SK-N-SH Cells
Cells (1 x 106/4 ml) medium were seeded in 6-well plates (Greiner,
Nuertingen, Federal Republic of Germany). After 24 h, 40 M' PBS
(control), 40 /il DFO (10 MM),and 40 M!FeSO4 (100 MM),respectively,
were added. Ferritin was determined in cell lysates after 24 h using a
commercial immunoassay (Tandem-Fer; Hybritech, Hürth-Effern,Fed
eral Republic of Germany).
Uptake of |'"C]Ascorbic Acid
SK-N-SH cells, grown as monolayers in 12-well plates (Greiner) at
a density of 1 x IO6cells, were incubated with 5-100 MM[14C]ascorbic
acid in phosphate-buffered saline plus 1% bovine serum albumin for 30
min at 37°C.Cells were then washed three times with cold buffer
solution, and cell pellets were lysed with 500 M!0.3 M NaOH and mixed
with 5 ml scintillation cocktail (Unisolve; Zinsser, Frankfurt, Federal
Republic of Germany) before counting.
For calculation of the intracellular AA concentration, cell volume
was determined by centrifugation of SK-N-SH cells in a hematocrit
centrifuge; IO6SK-N-SH cells correspond to 2.14 /»I.
Generation of SK-N-SH Cells with Different Ferritin Content
SK-N-SH cells were grown as monolayers using RPMI 1640 con
taining glutamine (2 ITIM),gentamicin (50 mg/liter), and 10% fetal calf
serum (Gibco) in 75-cm2 flasks. At about 80-90% confluency, flasks
were incubated with either 10 MMDFO or 100 MMFeSO4 for 24 h. The
medium was removed and filtered. After trypsinization, the detached
cells were resuspended in their corresponding media. After 2-3 h
incubation at room temperature the cell suspensions were centrifuged
and the cell pellets were resuspended in modified Gey's buffer (147 mM
NaCl-5 mM KC1-1.5 mM CaCl2-1.9 mM KH2PO4-0.3 niM MgCl2- 1.1
mM Na2HPO4-10 mM HEPES-5 mM glucose, pH 7.3) at a concentra
tion of 1-3 x 106 cells/ml. Cell viability was 90-95% (trypan blue
exclusion test). Subsequently, DFO- and FeSO4-pretreated cells were
both divided and left untreated or were treated with 1 mM AA for 30
min. Incubation with H2O2 and 6-OHDA was done for 15 or 30 min
at 37°C.Samples were further processed according to the test protocols
described below.
Determination of DNA strand breaks in SK-N-SH cells was carried
out according to the method of Birnboim and Jevcak (38) using the
protocol described in Ref. 39. In some experiments 1,10-phenanthroline
was added to the incubation mixture 10 min prior to the addition of
H202.
Determination of poly(ADP-ribose) polymerase (NAD+, ADP ribo-
syltransferase, EC 2.4.2.30) was carried out according to the protocol
6067
IN NEUROBLASTOMA
described in Ref. 39 with the exception that the permeabilized cells
were incubated for 12 min.
High Performance Liquid Chromatography Analysis of NAD* and ATP
Cells (about 2 x 106/rnl) were incubated for 30 min with 6-OHDA
and H2O2. Supernatants were removed and the cell pellets were treated
with 220 M!8% perchloric acid. After centrifugation, 200 M' of each
supernatant were neutralized with 200 M!KOH/Tris. Supernatants were
collected and frozen at —¿20°C
for high performance liquid chromatog-
raphy analysis. Two humdred n\ were injected and separation was
carried out on a Du Pont C,8 column. A gradient of 100% 100 mM
potassium phosphate, pH 6.0, to 60% buffer plus 40% methanol was
applied with a flow rate of 1.3 ml/min. Absorbance was measured at
256 nm; flow rate was 1.3 ml/min. The ATP signal appeared at about
5.4 min; the NAD* signal appeared at about 17 min.
RESULTS
SK-N-SH cells grown as monolayers were split and treated
either with 10 MMDFO, 100 /¿M FeSO4, or no further additions
for 24 h. As shown in Fig. 1, DFO treatment reduced and
FeSO4 treatment strongly enhanced the ferritin content in these
cells. The mean ferritin content of untreated cells was 19.1 ±
7.0 (SD) ng/106 cells. After treatment with DFO, ferritin con
tent dropped to 6.10 ±3.67 ng/106 (P < 0.01), after treatment
with FeSO4 it increased to 62.62 ±15 ng/106 cells (P < 0.01;
4 independent experiments in duplicate, Student's i test). Fer
ritin content in iron-pretreated cells was 13.5 ±8.7 times higher
than in the correspondent DFO-pretreated cells.
Fig. 2 shows the uptake of 5-100 MM[I4C] ascorbic acid after
30 min incubation time into SK-N-SH cells. Similar to neutro-
phils, the uptake system is composed of two compounds: a high
affinity transport system (in the example given: Km, 7.1 MM;
Vmax:480 pmol/106 cells/30 min, calculated according to meth
ods given in Ref. 40); and a less active uptake system which
proved to be linear for an AA concentration of at least 1 mivi
(data not shown). A more detailed analysis is given elsewhere
(41).
DNA Strand Breaks in Isolated PM2 Phage DNA. In order
to characterize the effects of H2O2,6-OHDA, and AA on DNA,
first experiments were carried out on isolated CCC PM2 DNA.
Fig. 3A shows that 6-OHDA as well as AA generate single
strand breaks in CCC PM2 DNA. H2O2 caused DNA strand
100
90
80
o
«D 70
O
60
50
40
et
e 30
20
10
DFO control FeSO4
Fig. 1. Ferritin levels in SK-N-SH cells after 24 h incubation with 10 ^M DFO
and 100 ^M FeSO4 compared to untreated controls [4 independent experiments,
each in duplicate (SEM, 5.1-8.9%)]. , changes of ferritin content in corre
spondent cells after treatment with DFO or FeSO4.
 ASCORBATE. FERRITIN. AND REACTIVE OXYGEN COMPOUNDS IN NEUROBLASTOMA

contrast to the cell-free system, AA alone had no effects in this

test system. However, AA enhanced the effects of 6-OHDA as
1600 well as of H2O2 in DFO and in FeSO4-pretreated cells. The
most distinct differences were observed in DFO-pretreated cells
using intermediate concentrations (100 MM)of H2O2 (P< 0.01)
I 1200 and 6-OHDA (P< 0.05); 3 independent experiments, 2-4-fold.
The amount of alkaline stable DNA ("control") was 70 ±5%
in DFO-pretreated cells compared to 61.9 ±11% in the FeSO4-
800
pretreated group (6 independent experiments, each 2-4-fold; P
> 0.05). Strand break rate after incubation with 75 ßMH2O2
400 was greatly reduced in both groups when the incubation was
carried out in the presence of the iron chelator 1,10-phenan-
throline (100 UM). It dropped from 37% to 10% in the DFO
group and from 42% to 5% in the FeSO4 group.
20 O 60 80 100 Poly (ADP-Ribose) Polymerase Activity after Treatment with
ascorbic acid
6-OHDA and H2O2. Fig. 5 shows the effects of H2O2 and 6-
Fig. 2. Uptake of |'"C] ascorbic acid into SK-N-SII cells. About IO6cells were OHDA on poly(ADP-ribose) polymerase activity in DFO- and
incubated at 37'C in 12-well plates for 30 min with 5-100 /iM ["CJascorbic acid. FeSO4-pretreated SK-N-SH cells. AA enhanced the effects in
Mean of triplicate measurements; bars, SEM. both groups. Similar to DNA strand break experiments, the
differences between incubations with and without AA were
more pronounced in DFO-pretreated cells. The basal enzyme
activity in FeSO4-pretreated cells was higher than in DFO-
pretreated cells (3.34 ±0.55 versus 2.49 ±0.93 pmol protein-
bound poly(ADP-ribose)/min/106 cells (5 independent experi
ments, 2 fold, P > 0.05).
Decrease of NAD* and ATP in SK-N-SH cells after Treatment
with 6-OHDA and H2O2. Fig. 6 shows the decrease of NAD+
levels in SK-N-SH cells after incubation with 6-OHDA and
H2O2. The effects were more intense in FeSO4-pretreated cells.

- so-

Fig. 3. A, strand break formation of CGC PM2 DNA. Lane I, control; Lane
2, H2O2; Lane 3, 6-OHDA; Lane 4, ascorbic acid; Lane 5, H2O2 plus ascorbic
acid; Lane 6. 6-OHDA plus ascorbic acid. Incubation was carried out for l h
using 250 MMH2O2 or 6-OHDA and 10 mM ascorbic acid. B, determination of
CGC PM2 DNA strand breaks after 20 min incubation with 1 m,\i ascorbic acid
and 100 MM6-OHDA in the absence and presence of mannitol (100 mM) and
desferrithiocine (10 mM). In parallel. DNA strand breaks were determined using
the fluorescence assay according to the method of Lown (numbers in parentheses,
percentage of DNA strand breaks measured in the fluorescence assay). Lane 1,
control (0%); Lane 2, 1 misi ascorbic acid (45.7%); Lane 3, 6-OHDA (100%);
Lane 4, 6-OHDA plus mannitol (22.9%); Lane S, 6-OHDA plus desferrithiocine
(10.5%); Lane 6, ascorbic acid plus mannitol (5.8%); Lane 7, ascorbic acid plus
desferrithiocine (5.8%).

breakage only in the presence of AA. Fig. 3Äshows that DNA

strand breaks caused by 6-OHDA or AA could be prevented by
adding mannitol (an OH radical scavanger) or desferrithiocine _ 8(

(an iron chelator), suggesting that in both cases DNA strand °60-
breakage is caused by ferrous iron-dependent formation of OH
radicals. Superoxide dismutase, previously purged by dialysis, J.40-
reduced strand break rates of AA and 6-OHDA (data not
shown). Essentially the same results could be observed using
the fluorescence assay (see legend to Fig. 3ß).
25 100 250
DNA Strand Breaks in SK-N-SH Cells after Treatment with
6-OHDA and H2O2. Treatment of SK-N-SH cells with H2O2 or Fig. 4. DNA strand breaks in DFO- and FeSO4-pretreated SK-N-SH cells by
6-OHDA led to dose-dependent DNA strand break formation I !.•().•
and 6-OHDA in the absence (•)and presence (•)of 1 mM ascorbic acid
(expressed in percentage of controls, set as 100%; n = 5). Number of independent
(Fig. 4). FeSO4-pretreated SK-N-SH cells were more sensitive experiments: 25 MM,n = 2; 100 MM,n = 3; 250 MM,n = 3. Experiments were
against 6-OHDA and H2O2 than DFO-pretreated cells. In carried out 2-fold or 4-fold; SEM (bars), 4.6-12.1%
6068
 ASCORBATE. FERRITIN. AND REACTIVE OXYGEN COMPOUNDS IN NEUROBLASTOMA

and AA induced strand breaks, whereas H2O2 was effective only

in the presence of AA. In agreement with other reports, DNA
strand break formation is apparently caused by iron-dependent
formation of OH radicals (48). Analogous processes may occur
within neuroblastoma cells. These cells contain iron-rich ferri
tin which is dispersed within the whole cell areas (13). There
fore, iron release by substances such as AA and 6-OHDA should
be feasible, in principle, on multiple regions. We supposed that
AA could inflict damage upon these iron-rich cells in the
presence of H2O2, which could be delivered by appropriate
drugs (e.g. 6-OHDA) or formed endogenously in critical
100 amounts under certain conditions (see below). The ability of
250 neuroblastoma cells to accumulate AA would be a presupposi
tion for this concept. SK-N-SH cells indeed incorporate it
effectively (Fig. 2). The range of AA in normal serum is about
50-150 /¿Mbut is often markedly reduced in cancer patients
(49). Uptake of large amounts of AA can enhance serum levels
2-3-fold (49). Under cell culture conditions, these amounts lead
to an intracellular AA concentration in neuroblastoma cells in
the millimolar range. Calculated from the uptake experiment
shown in Fig. 2, the intracellular ascorbate concentration in
SK-N-SH cells raised from 0.45 mM after incubation with 50
/¿M AA to 1.14 mM after 30 min incubation with 200 juMAA.
The basal levels in SK-N-SH cells are negligibly low since the
culture media commonly used do not contain this vitamin.
The present study showed that 6-OHDA and H2O2 caused
100! similar toxic effects in a concentration-dependent manner on
25 250 DNA and the metabolic pathways related to DNA strand break

Fig. 5. Poly(ADP-ribose) polymerase activity in DFO- and FeSO4-pretreated

SK-N-SH cells after treatment with 6-OHDA and H2O2 in the absence (•)or
presence (•)of 1 rtiMascorbate, expressed as percentage of controls (set as 100%;
n = 5). Single experiments in duplicate for 25 MMand 100 t¡M; two experiments
in duplicate for 250 ^M (SEM (bars), 2.1-8.3%).

AA clearly enhanced the effects of H2O2 in DFO-pretreated

cells whereas the effects of 6-OHDA were amplified in DFO-
as well as in FeSO4-pretreated cells. The basal levels of NAD*
of DFO-pretreated cells were 1.53 ±0.31 nmol/106 cells; those
of FeSO4-pretreated cells were 1.62 ±0.16 nmol/106 cells (4
independent experiments; P > 0.05).
The decrease of ATP levels in SK-N-SH cells after exposure 25
to H2O2 and 6-OHDA is shown in Fig. 7. Again, the effects
were more pronounced in FeSO4-pretreated cells. Similarly as
shown for NAD-I-, AA enhanced the ATP-depleting effects of
6-OHDA more strongly than those of H2O2. The control levels
of DFO-treated cells were 6.96 ±1.54 nmol/106 cells and of
FeSO4-treated cells 7.90 ±0.70 nmol/106 cells (4 experiments;
P > 0.05).

DISCUSSION

The aim of the present study was to investigate the effects of

AA on iron-dependent DNA strand break formation and related
processes in neuroblastoma cells after incubation with H2O2
and 6-OHDA. Numerous investigations document the interac 25 250 25
tions and cytotoxic effects of these substances in different
biological systems (e.g. Refs. 42-47). The biochemical features
of neuroblastoma cells (high ferritin content, endogenous H2O2
Fig. 6. Decrease of NAD* concentration in DFO- and FeSCVpretreated SK-
production) seem to render them especially vulnerable by AA.
N-SH cells by H2O2 and 6-OHDA in the absence (•)and presence (•)of 1 mM
Experiments with 6-OHDA, H2O2, and AA in a cell-free ascorbic acid, expressed as percentage of controls (set as 100%; n = 4); single
system using isolated CGC PM2 DNA showed that 6-OHDA experiments for 25 ^M and 100 ^M; two experiments for 250 ^M. Bars, SEM.
6069
 ASCORBATE. FERRITIN, AND REACTIVE OXYGEN COMPOUNDS
25C
Fig. 7. Decrease of ATP concentration in DFO- and FeSCVpretreated SK-N-
SH cells by H2O2 and 6-OHDA in the absence (•)and presence (•)of 1 mM
ascorbic acid, expressed as percentage of controls (set as 100%; n = 4). Single
experiments for 25 MMand 100 ^M; two experiments for 250 MM.Bars, SEM.
formation [activation of poly(ADP-ribose) polymerase and de
cline of NAD+ and ATP], Under the experimental conditions
described, cells with high ferritin content proved to be more
vulnerable to 6-OHDA and H2O2, especially in the intermediate
concentration range (100 fi\t). These results are in agreement
with published reports indicating that iron also enhanced the
DNA strand break rates caused by H2O2 in other cells (31).
The enhancing effect of elevated cellular iron content on oxi-
dative damaging processes is not restricted on DNA; e.g., the
rate of lipid peroxidation in different brain regions was found
to correlate closely with their respective ferritin contents (50).
AA constantly enhanced the effects of 6-OHDA and H2O:
on DNA strand breaks and the related processes in iron- as
well as in DFO-pretreated cells. These effects are of special
interest due to their potential therapeutic perspective. In the
following model an interpretation of the ways AA may interact
with iron and H2O2 in neuroblastoma cells is presented. In
these cells, endogeneous formation of H2O2 is feasible by dif
ferent pathways. First, H2O2 is produced during the autoxida-
tion of AA and this process is thought to be one reason for its
cytotoxicity (17). However, other effects being specific for
neuroblastoma cells may play a more important role; AA en
hances the biosynthesis of catecholamines in normal sympa
thetic cells (51, 52) as well as in neuroblastoma cells (53). In
normal sympathetic cells catecholamines are stored in vesicles
which protect them from oxidation. The important role of
storage vesicles for preventing H2O2 formation by catechol
amines was recently shown. The production of H2O2 was sub-
6070
IN NEUROBLASTOMA
stanially increased in striata! synaptosomes incubated with do-
pamine when this reaction was performed in the presence of
reserpine, an inhibitor of catecholamine uptake into storage
vesicles (54). It is therefore most important to point to the fact
that neuroblastoma cells contain only a very low number of
vesicles. Consequently, enhanced formation of H2O2 by the
monoamine oxidase reaction and by catecholamine autoxida-
tion processes occurring in the cytoplasm can be expected.
The cytotoxic effects of H2O2 may be facilitated in neuro
blastoma cells by several mechanisms. First, like many other
tumor cells, neuroblastoma cells are characterized by low levels
of defense enzymes against H2O2 [catatase, GSH peroxidase
(55)], thus decelerating its destruction. These data are in agree
ment with the observed long half-life time of H2O2 in SK-N-
SH cells (8-9 min when 250 MMH2O2 was added to 1.5 x 10"
cells). Additional factors may further impair the effects of both
enzymes. Most neuroblastoma cells lack cystathionase which
normally converts cystathionine to homoserine and cysteine.
Elevated amounts of cystathionine will be found in urine in
about 50-80% of neuroblastoma patients (56). Consequently,
only a reduced intracellular cysteine production occurs and
could limit GSH supply necessary for peroxidase activity and
other pathways of the cellular defense system (57). Further
more, AA is known being a catatase inhibitor in vitro (58). This
might also be true in vivo. Finally, an important aspect of
ascorbic acid toxicity is probably its capacity to release iron
from ferritin as shown in the cell-free system (27). Although
not yet proved experimentally, this reaction is assumed to occur
also intracellularely and should facilitate the formation of OH
radicals by the Fenton reaction. Thus, in neuroblastoma several
biochemical processes influenced by AA could act in concert
and be responsible for the H2O2-mediated destructive processes
described.
In order to support this interpretation, it is worth mentioning
that a very similar biochemical pattern is believed to be respon
sible for tissue destruction in Parkinson disease. According to
the "endogenous toxin hypothesis" (59, 60), an elevated dopa-
mine production occurs in some areas of the substantia nigra
which may compensate for loss of dopamine production in
already destroyed tissue compartments. This leads to an en
hanced intracellular H2O: formation due to an enhanced do
pamine turnover by monoamine oxidase. Since the substantia
nigra in this disease is characterized by an enhanced iron level
(61) and reduced GSH, catatase, and GSH peroxidase activity
(62), oxidative damaging processes are favored. Indeed, lipid
peroxidation products can be detected in the substantia nigra
of Parkinson's patients (63). Similar pathways may play a role
in neuroblastoma cells in the presence of ascorbic acid.
Neuroblastoma cells, like other tumor cells, are in a some
what unstable metabolic condition. Permanent high dose appli
cation of AA may continually irritate neuroblastoma cells by
forcing the described biochemical processes and finally allow a
break in this fragile metabolic balance in those cells that had
survived a previous therapy. Therefore, application of high
dosages of AA for an unlimited period of time after the end of
the conventional therapy may be a mild, beneficial, and easily
performable attempt to prevent relapses in neuroblastoma.
REFERENCES
1. Rosen, E. M.. Cassady. J. R., Frantz. C. N.. Kretschmar, C., Levey, R., and
Sallan, S. E. Neuroblastoma: the Joint Center for Radiation Therapy/Dana-
Farber Cancer Institute/Children's Hospital experience. J. Clin. Oncol., 2:
719-732, 1984.
 ASCORBATE. FERRITIN, AND REACTIVE OXYGEN COMPOUNDS
2. D'Angio, G. J., August, C, Elkins, W., Evans, A. E., Seeger, R., Feig, S.,
Lenarsky, C., Wells. J„Ramsay, N., Kim, T., Woods, W., Krivit, W.,
Strandjord. S., Coccia. P., and Novak, L. Metastatic neuroblastoma managed
by supralethal therapy and bone marrow reconstitution (BMRC). Results of
a four institution children's cancer study group pilot study. Prog. Clin. Biol.
Res., / 75: 565-568, 1985.
3. Hartmann, O., Bcnhamou, E.. Beaujean. F., Kalifa, C., Lejars, O., Patte. C.,
Behard. C.. Flamant, F., Thyss, A., Deville, A., Vannier, J. P., Pautard-
Muchemble. B.. and Lemerle, J. Repeated high-dose chemotherapy followed
by purged autologous bone marrow transplantation as consolidation therapy
in metastatic neuroblastoma. J. Clin. Oncol., 5: 1205-1211, 1987.
4. Pinkerton, R.. Philip, T., Bouffet, E.. Lashford, L., and Kemshead, J. T.
Autologous bone marrow transplantation in paediatric solid tumors. Clin.
Haematol., 15: 187-203. 1986.
5. Sandford, J.. Jr., Tobes, M. C., and Sisson, J. C. Sodium dependency of
uptake of norepinephrine and m-iodobenzylguanidine into cultured human
pheochromocytoma cells: evidence for uptake-one. Cancer Res., 47: 3920-
3928, 1987.
6. Gee, P., and Davison, A. J. Intermediates in the aerobic auto-oxidation of 6-
hydroxydopamine: relative importance under different reaction conditions.
Free Radical Biol. Med., 6: 271-284. 1989.
7. Angeletti, P. U., and Levi-Momalcini, R. Cytolytic effect of 6-hydroxydo-
pamine on neuroblastoma cells. Cancer Res., 30: 2863-2869, 1970.
8. Tiffany-Castiglioni, E., Saneto, P. H.. and Perez-Polo. J. R. Participation of
active oxygen species in 6-hydroxydopamine toxicity to a human neuro-
blastoma cell line. Biochem. Pharmacol., 31: 181-188, 1982.
9. Reynolds, C. P., Reynolds, D. A., Frenkel, E. P., and Smith, R. G. Selective
toxicity of 6-hydroxydopamine and ascorbate for human neuroblastoma in
vitro: a model for clearing marrow prior to autologous transplantation.
Cancer Res.. 42: 1331-1336. 1982.
10. Heikkila, R. E.. and Cohen. G. Further studies on the generation of hydrogen
peroxide by 6-hydroxydopamine. Potentiation by ascorbic acid. Mol. Phar
macol., Ä:241-248, 1972.
11. Bruchelt, G., Buck, J., Girgert, R., Treuner, J., and Niethammer, D. The
role of reactive oxygen compounds derived from 6-hydroxydopamine for
bone marrow purging from neuroblastoma cells. Biochem. Biophys. Res.
Commun., 130: 168-174, 1985.
12. Buck. J., Bruchelt. G.. Girgert. R., Treuner, J., and Niethammer, D. Specific
uptake of [12!I]mlBG in the human neuroblastoma cell line SK-N-SH. Cancer
Res., 45: 6366-6370. 1985.
13. lancu. T. C., Shiloh, H., and Kedar, A. Neuroblastomas contain iron-rich
ferritin. Cancer (Phila.), 61: 2497-2502, 1988.
14. Hann, H-W. L., Stahlhut, M. W., and Evans, A. E. Source of increased
ferritin in neuroblastoma; studies with concanavalin A-Sepharose binding. J.
Nati. Cancer Inst., 76: 1031-1033, 1986.
15. Diliberto, E. J., Heckman, G. D., and Daniels. A. J. Characterization of
ascorbic acid transport by adrenomedullary chromaffin cells. J. Biol. Chem.,
258: 12886-12894, 1983.
16. Spector, R., and Greene, L. A. Ascorbic acid transport by a clonal line of
pheochromocytoma cells. Brain Res., 136: 131-140. 1977.
17. Prasad, K. N., Sinha, P. K., Ramanujam, M., and Sakamoto, A. Sodium
ascorbate potentiates the growth inhibitory effect of certain agents on neuro
blastoma cells in culture. Proc. Nati. Acad. Sci. USA, 76: 829-832, 1979.
18. Park, C. H., Amare, M., Savin, M. A., and Hoogstraten. B. Growth suppres
sion of human leukemic cells in vitro by L-ascorbic acid. Cancer Res., 40:
1062-1065, 1980.
19. Pavelic, K., Kos, Z., and Spaventi, S. Antimetabolic activity of L-ascorbic
acid in human and animal tumors. Int. J. Biochem., 21: 931-935, 1989.
20. Levine, M. New concepts in the biology and biochemistry of ascorbic acid.
N. Engl. J. Med., 314: 892-901, 1986.
21. Frei. B., England. L., and Ames, B. N. Ascorbate is an outstanding antioxi-
dant in human blood plasma. Proc. Nati. Acad. Sci. USA. 86: 6377-6381,
1989.
22. Halliwell, B., and Gutteridge, J. M. Oxygen free radicals and iron in relation
to biology and medicine: some problems and concepts. Arch. Biochem.
Biophys.. 246: 501-514, 1986.
23. Higson. F. K.. Kohen. R., and Chevion, M. Iron enhancement of ascorbate
toxicity. Free Radical Res. Commun., 5: 107-115, 1988.
24. Samuni, A., Aronovitch, J.. Godinger, D., Chevion, M., and Czapski, G. On
the cytotoxity of vitamin C and metal ions. Eur. J. Biochem., 137:119-124,
1983.
25. Samokyszyn, V. M., Thomas, C. E., Reif, D. W., Saito, M., and Aust, S. D.
Release of iron from ferritin and its role in oxygen radical toxicities. Drug
Metab. Rev., 19: 283-303. 1988.
26. Monteiro, H. P., and Winterbourn, Ch. C. 6-Hydroxydopamine releases iron
from ferritin and promotes ferritin-dependent lipid peroxidation. Biochem.
Pharmacol., 38: 4177-4182, 1989.
27. Boyer, R. F., Grabill, T. W., and Petrovich, R. M. Reductive release of
ferritin iron: a kinetic assay. Anal. Biochem., 174: 17-22. 1988.
6071
IN NEUROBLASTOMA
28. Lode, H. N., Bruchelt, G., Rieth, A. G., and Niethammer, D. Release of iron
from ferritin by 6-hydroxydopamine under aerobic and anaerobic conditions.
Free Radical Res. Commun.. //: 153-158, 1990.
29. Halliwell. B., and Gutteridge, J. M. C. Role of iron in oxygen radical
reactions. Methods Enzymol.. 105: 47-57, 1984.
30. O'Connell. M.. Halliwell. B., Moorhouse, C. P., Aruoma, O. !.. Baum, H.,
and Peters, T. J. Formation of hydroxyl radicals in the presence of ferritin
and haemosiderin. Biochem. J., 234: 727-731, 1986.
31. Schraufstätter, I. U., Hyslop, P. A., Jackson, J. H., and Cochrane, C. G.
Oxidant-induced DNA damage of target cells. J. Clin. Invest., 82: 1040-
1050, 1988.
32. Hyslop, P. A.. Hinshaw, D. B., Halsey, Jr, W. A., Schraufstätter, I. U.,
Sauerheber, R. D., Spragg. R. G., Jackson, J. H., and Cochrane, C. G.
Mechanisms of oxidant-mediated cell injury. J. Biol. Chem., 263: 1665-
1675, 1988.
33. Schraufstätter. I. U., Hyslop, P. A., Hinshaw, D. B., Spragg, R. G.. Sklar, L.
A., and Cochrane, C. G. Hydrogen peroxide-induced injury of cells and its
prevention by inhibitors of poly(ADP-ribose) polymerase. Proc. Nati. Acad.
Sci. USA, 83: 4908-4912,1986.
34. Schraufstätter, I. U., Hinshaw, D. B., Hyslop, P. A., Spragg, R. G., and
Cochrane, C. G. Oxidant injury of cells. J. Clin. Invest., 77: 1312-1320,
1986.
35. Berger, N. A. Cancer chemotherapy: new strategies for success. J. Clin.
Invest., 78: 1131-1135, 1986.
36. Ueda. K., and Hayaishi. O. ADP-ribosylation. Annu. Rev. Biochem., 54: 73-
100, 1985.
37. Lown, J. W. Ethidium binding assay for reactive oxygen species generated
from reductively activated Adriamycin (doxorubicin). Methods Enymol., 705:
532-539, 1984.
38. Birnboim, H. C., and Jevcak, J. J. Flourimetric method for rapid detection
of DNA strand-breaks in human white blood cells produced by low dose
radiation. Cancer Res., 41: 1889-1892, 1981.
39. Schraufstätter, I. U., Halsey, W. A. Jr., Hyslop, P. A., and Cochrane, C. G.
In vitro models for the study of oxidant-induced injury of cells in inflamma
tion. Methods Enzymol.. 163: 328-339, 1988.
40. Washko. P., Rotrosen, D., and Levine, M. Ascorbic acid transport and
accumulation in human neutrophils. J. Biol. Chem., 264: 18996-19002.
1989.
41. Bruchelt. G., Klinkenberg, C. R., Klingebiel. T., and Niethammer, D. Uptake
and cytotoxic effects of ascorbic acid on neuroblastoma cells. Free Radical
Biol. Med., 9(Suppl. 1): 169. 1990.
42. Gutteridge. J. M. C., Halliwell, B., Treffry, A.. Harrison, P. M.. and Blake,
D. Effect of ferritin-containing fractions with different iron loading on lipid
peroxidation. Biochem. J., 209: 557-560, 1983.
43. Nienhuis, A. W. Vitamin C and iron. N. Engl. J. Med., 304:170-171, 1981.
44. Miller, D. M.. and Aust, S. D. Studies on ascorbate-dependent. iron-catalyzed
lipid peroxidation. Arch. Biochem. Biophys., 271: 113-119, 1989.
45. Aronovitch, J., Godinger, D., Samuni. A., and Czapski, G. Ascorbic acid
oxidation and DNA scission catalyzed by iron and copper chelates. Free
Radical Res. Commun., 2: 241-258, 1987.
46. Fink, R. M., and Lengfelder. E. Hyaluronic acid degradation by ascorbic acid
and influence of iron. Free Radical Res. Commun., 3: 85-92, 1987.
47. Sadrzadeh, S. M. H., and Eaton, J. W. Hemoglobin-mediated oxidant
damage to the central nervous system requires endogenous ascorbate. J. Clin.
Invest., 82: 1510-1515, 1988.
48. Schneider. J. E.. Browning. M. M., and Floyd, R. A. Ascorbate/iron media
tion of hydroxyl free radical damage to PBR322 plasmid DNA. Free Radical
Res. Commun., 5: 287-295, 1988.
49. Friedrich, W. Vitamins, pp. 931-1001. Berlin: Walter de Gruyter, 1988.
50. Floyd, R. A. Oxygen radical-mediated damage to brain tissue. In: M. G.
Simic, K. A. Taylor, J. F. Ward, and C. von Sonntag (eds.). Oxygen Radicals
in Biology and Medicine, pp. 1015-1023. New York: Plenum Publishing
Corp., 1988.
51. Levine. M. Ascorbic acid specifically enhances dopamine-rf-monooxygenase
in resting and stimulated chromaffin cells. J. Biol. Chem., 261: 7347-7356,
1986.
52. Menniti, F. S., Knoth, J., and Diliberto. E. J. Role of ascorbic acid in
dopamine-/J-hydroxylation. J. Biol. Chem., 261: 16901-16908, 1986.
53. Rosemeyer. S. Thesis, Faculty of Biology. University of Tuebingen, Federal
Republic of Germany, 1987.
54. Spina. M. B., and Cohen, G. Hydrogen peroxide production in dopamine
neurons. In: M. G. Simic, K. A. Taylor, J. F. Ward, and C. von Sonntag
(eds.). Oxygen Radicals in Biology and Medicine, pp. 1011-1014. New York:
Plenum Publishing Corp., 1988.
55. Steinkühler.C., Mavelli, I.. Melino. G., Piacentini, M.. Rossi, L., Weser, U.,
and Rotilio, G. Antioxygenic enzyme activities in differentiating human
neuroblastoma cells. Ann. NY Acad. Sci., 551: 137-140, 1988.
56. Klein, C. E., Roberts, B., Holcenberg, J., and Glode, L. M. Cystathionine
metabolism in neuroblastoma. Cancer (Phila.), 62: 291-298, 1987.
 ASCORBATE. FERRITIN. AND REACTIVE OXYGEN COMPOUNDS IN NEUROBLASTOMA

57. Arrick, B. A., and Nathan, C. F. Glutathionc metabolism as a determinant
of therapeutic efficacy. Cancer Res., 44: 4224-4232. 1984.
58. Davison. A. J.. Kettle. A. J.. and Fatur, D. J. Mechanism of the inhibition
of catalase by ascorbate. J. Biol. Chem., 261: 1193-1200. 1986.
59. Fahn, S. The endogenous toxin hypothesis of the etiology of Parkinson's
. . f ,. , , „¿
disease and a pilot trial of high dosage antioxidants in an attempt to slow
the progression of the illness. Ann. NY Acad. Sci., 57ft- 186-196. 1989.
60. Spina, M. B., and Cohen, G. Dopamine turnover and glutathione oxidation:
implications for Parkinson disease. Proc. Nati. Acad. Sci. USA, 86: 1398-
1400. 1989.
61. Dexter, D. T.. Wells, F. R., Lees, A. J.. Agid, F.. Agid, V., Jenner, P., and
Marsden, C. D. Increased nigral iron content and alterations in other metal
ions occurring in brain in Parkinson's disease. J. Neurochem.. A. 1830-1836,
,9g9
., ' , ., , ,. ,
62. Ambani. L. M.. Van \\oert. M. H.. and Murphv. S. Brain peroxidase and
catalase in Parkinson s d.sease. Arch. Neurol., ^2: 114-118. 1975.
6;t Rexter, D. T.. Carter, C. J.. Wells. F. R.. Javoy-Agid, F.. Agid, V „¿
Lees, A.,
Jenner, P., and Marsden, C. D. Basal lipid peroxidation in substantia nigra
is increased in Parkinson's disease. J. Neurochem.. ?2: 381-389, 1989.

6072