| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Department of Biochemistry and Molecular Biology, The George Washington University, School of Medicine and Health Sciences, Washington, D.C. 20037, USA
1Correspondence: Department of Biochemistry and Molecular Biology, The George Washington University, School of Medicine and Health Sciences, 2300 Eye St., N.W., Ross Hall 526, Washington, D.C. 20037, USA. E-mail: bcmvwh{at}gwumc.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: metabolic labeling • radioisotopes • DNA fragmentation • reactive oxygen species • p53
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. Because cells often appear viable immediately
after labeling or after short-term pulse-chase experiments, it has been
assumed that the radioisotope is innocuous at the levels typically
used. However, several recent studies, including ours, have shown that
cell-incorporated low-energy ß-emitters such as
[35S]methionine and
3H-thymidine can cause cell cycle arrest and
apoptosis (2
3
4
5
6)
. Cell cycle arrest is most obvious in
cultures of radiolabeled synchronized cells that fail to divide within
their normal division cycle and has been attributed, in at least two
cases, to increases in the level of p53 (3
, 4)
, a major
cell cycle checkpoint regulator that serves as a molecular sensor for
DNA damage (7)
. However, there is little substantiation of
the occurrence of DNA damage by these low-energy ß-emitters at the
doses used in metabolic labeling experiments, with the exception of two
studies in which hemopoietic cells were labeled with
3H-thymidine for 24 or more hours (2
, 6)
. Here, we report that exposure of cells to [35S]methionine induces not only DNA damage, but also generates a significant amount of reactive oxygen species (ROS), at least in the cell types studied. Furthermore, these two events can be dissociated temporally as well as differentiated by their respective response to inhibitors of ROS and their relative dependence on p53. Collectively, these studies establish a molecular mechanism for the initiation of cell cycle arrest and/or apoptosis by low-energy ß-emitting radioisotopes. Moreover, our finding that radiolabeling also induces ROS production mandates further study to evaluate the effect of this oxidative stress on cell metabolism and signal transduction.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. The
NIH 3T3 murine fibroblasts were obtained from Dr. Patricia Berg (The
George Washington University Medical Center) and were cultured in
Dulbecco’s minimum essential medium (DMEM) supplemented with 10% FBS.
The p53+/+ and p53-/-
human colorectal carcinoma cell lines were generously provided by Dr.
Bert Vogelstein (Johns Hopkins University School of Medicine and the
Howard Hughes Medical Institute) and were cultured in McCoy’s 5A
medium with 10% FBS and antibiotic/antimycotic solution at the
concentrations stated above. All media and buffers were purchased from
Fisher Scientific (Pittsburgh, PA).
Radiolabeling protocol
Cells in methionine-free DMEM (DMEM-)
supplemented with 10% dialyzed FBS and 2 mM glutamine were incubated
with 35S-Trans label (spec. act. >1000 mCi/mM;
ICN Radiochemicals, Costa Mesa, CA) at different doses and for various
periods. In some experiments, a postlabeling incubation in normal
medium followed washout of the radiolabel. Control (mock-labeled) cells
were incubated with equivalent volumes of matching carrier buffer (50
mM L-lysine, 10 mM beta-mercaptoethanol, pH 7.4) obtained from ICN.
Positive controls for the Comet assay consisted of cells treated with
100 µM peroxide for 20 min at 4°C. To evaluate the role of
oxygen-derived ROS in DNA fragmentation, cells were subjected to
hypoxic conditions before and during the 1.5 h labeling period
before Comet analysis. Cells cultured in screw cap flasks were
preequilibrated under nitrogen in supplemented
DMEM- medium buffered with 25 mM HEPES at pH
7.2–7.4 for 30 min before addition of the radiolabel or carrier
buffer. Hypoxia was maintained during incubation with
35S-Trans label by flushing the flasks with
nitrogen and capping the flasks tightly. Harvested cells were also
flushed with nitrogen to maintain a hypoxic atmosphere prior to
preparation of the Comet slides. In one experiment described here,
cells were labeled with 3H-thymidine (spec. act.
77 Ci/mmol; ICN Radiochemicals) for 2 h, washed, incubated in
normal medium for 1 h, and evaluated for DNA fragmentation by the
Comet assay. In other studies, 5 mM allopurinol (Sigma, St. Louis,
Mo.), a xanthine oxidase inhibitor, was included in the medium during
the labeling as well as postlabeling periods before harvesting of cells
for Comet analyses or before ROS assays.
Comet assay for DNA fragmentation
The Comet assay involves single-cell electrophoresis in agarose
as first described by Ostling and Johanson (8)
. Cells were
incubated with radiolabel or carrier buffer (for mock-labeled controls)
for various periods of time and at different doses of radiolabel and
analyzed for DNA fragmentation using a CometAssayTM kit (Trevigen,
Gaithersburg, MD), following procedures outlined by Trevigen for the
neutral Comet assay. Under neutral conditions, the Comet assay will
detect primarily double-stranded DNA breaks. In brief, radiolabeled and
control cells were harvested with trypsin-EDTA solution, washed, and
suspended in low melting agarose at 
105
cells/ml. The agarose was applied to CometSlidesTM and allowed to set
at 4°C. After lysis of the agarose-embedded cells in Lysis Solution
(2.5 M NaCl, 100 mM EDTA, pH 10, 10 mM Tris base, 1% sodium lauryl
sarcosinate, 0.01% Triton X-100), the slides were placed in a Bio-Rad
(Hercules, CA) submarine gel electrophoresis unit and electrophoresed
in TBE, pH 8 (0.089 M Tris/0.089 M boric acid/0.003 M EDTA) at 1 V/cm
for 30 min. The samples were then fixed in MeOH and EtOH (5 min. each)
at -20°C and dried overnight before staining with SyBr Green to
visualize cellular DNA.
Fluorescence microscopy and laser scanning cytometry
SyBr Green-stained samples were examined by fluorescence
microscopy using an Olympus IX70-inverted system microscope equipped
with an IX-FLA inverted reflected light fluorescence observation
attachment. Pictures were recorded with an Olympus PM-20
photomicrographic system using a 20x objective. For quantitation of
the relative amount of DNA fragmentation in different samples, the
stained slides were scanned with a Meridian ACAS570 Interactive Laser
Cytometer at an excitation wavelength of 488 nm using laser power of 60
mW and scan strength of 10%, with a 10% neutral density filter. The
pseudocolor fluorescence images (see Fig. 1B
) were analyzed using the Cell Image program to circumscribe
the ‘head’ and ‘tail’ regions of each Comet and the integrated
(total) fluorescence values of each defined area were recorded. The
ratio of tail/head fluorescence was used as a relative measure of DNA
fragmentation for each sample. An average of 25 individual Comets were
scored per sample.
|
Assay for ROS
The fluorogenic, peroxide-sensitive compound
5(6)-chloromethyl-2',7'-dichlorodihydrofluorescin
(CM-H2DCF) was used to detect the
presence of ROS in living cells in a manner described for the parent
compound, 2',7'-dichlorofluorescin (9
, 10)
. Cells were
first radiolabeled for 2 h with 35S-Trans
label, washed, and incubated for varying periods in normal cold medium
before staining with a cell-permeable diacetate (DA) form of the dye
CM-H2DCFDA (Molecular Probes, Eugene, OR). After
entry into the cell and cleavage of the diacetate group by
intracellular esterases, this dye becomes fluorescent only after
oxidation by peroxide/hydroperoxides to chloromethyldichlorofluorescein
(CM-DCF). Laser scanning cytometry using the ACAS570 interactive laser
cytometer was again used to quantitate and analyze cell-associated
fluorescence. Because of the enlargement of radiolabeled cells over a
24 h period, average fluorescence (integrated fluorescence/number
of pixels in area defining each respective cell) was recorded as it was
found to be independent of cell size, whereas integrated fluorescence
was size dependent (data not shown). At least 100 cells were analyzed
per sample. For some studies, cells were incubated with 5 mM
allopurinol during and after the labeling period, before staining the
cells with CM-H2DCFDA for the ROS assay.
Statistical analyses
All Comet data are expressed as means ± SE.
The Student’s unpaired t test was used to determine
statistical significance of the difference between labeled samples vs.
mock-labeled controls.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
3
4
5
6)
, the initiating mechanism(s) for
these effects is(are) unknown. Inasmuch as DNA is a prime target for
ionizing radiation (e.g., 
- or X-rays) (11
, 12)
, we
investigated the effects of cellular incorporation of
[35S]methionine on DNA integrity in
radiolabeled cells using Comet analyses. Rabbit smooth muscle cells
were labeled for 2 h with [35S]methionine
at 1, 10 and 100 µCi/ml and analyzed by Comet assay for DNA
fragmentation 2 h after the label was washed out and the cells
returned to complete medium. As shown in Fig. 1A
, DNA
fragmentation was indicated by the production of ‘tails’ in the
radiolabeled samples, which were longer and thicker with increasing
dose. Laser scanning cytometry (LSC) was used to quantitate the
relative amount of DNA fragmentation at the different doses as
described in Materials and Methods. Figure 1B
shows a
representative pseudocolor fluorescence image of Comets obtained by
LSC; Fig. 1C
shows quantitation of tail-to-head ratio as a
function of labeling dose. As shown, the degree of DNA fragmentation is
directly related to the dose of radiolabel. At 24 h after the
labeling period, there is still evidence of DNA fragmentation, although
the degree of fragmentation is attenuated relative to that observed
2 h after washout of the label, especially at the lower doses
(Fig. 1D
). This latter finding suggests there is some repair
of double-stranded DNA breaks in cells exposed to
[35S]methionine at 1 and 10 µCi/ml, but
relatively little repair at 100 µCi/ml, which is the dose of
35S-label most frequently used in protein
labeling protocols (1)
. Nevertheless, our previous studies
demonstrated that even at 0.1 µCi 35S-label/ml,
colony formation was still significantly inhibited in this cell type
(5)
, suggesting that genotoxic effects of
cell-incorporated 35S are still
manifested at very low doses despite repair activity. These combined
findings therefore call for additional studies to determine whether
there is a threshold dose of low-energy internal ß-emitters that
results in genotoxicity or sublethal genetic mutations and for
reevaluation of safety limits regarding occupational hazards and
radioactive waste disposal methods.
Time dependence of DNA fragmentation
To determine the minimum labeling period required for the
manifestation of DNA damage, cells were labeled for 0.5, 1, 1.5, and
2 h in [35S]methionine at 100 µCi/ml and
analyzed for DNA fragmentation immediately after the respective
labeling periods. The results in Fig. 2
show that DNA fragmentation is observed as early as 1 h after the
radiolabel is added to the cell culture and increases with increasing
time of incubation up to 2 h. These results suggest that a
significant amount of DNA damage occurs even before most
radiolabeling protocols are completed.
|
Radiolabel-induced DNA fragmentation is independent of cell type
Several different cell lines were examined for DNA fragmentation
in response to metabolic labeling with
[35S]methionine. Figure 3
shows that all cell lines tested, including a standard cell model, the
3T3 murine fibroblast, and both p53-positive and p53-negative human
colorectal carcinoma cell lines, exhibited significant DNA
fragmentation 2 h after a 2 h labeling period. Thus, RSMC
cells are not unique in their response to this form of low-energy
ß-radiation, suggesting that DNA fragmentation is a common
phenomenon associated with metabolic radiolabeling.
|
Tritiated thymidine also induces DNA fragmentation
We also evaluated the potential of another commonly used
low-energy ß-emitter, 3H-thymidine, to cause
DNA fragmentation. As shown in Fig. 4
, dose-dependent DNA damage was observed as early as 1 h after a
2 h incubation with 3H-thymidine at 2, 20,
and 100 µCi/ml. These results are consistent with those previously
reported (2
, 6)
, but demonstrate a much earlier time point
for manifestation of DNA damage by this very low-energy emitter
(average beta energy=0.006 MeV).
|
Radiolabeling with 35S-methionine induces ROS
formation
Because ionizing radiation is usually associated with the
formation of ROS under aerobic conditions, we monitored the formation
of ROS using a peroxide-sensitive fluorescent dye,
CM-H2DCF. Cells were radiolabeled with
[35S]methionine for 2 h, then washed and
incubated in complete medium for 2 h before the ROS assay. Cells
were loaded with the membrane-permeable form of the dye
(CM-H2DCFDA) and analyzed by laser scanning
cytometry to detect the formation of intracellular peroxides as
indicated by increased dye fluorescence. In contrast to mock-labeled
cells, the radiolabeled RSMC exhibited significantly higher levels of
fluorescence, as revealed by the fluorescence distribution profiles
shown in Fig. 5A
, B
. Furthermore, the observed increase in fluorescence
could be reduced almost to control levels by treatment of radiolabeled
cells with the antioxidant allopurinol (Fig. 5C
). A similar
inhibition of ROS was obtained with exogenous catalase at 1000 U/ml
during the labeling and postlabeling incubation periods (data not
shown). These results implicated ROS as a possible mediator of DNA
damage. Thus, the effects of antioxidants and hypoxia on DNA
fragmentation were investigated.
|
ROS inhibitors and hypoxia do not attenuate DNA fragmentation
Even though [35S]methionine could induce
antioxidant-inhibitable elevation of ROS in RSMC (Fig. 6
), these same antioxidants were not able to inhibit DNA fragmentation,
as detected by the Comet assay (data not shown). Because the specific
antioxidants used may not have eliminated all potentially damaging
forms of ROS, the radiolabeling experiments were also carried out under
nitrogen to induce hypoxia. As shown in Fig. 6
, DNA fragmentation was
observed even under hypoxic conditions 1.5 h after addition of
label. In contrast, the earliest detectable increase in CM-DCF
fluorescence in RSMC occurred at 2 h postlabeling. Taken together,
these results suggest that either ROS was present, but at a level
undetectable by the CM-DCF fluorescence assay at the earlier time
points at which DNA fragmentation was clearly observable, or that ROS
is not a necessary initiator of
[35S]methionine-induced DNA damage. The results
with hypoxic cells would support the latter case.
|
p53 is associated with ROS production but is not necessary for DNA
fragmentation
Because p53 has been implicated in the formation of ROS through
transcriptional induction of redox-related genes (13)
, we
examined p53+/+ and
p53-/- cells for
[35S]methionine-induced ROS formation and DNA
fragmentation. Figure 7
shows that ROS was detectable by CM-DCF fluorescence only in the
radiolabeled p53+/+ cells 24 h after
labeling. Earlier time points (e.g., 2 h after washout of the
radiolabel) showed no significant rise in ROS levels in either
p53+/+ or p53-/- cells,
despite clear evidence for DNA fragmentation 2 h after labeling
for both these cell types (see Fig. 3
). Thus, despite an association
between functional p53 expression and ROS formation, there was a p53 as
well as temporal dissociation between ROS production and DNA
fragmentation. These results further imply that at least initial
DNA fragmentation by low-energy internal ß-radiation
is independent of oxygen-derived ROS.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
3
4
5
6)
.
Aside from DNA fragmentation, metabolic labeling with
[35S]methionine was associated with a rise in
ROS as detected by the peroxide-sensitive fluorescent dye,
CM-H2DCF. However, the ROS detected cannot be
responsible for the radioisotope-induced DNA damage observed, because
DNA fragmentation precedes ROS formation and could not be inhibited by
antioxidants, which were nevertheless able to inhibit the rise of ROS
in radiolabeled cells. To rule out the possibility that other forms of
oxygen-derived ROS (e.g., superoxide) were responsible for DNA
fragmentation, the radiolabeling procedure was carried out under
nitrogen. The results in Fig. 6
show that DNA fragmentation was still
significant under hypoxic conditions. Although the reduced level of
fragmentation in hypoxic cells (68% of normoxic level) may be
regarded as evidence for partial ROS contribution to DNA damage; it can
also be explained by a measured 50–66% reduction in uptake of
[35S]methionine during hypoxia.
Another novel and significant finding of these studies is that
functional p53 expression correlates with radiolabel-induced ROS but
not DNA damage. This observation not only further dissociates ROS from
early induction of DNA fragmentation, but also is consistent with a
model for p53-induced apoptosis proposed by Polyak et al.
(13)
in which functional p53 expression induces
p53-inducible genes, which in turn leads to elevation of ROS in cells.
The relative time scale of DNA fragmentation and ROS production in our
studies appears to be consistent with this scenario; that is, DNA
fragmentation, which presumably initiates activation of p53, precedes
ROS formation by 1 h or longer. In fact, we observed an increase
in p53 protein in RSMC as early as 30 min after the label is added
(D. S. Heikka and V. W. Hu, unpublished results), at which
time DNA damage is already detectable in some cells by Comet assay. It
is also noteworthy that there is a longer lag time between DNA
fragmentation and ROS production in the p53+/+
human carcinoma cells, suggesting that the kinetics or mechanism of ROS
production in response to internal ß-radiation may be dependent on
cell type.
The mechanism of DNA damage induced by the low-energy ß-radiation
from [35S]methionine or
3H-thymidine is not known, although free radical
production is a common result of ionizing radiation (12)
.
In an aqueous intracellular environment, the most abundant free
radicals are likely to be water and hydroxyl radicals, which in turn
may ionize biological molecules, including DNA. We have indeed observed
some attenuation of DNA fragmentation in the presence of sulfhydryl
compounds (V. W. Hu, unpublished results), suggesting the
involvement of free radicals in DNA damage, but the chemical nature of
the radicals and the DNA damage remains to be identified. On a broader
scope, if ionization by water and hydroxyl radicals is a common
mechanism of DNA damage by cell-incorporated radio emitters, then the
application of all forms of radiolabels for in vivo labeling or
tracking (including 14C-labeled metabolites)
would need to be reevaluated in terms of their effect on cell function.
In fact, a recent study has shown that metabolic labeling with
32P-orthophosphate also activates p53, inducing
p53-dependent growth arrest in human fibroblasts (14)
.
These results collectively emphasize the need to find and use
alternative, nonradioactive tools for the in vivo analysis of cellular
activities.
Finally, although oxygen-derived ROS has been excluded as an
initiator of the low-energy ß-induced DNA fragmentation,
as seen by Comet analyses, its potential contribution to downstream
apoptotic events or other adverse cellular responses cannot be
overlooked. For example, it is becoming increasingly clear that ROS and
free radicals play a substantial role in signal transduction both
physiologically and pathologically (15
16
17)
. Even if
metabolic labeling with low-energy ß-emitters does not always lead to
cell cycle arrest, inhibition of cell proliferation, or apoptosis, as
we and others have reported (2
3
4
5
6)
, it is likely, given
our current findings, to alter the redox status of the cell and
associated processes. Thus, rather than serving as an experimental tool
to track metabolic and signaling activities in vivo, radiolabels may
actually be perturbing or undermining the system they were meant to
study.
|
ACKNOWLEDGMENTS |
|---|
Received for publication February 2, 2001. Accepted for publication March 26, 2001.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  H. Yuan, M. Kamata, Y.-M. Xie, and I. S. Y. Chen Increased Levels of Wee-1 Kinase in G2 Are Necessary for Vpr- and Gamma Irradiation-Induced G2 Arrest J. Virol., August 1, 2004; 78(15): 8183 - 8190. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. F. MARKO, P. B. DIEFFENBACH, G. YAN, S. CERYAK, R. W. HOWELL, T. A. MCCAFFREY, and V. W. HU Does metabolic radiolabeling stimulate the stress response? Gene expression profiling reveals differential cellular responses to internal beta vs. external gamma radiation FASEB J, August 1, 2003; 17(11): 1470 - 1486. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Schulenberg, R. Aggeler, J. M. Beechem, R. A. Capaldi, and W. F. Patton Analysis of Steady-state Protein Phosphorylation in Mitochondria Using a Novel Fluorescent Phosphosensor Dye J. Biol. Chem., July 11, 2003; 278(29): 27251 - 27255. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Roos-Mattjus, K. M. Hopkins, A. J. Oestreich, B. T. Vroman, K. L. Johnson, S. Naylor, H. B. Lieberman, and L. M. Karnitz Phosphorylation of Human Rad9 Is Required for Genotoxin-activated Checkpoint Signaling J. Biol. Chem., June 27, 2003; 278(27): 24428 - 24437. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
