- Email: [email protected]
Inhibition of DNA synthesis and cell death
154
Biochimica et Biopl(vsica A eta, 696 (1982) 154-162
Elsevier Biomedical Press BBA 91014
INHIBITION OF DNA SYNTHESIS AND CELL DEATH IAN R. RADFORD a, ROGER F. MARTIN a, LLOYD R. FINCH b and GEORGE S. HODGSON a " Biological Research Unit, Cancer Institute, 481 Little Lonsdale Street, Melbourne, 3000 and h R~zs'sell Grimwade School of Biochemistry, Universi(v of Melbourne, Parkville, 3052 Victoria (Australia)
(Received October 6th, 1981)
Key words: DNA synthesis inhibition," Cell death; Hydroa(vurea," Repetitive DNA sequence
The association between DNA synthesis inhibition and cell death in mouse L-cells was investigated using the drug hydroxyurea. This drug produces a preferential labelling of low molecular weight DNA and dose-response studies revealed a correlation between this effect and cytotoxicity. Investigation of the reassociation kinetics of DNA labelled during hydroxyurea inhibition showed an over-replication of middle repetitive sequences, but the concentration dependence of this effect was quite different to that of cytotoxicity.
Introduction
Materials and Methods
Anti-metabolites are the subject of intensive study due in part to their actual and potential use in cancer therapy. There have been m a n y studies at the biochemical level which have sought to elucidate the target enzymes (generally in the pathways for nucleic acid biosynthesis) for such drugs [1]. Similarly, there have been m a n y studies at the cellular level seeking to correlate cytotoxicity with factors such as drug dose, cell type and stage in the cell cycle [2]. However there is a gulf between these two approaches which has received little attention. In particular, the question of how inhibition of D N A synthesis results in cell death is essentially unanswered. This study concentrates on hydroxyurea, an inhibitor of D N A synthesis which is cytotoxic to S-phase cells [3], and attempts to establish a correlation between biochemical effects and cytotoxicity.
Chemicals. The sources for some chemicals have been outlined previously [4]. Ribonuclease TI (Aspergillus oryzae) was from Sigma, Pipes and pancreatic ribonuclease (bovine pancreas) were from Calbiochem. Unlabelled deoxyribonucleosides were purchased from Sigma or Calbiochem. Bio-Gel and hydroxyapatite ( D N A grade Bio-Gel G T P ) were obtained from Bio-Rad. S l nuclease ( A . oryzae) was from Miles and CsCI was from BDH. Solutions. Phosphate-buffered saline, p H 7.1, comprises 0.14 M NaCI, 3 m M KCI, 8 m M Na2HPO4, 1.5 m M K H 2 P O 4 and l m M glucose; EDTA-phosphate-buffered saline is phosphatebuffered saline supplemented with 0.6 m M EDTA; buffer A comprises 0.25 M NaC1/20 m M Tris-HC1 ( p H 7 . 5 ) / 2 m M EDTA; buffer B comprises 10 m M Tris-HC1 (pH 8.0)/0.25M s u c r o s e / 8 m M CaC12; p h e n o l / c h l o r o f o r m / i s o a m y l alcohol are in the ratio of 50:50:1 (v/v). Cells. The cells and growth conditions have been described previously [4]. Cells were counted with a Coulter Counter Model B.
Abbreviations: Pipes, 1,4-piperazinediethanesulphonic acid; SDS, sodium dodecyl sulphate; dThd, thymidine; BrdUrd. bromodeoxyuridine; dCyd, deoxycytidine; FdUrd, fluorodeoxyuridine. 0167-4781/82/0000-0000/$02.75 ~ 1982 Elsevier Biomedical Press
155
Cell labelling.
The methods for cell labelling and lysis have been described previously [4]. When uniformly labelled [3H]DNA was required, ceils were given three pulses of [3H]dThd (0.02 C i / m mol) at 40 n C i / m l ( 2 # M ) over a total labelling period of 24 h. Cell cloning studies. Log-phase cultures were incubated with a range of concentrations of the cytotoxic agent for 80 min. The drug was then removed by rinsing each flask four times with growth medium. The cells were rinsed with EDTA -phosphate-buffered saline then collected after incubation with 0.01% pronase and counted. Harvested cells were then diluted with growth medium and 5 ml aliquots containing 120 cells were dispensed into 50-mm-diameter plastic petri dishes. Control cultures were treated identically with respect to rinsing, harvesting and plating. The dishes were then incubated at 37°C in an atmosphere of 10% CO 2 in air for 7days. After this period the medium was gently removed and the colonies were washed with phosphate-buffered saline, fixed for 20 min in neutral formalin solution (containing 4% f o r m a l d e h y d e / 3 0 m M N a H 2 P O 4 / 4 5 mM Na2HPO4) and stained with 0.01% crystal violet solution for 1 h before counting. A cell was considered to have retained viability if it gave rise to a colony composed of 50 or more cells. The data are presented as cell viability, which is calculated by dividing the mean number of colonies, obtained from at least 20 replicate dishes, at a particular drug concen'tration, by the mean number of colonies obtained from untreated cells. A standard deviation was then calculated for this ratio. Control cloning efficiencies of around 70% were routinely obtained. Purification of DNA. Mouse DNA was purified from two sources. Unlabelled D N A for use as driver in D N A renaturation experiments was obtained from nuclei isolated from the livers of 3month-old male and female B a l b / c mice that had been starved for 24 h prior to killing, in order to decrease glycogen levels. The procedure used was a modification of the method of Prashad and Cutler [6]. After killing the mice by bleeding under anaesthesia, the livers were excised and rinsed in ice-cold buffer B and then carefully cut into small pieces, fibrous tissue being dissected out. The pieces were homogenized in ice-cold buffer B by
six passes in a homogenizer with a glass cup and Teflon pestle. The pestle was rotated at about 1200 rev./min and had a clearance of 0.006 inch. Triton X-100 detergent was then added to 1% and the homogenate was stirred gently with a magnetic stirrer at 4°C for 5 min. Nuclei were then collected by three centrifugations of the homogenate at 500 × g for 5 min at 4°C, and the nuclear pellet was resuspended in 50 mM Tris-HC1 (pH 8.0)/1 mM E D T A buffer before taking a sample for microscopic examination. SDS was then added to 0.2% to lyse the nuclei. Proteinase K (pre-incubated at 37°C for 0.5 h) was added to 0.1 m g / m l and the lysate was incubated at 37°C for 1 h. The lysate was then sheared by three passages through a 23-gauge needle, followed by two extractions with phenol/chloroform/isoamyl alcohol. The solution was then extracted three times with chloroform, to remove traces of phenol, flushed with nitrogen and then concentration by rotary evaporation. The concentrate was applied to a Biogel P-30 column and equilibrated with 0.25 M NaC1/10 mM Pipes (pH 6.8), and the DNA in the void volume fractions was pooled and diluted with distilled water to give a final NaC1 concentration of 0.1 M. Pancreatic RNAase and RNAase T1 (both previously incubated at 80°C for 2 m in in 0.1 M N a C I / 1 0 m M Tris-HC1 (pH 7.5)/5 mM EDTA solution to inactivate any contaminating DNAases) were added to 20 # g / m l and 25 units/ml, respectively, and the mixture was incubated for 1 h at 37°C. The efficacy of the RNAase digestion was monitored by running a parallel digest of a sample of the DNA preparation to which [3H]RNA had been added. The elimination of acid-insoluble counts from this digest was confirmed before proceeding. After RNAase digestion the DNA preparation was incubated with proteinase K at 0.1 m g / m l for 1 h at 37°C. SDS was then added to 0.5% and the preparation was extracted three times with phenol/chloroform/isoamyl alcohol, followed by chloroform extraction, concentration and chromatography as described above. The purity of the DNA preparation was assessed from the ratio of its absorbance at 260 nm to that at 280 nm. Ratios of 1.80 or better were obtained [7]. Labelled DNA was obtained from L-cells. The D N A was purified as outlined above except that cells were lysed directly with 0.5% SDS/25 mM
156
Tris-HC1 (pH 7.5)/5 mM EDTA solution. Sonication of DNA. DNA fragments of about 450 nucleotides in length were prepared by sonication with a Branson S125 sonifier equipped with a microtip. The D N A preparation, in 0.25 M NaC1/10 mM Pipes (pH 6.8), was flushed with nitrogen before sonication at power level 2 using 16 pulsed exposures of 15 s each with 30 s cooling in ice between exposures. DNA renaturation conditions. DNA renaturation was performed in siliconized glass test tubes and the solution was overlaid with paraffin oil. Duplicate samples were taken for each point and were either assayed immediately or snap-frozen in an ethanol/solid CO 2 bath. DNA samples, denatured by immersion in a boiling water bath for 3 mins, were incubated in 0.25 M N a C I / 2 0 mM Pipes (pH 6.8) at 60°C [8,9]. DNA was present at a final concentration of between 5 and 1000/~g/ml. To obtain Cot values of 102 and greater, mouse liver D N A (see above) was added as driver. The equivalent Cot was calculated using a factor of 1.77 to correct for the Na + concentration used [10]. Equivalent Cot is the Cot ((absorbance at 260 n m / 2 ) X time (h)) multiplied by the ratio of the reassociation rate in the solution used to that in 0.12 M phosphate buffer. S 1 nuclease digestion. Digestions were performed in the presence of 0.25 M NaC1, 30 mM sodium acetate (pH 4.6), 25 mM dithiothreitol, 30 /~M ZnSO 4 and 0.1 m g / m l of D N A (where necessary the D N A concentration was made up to 0. l m g / m l by addition of sonicated calf thymus DNA). Samples were incubated with S1 nuclease (10/~g/ml) at 37°C for 30 minutes. The reaction was terminated by the addition of EDTA and Na4P207 to 5 mM and 2.5 mM, respectively. The extent of digestion was assessed by gel filtration on columns of Biogel P-30 which were washed and run with 0.25 M N a C I / 1 0 mM Pipes (pH 6.8) buffer. Fractions were collected and counted directly. The percentage of renatured D N A in each sample was calculated as 100 X [cpm in columnexcluded fractions)/(cpm in column-included fractions + cpm in column-excluded fractions)]. Hydroxyapatite fractionation. Samples for fractionation were diluted into 50 mM phosphate buffer (pH 6.8) and where necessary sonicated calf thymus DNA was added to give a total of 0.1 mg
DNA. The sample was then pumped through a column containing 1 ml hydroxyapatite equilibrated at 60°C in 50 mM phosphate buffer and the column was washed with 50 mM phosphate buffer to give a total volume for sample and wash of 10.5 ml. Insignificant amounts of the total radioactivity were eluted in this wash. Singlestranded DNA was eluted by pumping 5 ml 0.125 M phosphate buffer through the column, followed by 5 ml 0.30 M phosphate buffer to elute double-stranded DNA. The fractions obtained from these two elutions were then counted directly. In order to determine the level of crossover (i.e., the percentage of double-stranded DNA eluted during the single-stranded DNA wash and vice versa), reference samples of double-stranded and single-stranded D N A were prepared. It was found that the 0.125 M phosphate buffer wash released 5-10% of the double-stranded DNA, whilst 1-2% of single-stranded D N A eluted in the 0.30 M phosphate buffer wash. The percentage of DNA containing doublestranded regions was determined as 100 X [(cpm in 0.30 M' phosphate buffer)/(cpm in 0.125 M phosphate buffer + cpm in 0.30 M phosphate buffer)], where the counts in each wash were corrected for crossover. Analysis of Cot curves. Cot curves were analyzed using a computer least-squares fit [11], minimized according to the criterion of Y.,(O,E~) 2 (where O = o b s e r v e d value, E = c a l c u l a t e d value), for a maximum of three second-order components. An unconstrained fit was used. Hydroxyapatite and S~ nuclease data were fitted to formulae 1 [8,10] and 2 [12,13] respectively. Percent DNA hybridized = 3
P ~ f / ( 1 - (1 +kiCot)-' )
(1)
i--I
Percent DNA hybridized = 3
P ~,, f / ( 1 - (1 + kiCot) -°'44)
(2)
i--1
where P is the scaling parameter for each curve; f the fraction of P for each component; k i the rate constant of the component.
157
Results
Synthesis of low-Mr DNA and cytotoxicity The results presented in the preceding paper [4] suggest that hydroxyurea induces a state of uncoordinated DNA synthesis in cells. It was therefore of interest to see whether the dose-dependent killing of S-phase cells, produced by exposure to the drug, could be correlated with the dose response for accumulation of radioactive precursor in lowM r DNA. Accordingly, cells were treated with a range of concentrations of hydroxyurea for 80 rain and assayed for survival, whilst replicate cultures were labelled with [3 H]dThd and the lysates centrifuged through alkaline sucrose gradients. The previously reported [4] pattern of accumulation of radioactivity in low-Mr DNA was observed (Fig. 1). The results of the cloning assays (Fig. 1) correlate with these data. Both killing and accumulation of radioactivity in low-Mr DNA plateau at a hydroxyurea concentration of around 2raM. The maximum level of killing achieved, about 40%, correlated with a labelling index of 43% for untreated cells.
BO
\
•
60 8o c
80
.....
6o
40
~0
20
20
ill
I
I
I
2
~
6
8
Hydroxyurea ( rnM )
Fig. I. Cell killing and accumulation of label in low-M r D N A following hydroxyurea treatment. Cells were treated with the indicated concentration of the drug for 80 rain and cytotoxicity was then measured using a colony-forming assay (for details see Materials and Methods). Rcplicate cultures were treated with the drug for 1 h and [3H]dThd was then added for 20 min. The S D S / p r o t e i n a s e K lysates were centrifuged through alkaline sucrose gradients and the percentage of the total incorporated radioactivity in low-M r DNA was determined (for further details see previous paper [4]). • . . . . . . • , cell survival: ." 0 , percentage of incorporated label no greater than 21 S; the bars represent mean--+S.D.
Renaturation analyses Replication of the eukaryotic genome is thought to proceed by the initiation of synthesis in clusters of replicons in a strict temporal order [14]. It has been found that treatment of cells with S-phasespecific cytotoxics alters the pattern of initiation [15] and such a disturbance may be responsible for cytotoxicity. As a means of detecting possible alterations to the pattern of replication it was decided to compare the sequence complexicity of DNA labelled in the presence and absence of drug by hybridization analysis. DNA was purified from untreated cells that had been labelled with [3H]dThd for 10s or 24h and from cells that had been pretreated and labelled in the presence of hydroxyurea. The pulse -labelled DNA was included as a control because the 10 s pulse of [3H]dThd would allow an amount of DNA synthesis comparable to that occurring in the hydroxyurea-treated cells [4] and also since a report in the literature had claimed that preferential incorporation of [3 H]dThd into middle repetitive sequences occurred during brief labelling of CHO cells [16]. The reassociation kinetics of these samples, as assayed by hydroxyapatite chromatography, are shown in Fig. 2. It can be seen that the reassociation kinetics of the 24-h-labelled and 10-s -labelled DNA samples are very similar. However, the two preparations of DNA labelled in the presence of hydroxyurea show a marked difference to the control samples, in having a much higher rate of renaturation of sequences in Cot range 0-100. Analysis of the curves showed that this difference corresponded to an increased percentage of middle repetitive DNA sequences and a correspondingly lower level of unique sequence DNA (Table I); middle repetitive sequences increased from 11-14% in the controls to 22-29% of the DNA synthesized in hydroxyurea-treated cells. These experiments were repeated using S~ nuclease digestion to measure the renaturation kinetics and similar results were obtained (data not shown). There was a higher rate of renaturation of sequences in Cot range 0-100, so that approx. 10% more of the labelled DNA from hydroxyurea-treated cells behaved as middle repetitive DNA (Table I). It was next of interest to investigate the hydroxyurea dose response for the 'over-production'
158
O-
20 ¢-
9 40. 0 r~
60-
80L_
100 -3
I -2
I -1
I 0
Log equivelent
I 1
t 2
3
4
of middle repetitive DNA sequences with a view to seeing whether this effect could be correlated with cell death. Accordingly, DNA was purified from cells that had been pretreated and labelled in the presence of a range of concentrations (both toxic and non-toxic) of hydroxyurea. These samples were analysed over the Cot range 0-100 (Fig. 3). The Cot curves show that the DNA synthesized in cells treated with 5 mM hydroxyurea renatures considerably more rapidly than the control DNA, as previously shown. However, the renaturation rate of the 1 and 2 mM hydroxyurea samples is only marginally greater than that of the control, as shown by the curve gradients (parameter B in Table II). This is obviously quite different to the dose response for cytotoxicity (see Fig. 1), where 1
Cot
Fig. 2. Renaturation kinetics of D N A synthesized in the presence or absence of hydroxyurea, as analyzed by hydroxyapatite chromatography. Control cultures were labelled with [3 H]dThd for 2 4 h ( O O ) or 10s ( A . . . . . ~ ) . Two separate batches of cultures were treated with 5 m M hydroxyurea for 1 h and then labelled with [3H]dThd for 20 min (preparations 1
( + . . . . . . + ) and 2 ( x . . . . . . x ) ) . The D N A from these cultures was then purified and sonicated to 0.45 kilobasepairs in length for renaturation analysis. The curve through the data points is a computer generated least squares fit for a m a x i m u m of three second order components. An analysis of the data is presented in Table I. For all experimental details see Materials and Methods.
TABLE I RENATURATION OF 0.45-KILOBASEPAIR DNA FRAGMENTS, FROM HYDROXYUREA-TREATED AND UNTREATED L-CELLS, A N A L Y Z E D BY H Y D R O X Y A P A T I T E C H R O M A T O G R A P H Y A N D S I N U C L E A S E D I G E S T I O N T h e table presents a statistical analysis of the data from Fig. 2. For further details see Materials and Methods. C o m p o n e n t I represents highly repetitive sequences, repeated 1 - 106 or more times and with a Cot of less than I • 10 2. C o m p o n c n t 2 represents middle repetitive sequences, repeated I - 102 to 1,105 times and with a Cot of around I • 10 2 to 10. C o m p o n e n t 3 represents 'Unique' sequences with a Cot of more than 10. Sample
Component
Fraction Hydroxyapatite
S I nuclease
24-h-labelled
I 2 3
0.148 ± 0.014 0.138±0.014 0.714 a
0.165 ÷ 0.006 0.113 ±0.012 0.722
10-s-labelled
I 2 3
0.169 ± 0.018 0.112+0.017 0.719
0. I 14 ± 0.006 0.167+0.010 0.719
Hydroxyurea-labelled ( 1)
1 2 3
0.188 ± 0.004 0.216-+0.003 0.596
Hydroxyurea-labelled (2)
I 2 3
0.179 + 0.005 0.286 -+ 0.008 0.535
0.127 + 0.006 0.258 + 0.009 0.615
a The value for c o m p o n e n t 3 was obtained by s u m m i n g the values for c o m p o n e n t s 1 and 2 and subtracting them from unity.
159 T A B L E II
20
A N A L Y S I S O F Cot C U R V E S OF D N A S Y N T H E S I Z E D IN T H E PRESENCE O F D I F F E R E N T H Y D R O X Y U R E A CONCENTRATIONS The table presents a statistical analysis of the data from Fig. 3. The data were fitted to a curve of the form Y=AX -B, where Y is the percentage of labelled D N A b o u n d to hydroxyapatite and X is the Cot value. The estimated values for the curve parameters A ( Y - a x i s intercept) and B (curve gradient) are shown with their 955 confidence limits.
3C 9 \ \ \ \ \ \
40
Sample (mM)
\ \ X~
I
0
!
1 Log equivolent Cot
2
71'80 ~
A 20
10
o. u
0 I
i
e
.]1.8o
~ 20 .-
.
-°
10
0
i
I
10
20
Fraction No.
A
Confidence limits
B
Confidence limits
Control
25.93
24.77 27.15
0.0341
0.0201 0.0482
0.5
26.57
25.26 27.94
0.0489
0.0334 0.0645
1
23.90
22.80 25.06
0.0725
0.0576 0.0873
2
24.28
23.06 25.56
0.0663
0.0494 0.0833
5
22.91
21.17 24.80
0.143
0.119 0.167
I
Fig. 3. Renaturation kinetics of D N A synthesized in the presence of different hydroxyurea concentrations. Replicate cultures were treated with different concentrations of hydroxyurea for l h and [3H]dThd was then added for 20 min. The cells were then lysed and their D N A was purified. D N A obtained from cells that had been labelled with [3H]dThd for 24 h was used as a control. D N A renaturation was monitored by hydroxyapatite chromatography. Data points shown represent the mean of two determinations. Hydroxyurea (mM): × . . . . . . × , 5 m M ; + . . . . . . + , 2 m M ; /x . . . . . /x, l m M ; [] . . . . . [3, 0.5 mM; © ©, control.
E
Curve parameters
I
30
Fig. 4. Test for double replication of D N A segments following release from hydroxyurea treatment. Replicate cultures were incubated in medium containing 20 # M BrdUrd, 2 # M F d U r d and 1.7 # M [3H]dCyd (l C i / m m o l ) for l h. The cells were then washed, incubated in medium containing 20 # M dThd and 20 # M dCyd for 30 min and in growth medium for I h. During the latter and all subsequent steps colcemid was present at 0. I /~g/ml. The cultures were then incubated in the absence (A) or presence of 5 m M hydroxyurea (B) for 80 min. Following extensive washing, the untreated culture was incubated in m e d i u m containing []4C]dThd (4.6 # M , 56 m C i / m m o l ) for 30 min whilst the drug-treated culture was incubated with label for 4 h. The cells were lysed in 0.25 M N a C l / 2 0 m M Tris-HCl (pH 7 . 5 ) / 2 m M E D T A / 0 . 5 % SDS and then digested with 100 # g / m l proteinase K. The lysates were sheared by passage six times through a 26-gauge needle and then extracted twice with a p h e n o l / c h l o r o f o r m / i s o a m y l alcohol mixture. Following reextraction with chloroform the D N A was re-dissolved in 1.2 ml 0.2 M Tris-HC1 (pH 7.5)/20 m M E D T A and CsC1 solution was then added to a final density of 1.74 g / m l . The solutions were then centrifuged at 36000 r e v . / m i n for 65 h in a Beckman 50 Ti rotor at 20°C. Fractions of 0.25 ml were collected and analyzed for acid-insoluble radioactivity by precipitation onto filters with trichloroacetic acid. • • , 3H cpm; [] rT, 14C cpm; • . . . . . . • , density.
160 and 2 mM hydroxyurea were as cytotoxic to Sphase cells as was 5 mM hydroxyurea. Hence the dose response for this effect does not correlate quantitatively with that for cytotoxicity.
Abnormal re-initiation of DNA synthesis The above studies have sought an explanation for cell death by examining the DNA synthesized during drug treatment. However, the critical event leading to cell death might be occurring after the release of the cell from drug inhibition. One possibility of interest to study was whether there was induction of a second round of replication within the same S-phase following drug treatment [17]. Cells prelabelled with BrdUrd and [3 H]dCyd were treated with a toxic dose of hydroxyurea for 80 min and then labelled with [laC]dThd. The D N A was then purified and analyzed on neutral CsC1 gradients (Fig. 4). The profile from hydroxyureatreated cells is identical to that obtained from a control culture; 3H label is present in hybriddensity DNA, whilst 14C label is found solely in normal-density DNA. Hence no evidence was found for the induction of double replication following drug removal. Discussion
The D N A synthesis that is detected during hydroxyurea inhibition is characterised by a disproportionate synthesis of low-M r DNA fragments [4]. This paper reports a further abnormality of D N A synthesis during hydroxyurea inhibition, namely preferential synthesis of middle repetitive sequences. However, the central aspect of this report is the result of dose-response studies which reveal a correlation between preferential synthesis of low-M r DNA and cytotoxicity in hydroxyureatreated cells. The D N A synthesized during treatment with 5 m M hydroxyurea was found to be markedly enriched for middle repetitive sequences. This difference could not be ascribed to an artefact of the [3H]dThd labelling time used, as D N A isolated from untreated cells labelled for 10 s or 24 h had comparable renaturation kinetics. The latter result differs from that obtained in a study using CHO cells [16], where overrepresentation of middle repetitive sequences was found after short labelling
times. However, it concurs with a subsequent study by the same group in which HeLa cells were used [18]. The conflicting results were interpreted as being due to differences between cell types. How, then, might the anomalous D N A synthesis, occurring in the presence of hydroxyurea, be accounted for? Hydroxyurea treatment has been found to induce the accumulation of early replicative intermediates of both polyoma virus [19] and a bacterial plasmid [20]. Similarly it has been found that the replicons are smaller in mammalian cells that have been released from inhibition with FdUrd than they are in untreated control cells [21,22]. These results would suggest that the activation of sites for the initiation of D N A synthesis has continued during drug treatment. What might then be happening is that the cell's programme for replicon initiation continues in the presence of hydroxyurea, leading to the accumulation of initiation sites, with that DNA synthesis which does occur being close to these sites. It is then necessary to postulate that replicon initiation sites are enriched for middle repetitive sequences, although a study investigating the sequences rephcated following Xirradiation of CHO cells obtained indirect evidence for the enrichment of initiation sites for unique sequences [7]. Our results, however, would be consistent with the view that there is some sequence identity [23] amongst the approx. 1 • 105 initiation sites in the cellular genome. Our results with hydroxyurea-treated cells are strikingly similar to those found when the sequence composition of the DNA synthesized by nuclei isolated from normal or mitomycin Ctreated HeLa cells was investigated [24]. In both systems an increased synthesis of middle repetitive sequences was found. The over-replication of middle repetitive DNA sequences was found not to correlate with the concentration-dependent induction of cell death by hydroxyurea. This result, however, does not rule out the possibility that cell death is the result of an alteration to the pattern of replication. At the lower drug levels replication may be able to proceed sufficiently far past initiation sites so that there would be no detectable enrichment for sequences associated with initiation sites. The possible induction of double replication following drug treatment [17,25], as a cause of
161
cytotoxicity, was also examined. However, no evidence was found for its occurrence after hydroxyurea exposure. This is in agreement with the conclusion of others that hydroxyurea treatment does not induce double replication [26,27]. The results of the dose-response study of hydroxyurea cytotoxicity are consistent with a previous report [3]. Similarly, the results presented in this and the preceding paper [4] concur with previous osbervations that DNA synthesis must be decreased to a very low level by this drug before cell killing occurs [26]. Even when DNA synthesis is decreased to about 2% of the control level, as with a concentration of 0.5 mM hydroxyurea, cell killing is still slight. More interestingly, a correlation was found between cell killing by hydroxyurea and the incorporation of labelled precursor into low-Mr DNA in the drug-treated cells. We suggest that this correlation may provide an insight into the mechanism of cell death. The DNA fragments produced by hydroxyureatreated cells are attached to their template and presumably separated from each other and high-Mr DNA by single-stranded regions, as these drugs have been shown to have no effect on DNA ligase activity [28]. This has been verified for an SV40synthesizing system treated with hydroxyurea, where it was found that the DNA fragments could be joined only. when incubated with ligase plus a DNA polymerase and not by ligase alone [29]. The single-stranded regions which would be especially vulnerable to nuclease attack, are repairable by DNA polymerase fl which is relatively insensitive to hydroxyurea [28,30], but some inhibition doubtless occurs at higher drug concentrations. We postulate that the cytotoxicity of hydroxyurea arises from inhibition of DNA synthesis by the following sequence of events: (i) accumulation of short DNA fragments and adjacent single-stranded regions in the vicinity of replication origins; (ii) repair of the single-stranded regions is incomplete due to the clusteri~ng or concentration of the gaps and some inhibition of DNA repair synthesis at higher drug concentrations; (iii) nucleases convert some of the single-stranded regions into DNA double-strand breaks; (iv) some of the DNA double strand-breaks remain unrepaired and are manifest as chromatid breaks and (v) cell death results from chromosome lesions. Indeed hydroxyurea has been shown to
produce chromatid aberrations in treated cells [31]. We intend now to try to obtain experimental evidence for the importance of DNA double-strand breaks in producing cell death.
Acknowledgement The authors gratefully acknowledge the assistance of Dr. Ken Koschel in analyzing and plotting the DNA renaturation data.
References 1 Stock, J.A. (1975) in Biology of Cancer (Ambrose, E.J. and Roe, F.J.C., eds.), pp. 279-312, John Wiley, New York 2 Drewinko, B. (I 975) in Cancer Chemotherapy-Fundamental Concepts and Recent Advances, (Proceedings of the Annual Clinical Conferences on Cancer, sponsored by M.D. Anderson Hospital) pp. 63-78, Year Book, Chicago 3 Bacchetti, S. and Whitmore, G.F. (1969) Cell Tissue Kinet. 2, 193-211 4 Radford, I.R., Martin, R.F. and Finch, L.R. (1981) Biochim. Biophys. Acta 696, 145-153 5 Martin, R.F., Radford, I. and Pardee, M. (1977) Biochem. Biophys. Res. Commun. 74, 9-15 6 Prashad, N. and Cutler, R.G. (1976) Biochim. Biophys. Acta 418, 1-23 7 Mattern, M.R. and Painter, R.B. (1977) Biophys. J. 19, 117-123 8 Wetmur, J.G. and Davidson, N. (1968) J. Mol. Biol. 31, 349-370 9 Britten, R,J., Graham, D.E., Eden, F.C., Painchaud, D.M. and Davidson, E.H. (1976) J. Mol. Evol. 9, 1-23 10 Britten, R.J., Graham, D.E. and Neufeld, B.R. (1974) Methods Enzymol. 29E, 363-418 11 Powell, M.J.D. (1965) Computer J. 6, 303-308 12 Smith, M.J., Britten, R.J. and Davidson, E.H. (1975) Proc. Natl. Acad. Sci. USA 72, 4805-4809 13 Davidson, E.H., Hough, B.R., Klein, W.H. and Britten, R.J. (1975) Cell 4, 217-238 14 Hamlin, J.L. (1978) Exp. Cell Res. 112, 225-232 15 Adegoke, J.A. and Taylor, J.H. (1977) Exp. Cell Res. 104, 47-54 16 Mattern, M.R. and Painter, R.B. (1978) EXp. Cell Res. 111, 167-173 17 Woodcock, D.M., Fox, R.M. and Cooper, I.A. (1979) Cancer Res. 39, 1418-1424 18 Mattern, M.R. and Kapp, L.N. (t979) Exp. Cell Res. 120, 95-100 19 Bjursell, G. and Magnusson, G. (1976) Virology 74, 249-251 20 Perlman, D. and Rownd, R.H. (1975) Mol. Gen. Genet. 138, 281-291 21 Ockey, C.H. and Saffhill, R. (1976) Exp. Cell Res. 103, 361-372 22 Taylor, J.H. (1977) Chromosoma 62, 291-300
162 23 Callan, H.G. (1972) Phil. Trans. R. Soc. B 181, 19-41 24 Yamada, M. and Weissbach, A. (1977) Biochem. Biophys. Res. Commun. 77, 642-649 25 Woodcock, D.M. and Cooper, I.A. (1979) Exp. Cell Res. 123, 157-166 26 Ramseier, H.P., Burkhalter, M. and Gautschi, J.R. (1977) Exp. Cell Res. 105, 445-453 27 Ockey0 C.H. and Fox, M. (1976/77) Annual Report, pp. 108-109, Paterson Laboratories, Manchester, U.K.
28 Erixon, K. and Ahnstrom, G. (1979) Mutation Res. 59, 257-271 29 Laipis, P.J. and Levine, A.J. (1973) Virology, 56, 580-594 30 Wawra, E. and Dolejs, I. (1979) Nucleic Acids Res. 7, 1675-1686 31 Yu, C.K. and Sinclair, W.K. (1968) J, Cell. Physiol. 72, 39-42
Biochimica et Biopl(vsica A eta, 696 (1982) 154-162
Elsevier Biomedical Press BBA 91014
INHIBITION OF DNA SYNTHESIS AND CELL DEATH IAN R. RADFORD a, ROGER F. MARTIN a, LLOYD R. FINCH b and GEORGE S. HODGSON a " Biological Research Unit, Cancer Institute, 481 Little Lonsdale Street, Melbourne, 3000 and h R~zs'sell Grimwade School of Biochemistry, Universi(v of Melbourne, Parkville, 3052 Victoria (Australia)
(Received October 6th, 1981)
Key words: DNA synthesis inhibition," Cell death; Hydroa(vurea," Repetitive DNA sequence
The association between DNA synthesis inhibition and cell death in mouse L-cells was investigated using the drug hydroxyurea. This drug produces a preferential labelling of low molecular weight DNA and dose-response studies revealed a correlation between this effect and cytotoxicity. Investigation of the reassociation kinetics of DNA labelled during hydroxyurea inhibition showed an over-replication of middle repetitive sequences, but the concentration dependence of this effect was quite different to that of cytotoxicity.
Introduction
Materials and Methods
Anti-metabolites are the subject of intensive study due in part to their actual and potential use in cancer therapy. There have been m a n y studies at the biochemical level which have sought to elucidate the target enzymes (generally in the pathways for nucleic acid biosynthesis) for such drugs [1]. Similarly, there have been m a n y studies at the cellular level seeking to correlate cytotoxicity with factors such as drug dose, cell type and stage in the cell cycle [2]. However there is a gulf between these two approaches which has received little attention. In particular, the question of how inhibition of D N A synthesis results in cell death is essentially unanswered. This study concentrates on hydroxyurea, an inhibitor of D N A synthesis which is cytotoxic to S-phase cells [3], and attempts to establish a correlation between biochemical effects and cytotoxicity.
Chemicals. The sources for some chemicals have been outlined previously [4]. Ribonuclease TI (Aspergillus oryzae) was from Sigma, Pipes and pancreatic ribonuclease (bovine pancreas) were from Calbiochem. Unlabelled deoxyribonucleosides were purchased from Sigma or Calbiochem. Bio-Gel and hydroxyapatite ( D N A grade Bio-Gel G T P ) were obtained from Bio-Rad. S l nuclease ( A . oryzae) was from Miles and CsCI was from BDH. Solutions. Phosphate-buffered saline, p H 7.1, comprises 0.14 M NaCI, 3 m M KCI, 8 m M Na2HPO4, 1.5 m M K H 2 P O 4 and l m M glucose; EDTA-phosphate-buffered saline is phosphatebuffered saline supplemented with 0.6 m M EDTA; buffer A comprises 0.25 M NaC1/20 m M Tris-HC1 ( p H 7 . 5 ) / 2 m M EDTA; buffer B comprises 10 m M Tris-HC1 (pH 8.0)/0.25M s u c r o s e / 8 m M CaC12; p h e n o l / c h l o r o f o r m / i s o a m y l alcohol are in the ratio of 50:50:1 (v/v). Cells. The cells and growth conditions have been described previously [4]. Cells were counted with a Coulter Counter Model B.
Abbreviations: Pipes, 1,4-piperazinediethanesulphonic acid; SDS, sodium dodecyl sulphate; dThd, thymidine; BrdUrd. bromodeoxyuridine; dCyd, deoxycytidine; FdUrd, fluorodeoxyuridine. 0167-4781/82/0000-0000/$02.75 ~ 1982 Elsevier Biomedical Press
155
Cell labelling.
The methods for cell labelling and lysis have been described previously [4]. When uniformly labelled [3H]DNA was required, ceils were given three pulses of [3H]dThd (0.02 C i / m mol) at 40 n C i / m l ( 2 # M ) over a total labelling period of 24 h. Cell cloning studies. Log-phase cultures were incubated with a range of concentrations of the cytotoxic agent for 80 min. The drug was then removed by rinsing each flask four times with growth medium. The cells were rinsed with EDTA -phosphate-buffered saline then collected after incubation with 0.01% pronase and counted. Harvested cells were then diluted with growth medium and 5 ml aliquots containing 120 cells were dispensed into 50-mm-diameter plastic petri dishes. Control cultures were treated identically with respect to rinsing, harvesting and plating. The dishes were then incubated at 37°C in an atmosphere of 10% CO 2 in air for 7days. After this period the medium was gently removed and the colonies were washed with phosphate-buffered saline, fixed for 20 min in neutral formalin solution (containing 4% f o r m a l d e h y d e / 3 0 m M N a H 2 P O 4 / 4 5 mM Na2HPO4) and stained with 0.01% crystal violet solution for 1 h before counting. A cell was considered to have retained viability if it gave rise to a colony composed of 50 or more cells. The data are presented as cell viability, which is calculated by dividing the mean number of colonies, obtained from at least 20 replicate dishes, at a particular drug concen'tration, by the mean number of colonies obtained from untreated cells. A standard deviation was then calculated for this ratio. Control cloning efficiencies of around 70% were routinely obtained. Purification of DNA. Mouse DNA was purified from two sources. Unlabelled D N A for use as driver in D N A renaturation experiments was obtained from nuclei isolated from the livers of 3month-old male and female B a l b / c mice that had been starved for 24 h prior to killing, in order to decrease glycogen levels. The procedure used was a modification of the method of Prashad and Cutler [6]. After killing the mice by bleeding under anaesthesia, the livers were excised and rinsed in ice-cold buffer B and then carefully cut into small pieces, fibrous tissue being dissected out. The pieces were homogenized in ice-cold buffer B by
six passes in a homogenizer with a glass cup and Teflon pestle. The pestle was rotated at about 1200 rev./min and had a clearance of 0.006 inch. Triton X-100 detergent was then added to 1% and the homogenate was stirred gently with a magnetic stirrer at 4°C for 5 min. Nuclei were then collected by three centrifugations of the homogenate at 500 × g for 5 min at 4°C, and the nuclear pellet was resuspended in 50 mM Tris-HC1 (pH 8.0)/1 mM E D T A buffer before taking a sample for microscopic examination. SDS was then added to 0.2% to lyse the nuclei. Proteinase K (pre-incubated at 37°C for 0.5 h) was added to 0.1 m g / m l and the lysate was incubated at 37°C for 1 h. The lysate was then sheared by three passages through a 23-gauge needle, followed by two extractions with phenol/chloroform/isoamyl alcohol. The solution was then extracted three times with chloroform, to remove traces of phenol, flushed with nitrogen and then concentration by rotary evaporation. The concentrate was applied to a Biogel P-30 column and equilibrated with 0.25 M NaC1/10 mM Pipes (pH 6.8), and the DNA in the void volume fractions was pooled and diluted with distilled water to give a final NaC1 concentration of 0.1 M. Pancreatic RNAase and RNAase T1 (both previously incubated at 80°C for 2 m in in 0.1 M N a C I / 1 0 m M Tris-HC1 (pH 7.5)/5 mM EDTA solution to inactivate any contaminating DNAases) were added to 20 # g / m l and 25 units/ml, respectively, and the mixture was incubated for 1 h at 37°C. The efficacy of the RNAase digestion was monitored by running a parallel digest of a sample of the DNA preparation to which [3H]RNA had been added. The elimination of acid-insoluble counts from this digest was confirmed before proceeding. After RNAase digestion the DNA preparation was incubated with proteinase K at 0.1 m g / m l for 1 h at 37°C. SDS was then added to 0.5% and the preparation was extracted three times with phenol/chloroform/isoamyl alcohol, followed by chloroform extraction, concentration and chromatography as described above. The purity of the DNA preparation was assessed from the ratio of its absorbance at 260 nm to that at 280 nm. Ratios of 1.80 or better were obtained [7]. Labelled DNA was obtained from L-cells. The D N A was purified as outlined above except that cells were lysed directly with 0.5% SDS/25 mM
156
Tris-HC1 (pH 7.5)/5 mM EDTA solution. Sonication of DNA. DNA fragments of about 450 nucleotides in length were prepared by sonication with a Branson S125 sonifier equipped with a microtip. The D N A preparation, in 0.25 M NaC1/10 mM Pipes (pH 6.8), was flushed with nitrogen before sonication at power level 2 using 16 pulsed exposures of 15 s each with 30 s cooling in ice between exposures. DNA renaturation conditions. DNA renaturation was performed in siliconized glass test tubes and the solution was overlaid with paraffin oil. Duplicate samples were taken for each point and were either assayed immediately or snap-frozen in an ethanol/solid CO 2 bath. DNA samples, denatured by immersion in a boiling water bath for 3 mins, were incubated in 0.25 M N a C I / 2 0 mM Pipes (pH 6.8) at 60°C [8,9]. DNA was present at a final concentration of between 5 and 1000/~g/ml. To obtain Cot values of 102 and greater, mouse liver D N A (see above) was added as driver. The equivalent Cot was calculated using a factor of 1.77 to correct for the Na + concentration used [10]. Equivalent Cot is the Cot ((absorbance at 260 n m / 2 ) X time (h)) multiplied by the ratio of the reassociation rate in the solution used to that in 0.12 M phosphate buffer. S 1 nuclease digestion. Digestions were performed in the presence of 0.25 M NaC1, 30 mM sodium acetate (pH 4.6), 25 mM dithiothreitol, 30 /~M ZnSO 4 and 0.1 m g / m l of D N A (where necessary the D N A concentration was made up to 0. l m g / m l by addition of sonicated calf thymus DNA). Samples were incubated with S1 nuclease (10/~g/ml) at 37°C for 30 minutes. The reaction was terminated by the addition of EDTA and Na4P207 to 5 mM and 2.5 mM, respectively. The extent of digestion was assessed by gel filtration on columns of Biogel P-30 which were washed and run with 0.25 M N a C I / 1 0 mM Pipes (pH 6.8) buffer. Fractions were collected and counted directly. The percentage of renatured D N A in each sample was calculated as 100 X [cpm in columnexcluded fractions)/(cpm in column-included fractions + cpm in column-excluded fractions)]. Hydroxyapatite fractionation. Samples for fractionation were diluted into 50 mM phosphate buffer (pH 6.8) and where necessary sonicated calf thymus DNA was added to give a total of 0.1 mg
DNA. The sample was then pumped through a column containing 1 ml hydroxyapatite equilibrated at 60°C in 50 mM phosphate buffer and the column was washed with 50 mM phosphate buffer to give a total volume for sample and wash of 10.5 ml. Insignificant amounts of the total radioactivity were eluted in this wash. Singlestranded DNA was eluted by pumping 5 ml 0.125 M phosphate buffer through the column, followed by 5 ml 0.30 M phosphate buffer to elute double-stranded DNA. The fractions obtained from these two elutions were then counted directly. In order to determine the level of crossover (i.e., the percentage of double-stranded DNA eluted during the single-stranded DNA wash and vice versa), reference samples of double-stranded and single-stranded D N A were prepared. It was found that the 0.125 M phosphate buffer wash released 5-10% of the double-stranded DNA, whilst 1-2% of single-stranded D N A eluted in the 0.30 M phosphate buffer wash. The percentage of DNA containing doublestranded regions was determined as 100 X [(cpm in 0.30 M' phosphate buffer)/(cpm in 0.125 M phosphate buffer + cpm in 0.30 M phosphate buffer)], where the counts in each wash were corrected for crossover. Analysis of Cot curves. Cot curves were analyzed using a computer least-squares fit [11], minimized according to the criterion of Y.,(O,E~) 2 (where O = o b s e r v e d value, E = c a l c u l a t e d value), for a maximum of three second-order components. An unconstrained fit was used. Hydroxyapatite and S~ nuclease data were fitted to formulae 1 [8,10] and 2 [12,13] respectively. Percent DNA hybridized = 3
P ~ f / ( 1 - (1 +kiCot)-' )
(1)
i--I
Percent DNA hybridized = 3
P ~,, f / ( 1 - (1 + kiCot) -°'44)
(2)
i--1
where P is the scaling parameter for each curve; f the fraction of P for each component; k i the rate constant of the component.
157
Results
Synthesis of low-Mr DNA and cytotoxicity The results presented in the preceding paper [4] suggest that hydroxyurea induces a state of uncoordinated DNA synthesis in cells. It was therefore of interest to see whether the dose-dependent killing of S-phase cells, produced by exposure to the drug, could be correlated with the dose response for accumulation of radioactive precursor in lowM r DNA. Accordingly, cells were treated with a range of concentrations of hydroxyurea for 80 rain and assayed for survival, whilst replicate cultures were labelled with [3 H]dThd and the lysates centrifuged through alkaline sucrose gradients. The previously reported [4] pattern of accumulation of radioactivity in low-Mr DNA was observed (Fig. 1). The results of the cloning assays (Fig. 1) correlate with these data. Both killing and accumulation of radioactivity in low-Mr DNA plateau at a hydroxyurea concentration of around 2raM. The maximum level of killing achieved, about 40%, correlated with a labelling index of 43% for untreated cells.
BO
\
•
60 8o c
80
.....
6o
40
~0
20
20
ill
I
I
I
2
~
6
8
Hydroxyurea ( rnM )
Fig. I. Cell killing and accumulation of label in low-M r D N A following hydroxyurea treatment. Cells were treated with the indicated concentration of the drug for 80 rain and cytotoxicity was then measured using a colony-forming assay (for details see Materials and Methods). Rcplicate cultures were treated with the drug for 1 h and [3H]dThd was then added for 20 min. The S D S / p r o t e i n a s e K lysates were centrifuged through alkaline sucrose gradients and the percentage of the total incorporated radioactivity in low-M r DNA was determined (for further details see previous paper [4]). • . . . . . . • , cell survival: ." 0 , percentage of incorporated label no greater than 21 S; the bars represent mean--+S.D.
Renaturation analyses Replication of the eukaryotic genome is thought to proceed by the initiation of synthesis in clusters of replicons in a strict temporal order [14]. It has been found that treatment of cells with S-phasespecific cytotoxics alters the pattern of initiation [15] and such a disturbance may be responsible for cytotoxicity. As a means of detecting possible alterations to the pattern of replication it was decided to compare the sequence complexicity of DNA labelled in the presence and absence of drug by hybridization analysis. DNA was purified from untreated cells that had been labelled with [3H]dThd for 10s or 24h and from cells that had been pretreated and labelled in the presence of hydroxyurea. The pulse -labelled DNA was included as a control because the 10 s pulse of [3H]dThd would allow an amount of DNA synthesis comparable to that occurring in the hydroxyurea-treated cells [4] and also since a report in the literature had claimed that preferential incorporation of [3 H]dThd into middle repetitive sequences occurred during brief labelling of CHO cells [16]. The reassociation kinetics of these samples, as assayed by hydroxyapatite chromatography, are shown in Fig. 2. It can be seen that the reassociation kinetics of the 24-h-labelled and 10-s -labelled DNA samples are very similar. However, the two preparations of DNA labelled in the presence of hydroxyurea show a marked difference to the control samples, in having a much higher rate of renaturation of sequences in Cot range 0-100. Analysis of the curves showed that this difference corresponded to an increased percentage of middle repetitive DNA sequences and a correspondingly lower level of unique sequence DNA (Table I); middle repetitive sequences increased from 11-14% in the controls to 22-29% of the DNA synthesized in hydroxyurea-treated cells. These experiments were repeated using S~ nuclease digestion to measure the renaturation kinetics and similar results were obtained (data not shown). There was a higher rate of renaturation of sequences in Cot range 0-100, so that approx. 10% more of the labelled DNA from hydroxyurea-treated cells behaved as middle repetitive DNA (Table I). It was next of interest to investigate the hydroxyurea dose response for the 'over-production'
158
O-
20 ¢-
9 40. 0 r~
60-
80L_
100 -3
I -2
I -1
I 0
Log equivelent
I 1
t 2
3
4
of middle repetitive DNA sequences with a view to seeing whether this effect could be correlated with cell death. Accordingly, DNA was purified from cells that had been pretreated and labelled in the presence of a range of concentrations (both toxic and non-toxic) of hydroxyurea. These samples were analysed over the Cot range 0-100 (Fig. 3). The Cot curves show that the DNA synthesized in cells treated with 5 mM hydroxyurea renatures considerably more rapidly than the control DNA, as previously shown. However, the renaturation rate of the 1 and 2 mM hydroxyurea samples is only marginally greater than that of the control, as shown by the curve gradients (parameter B in Table II). This is obviously quite different to the dose response for cytotoxicity (see Fig. 1), where 1
Cot
Fig. 2. Renaturation kinetics of D N A synthesized in the presence or absence of hydroxyurea, as analyzed by hydroxyapatite chromatography. Control cultures were labelled with [3 H]dThd for 2 4 h ( O O ) or 10s ( A . . . . . ~ ) . Two separate batches of cultures were treated with 5 m M hydroxyurea for 1 h and then labelled with [3H]dThd for 20 min (preparations 1
( + . . . . . . + ) and 2 ( x . . . . . . x ) ) . The D N A from these cultures was then purified and sonicated to 0.45 kilobasepairs in length for renaturation analysis. The curve through the data points is a computer generated least squares fit for a m a x i m u m of three second order components. An analysis of the data is presented in Table I. For all experimental details see Materials and Methods.
TABLE I RENATURATION OF 0.45-KILOBASEPAIR DNA FRAGMENTS, FROM HYDROXYUREA-TREATED AND UNTREATED L-CELLS, A N A L Y Z E D BY H Y D R O X Y A P A T I T E C H R O M A T O G R A P H Y A N D S I N U C L E A S E D I G E S T I O N T h e table presents a statistical analysis of the data from Fig. 2. For further details see Materials and Methods. C o m p o n e n t I represents highly repetitive sequences, repeated 1 - 106 or more times and with a Cot of less than I • 10 2. C o m p o n c n t 2 represents middle repetitive sequences, repeated I - 102 to 1,105 times and with a Cot of around I • 10 2 to 10. C o m p o n e n t 3 represents 'Unique' sequences with a Cot of more than 10. Sample
Component
Fraction Hydroxyapatite
S I nuclease
24-h-labelled
I 2 3
0.148 ± 0.014 0.138±0.014 0.714 a
0.165 ÷ 0.006 0.113 ±0.012 0.722
10-s-labelled
I 2 3
0.169 ± 0.018 0.112+0.017 0.719
0. I 14 ± 0.006 0.167+0.010 0.719
Hydroxyurea-labelled ( 1)
1 2 3
0.188 ± 0.004 0.216-+0.003 0.596
Hydroxyurea-labelled (2)
I 2 3
0.179 + 0.005 0.286 -+ 0.008 0.535
0.127 + 0.006 0.258 + 0.009 0.615
a The value for c o m p o n e n t 3 was obtained by s u m m i n g the values for c o m p o n e n t s 1 and 2 and subtracting them from unity.
159 T A B L E II
20
A N A L Y S I S O F Cot C U R V E S OF D N A S Y N T H E S I Z E D IN T H E PRESENCE O F D I F F E R E N T H Y D R O X Y U R E A CONCENTRATIONS The table presents a statistical analysis of the data from Fig. 3. The data were fitted to a curve of the form Y=AX -B, where Y is the percentage of labelled D N A b o u n d to hydroxyapatite and X is the Cot value. The estimated values for the curve parameters A ( Y - a x i s intercept) and B (curve gradient) are shown with their 955 confidence limits.
3C 9 \ \ \ \ \ \
40
Sample (mM)
\ \ X~
I
0
!
1 Log equivolent Cot
2
71'80 ~
A 20
10
o. u
0 I
i
e
.]1.8o
~ 20 .-
.
-°
10
0
i
I
10
20
Fraction No.
A
Confidence limits
B
Confidence limits
Control
25.93
24.77 27.15
0.0341
0.0201 0.0482
0.5
26.57
25.26 27.94
0.0489
0.0334 0.0645
1
23.90
22.80 25.06
0.0725
0.0576 0.0873
2
24.28
23.06 25.56
0.0663
0.0494 0.0833
5
22.91
21.17 24.80
0.143
0.119 0.167
I
Fig. 3. Renaturation kinetics of D N A synthesized in the presence of different hydroxyurea concentrations. Replicate cultures were treated with different concentrations of hydroxyurea for l h and [3H]dThd was then added for 20 min. The cells were then lysed and their D N A was purified. D N A obtained from cells that had been labelled with [3H]dThd for 24 h was used as a control. D N A renaturation was monitored by hydroxyapatite chromatography. Data points shown represent the mean of two determinations. Hydroxyurea (mM): × . . . . . . × , 5 m M ; + . . . . . . + , 2 m M ; /x . . . . . /x, l m M ; [] . . . . . [3, 0.5 mM; © ©, control.
E
Curve parameters
I
30
Fig. 4. Test for double replication of D N A segments following release from hydroxyurea treatment. Replicate cultures were incubated in medium containing 20 # M BrdUrd, 2 # M F d U r d and 1.7 # M [3H]dCyd (l C i / m m o l ) for l h. The cells were then washed, incubated in medium containing 20 # M dThd and 20 # M dCyd for 30 min and in growth medium for I h. During the latter and all subsequent steps colcemid was present at 0. I /~g/ml. The cultures were then incubated in the absence (A) or presence of 5 m M hydroxyurea (B) for 80 min. Following extensive washing, the untreated culture was incubated in m e d i u m containing []4C]dThd (4.6 # M , 56 m C i / m m o l ) for 30 min whilst the drug-treated culture was incubated with label for 4 h. The cells were lysed in 0.25 M N a C l / 2 0 m M Tris-HCl (pH 7 . 5 ) / 2 m M E D T A / 0 . 5 % SDS and then digested with 100 # g / m l proteinase K. The lysates were sheared by passage six times through a 26-gauge needle and then extracted twice with a p h e n o l / c h l o r o f o r m / i s o a m y l alcohol mixture. Following reextraction with chloroform the D N A was re-dissolved in 1.2 ml 0.2 M Tris-HC1 (pH 7.5)/20 m M E D T A and CsC1 solution was then added to a final density of 1.74 g / m l . The solutions were then centrifuged at 36000 r e v . / m i n for 65 h in a Beckman 50 Ti rotor at 20°C. Fractions of 0.25 ml were collected and analyzed for acid-insoluble radioactivity by precipitation onto filters with trichloroacetic acid. • • , 3H cpm; [] rT, 14C cpm; • . . . . . . • , density.
160 and 2 mM hydroxyurea were as cytotoxic to Sphase cells as was 5 mM hydroxyurea. Hence the dose response for this effect does not correlate quantitatively with that for cytotoxicity.
Abnormal re-initiation of DNA synthesis The above studies have sought an explanation for cell death by examining the DNA synthesized during drug treatment. However, the critical event leading to cell death might be occurring after the release of the cell from drug inhibition. One possibility of interest to study was whether there was induction of a second round of replication within the same S-phase following drug treatment [17]. Cells prelabelled with BrdUrd and [3 H]dCyd were treated with a toxic dose of hydroxyurea for 80 min and then labelled with [laC]dThd. The D N A was then purified and analyzed on neutral CsC1 gradients (Fig. 4). The profile from hydroxyureatreated cells is identical to that obtained from a control culture; 3H label is present in hybriddensity DNA, whilst 14C label is found solely in normal-density DNA. Hence no evidence was found for the induction of double replication following drug removal. Discussion
The D N A synthesis that is detected during hydroxyurea inhibition is characterised by a disproportionate synthesis of low-M r DNA fragments [4]. This paper reports a further abnormality of D N A synthesis during hydroxyurea inhibition, namely preferential synthesis of middle repetitive sequences. However, the central aspect of this report is the result of dose-response studies which reveal a correlation between preferential synthesis of low-M r DNA and cytotoxicity in hydroxyureatreated cells. The D N A synthesized during treatment with 5 m M hydroxyurea was found to be markedly enriched for middle repetitive sequences. This difference could not be ascribed to an artefact of the [3H]dThd labelling time used, as D N A isolated from untreated cells labelled for 10 s or 24 h had comparable renaturation kinetics. The latter result differs from that obtained in a study using CHO cells [16], where overrepresentation of middle repetitive sequences was found after short labelling
times. However, it concurs with a subsequent study by the same group in which HeLa cells were used [18]. The conflicting results were interpreted as being due to differences between cell types. How, then, might the anomalous D N A synthesis, occurring in the presence of hydroxyurea, be accounted for? Hydroxyurea treatment has been found to induce the accumulation of early replicative intermediates of both polyoma virus [19] and a bacterial plasmid [20]. Similarly it has been found that the replicons are smaller in mammalian cells that have been released from inhibition with FdUrd than they are in untreated control cells [21,22]. These results would suggest that the activation of sites for the initiation of D N A synthesis has continued during drug treatment. What might then be happening is that the cell's programme for replicon initiation continues in the presence of hydroxyurea, leading to the accumulation of initiation sites, with that DNA synthesis which does occur being close to these sites. It is then necessary to postulate that replicon initiation sites are enriched for middle repetitive sequences, although a study investigating the sequences rephcated following Xirradiation of CHO cells obtained indirect evidence for the enrichment of initiation sites for unique sequences [7]. Our results, however, would be consistent with the view that there is some sequence identity [23] amongst the approx. 1 • 105 initiation sites in the cellular genome. Our results with hydroxyurea-treated cells are strikingly similar to those found when the sequence composition of the DNA synthesized by nuclei isolated from normal or mitomycin Ctreated HeLa cells was investigated [24]. In both systems an increased synthesis of middle repetitive sequences was found. The over-replication of middle repetitive DNA sequences was found not to correlate with the concentration-dependent induction of cell death by hydroxyurea. This result, however, does not rule out the possibility that cell death is the result of an alteration to the pattern of replication. At the lower drug levels replication may be able to proceed sufficiently far past initiation sites so that there would be no detectable enrichment for sequences associated with initiation sites. The possible induction of double replication following drug treatment [17,25], as a cause of
161
cytotoxicity, was also examined. However, no evidence was found for its occurrence after hydroxyurea exposure. This is in agreement with the conclusion of others that hydroxyurea treatment does not induce double replication [26,27]. The results of the dose-response study of hydroxyurea cytotoxicity are consistent with a previous report [3]. Similarly, the results presented in this and the preceding paper [4] concur with previous osbervations that DNA synthesis must be decreased to a very low level by this drug before cell killing occurs [26]. Even when DNA synthesis is decreased to about 2% of the control level, as with a concentration of 0.5 mM hydroxyurea, cell killing is still slight. More interestingly, a correlation was found between cell killing by hydroxyurea and the incorporation of labelled precursor into low-Mr DNA in the drug-treated cells. We suggest that this correlation may provide an insight into the mechanism of cell death. The DNA fragments produced by hydroxyureatreated cells are attached to their template and presumably separated from each other and high-Mr DNA by single-stranded regions, as these drugs have been shown to have no effect on DNA ligase activity [28]. This has been verified for an SV40synthesizing system treated with hydroxyurea, where it was found that the DNA fragments could be joined only. when incubated with ligase plus a DNA polymerase and not by ligase alone [29]. The single-stranded regions which would be especially vulnerable to nuclease attack, are repairable by DNA polymerase fl which is relatively insensitive to hydroxyurea [28,30], but some inhibition doubtless occurs at higher drug concentrations. We postulate that the cytotoxicity of hydroxyurea arises from inhibition of DNA synthesis by the following sequence of events: (i) accumulation of short DNA fragments and adjacent single-stranded regions in the vicinity of replication origins; (ii) repair of the single-stranded regions is incomplete due to the clusteri~ng or concentration of the gaps and some inhibition of DNA repair synthesis at higher drug concentrations; (iii) nucleases convert some of the single-stranded regions into DNA double-strand breaks; (iv) some of the DNA double strand-breaks remain unrepaired and are manifest as chromatid breaks and (v) cell death results from chromosome lesions. Indeed hydroxyurea has been shown to
produce chromatid aberrations in treated cells [31]. We intend now to try to obtain experimental evidence for the importance of DNA double-strand breaks in producing cell death.
Acknowledgement The authors gratefully acknowledge the assistance of Dr. Ken Koschel in analyzing and plotting the DNA renaturation data.
References 1 Stock, J.A. (1975) in Biology of Cancer (Ambrose, E.J. and Roe, F.J.C., eds.), pp. 279-312, John Wiley, New York 2 Drewinko, B. (I 975) in Cancer Chemotherapy-Fundamental Concepts and Recent Advances, (Proceedings of the Annual Clinical Conferences on Cancer, sponsored by M.D. Anderson Hospital) pp. 63-78, Year Book, Chicago 3 Bacchetti, S. and Whitmore, G.F. (1969) Cell Tissue Kinet. 2, 193-211 4 Radford, I.R., Martin, R.F. and Finch, L.R. (1981) Biochim. Biophys. Acta 696, 145-153 5 Martin, R.F., Radford, I. and Pardee, M. (1977) Biochem. Biophys. Res. Commun. 74, 9-15 6 Prashad, N. and Cutler, R.G. (1976) Biochim. Biophys. Acta 418, 1-23 7 Mattern, M.R. and Painter, R.B. (1977) Biophys. J. 19, 117-123 8 Wetmur, J.G. and Davidson, N. (1968) J. Mol. Biol. 31, 349-370 9 Britten, R,J., Graham, D.E., Eden, F.C., Painchaud, D.M. and Davidson, E.H. (1976) J. Mol. Evol. 9, 1-23 10 Britten, R.J., Graham, D.E. and Neufeld, B.R. (1974) Methods Enzymol. 29E, 363-418 11 Powell, M.J.D. (1965) Computer J. 6, 303-308 12 Smith, M.J., Britten, R.J. and Davidson, E.H. (1975) Proc. Natl. Acad. Sci. USA 72, 4805-4809 13 Davidson, E.H., Hough, B.R., Klein, W.H. and Britten, R.J. (1975) Cell 4, 217-238 14 Hamlin, J.L. (1978) Exp. Cell Res. 112, 225-232 15 Adegoke, J.A. and Taylor, J.H. (1977) Exp. Cell Res. 104, 47-54 16 Mattern, M.R. and Painter, R.B. (1978) EXp. Cell Res. 111, 167-173 17 Woodcock, D.M., Fox, R.M. and Cooper, I.A. (1979) Cancer Res. 39, 1418-1424 18 Mattern, M.R. and Kapp, L.N. (t979) Exp. Cell Res. 120, 95-100 19 Bjursell, G. and Magnusson, G. (1976) Virology 74, 249-251 20 Perlman, D. and Rownd, R.H. (1975) Mol. Gen. Genet. 138, 281-291 21 Ockey, C.H. and Saffhill, R. (1976) Exp. Cell Res. 103, 361-372 22 Taylor, J.H. (1977) Chromosoma 62, 291-300
162 23 Callan, H.G. (1972) Phil. Trans. R. Soc. B 181, 19-41 24 Yamada, M. and Weissbach, A. (1977) Biochem. Biophys. Res. Commun. 77, 642-649 25 Woodcock, D.M. and Cooper, I.A. (1979) Exp. Cell Res. 123, 157-166 26 Ramseier, H.P., Burkhalter, M. and Gautschi, J.R. (1977) Exp. Cell Res. 105, 445-453 27 Ockey0 C.H. and Fox, M. (1976/77) Annual Report, pp. 108-109, Paterson Laboratories, Manchester, U.K.
28 Erixon, K. and Ahnstrom, G. (1979) Mutation Res. 59, 257-271 29 Laipis, P.J. and Levine, A.J. (1973) Virology, 56, 580-594 30 Wawra, E. and Dolejs, I. (1979) Nucleic Acids Res. 7, 1675-1686 31 Yu, C.K. and Sinclair, W.K. (1968) J, Cell. Physiol. 72, 39-42