0031-8655/90 $03.00+0.00 Copyright 0 1990 Pergamon Press plc

Photochemktty and Photobiology Vol. 52, No. 3, pp. 525-532, 1990 hinted in Great Britain. All rights reserved

EFFECT OF POLYAMINE DEPLETION ON DNA DAMAGE AND REPAIR FOLLOWING UV IRRADIATION OF HeLa CELLS RONALDD. SNYDER* and PRASADS. SUNKARA Merrell Dow Research Institute, 2110 E. Galbraith Rd., Cincinnati, O H 45215, USA (Received 21 September 1989; accepted 30 January 1990)

Abstract-Treatment of HeLa cells with the polyamine biosynthesis inhibitors, methylglyoxal bis(guany1hydrazone) (MGBG), difluoromethylornithine (DFMO) or a combination of the two, resulted in reduction in cellular polyamine levels. Analysis of UV light-induced DNA damage and repair in these polyamine depleted cells revealed distinct differences in the repair process relative to that seen in cells possessing a normal polyamine complement. Initial yield of thymine dimers and rate of removal of these lesions from cellular DNA appeared normal in polyamine-depleted cells. However, depleted cells exhibited retarded sealing of DNA strand breaks resulting from cellular repair processes, reduced repair synthesis and an increased sensitivity to UV killing. Incision at damaged sites was not affected since ara-C repair-dependent breaks accumulated in a normal fashion. Molecular analysis of inhibited repair sites by exonuclease I11 and T4 DNA ligase probes suggest that the strand interruptions consist of gaps rather than ligatable nicks, consistent with an interpretation of the repair defect being at the gap-filling stage rather than the ligation step. Observed patterns of differential polyamine depletion by DFMO and MGBG, and partial reversal of repair inhibition by polyamine supplementation, suggests that polyamine depletion per se, rather than some secondary effect of inhibitor treatment, is responsible for the inhibition of repair. INTRODUCTION

Polyamines are naturally occurring cellular cations present in all living organisms. The exact roles that polyamines play in support of normal cellular functions are not well understood. Much of what we know regarding roles of polyamines in mammalian cells has been derived from studies in which the naturally occurring polyamines-putrescine, spermidine and spermine-have been partially or completely depleted by treatment of those cells with inhibitors of polyamine biosynthesis. Thus, difluoromethylornithine (DFMO)t, a specific inhibitor of ornithine decarboxylase (Metcalf et al., 1978), reduces putrescine and spermidine levels, and inhibitors of the enzyme S-adenosyl methionine decarboxylase, such as methylglyoxal bis(giany1hydrazone) (MGBG), reduce spermine levels. A striking consequence of polyamine depletion is the loss of proliferative potential. This finding has led to the discovery that polyamine depletion has therapeutic effects against certain neoplastic conditions (Sunkara, 1987, for review). In addition, polyamine depletion has been shown to alter the sensitivity of cultured mammalian cells to DNA interactive chemotherapeutic agents (Porter and Janne, 1987, for review) allowing for the possibility that chemotherapy in combination with polyamine depletion may provide an even more efficacious route to cancer treatment.

In order to most effectively exploit polyamine depletion chemotherapeutically, it is necessary to understand how polyamine depletion is related to altered sensitivity to DNA reactive drugs. In previous publications, we have demonstrated polyaminedepleted HeLa cells to have altered chromatin structure (Snyder, 1989a) and defective DNA repair following x-irradiation (Snyder, 1989b). While the alterations in chromatin structure could conceivably alter sensitivity to DNA reactive drugs, it was concluded that for repair of x-ray damage, at least, chromatin structure did not seem to be a principle factor in the inhibition of repair. Rather, the results were more consistent with the interpretation that polyamines were required in support of the repair machinery itself. This followed from the almost immediate reversal of repair inhibition upon polyamine supplementation to cells and the differential results obtained from the various inhibitor treatments. The present studies were conducted in order to determine if the repair inhibitory phenotype of polyamine-depleted cells extends to UV light-irradiated cells. Since the nature and the repair of the damage produced by x-rays and UV light are markedly different, these studies should provide further insight into the mechanism of repair inhibition accompanying polyamine-depletion. MATERIALS AND METHODS

*To whom correspondence should be addressed. tAbbreviations: DFMO, a-difluoromethylornithine; MGBG, methylglyoxal bis(guany1hydrazone); UDS, unscheduled DNA synthesis. PAP 5 2 3 - F

Cell culture. HeLa cells were seeded and maintained in Modified Eagles’ medium containing 10% fetal calf sequn, or 10% horse serum (with negligible polyamine oxidase activity) in experiments where polyamines were added to

525

526

RONALD D. SNYDER and PRASADS. SUNKARA

cultures. Cells were grown at 37°C in a C0,-controlled incubator. In all studies, inhibitors were added to cultures in the following manner: DFMO, 2.5 mM, 48 h; MGBG, 5 pM 24 h; DFMO/MGBG combined treatment was at the above concentrations with MGBG added during the last 24 h of DFMO treatment. Polyamine supplementation was for 3 h with a combined treatment of putrescine, spermine (each at 50 pM), and spermidine (10 F M ) . Polyamine determinations. Quantitation of cellular polyamines was accomplished by dansylation followed by reverse phase high pressure liquid chromatography, as described in detail elsewhere (Bowlin et al., 1986). A Waters system with 2 model 510 pumps, a model 680 automated gradient controller and a model 710B WISP Autoinjector was used. Colony forming ability studies. The ability of appropriately treated cells to form colonies was monitored as follows. Treated or untreated control cells were harvested by trypsinization and reseeded at a density of 250-5000 cells per 60 m m tissue culture dish. Colonies were grown for 10 days with one medium change. Cultures were then fixed with methanol, stained with Dif-Kwik, dried, and the colonies counted. Untreated control cells exhibited between 50 and 65% cloning efficiency utilizing this protocol. Nucleoid sedimentation analysis. To determine DNA strand breaks, cells were subjected to sedimentation through neutral sucrose gradients as follows: cells were harvested in EDTA-saline and 2.5 X lo5 cells were immediately added to a 300 p.! lysis layer (composed of 1.9 M NaCI, 100 mM EDTA, 32 mM Tris-HCI, pH 8.0, and 0.5% Triton-X 100) on top of a 1530% linear neutral sucrose gradient. Gradients also contained Hoechst 33258 for DNA visualization. Following a 20 min lysis period, gradients were spun for 45 min at 17-19 K rpm in an SW 50.1 rotor at 18°C. The position of the nucleoid band was determined by DNNdye fluorescence under UV light. Variability was usually no greater than 10%. Pyrimidine dimer analysis. Cells were prelabeled for 3 days with 1 pCi/rn.! [3H]thymidine (60 Ci/mmol), appropriately drug-treated, UV-irradiated (20 J/m2 UV254;G E germicidal lamp delivering 1.2 J/m*/s at 19 in.), and harvested. Thymine-thymine dimer content was analyzed in DNA isolated by phenol : chloroform, formic acid hydrolysis, and then subjected to HPLC. Chromatography utilized a reverse phase C,, column, 0.1% tetrahydrofuran as mobile solvent and a flow rate of 0.5 mUmin. Dimer was resolved from monomer as two discrete peaks separated by approximately 3 min. Unscheduled DNA synthesis (UDS) studies. Repair synthesis following UV irradiation was measured autoradiographically. HeLa cells were grown, drug treated, and UV irradiated on compartmentalized microscope slides (LabTek). Following irradiation, cells were incubated with 3 FCi [3H]thymidine (60 Ci/mmol) for 1 h and were then processed for autoradiography by standard procedures. A minimum of 30 nuclei were scored per treatment group and the average number of nuclear grains determined. S phase nuclei were easily distinguished and excluded from analysis. Structure of inhibited repair sites. In order to determine the molecular nature of inhibited repair sites, a modification of the procedure of Cleaver (1983) was used. Briefly, HeLa cells were prelabeled with [14C]thymidine (0.25 FCi/dish; 55 mCi/mmol), UV irradiated, and allowed to repair in the presence of [3H]bromodeoxyuridine (3 FCi; 42 Cilmmol). Parental density double stranded DNA (non-replicated) was collected from isopycnic cesium chloride gradients (Cleaver, 1977), lyophilized and redissolved in 50 mM Tris-HC1, 50 mM MgCl,, pH 8.4. One kg aliquots were treated with 75 units of exonuclease 111 (Bethesda Research Labs) for 30 min in order to measure release of tritium labeled repaired DNA. Other

DNA aliquots were first incubated with 50 units T4 DNA ligase in buffer consisting of 100 mM Tris-HC1, 7 mM MgCI,, 1.0 mM ATP, pH 7.6 for 30 rnin at 37°C. This was followed by addition of exonuclease 111 and a further 30 min incubation at 37°C. In experiments not presented here, it was determined that this amount of T4 DNA ligase completely ligated Hind111 fragments of pBR322. After exonuclease digestion, the DNA was precipitated with 5% trichloroacetic acid and the amount of tritium in the soluble was determined.

RESULTS

Treatment of HeLa cells with DFMO, MGBG or a combination of the two (DFMOIMGBG) caused a reduction in cellular polyamine pools, the nature and extent of which is consistent with that observed by numerous previous investigators. Table 1 demonstrates that DFMO reduced putrescine and spermidine to undetectable levels; MGBG reduced spermine and spermidine levels substantially while markedly elevating putrescine content. The combined treatment resulted in reduction of all three polyamines. Polyamine content was shown to be partially restored by a brief (3 h) treatment with all three polyamines. Thus, this system in which polyamine levels can be artificially manipulated is useful in studying the effects of polyamines on cellular DNA repair. Figure 1 demonstrates that cells depleted of polyamines by DFMO or DFMO/MGBG treatment exhibit enhanced sensitivity to UV killing, whereas MGBG-treated cells showed a sensitivity similar to that of untreated controls. These observations suggested that, under certain conditions, inhibitortreated cells were defective in repair of UV-induced damage, or that they might have increased susceptibility to induction of that damage. An examination of the initial yield of UV-induced pyrimidine dimers reveals no significant difference between any of the treatment groups and the controls (Table 2). Moreover, removal of dimers from cellular DNA by the excision repair process appears similar in all groups when measured at 12 or 24 h post-irradiation. Measurement of small amounts of dimers removed at shorter times after irradiation (3-6 h) was not reproducible and could not be used for assessing repair at short times after irradiation. Analysis of repair at early times after UV insult was conducted in a number of ways. First, cells were examined for the appearance of DNA strand interruptions resulting from normal excision repair processing of damaged sites. Examination for such sites by the sensitive nucleoid sedimentation assay reveals that at 1 h post-UV, untreated control cells accumulate DNA strand breaks in a UV dosedependent manner. Thus, as shown in Fig. 2, a small but reproducible number of excision-related strand breaks occurs in control cells following 1.2 J/mz UV irradiation and a considerably increased number of such breaks results from irradiation at

527

UV repair and polyamine depletion Table 1. Effects of polyamine biosynthesis inhibitors on cellular polyamine levels % Control

Putrescine

Spermidine

Spermine

Untreated + polyamines 3 h

100 118

100 105

100 102

DFMO 2.5 mM, 48 h + polyamines 3 h

ND 31

ND 28

92 141

1405 1913

56 166

31 110

10 I8

8 36

30 85

MGBG 5.0 p M 24 h + polyamines DFMO (48 h)lMGBG (24 h) + polyamines 3 h

Polyamine additions were at 50 p M (putrescine, spermine) or 10 p M (spermidine) for 3 h prior to harvesting and determination of polyamine levels. Normal polyamine levels in untreated HeLa cells were approximately 450 (putrescine), 2400 (spermidine) and 3000 (spermine) pmol/106 cells. ND, not detected.

5 1 00.0 ..-

n

Q 0,

.-C

E 10.0 L

0 LL

x C

0

0

0

1.0

1.2

4

0Control

c

Q,

2Q, 0.1

I

5 10 15 20 25 30 35 40

UV Dose (J/M2) Figure 1 . Colony forming ability of UV-irradiated polyamine-depleted HeLa cells. Cells were inhibitor-treated, UV-irradiated with various UV doses, and tested for their ability to form colonies as described in Materials and Methods. Points with error bars (SD) were determined from 3 separate experiments. Other points were derived from one determination. The percentage colony forming ability in drug-treated, non-UV-irradiated cultures was set equal to 100 for each treatment group. Colony forming ability in the absence of UV-irradiation is approximately 60, 50, 55 and 25%. respectively, for Control, DFMO, MGBG and DFMO/MGBG treated cells. 0-0, Untreated controls; M, MGBG; A--A, DFMO; U, DFMOIMGBG.

6.0 J/m2. It is also shown in Fig. 2 that while MGBG-treated cells exhibit the same extent of DNA strand breakage as untreated controls, cells treated either with DFMO or DFMO/MGBG contain a significantly elevated number of repair related breaks. Inhibitor treatment in the absence of UVirradiation was shown to have no effect on DNA sedimentation in this assay (not shown).

UV Dose (J/M2) MGBG

6.0

DFMO

n

D/M

Figure 2. Accumulation of UV-dependent DNA strand breaks. Inhibitor-treated or untreated cells were UVirradiated with 1.2 or 6 J/mZ UV,,, and analyzed for DNA strand breaks by nucleoid sedimentation after 1 h incubation at 37°C in the absence of inhibitor. Error bars represent standard error of the mean from 4 determinations. *Significantly different from control at P = 0.05.

An examination of repair-related DNA strand breaks as a function of time after irradiation is shown in Fig. 3. Following 6 J/m2 UV irradiation, nucleoids derived from either untreated or MGBGtreated cells sediment to 60-70% of the unirradiated position after 4 h incubation. On the other hand, a large number of breaks persist in DFMO and DFMO/MGBG-treated cells, sedimentation not returning to control position until approximately 12 h post UV. In studies in which putrescine, spermidine and spermine were all added to inhibitortreated or untreated cells for 3 h prior to UV irradiation, this accumulation of breaks was much less marked (Fig. 4), suggesting a direct role of polyamines in the excision repair process. Attempts to determine which of the three polyamines might be the most important in this regard provided equivocal results. All three polyamines, when added

RONALDD. SNYDERand PRASADS. SUNKARA

528

Table 2. Effects of polyamine biosynthesis inhibitors on initial yield and removal of UV-induced pyrimidine dimers from HeLa cell DNA Treatment

TO

%T

IN TT

Control DFMO MGBG DFMOlMGBG

0.051 ? 0.008 (9) 0.045 ? 0.006

Ttzh

T IN TT Yo

0.037

?

0.010

(7) 0.036

?

0.006

(8)

(2)

0.051 ? 0.007 (5) 0.043 ? 0.005

0.034 -C 0.006 (6) 0.037 ? 0.010 (6)

(11)

T24h

%T

IN 'IT 0.029 5 0.006 (13) 0.031 2 0.003 (5) 0.027 5 0.006 (4) 0.033 ? 0.004 (8)

HeLa cells were prelabeled with L3H]thymidineand were then treated with inhibitor (DFMO-48 h, 2.5 mM; MGBG-24 h, 5 pM; DFMOMGBG, as above with MGBG added only during last 24 h). At zero time (To), 12 h (T12h)or 24 h (TZ4h)post irradiation, DNA was isolated, subjected to formic acid hydrolysis, and thymine (T) was separated from thymine dimer (TT) by reverse phase HPLC. Numbers are derived from the number of trials (in parentheses) and are presented with the standard error of the mean.

.- 90L

80--

w 70Y

CY

?

6050-

-0 0

40-

Time (h) Figure 3. Rate of return of normal sedimentation in UVirradiated HeLa cells. Appropriately treated cells were irradiated with 6 J/mZ UVZs4and then analyzed for DNA strand breaks by the nucleoid sedimentation assay after various periods of time in the absence of inhibitor. Error bars are standard error of the mean and are derived from 3 separate determinations. 0-0, Untreated control; M, MGBG; A-A, DFMO; W--O, DFMO/ MGBG.

individually, were capable of partially reversing the accumulation of UV dependent strand breaks, but the irreproducibility associated with these particular determinations precludes a convincing determination of this point. Sites undergoing repair are known to be sensitive to inhibitors of DNA synthesis such as arabinofuranosyl cytosine (ara-C) (Hiss and Preston, 1977; Dunn and Regan, 1979; Snyder et al., 1981) or aphidicolin (Snyder and Regan, 1981; Waters, 1981). This sensitivity is characterized by an accumulation of DNA strand breaks at repairing

Time (h) Figure 4. Effect of polyamine supplementation on the accumulation of UV-dependent DNA strand breaks. Experiments were conducted as in Fig. 3 except that 3 h prior to UV-irradiation, fresh media containing putrescine, spermidine and spermine were added to inhibitor-treated or control cells. DNA breaks were analyzed at various times by nucleoid sedimentation. Error bars are standard deviations from 3 to 5 separate determinations. 0-U, Untreated control; A-A, untreated control DFMOIMGBG; L M , plus polyamines; U, DFMOlMGBG plus polyamines.

sites, quantitation of which has been shown to provide an estimate of repair kinetics in mammalian cells (Snyder and Regan, 1981). Thus, as incision occurs at repairing sites, those sites are held open and give rise to long-lived strand breaks. In cells in which incision does not occur, such as in Xeroderma pigmentosum cells, no such breaks accumulated (Dunn and Regan, 1979). In order to determine

529

UV repair and polyamine depletion

suggest that incision proceeds normally in polyamine-depleted cells. Repair of UV-induced DNA damage may also be followed by monitoring repair synthesis. It is shown in Table 4 that repair synthesis, as measured autoradiographically, is highest in the first hour after irradiation and then progressively declines over the next 5 h in cells with normal polyamine contents. In polyamine-depleted cells this same pattern is observed except that cells treated with DFMO or DFMO/MGBG exhibited markedly reduced repair synthesis relative to controls. MGBG-treated cells were similar to controls in this regard. Analysis of uptake of tritiated thymidine into the acid-soluble compartment of cells indicated no differences between the treatment groups. Thus, inhibition of repair is observed both in DNA strand break accumulation and in repair synthesis studies and, in

if incision was occurring normally in polyaminedepleted cells, experiments were conducted in which ara-C-dependent breaks were assessed in UVirradiated cells. Table 3 demonstrates that whereas DNA strand breaks disappear with time in the absence of ara-C, in its presence there is little or no reduction in breaks over a 12 h period in all treatment groups. Moreover, ara-C-dependent breaks increase at least for the first 3 h in all treatment groups, consistent with previous reports showing saturation of ara-C sensitivity with time (Snyder ef al., 1981). Ara-C dependent strand breaks are slightly elevated in non-UV-irradiated DFMO and DFMO/MGBG treated cells. This increased basal level of breaks also contributes to the total number of breaks in UV-irradiated cells. At present, the nature of these ara-C sensitive lesions in polyamine depleted cells is not known. These studies, however,

Table 3. Accumulation of repair-dependent DNA strand breaks from cells incuhatcd in the presence of 2 p,M ara-C for various periods of time Rad equivalents Control Time No UV l h 3h 12 h

(-)AC 0 65 35 20

DFMO

(+)AC

(-)AC 0

20 130 210 195

125

85 35

MGBG

(+)AC

(-)AC

50 190 245 230

60 40 25

DFMOIMGBG

(+)AC

(-)AC

20 120 200 200

0 150 130 4(!

0

(+)*C 50 225 260 285

Cells were UV-irradiated with 1.2 J/mZand analyzed for DNA strand breaks after 1, 3 or 12 h in the presence or absence of 2 p,M ara-C (AC). X-ray doses from 20 to 300 Rads produce a linear decrease in sedimentation in neutral sucrose gradients which allows calibration of strand breaks in Rad equivalents. 300 Rads of x-irradiation results in approximately a 50% reduction in nucleoid sedimentation. Values are derived from only one set of experiments but are representative of results seen in a variety of independent experiments.

Table 4. Repair synthesis following UV-irradiation as detected by autoradiography Pulsc time (h)

Grainsinucleus

Yo Control

Control

0-1 1-2 2-3 5-6

45.7 i 13.2 (65) 39.8 i 15.6 (30) 31.6 f 9.9 (45) 18.8 f 8.4 (40)

100 100 100 100

DFMO

@1 1-2 2-3 5-6

32.5 f 11.4 (56) 28.8 f 13.5 (33) 23.2 f 8.3 (3Y) 11.2 t 5.2 (46)

71 72 60

DFMOiMGBG

0-1 5-6

24.4 4.7

9.8 (50) 4.8 (47)

54 52

MGBG

0- 1

42.6 f 18.7 (39) 20.4 i 10.3 (49)

93 109

Treatment

5-6

f f

73

Unscheduled DNA synthesis was determined by autoradiography following 1 h pulses of radioactive thymidine at various times after 20 Jim' UV irradiation. Grains were determined from a minimum of 30 nuclei ( n , number in parentheses) and the standard error of the mean was determined. Nuclei undergoing replicative synthesis were easily distinguished from repairing nuclei and werc excluded from analysis.

RONALD D. SNYDER and PRASADS. SUNKARA

530

both assays, only cells treated either with DFMO or DFMOlMGBG exhibit the effect. As a final line of investigation, it was of interest to see if the nature of the aborted repair sites in polyamine-depleted cells could be determined. It had previously been demonstrated that repair synthesis conducted in the presence of various inhibitors led to incomplete repair patches which were sensitive to exonuclease 111digestion. Further, those sites were deduced to be gaps rather than nicks since T4 DNA ligase was unable to close them (Cleaver, 1982, 1983). In the present studies, UVirradiated cells were allowed to repair in the presence of [ 3 H ] b r ~ m ~ d e ~ x y ~ r i dparental ine, density double stranded DNA was isolated from those cells via neutral cesium chloride sedimentation, and the release of tritium from that DNA was monitored before and after treatment with T4 DNA ligase. The data in Table 5 demonstrate that repair synthesis occurring in cells treated with DFMO or DFMO/ MGBG was much more sensitive to exonuclease 111 digestion than that in untreated control cells, consistent with the interpretation that repairing sites failed to seal properly in those cells. MGBG-treated cells, which did not accumulate breaks showed similar nuclease sensitivity to controls. T4 DNA ligase pretreatment did not decrease exonuclease sensitivity, suggesting that the aborted repair sites were not simple ligatable nicks. DISCUSSION

The primary purpose of the present study was to examine the UV-induced excision repair process in human cells which had been artificially depleted of polyamines. We have previously demonstrated that in this compromised condition, HeLa cells exhibited distinct deficiencies in their ability to reseal x-ray induced breaks (Snyder, 1989b). We further demonstrated that this effect was readily overcome by Table 5. Structure of inhibited repair sites in polyaminedepleted HeLa cell5 3H-radioactivity released by exonuclease I11 digestion (dprn) Treatment ~~

~

(- ) T4 Ligase

( +) T4 Ligase

147 (3) 1114 (23) 212 (6) 1448 (31)

211 (4) 1226 (24) 198 (6) 1627 (29)

~

Unt reatcd DFMO MGBG DFMOlMGBG

Parental density DNA from 1 X lo7 cells was isolated from neutral cesium chloride density gradients. Reaction tubes contained approximately SO00 cpm "-labeled DNA resulting from repair synthesis following 6 Jim2 UV irradiation and 2 h repair. Numbers in parentheses are the percent total 3H counts released into the acid soluble compartment by exonuclease digestion.

partial restoration of cellular polyamine pools. It was hypothesized that polyamines acted as catalysts for some step(s) in the x-ray repair process, but it was not possible to determine the exact step involved. An alternative explanation considered, however, was that polyamine depletion led to altered chromosome structure and that this was responsible for altering the cell's capacity for repair of x-ray breaks. Chromatin structure does, in fact, seem to be altered in HeLa cells upon treatment with polyamine biosynthesis inhibitors (Snyder, 1989a), but it seemed unlikely that this contributed to retarded repair since MGBG treatment resulted in as great a disruption of chromatin structure as did DFMO treatment, yet only the latter caused significant repair inhibition. The question of the mechanism of this altered response to x-irradiation is of some importance since polyamine depleted cells have been shown to have altered sensitivity to a variety of DNA reactive agents (Oredsson et al., 1982, 1984; Hung et al., 1983; Ducore and McNamara, 1986; Seidenfeld et al., 1986a.b; Ducore, 1987). Does this relate to chromatin structure, altered DNA repair, both, or neither? Our present studies demonstrate that the defect in DNA repair in polyamine-depleted cells extends to UV-irradiation. Ultraviolet light not only induces a totally different spectrum of DNA damage than x-irradiation, but the repair of that damage proceeds via a different pathway. Nevertheless, the alteration in repair appears similar in both situations. The sealing of x-ray induced DNA strand breaks and breaks resulting from the normal UV excision repair process is retarded as compared to untreated controls. In both instances, however, these breaks do eventually seal, and where measured, the lesion is removed from the DNA (Table 2). The retardation of this process following xirradiation is presumably not particularly detrimental to the cell since no marked increased sensitivity to x-ray killing is noted in polyamine-depleted cells (Snyder, 1989b). On the other hand, increased sensitivity to UV killing is observed in DFMO and DFMO/MGBG-treated cells (Fig. 1). Increased sensitivity is also seen following treatment of polyamine-depleted cells with alkylating agents, which also induce typical excision repair (DuCore and McNamara, 1986; Ducore, 1987; Oredsson et al., 1984). Elevated sensitivity to UV light as opposed to x-irradiation could be due in part to the fact that x-ray breaks seal in less than 1.5 h, whereas breaks resulting from aborted UV repair require up to 12 h to close. This difference in rate of sealing most likely reflects the different nature of repair at these two types of lesions. These studies extend our previous x-ray repair studies and indicate that the defect in repair is most likely in a step common to both short-patch (xray-like) and long-patch (UV-like) repair processes. Because the resulting strand interruptions are not

531

UV repair and polyamine depletion

susceptible to DNA ligase, they are most likely gaps. The 30-50% reduction in repair synthesis is also consistent with this notion. If the defect were at the ligation step, repair synthesis would not be affected particularly if, as seems to be the case, the number of sites initiating repair is normal. The observed reduction in repair synthesis most likely reflects interrupted repair, as evidenced by the increased release of repair label upon exonuclease digestion. A number of proteins might be envisioned to play a role in gap filling including DNA polymerases, topoisomerases and accessory proteins of both enzymes. DNA polymerase is known to be stimulated in vitru by polyamines, as are many other DNA metabolizing enzymes (Kleppe et al., 1981), but it is unknown if this occurs in vivo or whether this implies that a lack of polyamines would tend to inhibit DNA polymerase. Polyamine depletion of mammalian cells is known to retard DNA synthesis thereby lengthening S-phase. Replicon initiation rather than elongation is thought to be involved; however, this point requires verification (see Marton and Morris, 1987, for review). DNA topoisomerase activity has similarly been shown to be modulated by polyamines (Srivenugopal and Morris, 1985; Srivenugopal et a l . , 1987). Polyamine depletion has been demonstrated to enhance the cytotoxicity of DNA topoisomerase inhibitors (Zwelling et al., 1985; Dorr et a l . , 1986). Inhibition of DNA topoisomerase activity in polyaminedepleted cells may not be a likely mechanism for repair inhibition, however, since (1) DNA supercoiling does not seem to be affected in polyamine depleted cells (Snyder, 1989a) and (2) since inhibition of DNA topoisomerase apparently has no effect on the UV excision repair process (Snyder, 1987; Downes, 1988). Both DNA polymerase and DNA topoisomerase activity may also be regulated by ADP-ribosylation (Yoshihara et al., 1985; Darby et al., 1985). It has previously been demonstrated (Wallace et al., 1984) and we have confirmed (unpublished observations) that polyamine-depleted cells have a higher level of ADP-ribosyltransferase activity. We are exploring the possibility that elevated ADP-ribosylation of repair enzymes or chromosomal proteins may play a role in the inhibition of repair in polyamine-depleted cells. Acknowledgement-The authors wish to express their thanks to Greg Davis for conducting the HPLC analysis of pyrimidine dimers and polyamines. REFERENCES

of DNA damaged by physical and chemical mutagens. In Handbook of Mutagenicity Test Procedures (Edited by B. Kilbey, M. Legator, W. Nichols and C . Ravel), pp. 19-48. Elsevier, Amsterdam. Cleaver, J. E. (1982) Specificity and completeness of inhibition of DNA repair by novobiocin and aphidicolin. Carcinogenesis 3, 1171-1174. Cleaver, J. E. (1983) Structure of repaired sites in human DNA synthesized in the presence of inhibitors of DNA polymerases alpha and beta in human fibroblasts. Biochim. Biophys. Acta 739, 301-311. Darby, M. K., B. Schmitt, J. Jongstra-Bilen and H.-P. Vosberg (1985) Inhibition of calf thymus type I1 DNA topoisomerase by poly(ADP-ribosylation). EMBO J . 4, 2129-2134. Dorr, R. T., J.D. Liddil and E. W. Gerner (1986) Modulation of etoposide cytotoxicity and DNA strand scission in L1210 and 8226 cells by polyamines. Cancer Res. 46, , 3891-3895. Downes, C. S. and R. T. Johnson (1988) DNA topoisomerases and DNA repair. Bioessays 8, 179-184. Ducore, J. M. (1987) Alpha-difluoromethylornithine effects on nitrosourea-induced cytotoxicity and crosslinking in a methylation excision repair positive (mer+) human cell line. Biochem. Pharmacol. 36, 2169-2174. Ducore, J. M. and L. McNamara (1986) Effect of alphadifluoromethylornithine on L-phenylalanine mustardinduced cytotoxicity and DNA interstrand cross-linking in a human cell line in vifro. Cancer Res. 46, 1068-1072. Dunn, W. C . and J. D. Regan (1979) Inhibition of DNA repair in human cells by arabinofuranosyl cytosine: effects on normal and Xeroderma pigmentosum cells. Mol. Pharmacol. 15, 367-374. Hiss, E. A. and R. J. Preston (1977) The effect of cytosine arabinoside on the frequency of single-strand breaks in DNA of mammalian cells following irradiation or chemical treatment. Biochim. Biophys. Acta 478, 1-8. Hung, D. T., L. J. Marton, D. F. Deen and R. H. Shafer (1983) Depletion of intracellular polyamines may alter DNA conformation in 9L rat brain tumor cells. Science 221, 368-370. KkDDe. K.. A. Osland. V. Fosse. R. Male. I. Lossius. D.'Helland, J. R. Lillehaug, A. J. Raae, R. K. Kleppe and I. F. Nes (1981) Effect of polyamines on enzymes involved in DNA repair. Med. Biol. 59, 374-380. Marton, L J. and D. R. Morris (1987) Molecular and cellular functions of polyamines. In Inhibition of Polyamine Metabolism. Biological Significance and Basis for New Therapies (Edited by P. P. McCann, A. E. Pegg and A. Sjoerdsma), pp. 7%105. Academic Press, Orlando, FL. Metcalf, B. W., P. Bey, C. Danzin, M. J. Jung, P. Casera and J. P. Vevert (1978) Catalytic irreversible inhibition of mammalian ornithine decarboxylase by substrate and product analogs. J . Am. Chem. SOC. 100, 2551-2553. Oredsson, S. M., D. F. Deen and L. F. Marton (1982) Decreased cytotoxicity of cis-diamminedichloroplatinum(I1) by alpha-difluoromethylornithine depletion of polaymines in 9L rat brain tumor cells in vitro. Cancer Res. 42, 1296-1299. Oredsson, S. M., A. E. Pegg, L. Alhonen-Hongisto, D. F. Deen and L. F. Marton (1984) Possible factors in the potentiation of 1-(2-chloroethyl)-3-trans-4-methylcyclohexyl-1-nitrosourea cytotoxicity by alpha-difluoromethylornithine in 9L rat brain tumor cells. Eur. J . Cancer Clin OI@~ 20, 535-542. Porter, C . W. and J. Janne ( 1 m Modulation of antineoplastic drug action by inhibitors of polyamine biosynthesis. In Inhibition of Polyamine Metabolism. Biological Significance and Basis for New Therapies (Edited by P. P. McCann, A. E Pegg and A . Sjoerdsma), pp. 203-248. Academic Press, Orlando, FL. Seidenfeld, J., K. A. Kornar, M. F. Naujokas and A. ~~

Bowlin, T. L., B. J. McKown and P. S. Sunkara (1986) Ornithine decarboxylase induction and polyamine biosynthesis are required for the growth of interleukin2- and interleukin-3-dependent cell lines. Cell. Immunol. 98, 341-350. Cleaver, J. E. (1977) Methods of studying excision repair

532

S . SUNKARA RONALD D. SNYDER and PRASAD

L. Block (1986a) Effect of DFMO-induced polyamine depletion on human tumor cell sensitivity to antineoplastic DNA cross-linking drugs. Cancer Chemother. Pharmacol. 17, 16-20. Seidenfeld, J., K. A. Komar, M. F. Naujokas and A. L. Block (1986b) Reduced cytocidal efficacy for adriamycin in cultured human carcinoma cells depleted of polyamines by difluoromethylornithine treatment. Cancer Res. 46, 1155-1159. Snyder, R. D. (1987) Is DNA topoisomerase involved in the UV excision repair process? New evidence from studies with DNA intercalating and non-intercalating antitumor agents. Photochem. Photobiol. 45, 105-11 1. Snyder, R. D. (1989a) Polyamine depletion is associated with altered chromatin structure in HeLa cells. Biochem. J . 260, 697-704. Snyder, R. D. (1989b) Inhibition of X-ray-induced DNA strand break repair in polyamine-depleted HeLa cells. Int. J . Radiat. Biol. 55, 773-782. Snyder, R. D., W. L. Carrier and J. D. Regan (1981) Application of arabinofuranosyl cytosine in the kinetic analysis and quantitation of DNA repair in human cells after ultraviolet irradiation. Biophys. J . 35, 339-350. Snyder, R. D. and J. D. Regan (1981) Aphidicolin inhibits repair of DNA in UV-irradiated human fibroblasts. Biochem. Biophys. Res. Commun. 99, 1088-1094. Srivenugopal, K. S . and D . R. Morris (1985) Differential modulation by spermidine of reactions catalyzed by type 1 prokaryotic and eukaryotic topoisomerases. Biochemistry 24, 4766-4771.

Srivenugopal, K. S., D. E. Wemmer and D. R. Morris (1987) Aggregation of DNA by analogs of spermidine; enzymatic and structural studies. Nucleic Acids Res. 15, 2563-2580. Sunkara, P. S., S. B. Baylin and G. D. Luk (1987) Inhibitors of polyamine biosynthesis: cellular and in vivo effects on tumor proliferation. In Inhibition of Polyamine Metabolism: Biological Significance and Basis for New Therapies (Edited by P. P. McCann, A. E. Pegg and A. Sjoerdsma), pp. 121-140. Academic Press, Orlando, FL. Wallace, H. M., A. M. Gordon, H. M. Keir and C. K. Pearson (1984) Activation of ADP-ribosyltransferase in polyamine-depleted mammalian cells. Biochem. J . 219, 211-221. Waters, R. (1981) Aphidicolin: An inhibitor of DNA repair in human fibroblasts. Cardnogenesis 2, 795-797. Yoshihara, K., A. Itaya, K. Tanaka, Y. Ohashi, K. Ito, H. Teraoka, K. Tsukuda, A. Matsukage and T. Kamiya (1985) Inhibition of DNA polymerase a,DNA polymerase p, terminal deoxynucleotidyl transferase, and DNA ligase by poly(ADP-ribosy1)ation reactions in vifro. Biochem. Biophys. Res. Commun. 128, 61-67. Zwelling, L. A,, D. Kerrigan and L. J. Marton (1985) Effect of difluoromethylornithine, an inhibitor of polyamine biosynthesis, on the topoisomerase 11-mediated DNA scission produced by 4’-(9-acridinylamino)methansulfon-m-anisididein L1210 murine leukemia cells. Cancer Res. 45, 1122-1126.

Effect of polyamine depletion on DNA damage and repair following UV irradiation of HeLa cells.

Treatment of HeLa cells with the polyamine biosynthesis inhibitors, methylglyoxal bis(guanylhydrazone) (MGBG), difluoromethylornithine (DFMO) or a com...
704KB Sizes 0 Downloads 0 Views