mechanism of interaction of cc-1065 (nsc 298223) with dma...obtained on a gary model 15...

[CANCER RESEARCH 42, 2821-2828, July 1982]0008-5472/82/0042-OOOOS02.00

Mechanism of Interaction of CC-1065 (NSC 298223) with DMA

David H. Swenson,1 Li H. Li, Laurence H. Hurley,2 J. Stefan Rokem, Gary L. Petzold, Brian D. Dayton, Tanya

L. Wallace, Alice H. Lin, and William C. Krueger

The Upjohn Company. Kalamazoo. Michigan 49001 Â¡D.H. S., L. H. L., G. L. P., B. D. D., T. L. W., A. H. L., W. C. K.Jand College of Pharmacy. University of Texas,Austin. Texas 78712 [L. H. H. J. S. R]

ABSTRACT

CC-1065 (NSC 298223), a potent new antitumor antibiotic

produced by Streptomyces zelensis, interacts strongly withdouble-stranded DNA and appears to exert its cytotoxic effectsthrough disruption of DNA synthesis. We undertook this studyto elucidate the sites and mechanisms of CC-1065 interaction

with DNA.The binding of CC-1065 to synthetic and native DNA was

examined by differential circular dichroism or by Sephadexchromatography with photometric detection. The binding ofCC-1065 with calf thymus DNA was rapid, being completewithin 2 hr, and saturated at 1 drug per 7 to 11 base pairs. Theinteraction of CC-1065 with synthetic DNA polymers indicateda specificity for adenine- and thymine-rich sites. Agarose gelelectrophoresis of CC-1065-treated supercoiled DNA showedthat CC-1065 did not intercalate. Site exclusion studies usingsubstitutions in the DNA grooves showed CC-1065 to bind

primarily in the minor groove.CC-1065 did not cause DNA breaks; it inhibited susceptibility

of DNA to nuclease St digestion. It raised the thermal meltingtemperature of DNA, and it inhibited the ethidium-induced

unwinding of DNA. Thus, in contrast to many antitumor agents,CC-1065 stabilized the DNA helix. DNA helix overstabilizationmay be relevant to the mechanism of action of CC-1065.

INTRODUCTION

CC-1065 (Chart 1) is a fermentation product of Streptomyces

zelensis that was recently discovered and characterized in theresearch laboratories of The Upjohn Company (6, 16, 17). Thisagent is one of the most cytotoxic antitumor agents known(15-18).3

Although the cytotoxic mechanism of action of CC-1065 isnot known, previous work from these laboratories suggestedthat it may act through inhibition of DNA synthesis (15).3 Thus,

the values for the doses required to inhibit DNA and RNAsynthesis by 50% were between 4 and 6 and between 45 and60 ng/ml, respectively, while protein synthesis was relativelyunaffected at these doses. In addition, DNA polymerase wasmarkedly inhibited by this agent (15).

The inhibition of DNA synthesis is consistent with the strongbinding of CC-1065 to this molecule, based on studies of theTm,4 difference CD spectra, Sephadex chromatography, and

Received November 6, 1981; accepted April 6, 1982.' To whom requests for reprints should be addressed.2 Supported by USPHS Research Grants CA-30349 and CA-31232.3 B. K. Bhuyan, K. A. Newell, S. L. Crampton. and D. D. Van Hoff. CC-1065

(NSC 298223). a most potent antitumor agent: kinetics of inhibition of growth,DNA synthesis, and cell survival, submitted for publication.

4 The abbreviations used are: Tm, thermal melting temperature of DNA; CD,

circular dichroism; poly(dA-dT), alternating copolymer of deoxyadenylate anddeoxythymidylate; poly(dA)-poly(dT), noncovalent double-stranded DNA duplexes; poly(dG-dC), alternating copolymer of deoxyguanylate and deoxycytidyl-

inhibition of CC-1065-induced cytotoxicity in L1210 cells inthe presence of DNA (15). CC-1065 did not bind strongly toheat-denatured DNA or yeast RNA, and its interaction with

protein appeared to be weak and reversible (1 5).In order to elucidate the site and mechanism of binding of

CC-1065 to DNA, we studied the interaction of this drug withvarious natural and synthetic nucleic acids using spectro-

scopic, Chromatographie, and electrophoretic methods.

MATERIALS AND METHODS

Materials. CC-1065 was provided by Dr. D. G. Martin, The Upjohn

Company, Kalamazoo, Mich. Ethidium bromide and type I agarosewere from Sigma Chemical Co., St. Louis, Mo. Calf thymus DNA (Agrade; Calbiochem-Behring Corp., La Jolla, Calif.), poly (dA-dT),poly(dA).polyWT), poly(dG-dC), poly(dG)-poly(dC), poly dA, poly(dT),T-4 DNA (Calbiochem-Behring or Miles Laboratories, Inc., Elkhart,

Ind.) and supercoiled 0X174RF DNA (Bethesda Research Laboratories,Bethesda, Md.) were commercial products.

[3H]MNU was a product of New England Nuclear, Boston, Mass.,and [3H]ethylamine and ['4C]MMS were purchased from Amersham-Searle Corp., Arlington Heights, III. [3H]Ethylnitrosourea was preparedfrom [3H]ethylamine as described previously (22).

Measurement of CC-1065 Interaction with Macromolecules UsingCD. The interaction between CC-1065 and selected macromolecules

was carried out as described previously (13, 14). UV spectra wereobtained on a Gary Model 15 spectrophotometer, and CD spectra wereobtained on a Gary Model 60 spectropolarimeter equipped with aModel 6003 CD attachment (Varian Associates, Inc., Palo Alto, Calif.).The Gary 60 was calibrated with 10-camphorsulfonic acid. DifferenceCD (AGO) spectra were calculated by subtracting the CD of CC-1065from the CD of the CC-1065:macromolecule mixtures. The CD of DNA

in the wavelength range of interest in this report was zero. For allexperiments, the concentration of CC-1065 in 0.01 M sodium phosphate buffer, pH 7.2, was 3.7 x 10~6 M. These concentrations were

achieved by injecting into the buffer the correct amount of a concentration of CC-1065 in DMF (usually 0.06 ml CC-1065 solution in 25 ml

buffer).Thermal Melting of DNA. The thermal melting curves of control or

CC-1065-treated calf thymus DNA or poly(dA-dT) in 0.01 M sodium

phosphate, pH 7.2:1 % DMF were obtained on a Gilford 2400 spectrophotometer equipped with a Model 2527 thermoprogrammer (13, 14).The drug:nucleotide ratio in the CC-1065-treated DNA was 0, 1:7,

1:14, 1:28, 1:56, or 1:112.UV Measurements of the CC-1065 and DNA Interactions. CC-

1065 was caused to react with the DNAs in 5% DMF:0.01 M sodiumphosphate buffer at pH 7.2 and 37Â°. The nucleic acid concentration

was equivalent to 60 nmol DNA-P per ml, and CC-1065 concentrations

in the final reaction ranged from 0 to 24 nmol/ml. After incubation for4 hr, 0.02-ml aliquots of the reaction were applied to a 0.46- x 14-cmSephadex LH-20 column and were eluted with 0.01 M sodium phosphate buffer, pH 7.2, using an Altex Model 322 MP liquid chromato-

ate; poly(dG)-poly(dC), noncovalent double-stranded DNA duplexes; poly(dA).polydeoxyadenylate; poly(dT), polydeoxythymidylate; MNU, methylnitrosourea;MMS, methyl methanesulfonate; DMF, dimethylformamide; DNA-P. DNA-phos-

phate.

JULY 1982 2821

on July 4, 2021. © 1982 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

D.H.Sivensoneial.

CH3

OHOCH3 OCH3

Chart 1. Structure of CC-1065.

graph. The UV absorption of the eluate was measured with WatersModel 440 dual-wavelength photometer. The amount of DNA-boundCC-1065 was estimated from its absorbance at 365 nm assuming an

extinction of 48,000, and the nucleic acid was estimated from itsabsorption at 254 nm. In some experiments, CC-1065 (0 to 7 mM inDMF) was incubated with DNA (1 Â¿trnolDNA-P per ml in 0.01 M sodium

phosphate buffer, pH 7.2). The nucleic acid was precipitated withsodium chloride and ethanol and redissolved in 0.1 M phosphate buffer,and the UV-visible spectrum was determined. The average base molar

extinction values for calf thymus or 0X1 74RF DNA were taken to be6500 M~' cm, and the comparable extinction values for the AT and GC

polymers were taken to be 6600 and 7200, respectively.In some cases, samples of the reaction mixture were chromato-

graphed at various times after mixing to determine the time course ofthe interaction.

It should be noted that the Sephadex LH-20 method did not separateunbound CC-1065 from DNA solely on the basis of molecular sizing,since CC-1065 did not elute from the column under the conditionsused. When large-scale incubations were used, it was found that

precipitation of DNA from the incubation mixture with NaCI and ethanolalso separated unbound CC-1065 from calf thymus DNA. SephadexLH-20 chromatography of the ethanol-precipitated CC-1065-DNA didnot further reduce the amount of DNA-bound drug, nor did NaCI/ethanol precipitation of the Sephadex LH-20-chromatographedDNA:CC-1065 complex alter the drug:DNA binding ratio.

Agarose Electrophoresis. Electrophoresis was carried on 1-mm-

thick slabs of 1% agarose. The gel was prepared by melting 1.0 g ofagarose (type I, low electroendosmosis; Sigma) in 100 ml of buffercontaining 0.08 M Tris-HCI, 0.01 M sodium acetate, and 2 nw EDTA,pH 8.1, at 25Â° (4, 5). Samples of X174RF DNA with or without CC-

1065 pretreatment were treated with 0.2 volume of an aqueous solutionof 20% sucrose:0.25% bromphenol blue, and 10- to 20-/il samples

were applied to separate wells and were electrophoresed at 30 to 35ma until the bromphenol blue dye migrated about halfway down thegel. The slabs were stained for 15 to 30 min with ethidium bromide(0.5 fig/ml) and were photographed under UV with a Polaroid MP-4camera and a yellow filter using Polaroid type 55 P/N film. Densitom-

etry on the negatives was carried out using a Schoeffel scanningdensitometer.

Ethidium bromide titration of CC-1065-treated X174RFDNA was

carried out as described by Espejo and Lebowitz (5), using 1% agarosein 5 x 75-mm tubes.

Alkylation of Calf Thymus DNA and the CC-1065-DNA Adduci.Three ml of calf thymus DNA (1 /xmol DNA-P per ml in 0.01 M sodium

phosphate buffer, pH 7.2) were mixed with dimethylacetamide (0.13ml) or with CC-1065 (0.6 /imol in 0.13 ml dimethylacetamide). Afterincubation for 2 hr at 37Â°,the samples were precipitated with 0.3 ml of

5.2 M NaCI and 6 ml ethanol and were redissolved in 0.01 M sodiumphosphate buffer, pH 7.2, and adjusted to equal DNA concentrations(1.1 to 1.3 /imol DNA-P per ml) with buffer.

The CC-1065-treated and control DNA samples (1.0 ml each) werecaused to react with [3H]ethylnitrosourea, (0.8 Â¿imol,100 /iCi in 0.02ml ethanol), with [3H]MNU (0.09 Â¡imo\,100 /iCi in 0.1 ml ethanol), orwith [14C]MMS (0.03 jumol, 2 Â¿Â»Ciin 0.022 ml ethanol). The DNA was

isolated by precipitation with NaCI/ethanol and was hydrolyzed at100Â° and neutral pH to release 3- and 7-alkyl purines and O2-alkylcy-

tosine. The residual DNA was precipitated with acid and digested in0.1 M HCI at 70Â°for 30 min to liberate purines (1). The hydrolysates

were fractionated by high-pressure liquid chromatography, as de

scribed previously (1 ).Si Nuclease Experiments. These were carried out as described

previously (7) by an adaptation of the method described by Vogt (23).Saturation Binding of CC-1065 to Anthramycin-DNA Adducts.

Anthramycin:DNA adducts of various degrees of binding were performed as described previously by Kaplan and Hurley (9). Excess CC-1065 in 0.5% aqueous DMF was added to 1-ml samples of the

anthramycin:DNA adducts (DNA concentration, 165 /jg/ml). After reacting overnight, the CC-1065:DNA adducts were purified by chromatography on Sephadex G-60 using disodium citrate as an eluant (9).Fractions (1 ml) were collected and assayed for CC-1065 by absorbance at 365 nm. The comparative amounts of CC-1065 bound to the

various complexes were assayed using the absorbance ratio at 365and 259 nm as a quantitation of CC-1065 binding to DNA.

RESULTS

CD Studies. The ACD values were determined at 385 nm,which is the wavelength maximum of the CD induced in the CC-

1065 electronic transition upon interaction with DNA polymers.The ACD for several different DNA samples is given in Table 1.CC-1065, dissolved in DMF, forms a colloidal suspension uponinjection into an aqueous phosphate buffer solution. Macro-molecules in the buffer that bind CC-1065 slowly solubilize thedrug. The data in Table 1 were obtained after 1 day of incubation to allow the CC-1065 to dissolve and bind. The inducedCD spectrum after 2 days of incubation was the same as the 1-

day spectrum. We assume that the induced ellipticity in theCC-1065 electronic transition, as it interacts with the asymmetrical environment of the macromolecule, is directly proportional to its binding affinity with the macromolecule for at leastone mechanism of binding. As Table 1 shows, the AT polymersreacted with CC-1065 to a much greater extent than did theGC polymers. Calf thymus and phage T-4 DNA (65% glycosyl-

ated in the major groove) reacted nearly as well as did the ATpolymers. However, single-strand DNA does not appear toefficiently bind CC-1065, since poly(dA) and poly(dT) did notinteract with CC-1065 to produce a large induced CD. Therewas no indication from the CD spectrum that the CC-1065adduct, once formed, was unstable or decomposed over a 10-day period. Moreover, CD analysis of the CC-1065:DNA adduct

indicated its resistance to venom phosphodiesterase in contrast to control DNA.5

Tm Studies. We found that the interaction of CC-1065 withcalf thymus DNA increased the Tm of the DNA in a dose-dependent manner. At the highest CC-1065 concentration, the

Tm could only be estimated for calf thymus DNA from the initialslope of the curve, which was below 100Â°.Based on the higher

Tm for poly(dA-dT), CC-1065 appeared to interact morestrongly with poly(dA-dT) than with calf thymus DNA (Table 2).

UV Measurements of the CC-1065 and DNA Interactions.The reaction of CC-1065 with calf thymus DNA gave rise to an

adduct with a UV absorption maximum at 360 to 380 nm inaddition to the expected DNA absorption at 259 nm. TheDNA:drug complex could be isolated from the reaction mixtureby NaChethanol precipitation or by chromatography on Sephadex LH-20. Further precipitation on Sephadex LH-20 chromatography of the isolated CC-1065:DNA adduct did not reduce the level of CC-1065 coisolated with the DNA (data not

5 W. C. Krueger. unpublished observations.

2822 CANCER RESEARCH VOL. 42



Interaction of CC- 7065 with DNA

shown). There was no evidence of a large extinction changefor CC-1065 upon interaction with DNA.

Using the Sephadex LH-20 system, we found that CC-1065

reacted rapidly with calf thymus DNA. Within 5 min of mixing,the binding was about 85% of maximum. After 2 hr, little or nofurther reaction was noted. Further incubation at 37Â°for up to

23 hr did not result in a loss of CC-1065 from the DNA. CC-1065 bound to calf thymus DNA in a dose-dependent manner

that saturated at about 1 drug bound per 22 base residues atan initial ratio of 1 drug per 10 base residues in the reaction(Chart 2).

The relative reactivity of CC-1065 toward other DNA polymers at 3 CC-1065 concentrations is given in Chart 3. Theorder of reaction is poly dA-poly(dT), poly(dA-dT) > calf thymus DNA > 4>X174RF DNA Â» poly(dG-dC), poly(dG)-(dG)-

poly(dC).Saturation Binding of CC-1065 to Anthramycin:DNA Com

plexes. Chart 4 shows the relative amount of CC-1065 boundto anthramycin-DNA complexes at varying ratios of anthramy-

cin:DNA nucleotide. At low anthramycin bound:DNA nucleotideratios (

D. H. Swenson et al.

Charts. Binding of CC-1065 with nucleic acids.CC-1065 was mixed with samples of double-strandednucleic acids at 3 levels of drugiDNA nucleotide, andbinding was detected as described for Chart 2. Experiment 3. S.D., 1 to 9% of the means.

Ill 0.079O 0.06

9 0.03 -

Uu

INITIAL CC-1065/DNA NUCLEOTIDEIN REACTION

D 0.01| 0.08

I >0.2

POLY (dA-dT) POLY dA-POLY dT CALF THYMUSDNA

1.00Â»

0.000.02 0.04 0.06

ANTHRAMYCIN BOUND/DNA BASE0.08

Chart 4. Binding of CC-1065 to anthramycin-treated DMA. Anthramycin wasbound to DNA at various levels of substitution. The DNA:anthramycin adductswere then reacted with CC-1065, and the binding of CC-1065 to DNA was

estimated after Sephadex chromatography. See text for details.

(apparent first-order process) under these conditions to be

approximately 4 hr.Fig. 2 shows the ethidium bromide-induced intercalative

unwinding of a sample of 0X174RF DNA that was preincubatedwith CC-1065 at an initial drug:DNA nucleotide ratio of 1:50.Compared to unreacted DNA (Fig. ~\A), it required about 20

times the concentration of ethidium bromide to achieve electro-phoretic comigration of Forms I and II for that CC-1065-treated

DNA (Fig. 2) compared to untreated DNA (Fig. 1,4).We found the concentration of ethidium required to cause

electrophoretic comigration of Forms I and II of X174RFDNAwas dependent on the amount of CC-1065 bound to the DNA.Chart 6 is a plot of the distance between Forms I and II DNA asa function of ethidium bromide concentration in the electropho-resis gel, for various initial CC-1065:DNA nucleotide ratios. Asthe drug:nucleotide ratio increased, the amount of ethidiumrequired to cause comigration of Forms I and II increased, anda broader concentration range was required to effect thecomigration and continued unwinding to the supercoil of theopposite polarity.

Suppression of DNA Alkylation by CC-1065. We found(Table 3) that alkylations at N-3 of guanine, N-3 of adenine,and O2 of cytosine (all in the minor groove) were suppressed50 to 80% in DNA samples pretreated with CC-1065. In contrast, alkylation at N-7 and O6 of guanine and at the N-7 of

adenine (major groove) were not reduced by pretreatment ofDNA with CC-1065. Similar results were obtained with MMS(data not shown).

|(>X174 RFSUPERCOILED

DNA

POLY (dG-dC)POLY dG-POLY dC

120

Chart 5. Effect of CC-1065 on S, nuclease sensitivity of calf thymus DNA.Calf thymus DNA, heat-denatured DNA, or CC-1065-treated DNA were incubatedwith S, nuclease, and the digestion was followed by release of acid-solublenucleotides. See text for details. DNA concentration was 135 /ig/ml. Enzymeconcentration was 6 (il/ml (8). CC-1065:DNA nucleotide ratios were 0.031 (â€¢)and 0.043 (O). A, calf thymus DNA; A, heat-denatured DNA.

0.0

CONCENTRATION OF ETHIDIUM BROMIDE IN GEL ,..q/ ml,

06 09 1.2 1.5 1.8 21 24 2.7 3.0

Chart 6. Dose-dependent effects of CC-1065 on ethidium-induced unwindingof supercoiled DNA. 0X174RF DNA was reacted with CC-1065 at variousdrugrnucleotide ratios as outlined on the chart. The ethidium-induced unwindingcurve was determined for each DNA sample as outlined in the text. CC-1065:DNAnucleotide: O, 0; â€¢,0.01; A, 0.015; A, 0.025; D, 0.33.

DISCUSSION

We have found that CC-1065 interacts strongly with double-

stranded DNA (3, 15) on the basis of Tm and CD measurements. CD has been used previously to assess the interactionof both chiral and nonchiral (4, 13, 26, 27) agents with DNA.Since little or no interaction between CC-1065 and poly(dA) orpoly(dT) was detected (Table 1) and since only the complexes




Interaction of CC-1065 with DNA

Table 3Suppression of minor groove DNA alkylation by CC- ÃŒ065

Calf thymus DNA (3 /imol DNA-P in 3 ml 0.01 M sodium phosphate, pH 7.2)was mixed with dimethylacetamide or CC-1065 (0.6/imol) in 0.13 ml dimethyla-cetamide and incubated at 37" for 2 hr. The samples were precipitated with

NaCI: ethanol and redissolved in 0.01 M sodium phosphate buffer. The controland CC-1065-treated samples were adjusted to equal DNA concentrations andwere treated equally with [3H]MNU or [3H]ethylnitrosourea as described in"Materials and Methods.' The alkylated samples were isolated and digested,

and alkylated base residues were separated by high-pressure liquid chromatog-raphy (approximately 1 /imol DNA-P per sample) as described in the text. Thevalues shown are from a single experiment. The study was repeated at differentlevels of alkylation with similar responses.

/imol alkyl/molDNA-PSite

ofalkylation[3H]MNU-methylatedDNA7-Adenine7-GuanineO6-Guanine3-Adenine3-Guanine[3H]ENU-ethylated

DNA7-Adenine7-GuanineO6-Guanine3-Adenine3-GuanineO2-CytosineGroove

locationMajorMajorMajorMinorMinorMajorMajorMajorMinorMinorMinorControlDNA0.39.51.11.40.20.613.77.76.01.53.4CC-1

065-treatedDNA0.611.31.40.40.11.213.17.91.10.71.0

of CC-1065 and certain double-stranded DNA polymers butnot that of CC-1065 and yeast RNA (data not presented) wereisolated by Sephadex LH-20 chromatography (Charts 2 and 3),we confirm that double strandedness was required for a strongasymmetrical interaction of CC-1065 with DNA (15). The Tm ofthese drug:DNA adducts increased sharply in a dose-relatedfashion (Table 2). CC-1065 apparently binds to DNA much

stronger than do Adriamycin, distamycin A, netropsin, andnogalamycin (14, 15, 24-27).

In the CD studies, nucleic acid appeared to solubilize theCC-1065, but this initial solubilization did not cause a large CD

increase (type I binding). A second binding phenomenon (typeII) was characterized by a large ACD (Table 1) and occurred inthe first 4 to 6 hr. Over the next 24 hr, the wavelength ofmaximum ellipticity for the CC-1065 chromophore slowly underwent a 5- to 20-nm hypsochromic shift (type III binding).Type I binding occurred with all macromolecules examined,including single-stranded DNAs, RNA, and serum albumin,

whereas type II and III binding occurred mainly at AT sites indouble-stranded DNA (Table 1). It is likely that type I binding

was eliminated by Sephadex chromatography whereas typesII and III binding were Sephadex resistant. Furthermore, aportion of the DNA-bound CC-1065 could be removed byextraction with phenol:cresol (8), but by 24 hr the CC-1065:DNA complex was resistant to this treatment. Thesepresumably represent the type II and III binding.

Generally, the CD and Sephadex methods yielded parallelresults (Fig. 1, Chart 3) with the exception of poly(dG-dC),which did not show binding to CC-1065 by the Sephadexmethod (Chart 3). This may be due to the shorter incubationtime used in that study (4 hr) compared to the 24 hr used in theCD study (Fig. 1).

On the basis of the 3 methods (Tables 1 and 2, Chart 3, andRef. 15), the order of reactivity of the DNA polymers was

poly(dA-dT) > poly(dA)-poly dT > calf thymus DNA ~ T-4phage DNA > X174RF DNA > poly(dG-dC) > heat denaturedDNA, poly(dG)-poly(dC), poly(dA), and poly(dT). From this, weconclude that CC-1065 reacts preferentially at AT-rich sites inthe double-stranded DNAs. This conclusion was further supported by our observation that anthramycin, a reagent thatpreferentially binds GC-rich sites in DNA (7, 19), only partiallyblocked the binding of CC-1065 (Chart 4), while the AT-selec-tive reagent netropsin strongly interfered with CC-1065 binding(3). Supercoiled X174RFDNA bound less CC-1065 than did

calf thymus DNA (Chart 3). It is not clear whether this is due todifferent AT distributions in the DNAs or to supercoiling. However, we found in preliminary experiments that double-stranded

linear Form III DNA (produced by treatment of supercoiledForm I with Pstl restriction enzyme) did not significantly differfrom Form I DNA in amounts of CC-1065 bound. Further work

with various supercoiled and linear DNAs will be required toconfirm this observation.

The interaction between CC-1065 and DNA did not involveintercalation, as evidenced by the inability of CC-1065 to

unwind supercoiled X174RFDNA, whereas ethidium bromide,an intercalating agent, clearly unwound the supercoiled DNA(Fig. 1). The mechanism for this process has been discussedelsewhere (5).

Four lines of evidence indicate that CC-1065 binds primarilyin the minor groove of DNA. (a) The binding of CC-1065 withT-4 phage DNA (Table 1) and the kinetics of binding, followed

by CD, were approximately the same as with calf thymus DNA.Since T-4 phage DNA is glycosylated at hydroxymethylcytosine

residues in the major groove, one would expect reduced binding with T-4 DNA if the major groove were involved (3).

(b) Netropsin, a reagent that binds to DNA at AT sites in theminor groove (10, 24, 26, 27), inhibited the binding of CC-1065 to calf thymus DNA (3). CC-1065 could slowly replacenetropsin from the DNA, but CC-1065 once bound could not

be displaced by netropsin (3). These results also indicate thatCC-1065 binds to DNA more strongly than does netropsin.

(c) The binding of CC-1065 to DNA could be partially in

hibited by preincubation of DNA with anthramycin (Chart 4), areagent that covalently reacts with the N-2 of guanine in theminor groove (7). The failure of anthramycin to inhibit CC-1065

binding completely, even at high anthramycin:nucleotide ratios(Chart 4), may reflect the differing base specificity of the 2reagents.

Finally, methylation and ethylation at minor but not majorgroove sites was strongly inhibited by CC-1065 (Table 3). This

approach has been used to characterize the minor groovespecificity of distamycin A and netropsin (10). More recentwork has shown, however, that alkylation of major groove sites(N-7 and O6 of guanine) by MNU was inhibited by pretreatment

of DNA with distamycin A (20); hence, the specificity of CC-1065 for the minor DNA groove may exceed that of distamycinA.

Although the evidence weighs heavily in favor of the minorDNA groove as the major binding site for CC-1065, the datacannot totally exclude the possibility of low levels of majorgroove binding. The X-ray crystallographic structure of CC-1065 shows a twist and pitch in the molecule that wouldprovide a good match for the major groove (3).

The complex formed between CC-1065 and calf thymus DNAwas either very strong noncovalent or covalent, and this com-

JULY1982 2825



O. H. Swenson eÃ-al.

plex appeared to be stable. Thus, we could not separate boundCC-1065 from DNA by Sephadex LH-20 chromatography and

repeated precipitation with NaChethanol. However, at earlytimes, it could be partially dissociated by a phenolrcresolextraction. In addition, CC-1065 inhibited the ethidium-induced

intercalative unwinding of X174RFDNA as determined by gelelectrophoresis (Fig. 2; Chart 6). Had the complex been weaklyassociated, it would have dissociated in the electrophoreticgradient, and little or no effect on ethidium interaction withDNA would have been noted.

The CC-1065:DNA complex appears to be stable at 37Â°and

neutral pH and did not cause single- or double-stranded breaks

in DNA (Fig. 1). Also, we did not observe any changes in theCD spectrum over a 1-week period that would suggest decom

position of the complex (e.g., ACD or wavelength shifts). Inaddition, Sephadex chromatography showed no loss of CC-

1065 from the complex. This was consistent with our previousobservation that CC-1065 when complexed with DNA was notcytotoxic to L1210 cells in culture, even during a 3-day incubation at 37Â°(15).

CC-1065 appears to stabilize the DNA helix. Evidence forthis comes from 3 sources, (a) Tm studies show that CC-1065causes a dose-dependent increase in the melting transition forcalf thymus DNA (Table 2). (b), CC-1065 inhibits the S, nu-

clease susceptibility of calf thymus DNA (Chart 5). S, nucleasedegrades single-stranded and locally denatured regions ofDNA. We interpret this to indicate that CC-1065 suppresses

the local denaturation of DNA in much the same way as itsuppresses thermal denaturation of DNA. (c), CC-1065-treated4>X174RF DNA was resistant to ethidium-induced intercalative

unwinding (Fig. 2; Chart 6). We interpret this to indicate astiffening of the helix by the CC-1065 that is similar to that

found for distamycin A and netropsin (26). Minor groove reagents such as distamycin A reduce binding of ethidium bromide to DNA (2). The mechanism for this is unclear but is likelyto be a consequence of helix stiffening that in turn inhibits theDNA helix unwinding necessary to accommodate intercalativebinding of ethidium. The direct displacement of ethidium fromthe helix may also occur, since the phenyl ring of ethidiumbromide sits in the minor DNA groove (2).

We speculate that one mechanism of CC-1065 cytotoxicity

is that it inhibits the normal melting and unwinding processesthat are required for DNA synthesis and may act as a stericblock to the enzymes that operate on DNA during replicationand transcription (15).

Whether the nature of the interaction of CC-1065 in the

minor DNA groove involves covalent bond formation has notyet been resolved. Braithwaite and Baguley (2) noted that, fora series of antitumor compounds with affinity for AT sites in theminor DNA groove, hydrogen bond formation may contributeto the selectivity of these reagents but is not essential. Chides-ter ef al. (3) noted that CC-1065 can form several hydrogen

bonds while in one of the DNA grooves.In general, CC-1065 appears to share some DNA-binding

characteristics with distamycin A or netropsin. However, CC-

1065 has a cyclopropyl function that may serve as an alkylatingfunction and tie the molecule firmly in the groove. Indirectevidence for the suggestion comes from the observation thatCC-1065 can displace netropsin from DNA but once boundcannot be displaced by netropsin (3). The apparent stability ofthe adduct, as shown by electrophoresis, and the resistance of

CC-1065 to separation from DNA by repeated solvent precipi

tation or by Sephadex chromatography are also suggestive ofcovalent bond formation. If a covalent bond is formed at ATsites in the minor DNA groove, 2 sites may be considered majorpotential targets: the N-3 of adenine and the O2 of thymine (1,

12, 21). The formation of single-strand breaks upon heatingCC-1065-treated X174RF DNA is consistent with alkylation

at either of these sites (21). Further work is in progress in thislaboratory to characterize more fully the nature and sites ofCC-1065 binding to DNA and to determine whether alkylationof DNA by CC-1065 occurs.

ACKNOWLEDGMENTS

The authors wish to acknowledge the technical assistance of D. J. Kaplan.

REFERENCES

1. Beranek, D. T., Weis, C. C., and Swenson, D. H. A comprehensive quantitative analysis of methylated and ethylated DNA using high pressure liquidchromatography. Carcinogenesis. 1: 595-606, 1980.

2. Braithwaite, A. W., and Baguley. B. C. Existence of an extended series ofantitumor compounds which bind to DNA by nonintercalative means. Biochemistry, 19: 1101-1106, 1980.

3. Chidester. C. G., Krueger, W. C.. Mizsak, S. A., Duchamp, D. J.. and Martin.D. G. The structure of CC-1065, a potent anti-tumor agent and its binding toDNA. J. Am. Chem. Soc., 103: 7629-7635. 1981.

4. Drinkwater, N. R.. Miller. J. A., Miller, E. C., and Yang, N.-C. Covalentintercalative binding to DNA in relation to the mutagenicity of hydrocarbonepoxides and W-acetoxy-2-acetylaminofluorene. Cancer Res.. 38. 3247-

3255. 1978.5. Espejo, R. T., and Lebowitz, J. A simple electrophoretic method for the

determination of superhelix density of closed circular DNAs and for observation of their superhelix density heterogeneity. Anal. Biochem., 72. 95-103, 1976.

6. Hanka, L. J., Dietz, A.. Gerpheide, S. A., Kuentzel, S. L., and Martin, D. G.CC-1065 (NSC-298223), a new antitumor antibiotic. J. Antibiot. (Tokyo),31: 1211-1217, 1978.

7. Hurley, L. H., and Petrusek, R. Proposed structure of the anthramycin-DNAadduct. Nature (Lond.), 282. 429-531, 1979.

8. Irving, C. C., and Veazey, R. A. Isolation of deoxyribonucleic acid andribosomal ribonucleic acid from rat liver. Biochim. Biophys. Acta, 166: 246-

248, 1968.9. Kaplan, D. J., and Hurley, L. H. Anthramycin binding to DNA-mitomycin C

complexes. Evidence for drug induced DNA conformational change andcooperativity in mitomycin C binding. Biochemistry. 20 7572-7580, 1981.

10. Kolchinskii, A. M., Mirazabekov. A. D.. Zasadatelev, A. S., Gurskii, G. V.,Grakhouskii, S. L., Zhuze, A. L., and Gottikh, B. P. Structure of complexesof antibiotics of the distamycin type and actinomycin D with DNA: newexperimental data on the location of the antibiotics within the narrow grooveOf DNA. Mol. Biol. (Mosc.). 9. 19-27. 1975.

11. Kuhnlein, U., Tsang, S. S.. and Edwards, J. Characterization of DNA damages by filtration through nitrocellulose filters: a simple probe for DNA-modifying agents. MutÃ¢t.Res., 64: 167-182, 1979.

12. Lawley, P. D. Methylation of DNA by carcinogens: some applications ofchemical analytical methods. In: R. Montesano, H. Bartsch, and L. Tomatis(eds.) Screening Tests in Chemical Carcinogenesis. pp. 181-208. Lyon.France: International Agency for Research on Cancer, 1976.

13. Li, L. H., Clark, T. D., Murch, L. L., Wooden. J. M., Pschigoda, L. M.. andKrueger. W. C. Biological and biochemical effects of chartreusin on mammalian cells. Cancer Res., 38. 3012-3018, 1978.

14. Li, L. H., Kuentzel, S. L., Murch, L. L., Pschigoda, L. M.. and Krueger, W. C.Comparative biological and biochemical effects of nogalamycin and itsanalogs on L1210 leukemia. Cancer Res., 39. 4816-4822, 1979.

15. Li, L. H., Swenson, D. H.. Schpok, S. L. F., Kuentzel. S. L., Dayton. B.. andKrueger, W. C. CC-1065 (NSC 298223). a novel antitumor agent thatinteracts with strongly double-stranded DNA. Cancer Res.. 42. 999-1004,

1982.16. Martin, D. G., Biles, C., Gerpheide, S. A., Hanka, L. J., Krueger, W. C.,

McGovren, J. P., Mizsak, S. A., Neil, G. L., Stewart, J. C. and Visser, J. CC-1065 (NSC-298223), a potent new antitumor agentâ€”improved production

and isolation, characterization and antitumor activity. J. Antibiot. (Tokyo).34: 1119-1125, 1981.

17. Martin, D G., Chidester, C. G., Duchamp. D. J.. and Mizsak, S. A. Structureof CC-1065 (NSC-298223). a new antitumor antibiotic. J. Antibiot. (Tokyo).33. 902-903, 1980.

18. Neil, G. L., Clarke, G. L., and McGovren, J. P. Antitumor activity and acutetoxicity of CC-1065 (NSC-298223) in the mouse. Proc. Am. Assoc. Cancer




Interaction of CC-1065 with DNA

Res.. 22. 244. 1981.19. Petrusek. R. L., Anderson, G. L., Games, T. F., Fannin, Q. L, Kaplan, D. J.,

Zimmer. S. G., and Hurley, L. H. Pyrrolo(1,4)-benzodiazepine antibiotics.Proposed structures and characteristics of the in vitro DNA adducts ofanthramycin. tomaymycin, sibiromycin and neothramycins A & B. Biochemistry. 20. 1111-1119, 1981.

20. Rajalakshmi, S.. Rao, P. M.. and Sarma, D. S. R. Studies on carcinogenchromatin-DNA interaction: inhibition of MNU-induced methylation of chro-matin-DNA by spermine and distamycin A. Biochemistry, 17: 4515-4518,

1978.21. Singer, B., Kroger, M., and Carano. M. O2 and O4-alkyl pyrimidine nucleo-

sides: stability of the glycosyl bond and of the alkyl group as a function ofpH. Biochemistry, 17: 1246-1250, 1978.

22. Swenson. D. H., and Lawley, P. D. Alkylation of deoxyribonucleic acid bycarcinogens dimethyl sulphate, ethyl methanesulphonate, N-ethyl-N-nitro-

sourea and N-methyl-W-nitrosourea. Relative reactivity of the phosphodiestersite thymidylyl(3'-5')thymidine. Biochem. J., / 71: 575-587. 1978.

23. Vogt. V. M. Purification and further properties of single-strand-specificnuclease from Aspergillus oryzae. Eur. J. Biochem.. 33 192-200, 1973.

24. Wartell, R. M., Larson, J. E., and Wells, R. D. Netropsin. J. Biol. Chem.,249: 6719-6731, 1974.

25. Wartell, R. M., Larson, J. E.. and Wells. R. D. The compatibility of netropsinand actinomycin binding to natural DNA. J. Biol. Chem., 250. 2698-2702,

1975.26. Zimmer, C. Effects of the antibiotics netropsin and distamycin A on the

structure and function of nucleic acids. Prog. Nucleic Acid Res. Mol. Biol.,15: 285-313, 1975.

27. Zimmer, C., Luck, G., Thrum, H., and Pitra, C. Binding of analogs of theantibiotics distamycin A and netropsin to native DNA. Eur. J. Biochem., 26:81-89, 1972.

Fig. 1. CC-1065 does not unwind or break supercoiled DNA. A. typical ethidium-induced intercalative unwinding of supercoiled X174RFDNA as described byEspejo and Lebowitz (5). B, CC-1065 does not cause intercalative unwinding of X174RFDNA when preincubated with DNA at druginucleotide ratios ranging from10~4 to 2 x 10~' (approximately 1 to 2000 drug molecules per DNA molecule). See text for details.

Fig. 2. Effect of CC-1065 on ethidium-induced intercalative unwinding of supercoiled DNA. X174RFDNA was treated with CC-1065 at a level of 1 drug moleculeper 50 DNA nucleotides. The DNA was electrophoresed on agarose gels containing ethidium bromide as described in the text.

JULY 1982 2827



O 0.02 0.04 0.06 0.08 0.10

CONCENTRATION OF ETHIDIUM BROMIDE ^g/ml)

4 16 64 256 1024

INITIAL CC-1065/DNA MOLECULE

0.4 0.8 1.2 1.6

ETHIDIUM BROMIDE IN GEL

2828



1982;42:2821-2828. Cancer Res David H. Swenson, Li H. Li, Laurence H. Hurley, et al. Mechanism of Interaction of CC-1065 (NSC 298223) with DNA

Updated version

http://cancerres.aacrjournals.org/content/42/7/2821

Access the most recent version of this article at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/42/7/2821To request permission to re-use all or part of this article, use this link


http://cancerres.aacrjournals.org/content/42/7/2821http://cancerres.aacrjournals.org/cgi/alertsmailto:[email protected]://cancerres.aacrjournals.org/content/42/7/2821http://cancerres.aacrjournals.org/

mechanism of interaction of cc-1065 (nsc 298223) with dma...obtained on a gary model 15...

Documents