- Email: [email protected]
Characterization of VP-16-induced DNA damage in isolated nuclei from L1210 cells
74
Biochimica et BiophysicaActa, 783 (1984) 74-79
Elsevier BBA 91390
C H A R A C T E R I Z A T I O N O F V P - 1 6 - 1 N D U C E D D N A D A M A G E IN I S O L A T E D N U C L E I FROM LI210 CELLS BONNIE S. GLISSON, SHERIN E. SMALLWOOD and WARREN E. ROSS * Departments of Pharmacology and Medicine, University of Florida, Gainesville, FL 32610 (U.S.A.)
(Received April 9th, 1984)
Key words: DNA damage; DNA-drug interaction; VP-16; Enzyme dependence," Drug activation; Strand break
Based on the observation that VP-16-induced D N A damage can be demonstrated in isolated nuclei but not in purified DNA, and that this effect is temperature-dependent, it is postulated that the mechanism of action of VP-16 involves an essential intranuclear event, perhaps enzyme-mediated, which is a prerequisite for the cleavage of DNA. Using alkaline elution to assay single-strand breaks in isolated L1210 nuclei, we have further characterized conditions influencing this putative intranuclear reaction. We have found drug activity to be dependent on magnesium and pH and to be stimulated by low concentrations of ATP (0.05-1 mM), an effect which was not observed with a nonhydrolyzable analog of ATP. Heat-labile activity in a nuclear non-histone protein extract was critical to VP-16-mediated D N A damage. This new evidence lends further credence to the hypothesis that activity of an intranuclear enzyme, possessing characteristics consistent with a type II D N A topoisomerase, is a prerequisite for the cleavage of D N A by VP-16.
Introduction VP-16 (etoposide), a semi-synthetic derivative of podophyllotoxin, has become an important cancer chemotherapeutic agent in the clinical a r m a m e n t a r i u m against several tumors, most notably germ cell tumors of the testis, small cell lung cancer, and lymphoma. Although the precise mechanism of action of VP-16 remains unknown, m u c h evidence now exists to implicate D N A d a m a g e as its major effect. W o r k from this laboratory [1] and others [2,3] indicates that VP-16 or its congener VM-26 induces single-strand breaks in * To whom correspondence should be addressed. Abbreviations: AMP-PNP, (fl,'t-imido)adenosine triphosphate; NTPs and dNTPs, ribonucleoside and 2'-deoxyribonucleoside 5'-triphosphates; VP-16, 4'-demethylepipodophyllotoxin-4-(4,6O-ethylidene-fl-D-glucopyranoside)(etoposide); VM-26, 4'-demethylepipodophyllotoxin-4-(4,6-O-thenylidene-fl-D-glucopyr-
anoside) (teniposide); m-AMSA, 4'-9-acridinyl methanesulfon-m-anisidide.
amino-
0167-4781/84/$03.00 © 1984 Elsevier Science Publishers B.V.
the D N A of m a m m a l i a n cells. We have also shown that D N A double-strand breaks, a more lethal f o r m of damage, and D N A protein cross-links are p r o d u c e d when cells are exposed to VP-16. Further, inhibition of strand-breaking activity by disulfiram and other agents is accompanied by a reduction in cytotoxicity [4]. This suggests that the activity of VP-16 as a chemotherapeutic agent is directly related to its effects on D N A . The mechanism by which VP-16 induces D N A d a m a g e is currently undefined, although the work of Loike and Horwitz [2] supplied the initial clue. T h e y observed no strand-breaking activity when VP-16 was incubated with purified adenovirus or H e L a cell D N A . We were subsequently able to demonstrate, however, that drug-induced D N A d a m a g e does occur in isolated L1210 nuclei and that this effect is temperature-dependent [1]. Thus, we have hypothesized that an intranuclear enzyme, e.g., one mediating drug activation or perhaps one with nuclease activity, is a prerequisite for the
75 cleavage of D N A by VP-16. The possibility that this enzyme is actually a type II topoisomerase was investigated by Long and Minocha [5]. Using a partially purified enzyme preparation, they observed inhibition of catenation of PM2 D N A by VP-16. However, they did not observe strand scission in their system. Thus, it was of interest to characterize further the intranuclear events underlying VP-16-induced D N A damage. During the course of this work, we have also recently found that both VP-16 and its congener VM-26 stimulate D N A cleavage by a highly purified calf thymus type II topoisomerase [6]. The evidence presented herein, which more fully defines intranuclear conditions necessary for maximal drug activity, lends further credence to the hypothesis that type II topoisomerase is the intracellular target for the anti-tumor effects of VP-16. Using alkaline elution to assay single-strand breaks in isolated L1210 nuclei, we found that drug activity was dependent on p H and magnesium concentration and was enhanced by low concentrations of ATP, an effect which appeared to require hydrolysis of the ATP. Finally, we found that heat-labile activity located in a 0.35 M NaC1 extract was critical to VP-16-induced D N A damage. Materials and Methods
Murine leukemia L1210 cells, with a doubling time of approx. 12 h, were grown in Roswell Park Memorial Institute Medium 1630 containing 20% fetal calf serum, penicillin and streptomycin. Cells were labeled with [2-14C]thymidine (53 m C i / mmol; 0.01 # M C i / m m o l ; New England Nuclear, Boston, MA) approx. 20 h before their use in experimentation. VP-16, which was dissolved in dimethyl sulfoxide, was a gift from Bristol Laboratories (Syracuse, NY). NTPs, dNTPs, AMP, A D P and (fl,),-imido)adenosine triphosphate (AMP-PNP) were obtained from Sigma Chemical Co. (St. Louis, MO), dissolved in buffer A (see below), and the p H was corrected to 6.4. Incubation with these compounds occurred simultaneously with that of VP-16 treatment. Isolated L1210 nuclei were used in all experiments. These were prepared by first washing 40 ml
of 14C-labeled whole cells at 5 • 105 c e l l s / m l with cold buffer A (1 m M K H E P O 4 / 5 m M MgC12/150 m M N a C l / 1 m M EGTA) at p H 6.4. The cells were then resuspended in 1 ml of buffer A before lysis with 9 ml buffer B (buffer A plus 0.3% Triton X-100; Eastman K o d a k Co., Rochester, NY). The suspension was kept on ice for 30 min, after which 40 ml buffer A were added and the nuclei sedimented by centrifugation at 1000 rpm for 10 min. Phase microscopy confirmed the presence of isolated nuclei only. After resuspension in buffer A at a density of 10 6 nuclei/ml, the nuclei were treated with drug for 60 rain at 37°C. In experiments utilizing a range of p H or magnesium concentrations, nuclei were suspended in buffer A at the specified p H or magnesium concentration. Nuclear extract containing non-historic proteins was prepared by suspending isolated nuclei in buffer A contianing 0.35 M NaC1 and 0.2 m M dithiothreitol. Centrifugation of the suspension served to separate non-histone proteins in the supernatant from D N A and histone proteins in the pellet. Extracted nuclei were prepared by exposing isolated nuclei on the elution filters to buffer A containing 0.35 M NaCl. This solution was removed by gravity drip, leaving high molecular weight D N A bound to histone proteins on the filter. Treatment with drug and protein extract then followed resuspension of the extracted nuclei in warm buffer A (0.15 M NaC1). High-frequency D N A single-strand breaks were assayed by alkaline elution, as previously described [7]. 14C-labeled nuclei were layered into poly(vinyl chloride) filters (pore size 2 #m; Millipore Corp., Bedford, MA) and lysed with a solution of 2% SDS, 10 m M disodium E D T A and 0.5 m g / m l proteinase K (Merck, Darmstadt, F.R.G.). D N A was then eluted from the filters with tetrapropylammonium hydroxide (RSA Corp., Ardsdale, NY) at an elution flow rate of 0.16-0.20 m l / m i n , collecting fractions at 5-min intervals over a total elution time of 30 rain. The technique in experiments with the nuclear protein extract differed only in the use of polycarbonate filters (pore size 2 /~m), to minimize protein adsorption, and an increase in elution time to 60 min. The alkaline elution assay for D N A double-strand breaks has been previously described in detail [8].
76 Results
I0
Given our hypothesis that interaction of VP-16 with an intranuclear enzyme was necessary for the production of D N A damage, we believed it important to define intranuclear conditions that might influence enzymatic activity and, thus, the degree of drug-induced D N A damage. The effect of pH was investigated in the range 5.0-8.0. The optimal pH range for drug activity was 5.5-6.4, with marked reduction in activity seen at pH 7.0 and higher (Fig. 1). Controls were run at each pH point tested; however, only at pH 5.0 did pH alone influence the rate of elution. At this pH, D N A damage in the control was so excessive as to preclude interpretation of drug-mediated damage (data not shown). Magnesium dependence of the reaction was demonstrated by complete loss of drug activity in the absence of magnesium (Fig. 2). Stepwise restoration of drug activity was seen as
0.8 IOL * I0 n'~ Mg'* tOL~ 0 Mg*÷ 0 Mg**
0.6
I0 mM Mg *° Z
0.4 • I 5 mM M9 .6
3 mM MIG~*
b tL 02
[VP-16] = 4 0 uM
I
2 3 FRACTION
4 NUMBER
5
6
Fig. 2. The effect of magnesium concentration on VP-16 activity in isolated nuclei. Nuclei were suspended in buffer A 1.0 Control 0.8
containing the specified concentrations of magnesium and treated with drug for 60 min. Nuclei were then lysed and DNA single-strand breaks were assayed by alkaline elution at pH 12.1.
VP-16 pH 7 5 VP-16 pH 8 0 VP-16 pH 7 0 06
C
"~
0.4
I~ <~
VP- 16 pH 5.5 VP" 16 pH 6.0
g
O~
V P - 1 6 pH 6.4
o
,,= 02
[ V P - 1 6 ; ' 4 0 pM
II
I 2 Fraction
t 3
I 4
I 5
L 6
Number
Fig. 1. The effect of pH on VP-16 activity in isolated nuclei. 14C-labeled nuclei were suspended in buffer A at the specified pH points and treated with VP-16 for 60 min. After dilution in cold buffer, the nuclei were lysed and DNA single-strand breaks were assayed by alkaline elution at pH 12.1.
magnesium concentration was increased to its optimal 5 mM level. The addition of ATP (0.05-1 mM) resulted in marked enhancement of strand-breaking activity, an effect which gradually diminished at ATP concentrations of 2 mM and higher (Fig. 3). The results of several experiments testing the effect of A T P are represented in Fig. 4. Consistent stimulation of drug activity was seen with concentrations of ATP as low as 0.05 mM, increasing in a dose-related way until maximal effect was achieved with 1 mM ATP. More variability in the dose-response relationship was seen with higher ATP concentrations (2-5 mM), although generally there was gradual loss of this enhancement and, occasionally, inhibition of drug activity. The frequency of drug-induced D N A double-strand breaks was also enhanced by ATP to an extent similar to that observed for single-strand breaks. Though less potent than ATP, stimulation of strand-breaking activity was also seen with dATP (0.5-2 mM). GTP, CTP, UTP, TTP and their deoxy-derivatives
1.0
0.8
0.6
77
~
1.0
AI"P 3 mM AMP-PNP I mM
0.8
0.6
VP-16*
~ P - P N P I mM
i
VP-16
VP-16* ATP 3ram
o
~.
Control
- t - - - - - - - - - - - . 4 k ~ . , ~ a m . 4 Control ATP 3raM
0.4
VP-16
o
e 0.4 ri-
VP-16 * ATP .O5 mM
Q
VP'I6* ATP 2 mM
g
g
g
VP- 16 * ATP .25mM
u-
0.2
0.2
VP-16* ATP .5 mM
VP-16 t ATP I mM
VP'I6* ATP I mM
[.VP- 16] " I0 MM
I
I
I
2
L
I
3
Fraction
I
4
5
t
I
6
I
I
i
I
L
I
2
3
4
5
Number
Fraction
Fig. 3. The effect of ATP on VP-16 activity in isolated nuclei. Nuclei were treated with ATP and VP-16 simultaneously for 60 min. Alkaline elution was then carried out as in Figs. I and 2.
were without effect, as were AMP and A D P (data not shown). To determine whether the effect of ATP was linked to its hydrolysis or, instead, a result of steric modulation, a non-hydrolyzable analog of ATP, AMP-PNP, was tested and also found to be ineffective (Fig. 5).
I 6
Number
Fig. 5. The effect of A M P - P N P as compared to ATP on drug activity in isolated nuclei. Nuclei were incubated with VP-16, ATP and AMP-PNP for 60 min before lysis and alkaline elution at pH 12.1. I.O
08 Ceetrol
VP'16÷ H.L Extract VP-I6
0.6 Extract
.
e
g 04
g
+ V P ' I 6 • Extract
a
==
6
m
0.2
|
EVP' I I ~ • 80 HM
i i ,I 2 5 .5 £)5
i
I
i 2
i 3
i 4
D i 5
mM ATP
Fig. 4. The results of several experiments studying the effect of ATP on drug activity. One on the ordinate represents that amount of damage seen with 10 ~aM VP-16 alone. Nuclei were incubated with drug and ATP as in Fig. 3. Alkaline elution was carried out at pH 12.1.
I
£
I
I
I
I
l
2
3
4
5
6
Fraction
Number
Fig. 6. VP-16 activity in nuclei subjected to 0.35 M NaCI extraction, with addition of the 0.35 M NaCI extract, and with addition of heat-inactivated (HI) extract. Extracted nuclei were treated with drug and extract for 60 min and then exposed to proteinase K before alkaline elution at pH 12.1.
78 In an effort to characterize more fully the putative enzyme, isolated nuclei were subjected to 0.35 M NaC1 extraction, removing non-histone proteins. No VP-16-induced D N A damage could be demonstrated in nuclei so treated. However, addition of the 0.35 M NaC1 protein extract to these nuclei, in the presence of VP-16, reconstituted strand-breaking activity (Fig. 6). Extract exposed to heating for 15 min at 60°C was without effect. Drug-induced strand breaks induced in this system were completely reversed after incubation in drug-free buffer A with 0.5 M NaC1 at 37°C for 60 m i n (data not shown). Discussion
Our data indicate that VP-16-induced D N A damage is dependent upon pH and magnesium concentration and is markedly stimulated by low concentrations of ATP. More importantly, we have shown that heat-labile activity located in a 0.35 M NaC1 extract is critical to drug effect. These data are most consistent with a mechanism of action requiring activity of an intranuclear protein, in all likelihood an enzyme. While it is conceivable that the putative enzyme represents a heretofore undescribed ATP-stimulated endonuclease or, less likely, a drug-activating enzyme, we have adopted the hypothesis that this reaction is mediated by a type II D N A topoisomerase. This is supported by the observations that VP-16 induces both singleand double-strand breaks as well as DNA-protein cross-links [1], and that this activity is stimulated by ATP. Further, D N A damage resulting from interaction of a heat-labile intranuclear component extractable in 0.35 M NaC1 with VP-16 is completely reversible by incubation in 0.5 M NaC1. Given that this reversal occurs under conditions that disperse non-histone proteins from chromatin, it is unlikely that this reaction could be mediated by a typical repair endonuclease. These same phenomena have, however, been previously observed to characterize the interaction of mammalian type II topoisomerase with isolated D N A [9]. The biological roles of eukaryotic type II topoisomerases have yet to be fully defined. In vitro they are characterized by their ability to catalyze the topological passing of two doublestranded D N A segments by transiently introduc-
ing a a reversible double-strand break in one of the crossing segments [10]. Under protein-denaturing conditions, it is possible to stop this reaction in the middle of the nicking-closing cycle and demonstrate a protein-associated D N A strand break which consists of a topoisomerase subunit covalently linked to the 5' end of the nicked D N A strand [9]. Activities conferred upon the enzyme by virtue of its ability to pass an intact doublestranded D N A segment through this transient nick require hydrolysis of ATP [10]. While the cleavage reaction per se is not dependent upon ATP, stimulation of cleavage activity by ATP has been observed (Liu, L.F., personal communication). Notably, it has also been shown that dATP can fully substitute for ATP in topoisomerase-catalyzed reactions [10]. Recently we have found that both VP-16 and VM-26 stimulate cleavage of D N A by a highly purified calf thymus type II topoisomerase [6]. VM-26 is 5-10-fold more active in this assay than is VP-16, a relative potency difference which is also observed for induction of strand breaks as measured by alkaline elution and cytotoxicity after drug treatment in L1210 cells [6]. Type II topoisomerase-mediated D N A cleavage is also stimulated by m-AMSA, an intercalating agent and effective antitumor drug [11]. Much evidence now exists to indicate that this enzyme represents the major intracellular target for m-AMSA, as well as several other intercalating agents [12]. However, these agents by definition bind to D N A by intercalation, making a precise interpretation of their interaction with the enzyme problematic. Using equilibrium dialysis and radiolabeled drug, we have observed no binding of VP-16 to purified D N A [6]. These observations raise the intriguing possibility that VP-16-induced D N A damage may result entirely from direct interactions with the enzyme or an enzyme-DNA complex. More specifically, VP-16 may inhibit strand rejoining by competing with the 3' end of the nicked D N A for binding on the enzyme or may, more simply, act as an allosteric inhibitor to the closing reaction. If experimentally verified in whole cells, VP-16 would be the first chemotherapeutic agent whose effects on D N A are mediated by type II topoisomerase in the absence of drug binding to DNA. Interestingly, oxolinic acid, a compound bearing some
79
H
°~°-~o~/H
HaCO" ~ O C Ha OH
d r u g does n o t b i n d to D N A , its i n t e r a c t i o n with the e n z y m e m a y r e p r e s e n t a novel m e c h a n i s m of a c t i o n for a c h e m o t h e r a p e u t i c agent. F u r t h e r efforts to isolate the essential p r o t e i n from the n u c l e a r extract, c o m p a r e its activities to t y p e II t o p o i s o m e r a s e and, subsequently, directly investigate d r u g - e n z y m e i n t e r a c t i o n s will m o r e clearly delineate the m o l e c u l a r basis for VP-16's efficacy as an a n t i t u m o r agent a n d eventually aid in the d e v e l o p m e n t of new d r u g analogs.
VP-16 O
Ni CH2CHz, OXOLINIC ACID Fig. 7. Molecular structures of VP-16 and oxolinic acid. structural r e s e m b l a n c e to VP-16 (Fig. 7), behaves in an a n a l o g o u s m a n n e r in a p r o k a r y o t i c system, exerting its b a c t e r i c i d a l effect b y b i n d i n g to E. coli D N A gyrase and, thus, p r e v e n t i n g nick closure a n d dissociation of the e n z y m e f r o m the D N A t e m p l a t e [13]. W e have previously shown [4] that d i s u l f i r a m is a p o t e n t i n h i b i t o r o f VP-16-induced s t r a n d b r e a k s a n d cytotoxicity. W e originally i n t e r p r e t e d these d a t a to indicate the i n v o l v e m e n t of a d e h y d r o genase in d r u g activation. M o r e recently we have f o u n d that d i s u l f i r a m also inhibits t y p e II topois o m e r a s e - m e d i a t e d D N A cleavage s t i m u l a t e d b y i n t e r c a l a t i n g agents (W. Ross, T. R o w e et al., u n p u b l i s h e d data). Thus, we n o w consider it unlikely that VP-16 requires activation. In s u m m a r y , VP-16-induced D N A d a m a g e app e a r s to b e m e d i a t e d b y an e n z y m e possessing characteristics consistent with a type II topoisomerase. This finding, in c o n j u n c t i o n with that of Ross et al. [6] d e m o n s t r a t i n g V P - 1 6 - i n d u c e d stimul a t i o n of t y p e II t o p o i s o m e r a s e D N A cleavage in vitro, strongly i m p l i c a t e s type II t o p o i s o m e r a s e as the critical e n z y m e involved in vivo. G i v e n that the
Acknowledgements W e wish to t h a n k A n g e l i q u e F a i r for her secretarial assistance. B.S.G. is a P M A Clinical Pharm a c o l o g y Fellow. W . E . R . is s u p p o r t e d b y g r a n t R C D A CA-00537 from the N a t i o n a l C a n c e r Institute a n d g r a n t CH-261 f r o m the A m e r i c a n C a n c e r Society, F l o r i d a Division. References
1 Wozniak, A.J. and Ross, W.E. (1983) Cancer Res. 43, 120-124 2 Loike, J.D. and Horwitz, S.B. (1976) Biochemistry 15, 5443-5448 3 Roberts, D., Hilliard, S. and Peek, C. (1980) Cancer Res. 40, 4225-4231 4 Wozniak, A.J., Glisson, B.S., Hande, K.R. and Ross, W.E. (1984) Cancer Res. 44, 626-632 5 Long, B.H. and Minocha, A. (1983) Proc. Am. Assoc. Cancer Res. 24, 321 6 Ross, W., Rowe, T., Glisson, B., Yalowieh J. and Liu, L. (1984) Cancer Res., in the press 7 Zwelling, L.A., Michaels, S., Erickson, L.C., Ungerleider, R.S., Nichols, M. and Kohn, K.W. (1981) Biochemistry 20, 6553-6563 8 Ross, W.E. and Bradley, M.O. (1981) Biochim. Biophys. Aeta 654, 129-134 9 Liu, L.F., Rowe, T.C., Yang, L., Tewey, K.M. and Chen, G.L. (1983) J. Biol. Chem. 258, 15365-15370 10 Miller, K.G., Liu, L.F. and Englund, P.T. (1981) J. Biol. Chem. 256, 9334-9339 11 Nelson, E.M., Tewey, K.M. and Liu, L.F. (1984) Proc. Natl. Aead. Sei. USA 81, 1361-1365 12 Tewey, K., Chen, G., Nelson, E. and Liu, L. (1984) J. Biol. Chem. 259, 9182-9187 13 Gellert, M. (1981) Annu. Rev. Biochem. 50, 879-910
Biochimica et BiophysicaActa, 783 (1984) 74-79
Elsevier BBA 91390
C H A R A C T E R I Z A T I O N O F V P - 1 6 - 1 N D U C E D D N A D A M A G E IN I S O L A T E D N U C L E I FROM LI210 CELLS BONNIE S. GLISSON, SHERIN E. SMALLWOOD and WARREN E. ROSS * Departments of Pharmacology and Medicine, University of Florida, Gainesville, FL 32610 (U.S.A.)
(Received April 9th, 1984)
Key words: DNA damage; DNA-drug interaction; VP-16; Enzyme dependence," Drug activation; Strand break
Based on the observation that VP-16-induced D N A damage can be demonstrated in isolated nuclei but not in purified DNA, and that this effect is temperature-dependent, it is postulated that the mechanism of action of VP-16 involves an essential intranuclear event, perhaps enzyme-mediated, which is a prerequisite for the cleavage of DNA. Using alkaline elution to assay single-strand breaks in isolated L1210 nuclei, we have further characterized conditions influencing this putative intranuclear reaction. We have found drug activity to be dependent on magnesium and pH and to be stimulated by low concentrations of ATP (0.05-1 mM), an effect which was not observed with a nonhydrolyzable analog of ATP. Heat-labile activity in a nuclear non-histone protein extract was critical to VP-16-mediated D N A damage. This new evidence lends further credence to the hypothesis that activity of an intranuclear enzyme, possessing characteristics consistent with a type II D N A topoisomerase, is a prerequisite for the cleavage of D N A by VP-16.
Introduction VP-16 (etoposide), a semi-synthetic derivative of podophyllotoxin, has become an important cancer chemotherapeutic agent in the clinical a r m a m e n t a r i u m against several tumors, most notably germ cell tumors of the testis, small cell lung cancer, and lymphoma. Although the precise mechanism of action of VP-16 remains unknown, m u c h evidence now exists to implicate D N A d a m a g e as its major effect. W o r k from this laboratory [1] and others [2,3] indicates that VP-16 or its congener VM-26 induces single-strand breaks in * To whom correspondence should be addressed. Abbreviations: AMP-PNP, (fl,'t-imido)adenosine triphosphate; NTPs and dNTPs, ribonucleoside and 2'-deoxyribonucleoside 5'-triphosphates; VP-16, 4'-demethylepipodophyllotoxin-4-(4,6O-ethylidene-fl-D-glucopyranoside)(etoposide); VM-26, 4'-demethylepipodophyllotoxin-4-(4,6-O-thenylidene-fl-D-glucopyr-
anoside) (teniposide); m-AMSA, 4'-9-acridinyl methanesulfon-m-anisidide.
amino-
0167-4781/84/$03.00 © 1984 Elsevier Science Publishers B.V.
the D N A of m a m m a l i a n cells. We have also shown that D N A double-strand breaks, a more lethal f o r m of damage, and D N A protein cross-links are p r o d u c e d when cells are exposed to VP-16. Further, inhibition of strand-breaking activity by disulfiram and other agents is accompanied by a reduction in cytotoxicity [4]. This suggests that the activity of VP-16 as a chemotherapeutic agent is directly related to its effects on D N A . The mechanism by which VP-16 induces D N A d a m a g e is currently undefined, although the work of Loike and Horwitz [2] supplied the initial clue. T h e y observed no strand-breaking activity when VP-16 was incubated with purified adenovirus or H e L a cell D N A . We were subsequently able to demonstrate, however, that drug-induced D N A d a m a g e does occur in isolated L1210 nuclei and that this effect is temperature-dependent [1]. Thus, we have hypothesized that an intranuclear enzyme, e.g., one mediating drug activation or perhaps one with nuclease activity, is a prerequisite for the
75 cleavage of D N A by VP-16. The possibility that this enzyme is actually a type II topoisomerase was investigated by Long and Minocha [5]. Using a partially purified enzyme preparation, they observed inhibition of catenation of PM2 D N A by VP-16. However, they did not observe strand scission in their system. Thus, it was of interest to characterize further the intranuclear events underlying VP-16-induced D N A damage. During the course of this work, we have also recently found that both VP-16 and its congener VM-26 stimulate D N A cleavage by a highly purified calf thymus type II topoisomerase [6]. The evidence presented herein, which more fully defines intranuclear conditions necessary for maximal drug activity, lends further credence to the hypothesis that type II topoisomerase is the intracellular target for the anti-tumor effects of VP-16. Using alkaline elution to assay single-strand breaks in isolated L1210 nuclei, we found that drug activity was dependent on p H and magnesium concentration and was enhanced by low concentrations of ATP, an effect which appeared to require hydrolysis of the ATP. Finally, we found that heat-labile activity located in a 0.35 M NaC1 extract was critical to VP-16-induced D N A damage. Materials and Methods
Murine leukemia L1210 cells, with a doubling time of approx. 12 h, were grown in Roswell Park Memorial Institute Medium 1630 containing 20% fetal calf serum, penicillin and streptomycin. Cells were labeled with [2-14C]thymidine (53 m C i / mmol; 0.01 # M C i / m m o l ; New England Nuclear, Boston, MA) approx. 20 h before their use in experimentation. VP-16, which was dissolved in dimethyl sulfoxide, was a gift from Bristol Laboratories (Syracuse, NY). NTPs, dNTPs, AMP, A D P and (fl,),-imido)adenosine triphosphate (AMP-PNP) were obtained from Sigma Chemical Co. (St. Louis, MO), dissolved in buffer A (see below), and the p H was corrected to 6.4. Incubation with these compounds occurred simultaneously with that of VP-16 treatment. Isolated L1210 nuclei were used in all experiments. These were prepared by first washing 40 ml
of 14C-labeled whole cells at 5 • 105 c e l l s / m l with cold buffer A (1 m M K H E P O 4 / 5 m M MgC12/150 m M N a C l / 1 m M EGTA) at p H 6.4. The cells were then resuspended in 1 ml of buffer A before lysis with 9 ml buffer B (buffer A plus 0.3% Triton X-100; Eastman K o d a k Co., Rochester, NY). The suspension was kept on ice for 30 min, after which 40 ml buffer A were added and the nuclei sedimented by centrifugation at 1000 rpm for 10 min. Phase microscopy confirmed the presence of isolated nuclei only. After resuspension in buffer A at a density of 10 6 nuclei/ml, the nuclei were treated with drug for 60 rain at 37°C. In experiments utilizing a range of p H or magnesium concentrations, nuclei were suspended in buffer A at the specified p H or magnesium concentration. Nuclear extract containing non-historic proteins was prepared by suspending isolated nuclei in buffer A contianing 0.35 M NaC1 and 0.2 m M dithiothreitol. Centrifugation of the suspension served to separate non-histone proteins in the supernatant from D N A and histone proteins in the pellet. Extracted nuclei were prepared by exposing isolated nuclei on the elution filters to buffer A containing 0.35 M NaCl. This solution was removed by gravity drip, leaving high molecular weight D N A bound to histone proteins on the filter. Treatment with drug and protein extract then followed resuspension of the extracted nuclei in warm buffer A (0.15 M NaC1). High-frequency D N A single-strand breaks were assayed by alkaline elution, as previously described [7]. 14C-labeled nuclei were layered into poly(vinyl chloride) filters (pore size 2 #m; Millipore Corp., Bedford, MA) and lysed with a solution of 2% SDS, 10 m M disodium E D T A and 0.5 m g / m l proteinase K (Merck, Darmstadt, F.R.G.). D N A was then eluted from the filters with tetrapropylammonium hydroxide (RSA Corp., Ardsdale, NY) at an elution flow rate of 0.16-0.20 m l / m i n , collecting fractions at 5-min intervals over a total elution time of 30 rain. The technique in experiments with the nuclear protein extract differed only in the use of polycarbonate filters (pore size 2 /~m), to minimize protein adsorption, and an increase in elution time to 60 min. The alkaline elution assay for D N A double-strand breaks has been previously described in detail [8].
76 Results
I0
Given our hypothesis that interaction of VP-16 with an intranuclear enzyme was necessary for the production of D N A damage, we believed it important to define intranuclear conditions that might influence enzymatic activity and, thus, the degree of drug-induced D N A damage. The effect of pH was investigated in the range 5.0-8.0. The optimal pH range for drug activity was 5.5-6.4, with marked reduction in activity seen at pH 7.0 and higher (Fig. 1). Controls were run at each pH point tested; however, only at pH 5.0 did pH alone influence the rate of elution. At this pH, D N A damage in the control was so excessive as to preclude interpretation of drug-mediated damage (data not shown). Magnesium dependence of the reaction was demonstrated by complete loss of drug activity in the absence of magnesium (Fig. 2). Stepwise restoration of drug activity was seen as
0.8 IOL * I0 n'~ Mg'* tOL~ 0 Mg*÷ 0 Mg**
0.6
I0 mM Mg *° Z
0.4 • I 5 mM M9 .6
3 mM MIG~*
b tL 02
[VP-16] = 4 0 uM
I
2 3 FRACTION
4 NUMBER
5
6
Fig. 2. The effect of magnesium concentration on VP-16 activity in isolated nuclei. Nuclei were suspended in buffer A 1.0 Control 0.8
containing the specified concentrations of magnesium and treated with drug for 60 min. Nuclei were then lysed and DNA single-strand breaks were assayed by alkaline elution at pH 12.1.
VP-16 pH 7 5 VP-16 pH 8 0 VP-16 pH 7 0 06
C
"~
0.4
I~ <~
VP- 16 pH 5.5 VP" 16 pH 6.0
g
O~
V P - 1 6 pH 6.4
o
,,= 02
[ V P - 1 6 ; ' 4 0 pM
II
I 2 Fraction
t 3
I 4
I 5
L 6
Number
Fig. 1. The effect of pH on VP-16 activity in isolated nuclei. 14C-labeled nuclei were suspended in buffer A at the specified pH points and treated with VP-16 for 60 min. After dilution in cold buffer, the nuclei were lysed and DNA single-strand breaks were assayed by alkaline elution at pH 12.1.
magnesium concentration was increased to its optimal 5 mM level. The addition of ATP (0.05-1 mM) resulted in marked enhancement of strand-breaking activity, an effect which gradually diminished at ATP concentrations of 2 mM and higher (Fig. 3). The results of several experiments testing the effect of A T P are represented in Fig. 4. Consistent stimulation of drug activity was seen with concentrations of ATP as low as 0.05 mM, increasing in a dose-related way until maximal effect was achieved with 1 mM ATP. More variability in the dose-response relationship was seen with higher ATP concentrations (2-5 mM), although generally there was gradual loss of this enhancement and, occasionally, inhibition of drug activity. The frequency of drug-induced D N A double-strand breaks was also enhanced by ATP to an extent similar to that observed for single-strand breaks. Though less potent than ATP, stimulation of strand-breaking activity was also seen with dATP (0.5-2 mM). GTP, CTP, UTP, TTP and their deoxy-derivatives
1.0
0.8
0.6
77
~
1.0
AI"P 3 mM AMP-PNP I mM
0.8
0.6
VP-16*
~ P - P N P I mM
i
VP-16
VP-16* ATP 3ram
o
~.
Control
- t - - - - - - - - - - - . 4 k ~ . , ~ a m . 4 Control ATP 3raM
0.4
VP-16
o
e 0.4 ri-
VP-16 * ATP .O5 mM
Q
VP'I6* ATP 2 mM
g
g
g
VP- 16 * ATP .25mM
u-
0.2
0.2
VP-16* ATP .5 mM
VP-16 t ATP I mM
VP'I6* ATP I mM
[.VP- 16] " I0 MM
I
I
I
2
L
I
3
Fraction
I
4
5
t
I
6
I
I
i
I
L
I
2
3
4
5
Number
Fraction
Fig. 3. The effect of ATP on VP-16 activity in isolated nuclei. Nuclei were treated with ATP and VP-16 simultaneously for 60 min. Alkaline elution was then carried out as in Figs. I and 2.
were without effect, as were AMP and A D P (data not shown). To determine whether the effect of ATP was linked to its hydrolysis or, instead, a result of steric modulation, a non-hydrolyzable analog of ATP, AMP-PNP, was tested and also found to be ineffective (Fig. 5).
I 6
Number
Fig. 5. The effect of A M P - P N P as compared to ATP on drug activity in isolated nuclei. Nuclei were incubated with VP-16, ATP and AMP-PNP for 60 min before lysis and alkaline elution at pH 12.1. I.O
08 Ceetrol
VP'16÷ H.L Extract VP-I6
0.6 Extract
.
e
g 04
g
+ V P ' I 6 • Extract
a
==
6
m
0.2
|
EVP' I I ~ • 80 HM
i i ,I 2 5 .5 £)5
i
I
i 2
i 3
i 4
D i 5
mM ATP
Fig. 4. The results of several experiments studying the effect of ATP on drug activity. One on the ordinate represents that amount of damage seen with 10 ~aM VP-16 alone. Nuclei were incubated with drug and ATP as in Fig. 3. Alkaline elution was carried out at pH 12.1.
I
£
I
I
I
I
l
2
3
4
5
6
Fraction
Number
Fig. 6. VP-16 activity in nuclei subjected to 0.35 M NaCI extraction, with addition of the 0.35 M NaCI extract, and with addition of heat-inactivated (HI) extract. Extracted nuclei were treated with drug and extract for 60 min and then exposed to proteinase K before alkaline elution at pH 12.1.
78 In an effort to characterize more fully the putative enzyme, isolated nuclei were subjected to 0.35 M NaC1 extraction, removing non-histone proteins. No VP-16-induced D N A damage could be demonstrated in nuclei so treated. However, addition of the 0.35 M NaC1 protein extract to these nuclei, in the presence of VP-16, reconstituted strand-breaking activity (Fig. 6). Extract exposed to heating for 15 min at 60°C was without effect. Drug-induced strand breaks induced in this system were completely reversed after incubation in drug-free buffer A with 0.5 M NaC1 at 37°C for 60 m i n (data not shown). Discussion
Our data indicate that VP-16-induced D N A damage is dependent upon pH and magnesium concentration and is markedly stimulated by low concentrations of ATP. More importantly, we have shown that heat-labile activity located in a 0.35 M NaC1 extract is critical to drug effect. These data are most consistent with a mechanism of action requiring activity of an intranuclear protein, in all likelihood an enzyme. While it is conceivable that the putative enzyme represents a heretofore undescribed ATP-stimulated endonuclease or, less likely, a drug-activating enzyme, we have adopted the hypothesis that this reaction is mediated by a type II D N A topoisomerase. This is supported by the observations that VP-16 induces both singleand double-strand breaks as well as DNA-protein cross-links [1], and that this activity is stimulated by ATP. Further, D N A damage resulting from interaction of a heat-labile intranuclear component extractable in 0.35 M NaC1 with VP-16 is completely reversible by incubation in 0.5 M NaC1. Given that this reversal occurs under conditions that disperse non-histone proteins from chromatin, it is unlikely that this reaction could be mediated by a typical repair endonuclease. These same phenomena have, however, been previously observed to characterize the interaction of mammalian type II topoisomerase with isolated D N A [9]. The biological roles of eukaryotic type II topoisomerases have yet to be fully defined. In vitro they are characterized by their ability to catalyze the topological passing of two doublestranded D N A segments by transiently introduc-
ing a a reversible double-strand break in one of the crossing segments [10]. Under protein-denaturing conditions, it is possible to stop this reaction in the middle of the nicking-closing cycle and demonstrate a protein-associated D N A strand break which consists of a topoisomerase subunit covalently linked to the 5' end of the nicked D N A strand [9]. Activities conferred upon the enzyme by virtue of its ability to pass an intact doublestranded D N A segment through this transient nick require hydrolysis of ATP [10]. While the cleavage reaction per se is not dependent upon ATP, stimulation of cleavage activity by ATP has been observed (Liu, L.F., personal communication). Notably, it has also been shown that dATP can fully substitute for ATP in topoisomerase-catalyzed reactions [10]. Recently we have found that both VP-16 and VM-26 stimulate cleavage of D N A by a highly purified calf thymus type II topoisomerase [6]. VM-26 is 5-10-fold more active in this assay than is VP-16, a relative potency difference which is also observed for induction of strand breaks as measured by alkaline elution and cytotoxicity after drug treatment in L1210 cells [6]. Type II topoisomerase-mediated D N A cleavage is also stimulated by m-AMSA, an intercalating agent and effective antitumor drug [11]. Much evidence now exists to indicate that this enzyme represents the major intracellular target for m-AMSA, as well as several other intercalating agents [12]. However, these agents by definition bind to D N A by intercalation, making a precise interpretation of their interaction with the enzyme problematic. Using equilibrium dialysis and radiolabeled drug, we have observed no binding of VP-16 to purified D N A [6]. These observations raise the intriguing possibility that VP-16-induced D N A damage may result entirely from direct interactions with the enzyme or an enzyme-DNA complex. More specifically, VP-16 may inhibit strand rejoining by competing with the 3' end of the nicked D N A for binding on the enzyme or may, more simply, act as an allosteric inhibitor to the closing reaction. If experimentally verified in whole cells, VP-16 would be the first chemotherapeutic agent whose effects on D N A are mediated by type II topoisomerase in the absence of drug binding to DNA. Interestingly, oxolinic acid, a compound bearing some
79
H
°~°-~o~/H
HaCO" ~ O C Ha OH
d r u g does n o t b i n d to D N A , its i n t e r a c t i o n with the e n z y m e m a y r e p r e s e n t a novel m e c h a n i s m of a c t i o n for a c h e m o t h e r a p e u t i c agent. F u r t h e r efforts to isolate the essential p r o t e i n from the n u c l e a r extract, c o m p a r e its activities to t y p e II t o p o i s o m e r a s e and, subsequently, directly investigate d r u g - e n z y m e i n t e r a c t i o n s will m o r e clearly delineate the m o l e c u l a r basis for VP-16's efficacy as an a n t i t u m o r agent a n d eventually aid in the d e v e l o p m e n t of new d r u g analogs.
VP-16 O
Ni CH2CHz, OXOLINIC ACID Fig. 7. Molecular structures of VP-16 and oxolinic acid. structural r e s e m b l a n c e to VP-16 (Fig. 7), behaves in an a n a l o g o u s m a n n e r in a p r o k a r y o t i c system, exerting its b a c t e r i c i d a l effect b y b i n d i n g to E. coli D N A gyrase and, thus, p r e v e n t i n g nick closure a n d dissociation of the e n z y m e f r o m the D N A t e m p l a t e [13]. W e have previously shown [4] that d i s u l f i r a m is a p o t e n t i n h i b i t o r o f VP-16-induced s t r a n d b r e a k s a n d cytotoxicity. W e originally i n t e r p r e t e d these d a t a to indicate the i n v o l v e m e n t of a d e h y d r o genase in d r u g activation. M o r e recently we have f o u n d that d i s u l f i r a m also inhibits t y p e II topois o m e r a s e - m e d i a t e d D N A cleavage s t i m u l a t e d b y i n t e r c a l a t i n g agents (W. Ross, T. R o w e et al., u n p u b l i s h e d data). Thus, we n o w consider it unlikely that VP-16 requires activation. In s u m m a r y , VP-16-induced D N A d a m a g e app e a r s to b e m e d i a t e d b y an e n z y m e possessing characteristics consistent with a type II topoisomerase. This finding, in c o n j u n c t i o n with that of Ross et al. [6] d e m o n s t r a t i n g V P - 1 6 - i n d u c e d stimul a t i o n of t y p e II t o p o i s o m e r a s e D N A cleavage in vitro, strongly i m p l i c a t e s type II t o p o i s o m e r a s e as the critical e n z y m e involved in vivo. G i v e n that the
Acknowledgements W e wish to t h a n k A n g e l i q u e F a i r for her secretarial assistance. B.S.G. is a P M A Clinical Pharm a c o l o g y Fellow. W . E . R . is s u p p o r t e d b y g r a n t R C D A CA-00537 from the N a t i o n a l C a n c e r Institute a n d g r a n t CH-261 f r o m the A m e r i c a n C a n c e r Society, F l o r i d a Division. References
1 Wozniak, A.J. and Ross, W.E. (1983) Cancer Res. 43, 120-124 2 Loike, J.D. and Horwitz, S.B. (1976) Biochemistry 15, 5443-5448 3 Roberts, D., Hilliard, S. and Peek, C. (1980) Cancer Res. 40, 4225-4231 4 Wozniak, A.J., Glisson, B.S., Hande, K.R. and Ross, W.E. (1984) Cancer Res. 44, 626-632 5 Long, B.H. and Minocha, A. (1983) Proc. Am. Assoc. Cancer Res. 24, 321 6 Ross, W., Rowe, T., Glisson, B., Yalowieh J. and Liu, L. (1984) Cancer Res., in the press 7 Zwelling, L.A., Michaels, S., Erickson, L.C., Ungerleider, R.S., Nichols, M. and Kohn, K.W. (1981) Biochemistry 20, 6553-6563 8 Ross, W.E. and Bradley, M.O. (1981) Biochim. Biophys. Aeta 654, 129-134 9 Liu, L.F., Rowe, T.C., Yang, L., Tewey, K.M. and Chen, G.L. (1983) J. Biol. Chem. 258, 15365-15370 10 Miller, K.G., Liu, L.F. and Englund, P.T. (1981) J. Biol. Chem. 256, 9334-9339 11 Nelson, E.M., Tewey, K.M. and Liu, L.F. (1984) Proc. Natl. Aead. Sei. USA 81, 1361-1365 12 Tewey, K., Chen, G., Nelson, E. and Liu, L. (1984) J. Biol. Chem. 259, 9182-9187 13 Gellert, M. (1981) Annu. Rev. Biochem. 50, 879-910