hyperoxia induces dna damage in mammalian cells

Free Radical Biology & Medicine. Vol. 14, pp. 267-276, 1993 0891-5849/93 $6.00 + .00 Printed in the USA. All fights reserved. Copyright © 1993 Pergamon Press Ltd.

Original Contribution

H Y P E R O X I A I N D U C E S D N A D A M A G E I N M A M M A L I A N C E L L S

MARCO A. CACCIUTTOLO, LOC TR1NH, JANICE A. LUMPKIN, and GOVIND RAO*

Chemical and Biochemical Engineering Program, University of Maryland Baltimore County, Baltimore, MD 21228, USA; and *Medical Biotechnology Center of the Maryland Biotechnology Institute,

618 West Lombard St., Baltimore, MD 21201, USA

(Received 5 May 1992; Revised 17 July 1992; Re-revised and Accepted 22 September 1992)

Abstract--There is mounting evidence on the role of oxygen-derived free radicals in causing damage to various cellular components. However, most studies reported in the literature have been conducted under conditions where cells were chal- lenged with chemical free radical generating systems. In contrast, we measured DNA strand breaks, through a relatively simple and sensitive technique, as a function of the dissolved oxygen tension in a bioreactor. Cells were exposed to a step change in oxygen tension at mid-exponential growth phase. Several levels of oxygen were tested (200, 300, and 476% dissolved oxygen with respect to air saturation at 1 atmosphere) and compared against a control ( 10% dissolved oxygen). Hyperoxia was found to cause monotonically increasing DNA strand breakage at all the oxygen levels. In addition, hyperoxia was found to affect other metabolic functions such as the glucose consumption rate, lactate production rate, and cell growth. When hyperoxia-induced DNA strand breakage was compared to that induced by exposure to hydrogen peroxide, a similar response was observed. Exposure to a dissolved oxygen level of 200% induced DNA strand breakage comparable to a bolus of 4.2 #M hydrogen peroxide. Our results show that there is an association between hyperoxia and DNA damage.

Keywords--Oxygen free radicals, Hyperoxia, Mammalian cells, DNA damage, Free radicals

INTRODUCTION

There is considerable evidence on the role of oxygen and its derivatives in causing cellular damage ~ and on their involvement with the development of several diseases in humans. 2 In all aerobic organisms, 02 serves as the terminal electron acceptor during the

Net:

M "+l + 02- --~ M "+ + 0 2

M "+ + H~O7 -'* "OH + OH- + M n+l

H202 + 02- ~ "OH + OH- + 02

process of respiration by accepting four electrons and forming H20 (Ref. 3). However, if the reduction of O2 is incomplete, then the result is the formation of either the superoxide (02-) radical or hydrogen peroxide (H202). 4 There is yet another oxygen-derived species even more reactive than either 02- or H202. This is the hydroxyl radical ("OH), and it is generated by the following (simplified) sequential reactions:

(02- reduces M n+l)

(Fenton reaction)

(Metal ion catalyzed Haber-Weiss reaction).

In the preceding reactions, M "÷ is a metal ion (usually Fe 2÷) produced by the action of 02-. It is generally believed that most of the destructive effects of oxygen- derived species are primarily caused by "OH. Hy- droxyl radicals have been implicated in DNA damage, lipid peroxidation of unsaturated fatty acids lead-

Address correspondence to: Govind Rao, Chemical and Bio- chemical Engineering Program, University of Maryland Baltimore County, Baltimore, MD 21228.

ing to membrane damage, and in the modification of amino acids that are largely responsible for the struc- tural integrity of proteins. 5-7 All these and other studies have used superoxide-generating chemicals or directly added H202 in their experiments to evaluate their effects.

While the preceding scheme is a useful theoretical framework to explain the role of oxygen in causing cell injury, there are very few studies available that

267

268 M.A. CACCIUTTOLO el al.

show quantitative relationships between oxygen tensions and specific cellular responses. Studies conducted with systems such as hydrogen peroxide have shown that DNA is very sensitive to the action of oxi- dants, 8 but no data are available regarding the extent of DNA damage as a function of external oxygen tension levels. Among the few examples of studies show- ing the relationships between dissolved oxygen levels and cellular damage is a recent report that shows the dependence of lipid peroxidation with oxygen tension. 9

Boveris and co-workers l° have reported that the rate of hydrogen peroxide generation by mitochondria is a function of the oxygen tension. By increasing the oxygen level from 100% saturation with respect to air to 476% (pure oxygen) at atmospheric pressure, they showed that the rate of hydrogen peroxide generation increases about 50%. Since it has been reported that hydrogen peroxide leaks from mitochondria, l° it is expected that under increasing oxygen tension, increased intracellular hydrogen peroxide production will occur, thus leading to increased cellular damage.

We have chosen to quantitate the effect of hyperoxia on DNA integrity. The primary reason for this is that even sublethal damage to the genome can lead to culture instability or death. ~ Oxidative DNA damage is one of the leading suspects for instability in hetero- karyotypic cells, and there is strong circumstantial evidence for its role as a possible causative agent of cancer.lZ In terms of the biotechnology industry, culture stability is a major problem, 13 and its roots may well lie in the stability at the genetic level. Although cellular repair enzymes are available to repair several kinds of DNA damage,14 there is always a probability of misrepair. This may lead to altered cellular function or, if the damage is at a critical point, death.

As mentioned earlier, there is a lack of experimental evidence relating oxygen partial pressures and DNA damage. Part of the reason for this lack of infor- mation has been the difficulties in maintaining controlled environments in which to conduct studies and the unavailability of sufficiently sensitive analytical techniques. Our approach has considered both of these factors: First, we have used suspension culture in a bioreactor which permits precise control over en- vironmental conditions. Dissolved oxygen tensions (DO levels) are easily changed and maintained by bubbling different oxygen/nitrogen gas mixtures into the bioreactor. Second, we have employed a modified fluorimetric assay for DNA unwinding which is a relatively easy, rapid, and inexpensive technique to quantify DNA strand breaks at very sensitive levels (with a reported detection limit of up to one strand break per chromosome). 15

MATERIALS AND METHODS

Chemicals

All cell culture medium components (DME/F-12 mixture, sodium bicarbonate, glutamine, gentamicin, and fetal bovine serum) were purchased from Sigma Chemical Co. (St. Louis, MO). Glucose, mercaptoethanol, sodium dodecyl sulphate, myo-inositol, hydrogen peroxide, cyclohexanediaminetetraacetate, and ethidium bromide were purchased from Sigma. Magnesium chloride, urea, and sodium hydroxide were obtained from Fisher Scientific (Fairlawn, N J). Sodium phosphate monobasic and trypan blue dye were purchased from EM Chemicals (Cherry Hill, N J).

Cells

Mouse-mouse hybridomas, cell line HyHEL-10 (Courtesy of Dr. Sandra J. Smith-Gill, National Cancer Institute), producing antibodies against lyso- zyme were used in this study.

Culture medium

Dulbecco's Modified Eagle's Medium (DMEM) was prepared by mixing DME/F-12 mixture, 1.2 g of sodium bicarbonate, 17.5 ml of 200 mM solution of glutamine, 0.12 ml of mercaptoethanol, and 1 ml of gentamicin (all reagents purchased from Sigma) for each liter of final medium. After mixing, the final solution was sterilized by filtration and stored at 4°C in sterile media bottles. Immediately prior to use, the final formulation was obtained by adding fetal bovine serum (FBS) up to 4% vol/vol using sterile technique.

Master cell bank

Master cells are stored in liquid nitrogen. Every 2 months (or about 50 generations), new cells are thawed and propagated in DMEM plus 10% FBS for about 2 weeks. After that time, cells are cultured in DMEM plus 4% FBS and become part of a working cell bank.

Transfer protocol

To obtain a constant supply of actively growing cells, a working cell stock was maintained by periodic subculturing of cells in DMEM plus 4% FBS. Cells were transferred every 48 h, 10% inoculum in each transfer, placed into an incubator (VWR Scientific, Philadelphia, PA), and cultured under controlled conditions of CO2 (5%), temperature (37°C), and hu- midity.

Hyperoxia-induced DNA damage 269

Bioreactor

A 2.3 liter working volume Applikon bioreactor (total volume 3 liters, Applikon, Foster City, CA) was used for these experiments. Temperature control at 37°C was achieved through a thermostatic bath. Dis- solved oxygen was controlled by connecting a dissolved oxygen probe to a CeUigen controller (New Brunswick Scientific, Inc., Edison, N J), which pos- sesses a gas delivery control microprocessor. DO levels can be controlled at any desired level by the appro- priate mix of inlet gases, which is automatically done by the microprocessor within the CeUigen controller. The gas supply was initially done through head space aeration and was switched to sparging at the onset of DO step changes. This was also done for the controls to eliminate artifacts that may have resulted by sparging. Cells were gently agitated at a rotational speed of 100 rpm. pH was monitored and controlled to 6.9 __ 0.1 by addition of CO2 in the gas mixture and by sterile manual additions when required of 1 N NaOH through a septum. Cell density was measured by counting cells directly in a hematocytometer chamber. Cell viability was measured by trypan blue exclusion.

Exposure to hyperoxia

Cells were grown under the conditions described above. At mid-exponential phase, the oxygen level was changed by sparging oxygen until a desired preset value was reached. Subsequently, the set point was maintained by automatic control through sparging. For the control batch, 10% DO was maintained during the whole run.

DNA strand break assay

The principle of the strand break assay is based on DNA unwinding in an alkaline environment. 15 If strand breaks are present, then each break acts as a pivot and increases the rate of unwinding. A blank (B) measurement is made of background fluorescence in a sonicated sample caused by free dye (ethidium bromide) and cell components other than double stranded DNA. A second sample (T) is used for mea- suring total fluorescence of double stranded DNA plus contaminants. The difference between T and B provides an estimate of total double stranded DNA. A third sample (P) is used to estimate the unwinding rate of the DNA by exposing it to an alkaline environment. The difference between the P and B values is a measure of double stranded DNA remaining. The results are reported as (P - B)/(T - B) X 100%. Specifi- cally, a 9 ml sample was drawn from the bioreactor

and divided in nine tubes marked T 1, T2, T3, B 1, B2, B3, P1, P2, and P3 (1 ml per tube). All the tubes were centrifuged at 3000 rpm for 5 min at 4°C. The super- natant was discarded, and the tubes containing the cell pellet were placed in an ice bath. In darkness (only subdued incandescent light was allowed at this stage), 0.2 ml of solutions B (myoinositol 0.25 M, sodium phosphate monobasic 10 mM, magnesium chloride 1 mM, pH 7.2) and C (urea 9 M, NaOH 10 mM, cyclohexanediaminetetraacetate 2.5 mM, sodium dodecyl sulphate 0.1%, pH 12) were added to all the tubes, and 0.4 ml of solution F (glucose 1 M, mercaptoethanol, 14 mM, pH 2.5) was added to all three T tubes only. After 10 min, 0.1 ml of solutions D (0.45 volumes of solution C and 0.55 volumes of NaOH 0.2 N, pH 13.5) and E (0.4 volumes of solution C and 0.6 volumes of NaOH 0.2 N, pH 13.5) were added to all tubes. Then all three B tubes were sonicated for 1 min and placed back into the ice bath. After 30 min of adding solutions D and E, all tubes were very carefully (without shaking) placed in a 15 °C water bath for ex- actly I h. After this time, all tubes were again transferred very carefully and placed back into the ice bath, and immediately 0.4 ml of solution F was added to all B and P tubes. After 15 min, 1.5 ml of solution G (ethidium bromide 6.7 lzg/lt and NaOH 13.3 mM) was added, and all tubes were well mixed by vortexing and stored at 4°C. Fluorescence was measured the following day at 545 nm excitation and 575 nm emis- sion in a fluorescence microplate reader. This mini- mizes the error in the measurement of fluorescence from the sample. Ethidium bromide tends to form clusters as it binds with DNA. These clusters tend to cause errors in the fluorescence readings taken in a spectrofluorimeter cuvette. Therefore, as a way to eliminate any instability of the signal during the read- ing, each one of the sample tubes is split into eight to nine microplate wells. An average of the readings for each well thus yields superior results by accounting for sample inhomogeneities.

Lactate

Assay was on an enzymatic analyzer (Model 2700, Yellow Springs Instruments Co., Yellow Springs, OH).

RESULTS AND DISCUSSION

Immediately after elevating the DO level from 10% to any other value of hyperoxia used in our experiments (200, 300, and 476%), cell growth was affected to extents dependent on DO (Figs. 1 and 2). In only one case, 200% DO, cells continued growing although

270 M.A. CACCIUTTOLO et al.

u

u u

u u

2.0¢+6

1.6e+6 •

1.4e+6 "

1.2e+6 •

1.0e+6 •

8 . 0 e + 5 •

6.0e+5

4.0e+5 •

2.0e+5 •

0.0e+0 •

0

m

200% DO

~ I ~ 300% DO

4 7 6 % D O

Control (10% D O )

20 4 0 60 80 100 120 140 160 180

Time (hrs)

Fig. 1. Cell growth profiles under different dissolved oxygen (DO) levels.

at a slower rate than that of the control. In all other cases, once cells ceased to grow, the final outcome after a plateau period was cell death. Concomitant with the observed effect on cell growth, there was a

marked effect in the rates of both lactate production (Fig. 3) and glucose consumption (not shown). As other workers have also observed, 9'16 it is evident that hyperoxia induces major alterations in cell metabo-

.O °~

100

90

80

70

60

50

40

30

20

10

0

0 20 40 60 80 I00 120 140 160 180

Time (hrs)

Fig. 2. Viability profiles under different dissolved oxygen (DO) levels.


3.0

t~ .2

2 .5

2.0

1.5

1.0

0.5

0.0

Do~

0 - - - Control (10% DO) s 200% DO ~- 300% DO a 476% DO

0 20 40 60 80 100 120 140 160 180

Time (hrs)

Fig. 3. Lactate production profiles at different DO levels. The transition to different DO levels was at the time shown in Fig. 1.

lism. Lactate production ceased immediately after exposure to hyperoxia, independently of the level ofhyperoxia. This is not, however, an indication of altered glucose metabolism but rather a consequence of the

arrest in cell division. As Fig. 4 shows, cells maintain a similar ratio of lactate produced per viable cell well beyond (about 60 h) the time of exposure to hyperoxia. Only at 476% DO there is a sharp increase in

O

¢U

tJ

4.o i ,,

~.~ ~-~o~ i,o .~, ' - T - ~ o ~ , , o ~ / " :: / 3 / i /

{ ' ! !

0.S i I I I

0 . 0 I , t i i ! | | i

0 20 40 60 80 100 120 140 160 180

Time (hrs)

Fig. 4. Specific lactate production profiles under different DO levels. The values shown correspond to the ratio between the data provided in Figs. l and 3.

272 M . A . CACCIUTTOLO et al.

specific lactate production 10 h after exposure, but in this case it is likely due to the rapid loss in viability and not due to an increased rate of glycolysis.

The effect of hyperoxia on DNA integrity was ana- lyzed in greater detail than that on growth or other metabolic quotients. To follow the time course of DNA damage, samples were taken at 15-min intervals (this being the minimum sampling time required for sample processing) after the change in oxygen tension. At the highest DO level tested in this work (476% with respect to air saturation), a drastic effect on DNA integrity was observed (Fig. 5). Since it is difficult to observe the immediate consequences of the exposure to hyperoxia in this plot, the time scale right after the change in oxygen tension was expanded to observe in greater detail the response of DNA integrity (Fig. 6). A minimum DNA integrity (maximum extent of strand breakage) was observed after approximately 90 min of exposure to pure oxygen. Subsequently, DNA integrity appeared to recover to the initial level rather quickly, in about 30 min or less, presumably as a result of the action of repair enzymes that act once damage has been produced.

It is interesting to note that at all DO levels, a tran- sient response at the DNA level was observed (data not shown for other levels). A distinctive and repeti- tive minimum in DNA integrity (maximum in the extent of the damage) was observed, followed by

short-term recovery, presumably as a result of both enzymatic repair and the establishment of a new level ofantioxidants. Our results appear to indicate that the sustained oxidative challenge exhausts cellular de- fenses leading to growth inhibition. A possible expla- nation can be found in the action of the antioxidant enzyme glutathione peroxidase. Glutathione peroxidase is a well-known first line of defense against oxidative stress, which in turn requires glutathione as a cofactor. Among the many functions ofglutathione is the generation of the nucleotide precursors of DNA via the reduction of ribonucleotides to deoxyribonu- cleotides.17 Presumably glutathione levels are reduced because of its role as an antioxidative cofactor, which would result in altered DNA synthesis. Subsequent cell death apparently is the result of the steady exposure to hyperoxia, which causes metabolic dysfunc- tion at several levels.~'9 The fluctuations in DNA integrity observed within the first 5 min of exposure to hyperoxia (Fig. 6) are presumably the result of tran- sient adjustments of cell metabolism to a new steady state of free radical generation. Possible reasons for the observed fluctuations could be either activation of gene transcription and/or simply damage/repair ki- netics. However, it is not possible at this time to iden- tify the exact cause.

To quantify the extent of DNA damage and relate it with the external oxygen tension, the maximum net

m

° m

2e+6

2e+6 Intact DNA

l e + 6 -

l e + 6 •

8e+5 '

6e+5 •

j 4e+5 • Viable cells

2e+5 -

0e+0 |

0 20 ! | | |

4 0 6 0 8 0 1 0 0

100

90

80

70

60

- 50

- 40

- 3 0

- 20

- 10

0 1 2 0

Z

,=

T i m e ( h r s )

Fig. 5. D N A in tegr i ty and cell v iab i l i ty as a func t ion o f t i m e u p o n exposure to pure oxygen. The er ror bars represent e s t ima tes o f the s t andard dev i a t i on and the t va lues for a 90% conf idence level (descr ibed in legend to Fig. 6).


t_

lO1)

80-

7 0 - _

60

50

40 ..

20

10

0 3,t

' ~ I I I I I I

~ +/- 2.9 !

39.7 +/- 8.1 ~ IntactDNA . . . . . . . . . . . . . . . . . ~ . . . . . . . . . . . . ,b

D = 28.3 +/- 4.9

t = 90 min

I I I I • I I I i

35 36 37 38 39 40 41 42 43 44

Time (hrs)

Fig. 6. Maximum net DNA damage (D) is understood as the difference between DNA integrity levels from the time immediately before the change in oxygen level and the first min imum value of DNA integrity immediately after the challenge. We decided to use a time-averaged mean value of the DNA integrity prior to the challenge, which is the average of all DNA integrity measurements taken within 4 h of the challenge. By using the time-averaged mean value, rather than the instantaneous mean calculated from the data collected just prior to hyperoxia induction, the calculation is not biased by natural fluctuations in the baseline. The use of the time-averaged mean DNA integrity level is justified by examination of the error bars for the instantaneous means (calculated for a 90% confidence interval) collected over the 4-h interval. No statistically significant difference existed between these values. After the challenge, it is no longer possible to assume that the DNA integrity is constant with time, and data collected for a given sampling time must be considered separately.

The error bars for the initial (time-averaged) mean DNA integrity and the min imum DNA integrity values were determined with the Student t calculation for a 90% confidence level, with the standard relationship for the mean, ~t = x , ,~ +_ s t n _ l J n ~/2. An averaged standard deviation, s, was determined for the DNA integrity measurements (strand break assay) in a given experiment by determining the joint estimate of the variance, s 2. The value of s 2 was determined from the individual estimators, s~, calculated at each sample time. The integrity measurements done at each sample time i were done in triplicate. A min imum of 10 and a maximum of 20 sampling times were recorded in each experiment (i = 10 to 20); thus, s 2 was calculated with a min imum of 10 sets of triplicate data. The t values were chosen based on the number of data points, n, used in the calculation of the average, xave, (for the initial mean DNA integrity calculations, n = 9 or 12; for the min imum DNA integrity values, n = 3).

DNA damage was determined in each experiment. Rather than compare the absolute DNA integrity levels, which vary with the initial basal level of DNA breakage, we determined the difference between the initial level of DNA integrity and the point of minimum DNA integrity after exposure to hyperoxia. The error associated with the maximum net damage values was calculated using standard statistical methodol- ogies (described under Fig. 6). In this way one can rigorously determine whether a statistically meaningful difference between the basal and maximum DNA damage exists (for a specified confidence interval) in a given experiment. Additionally, one can unambigu- ously determine whether statistically relevant differences exist between the net DNA damage values determined from experiments conducted at different DO levels.

Following this rationale, the maximum net DNA damage corresponding to 476% DO is found to be 28.3 + 4.9%; the calculation is detailed in Fig. 6. The same analysis was repeated for every level of hyperoxia tested in this work (10, 200, and 300%). The re- suits of duplicate batch experiments showed that the net DNA damage was reproducible, and comparison between DO levels shows measurable, statistically meaningful differences. Figure 7 summarizes our ob- servations. There is a monotonically increasing relationship between the level ofhyperoxia and the extent of DNA damage (Fig. 7).

By comparing the results shown in Figs. 1, 2, and 7, it appears that although 200% DO causes measurable DNA strand breakage, it is sublethal since cellular defense mechanisms appear to be able to cope with DNA damage and cells continue to grow, albeit to a

274 M.A. CACCIUTTOLO et al.

z

50

40

30

20 ¸

10 t N " , a u ! u I

0 100 200 300 400 500 600

Dissolved oxygen (%)

Fig. 7. The effect of DO level on the extent of DNA damage shows a monotonic increase in damage with increasing DO. Each point represents duplicate measurements from two independent experiments. The error bars are determined from estimates of the standard deviation and the t values for a 65% (one standard deviation) confidence level. The confidence interval at which statistically relevant differences are noted is likely to rise as the number of repeat experiments is increased.

lesser extent than the control. At 300% DO, cells appear to cease to grow once the DO challenge begins. After an approximate plateau of 40 h, cells begin to die. At the highest level of DO tested (476%), the great- est extent of damage to DNA was observed and cells appeared to begin to die within 2 h of the challenge. It therefore appears that the sublethal DO threshold for this cell line is around 300%.

In contrast to oxidative stress induced by hyperoxia, cells have a qualitatively different response when exposed to a bolus of hydrogen peroxide. As shown in Fig. 8, the DNA strand breakage levels at an exposure of 4.2 #M and 16 #M H 2 0 2 a re comparable to those at a DO challenge of 200% and 476%, respectively (Fig. 7). However, the induction time for maximal breakage is on the order of 10 rain for H202 as compared to around 90-180 min for hyperoxia (data not shown). The response of DNA against a challenge with hydrogen peroxide is invariably rapid. Maximal damage is usually observed within the first 15 min of exposureJ 8 The differential response is probably due to the vastly different nature of the oxidative stress imposed upon the cells. Whereas added H 2 0 2 diffuses directly into the cell and damages DNA, hyperoxia- induced production ofH202 is expected to take a little longer to exert its effects. Our results appear to agree with this mechanism. Recent evidence also suggests

that the damaging effects of H202 are potentiated by carders that can extend the diffusion range of H202 (Ref. 19). Our results agree with those obtained previ- ously in literature for similar levels of hydrogen peroxide exposure. 2°

Cell viability did not change during the immediate time of exposure (within the first l0 h) neither for hyperoxia (200, 300, and 476%; Fig. 2) nor for sublethal hydrogen peroxide (up to 40 #M; Fig. 9), sug- gesting that (l) cell viability does not contribute to any variability in the DNA integrity data, and (2) that DNA damage may or may not cause cell death in the short term (within the first l0 h of exposure). Cells are able to repair DNA even at elevated hyperoxia levels such as 476% (as shown in Fig. 6), and therefore later cell death is presumably due to more widespread metabolic dysfunctions rather than a direct effect of DNA damage.

We have shown here clear evidence of DNA damage in cells associated with exposure to a series of hy- peroxic DO levels. Furthermore, such stress can result in damage that can be lethal or sublethal, and its extent can be quantitatively measured. The methodol- ogy described here is a useful paradigm for under- standing and in assessing the mechanisms by which cellular damage is likely to occur in vivo. Although we acknowledge that our preliminary data do not allow


80

70

60

50 tl

40 Z

'~ 30

~ 20 r~

10-

!

0 "1 | | i | ! i e ! !

0 5 10 15 20 25 30 35 40 45 50

H y d r o g e n p e r o x i d e c o n c e n t r a t i o n (p~M)

Fig. 8. DNA integrity upon exposure to a bolus of H202. The error bars represent estimates of the standard deviation and the t values for a 65% (one standard deviation) confidence level. These experiments were carried out in T-flasks. Sixty milliliters of mid-exponential phase cells were collected from the incubator, centrifuged, and resuspended in 60 ml of prewarmed DMEM plus 4% FBS. To this suspension, H202 was added to give the final concentration. An equivalent volume of distilled water was added to control samples. Each point represents the minimum DNA integrity observed after the addition ofH202. Cells exposed to 4.2 #M showed recovery from the damage to pretreatment levels within an hour, while those exposed 16 #M and 40 uM H202 respectively did not (data not shown).

.o .a

lO0

90-

$o-

70-

60-

50-

40-

30-

20-

1o

Cells (Control) I Cdls (4.2 pM) ceus (ts ~ ) Cells (40 ~,vl)

0 I I I I I

-100 0 100 200 300 400 500 600

T i m e (min)

Fig. 9. Viability profiles upon exposure to a bolus of hydrogen peroxide.

276 M.A. CACCIUTTOLO et at.

us to establish a direct relationship between hyperoxia and DNA damage since further mechanistic studies are required to do so, this has a potential relevance for the elucidation of the link between DNA damage and cellular malfunction/death caused by oxidative stress.

Acknowledgements - - This work was supported by National Insti- tutes of Health grant R01 RR06562-01 and by a Presidential Young Investigator Award from the National Science Foundation to G.R. We thank Dr. Amy Fulton and Mr. Keith Lee for their generous assistance with the DNA strand break assay and Dr. Ted Cadman for his useful suggestions on statistical analysis of the data.

REFERENCES

1. Halliwell, B.; Gutteridge, J. M. C. Free radicals in biology and medicine. 2nd ed. Oxford, England: Clarendon Press; 1989.

2. Floyd, R. A. Role of oxygen free radicals in carcinogenesis and brain ischemia. FASEB J. 4:2587-2597; 1990.

3. Stryer, L. In: Biochemistry. New York: W. H. Freeman Pub- lishers; 1988:401.

4. Pryor, W. Oxy-radicals and related species: Their formation, lifetimes and reactions. Ann. Rev. Physiol. 48:657-667; 1986.

5. Dizdaroglu, M.; Nackerdien, Z.; Chao, B. C.; Gajewski, E.; Rao, G. Chemical nature of in vivo DNA base damage in hydrogen peroxide-treated mammalian cells. Archiv. Biochem. Biophys. 285(2):388-390; 1991.

6. Borg, D. C.; Schaich, K. M. Iron and hydroxyl radical in lipid peroxidation: Fenton reactions in lipid and nucleic acids co-ox- idized with lipid. In: Cerutti, P. A.; Fridovich, I.; McCord, J. M., eds. Oxy-radicals in molecular biology and pathology. New York: Alan R. Liss, Inc.; 1988:427-441.

7. Davies, K. J. A. Protein damage and degradation by oxygen radicals. I General Aspects. Z Biol. Chem. 262(20):9895-9901; 1987.

8. Cantoni, O.; Brandi, G.; Cerutti, P.; Meyn, R. E.; Murray, D. Mechanisms of hydrogen peroxide cytotoxicity in mammalian and bacterial cells. In: Castellani, A., ed. DNA damage and repair. New York: Plenum Press; 1989:281-290.

9. De Groot, H.; Noll, T. The role of physiological oxygen partial pressures in lipid peroxidation. Theoretical considerations and experimental evidence. Chem. Phys. Lipids 44:209-226; 1987.

10. Boveris, A.; Chance, B. The mitochondrial generation of hydrogen peroxide. Biochem. J. 134:707-716; 1973.

11. Bailey, J.; Ollis, D. F. Biochemical engineering fundamentals. New York: McGraw-Hill; 1986:322.

12. Floyd, R. A. Role of oxygen free radicals in carcinogenesis and brain ischemia. FASEB 3. 4:2587-2597; 1990.

13. Lubiniecki, A. S.; Wiebe, M. E.; Builder, S. E. Process valida- tion for cell culture-derived pharmaceutical proteins. In: Lu- biniecki, A. S., ed. Large-scale mammalian cell culture technol- ogy. New York: Marcel Dekker; 1990:533.

14. Doetsch, P. W.; Hamilton, K. K.; Rapkin, L.; Okenquist, S. A.; Lenz, J. Base excision repair enzymes from eukaryotic sources. In: Wallace, S. S.; Painter, R. B., eds. Ionizing radiation damage to DNA: Molecular aspects. New York: Wiley-Liss, Inc.; 1990:109-125.

15. Birnboim, H. C.; Jevcak, J. J. Flourometric method for rapid detection of DNA strand breaks in human white blood cells produced by low doses of radiation. Cancer Res. 41:1889- 1892; 1981.

16. Jamieson, D. Oxygen toxicity and reactive oxygen metabolites in mammals. Free Radic. Biol. Med. 7:87-108; 1989.

17. Meister, A. Glutathione deficiency produced by inhibition of its synthesis, and its reversal; applications in research and ther- apy. Pharmac. Ther. 51:155-194; 1991.

18. Baker, M. A.; He, S. Elaboration of cellular DNA breaks by hydroperoxides. Free Radic. Biol. Med. 11:563-572; 1991.

19. Schubert, J.; Wilmer, J. W. Does hydrogen peroxide exist "free" in biological systems? Free Radic. Biol. Med. 11:545- 555; 1991.

20. Schraufstatter, I. U.; Hinshaw, D. B.; Hyslop, P. A.; Spragg, R. G.; Cochrane, A. G. Oxidant injury of cells. J. Clin. Invest. 77:1312-1320; 1986.

hyperoxia induces dna damage in mammalian cells

Documents