Indian Journal of Biochemistry & Biophysics Vol. 38, February & April 2001, pp. 64-70

Interactions of chromomycin A3 and mithramycin with the sequence d(TAGCTAGCTA)2

Sukanya Chakrabarti" and Dipak Dasgupta*

*Biophysics Division, Saha Institute of Nuclea r Physics, 37, Belgachhia Road, Calcutta 700 037, India

"Chemistry Department , Lady Brabourne College, PI/2 Suhrawardy Avenue, Calcutta 700 017, India

Accepted 3 October 2000

Anti-cancer antibiotics, chromomycin A3 (CHR) and mithramycin (MTR) inhibit DNA directed RNA synthesis in vil'O by binding reversibly to template DNA in the minor groove with GC base specificity, in the presence of divalent cations like Mg2+. Under physiological conditions, (drug)2Mg2+ complexes formed by the antibiotics are the potenlial DNA binding li gands. Structures of CHR and MTR differ in their saccharide residues. Scrutiny of the DNA binding properties reveal sign ificant differences in their sequence se lecti vity, orientation and stoichiometry of binding. Here, we have analyzed binding and thermodynamic parameters for the interac tion of the antibiotics with a model oligonucleotide sequence, d(TAGCTAGCTAh to understand the role of sugars. The oligomer contains two poten ti al binding sites (GpC) for the ligands. The study illustrates that the drugs bind differently to the sequence. (MTRhMg2

+ binds to both sites whereas (CHR)2Mg2+ binds to a single site. UV melting profiles for the decanucleotide saturated with the ligands show that MTR bound oligomer is highly stabilized and melts symmetrically. In contrast, with CHR, loss of symmetry in the oligomer following its association with a single (CHR)zMg2+ complex molecule leads to a biphasic melting curve. Results have been interpreted in the light of saccharide dependent differences in li gand nexibility between the two antibiotics.

Introduction Chromomycin A3 (CHR) and Mithramycin (MTR) are two anti-cancer antibiotics I of the aureolic acid group that potentially inhibit DNA directed RNA synthesis, in viv02.3 (Fig. I). They do so by binding reversibly to template DNA in the minor groove with GC base specificity4-7. At and above physiological pH the above association takes place only in the presence of divalent cations5, like Mg2+ . We have earlier shown that in the absence of DNA, the antibiotics have the potential to bind to Mg2+, to form two different types of drug-Mg2+ complexes. The stoichiometries of the complexes formed are 1: 1 (Complex J) and 2: 1 (Complex II) in terms of drug:Mg2+ rati08

.11 . These

complexes bind to DNA at and above physiological pH. Studies from different laboratories including ours have shown that under normal physiological conditions, in the presence of millimolar concentration of the metal ion, (drug)2Mg2+ complex or complex II is the DNA binding ligand8

.9

.I2

.

Structurally, the two antibiotics are closely related, each conslstlOg of a planar aglycone, the chromomycinone moiety and five hexapyranoses

*Author iorcorrespondence. Fax: 09 1-033-337 4637; Phone: 091-033-556 5611; E-mail : dipak@biop.saha.ernet.in

attached to it (Fig. 1). The antibiotics differ only in a small number of substitutions in the saccharide residues. Unlike CHR, MTR does not contain acetoxy or methoxy substituents on its sugar residues. These differences lead to significant dissimilarities in their chemical properties. As for example, MTR has a pKa of 5.0 whereas for CHR, the valuel3 is 7.0.

Gross structural similarities between the two antibiotics lead to GC base specific association for both the (drughMg2+ complexes with DNA. The association originates from ligand induced alteration in the base specific groove width dimension, finally leading to a local di stortion of the B-DNA double helix to a A-type conformation. However, a closer scrutiny of the DNA binding properties also indicate significant differences which are pri marily the result of chemically distinctive sugar residues. Footprinting studies with the drug-DNA complexes show that the two antibiotics exhibit differences in their sequence selectivit/·14, though the mjnimum requirement of two contiguous GC base pairs for association is observed for both an tibiotics6. NMR studi es indicate a difference in the geometry and orientation of the two antibiotics in their complex with

I· I'd 11 1516 Th' I d d' ff . o Igonuc eotl es -. ' . IS ea s to I erences In their binding geometry and binding stoichiometry with the same DNA sequence. However, the role of

CHAKRABARTI & DASGUPTA: INTERACTIONS OF CHROMOMYCIN A3 AND MITHRAMYCIN 65

R'O

OH

4' CH;)

OH

R'O

CH3 CH3

./\~\HO~--O\ HO.j()~O~~

HO B A

OH OC~

Z 3' 4' CHJ

OH

Fig. I- Structures of the antibiotics: rea): Ch ro llloillycin A3; (b): Mithralllycin .J

sugar residues in the two antibiotics leading to different DNA binding characteristics of the antibiotic is still not clearly understood. It is necessary to understand this aspect in order to explain their pharmacological properties which are attributed to the DNA binding potentials, e.g. in terms of transcription inhibition potential under in vivo conditions, CHR is found to be more effective than MTR.

As a sequel to our studies to understand the role of sugar residues in the base specific recognition of these antibiotics with their prime cellular target DNA, we have examined the effects of ligand flexibility and concomitant ON A di stortion upon the association process using the model oligonucleotide sequence d(TAGCTAGCTAh. While the GpC base step in the sequence is sufficient fo r the association of the (drughMg2+ complexes with it, the characteristic feature of the oligomer is the presence of two GpC binding sites separated by one TpA base step. Since the binding site per ligand molecule is approximately around 6-7 bases, the two potential binding sites are partially overlapping. Spectroscopic techniques such as absorbance and fluorescence have been used to monitor the interaction because they permit us to work at the micromolar range of antibiotic concentration ( 10-50 f..lM) where aggregation of the free ligand is absent as detected by the optical spectroscopic tools 17. The interactions were further characterized from analysi s of DNA melting studies and thermodynamic parameters associated with the interactions. The results have been compared with

structural information known from previous NMR 16. IS

and optical spectroscopic studies l 9.

Materials and Methods

Materials Chromomycin A3, mithramycin, Tris and

magnesium chloride solution (4.9 M) were fro m Sigma Chemical Company, USA. The oligonucleotide, d(TAGCTAGCTAh was synthesized using Gene Assembler Special from Pharmacia Biotech Ltd ., Sweden, using the phosphoramidite method. Purity of the oligomer was checked from the appearance of a single band in 25 % native polyacrylarnide gel electrophoresis and a single peak in reverse phase HPLC column chromatography using proRPC™ HR 5/10 column. Concentrations of the antibiotics were determined from their known molar extinction coefficientss.9. Concentration of the oligomer was determined by phosphate estimation . Buffers used were 20 mM Tris-HCl buffer, plus 10 mM MgCb, pH 8.0 for binding studies and 5 mM Hepes buffer, plus 10 mM MgCI2, pH 8.0 for UV DNA melting experiments. Buffers were prepared in deionized, all quartz distilled water.

Binding studies Interaction of ligand with oligomer was studied by

pre-incubating the antibiotic(s) with Mg2+ for an hour at the desired temperature to ensure complete complex formation . Small aliquots of oligomer were then added to the complex and the equilibri um

66 INDIAN J BIOCI-IEM BIOPI-IYS, VOL. 38, FEBRUARY & APRIL 2001

spectrum recorded. Absorption and fluorescence spectra were recorded with Hitachi U-2000 spectrophotometer and Shimadzu RF-540 spectrofl uorometer, respecti vel y.

UV Melting studies Ultrav iolet DNA melting curves were determined

using Hitachi U-3300 spectrophotometer, equ ipped with a programmable thermoelectric temperature controller. Oligonucleotide samples, pre-incubated with 10 mM MgCb were used for the study, in 5 mM Hepes buffer, plus 10 mM MgCI2, pH 8.0. To a final concentration of 100 J..lM of the oligomer, 40 J..lM (in terms of antibiotic) of (drughMg2+ complexes were added to ensure saturation of the DNA molecule with the ligand. The samples were then further incubated for 30 min for eq uilibrium to be established. During melting experiment, the samples were heated rapidly till 26°C at the rate of 3°C per min. Further heating was carried out slowly, at the rate of 1°C per min, while continuously monitoring the absorbance at 260 nm. Reversibility of the melting profile was checked by slow cooling of the melted sample.

Analysis of binding data All spectrophotometric and spectrofluoremetric

tilrations were carried out at least lOoC below the melting temperature of lhe oligomers. Results from fluoremetric titrations were analyzed by the followin g methods. Apparent binding constant (K"p) was determined using non-linear curve fitting analysis vide equations (I) and (2). All experimental points for binding isotherms were fitted by least-square analysis.

Kd = [Co-(L'..FIL'..Flllax)-CoHCp-(L'..FIL'..FIll"x)·Cul I x[(L'..FIL'..Flllax)·Col . .. (I)

... (2)

where, L'..F is the change in fluorescence emission intensity at 540 nm (Aex = 470 nm) for each point of titration curve, L'..Fmax is the same parameter when the ligand is totally bound to oligomer, Cp is the concentration of the ol igomer and Co is the initial concentration of the antibiotic. Double reciprocal plot20 was used for determination of L'..Flllax using equation (3).

... (3)

A linear plot of lIL'..F against lI(Cp-Co) is extrapolated to the ordinate and the value of L'..Flllax obtained from the intercept. The approach is based on the assumption that emission intens ity is proportional to the concentration of the ligand. T is is found to be true for the concentration range of 10-50 J..lM of the ligands employed and under the condition Cp» Co which was followed by keepi ng at least eight-fold excess of oligomer with respect to antibiotic concentration.

The second method uses the Sc tchard equation to estimate the intrinsic binding constant (Ko) and binding stoichiometry (n).

r/Cf = Ko (n-r) ... (4)

r/Cf was plotted against r and the best-fit straight line of the experimental points was drawn. In the above equation, r = Cr/Cp, where Cb is the concentration of the bound ligand and Cp is the concentration of oligomer. Kap is given by Ko.n. The concentration of the bound ligand was determined from f1uoremetric titration using the expression, Cb = (L'..FIL'..Flllax).CO

Binding stoichiometry determined from Scatchard equation gives the value of 'n' in terms of number of drug molecules bound per nucleotide bases . Binding stoichiometry was also determined from the intersection of the two straight lines in the plot of normalized increase in fluorescence against the ratio of the input concentrations of the oligomer and I· 117 Igan( .

Analysis ojtherlllodynal11ic paraJlleters Thermodynamic parameters, L'..H (van't Hoff

enthalpy), L'..S (entropy) and f"..G (free energy) were determined using the following equations21

,

In K ap = -L'..HIRT + L'..SIR ... (5)

L'..G = L'..H-Tf"..S ... (6)

where Rand T are the universal gas constant and absolute temperature respectively . K:q was determined at 15, 20 and 25°C, respectively to evaluate M and L'..S. These values were incorporated in equation (6) to obtain the value of L'..G·van't Hoff ,'}.H values for the association were also calculated from the DNA melting curves. Enthalpy changes (Mf) for the following equilibria were considered,

CHAKRABARTI & DASGUPTA: INTERACTIONS OFCHROMOMYCIN A3 AND MITHRAMYCIN 67

2 single strands ~ duplex (drug)2Mg2++2 single strands ~

(drughMg2+ -duplex

Combining these two reactions, the binding reaction and its 6.H3 may be obtained,

(drughMg2+ + duplex ~ (drughMg2+ - duplex (6.H3 = 6.H2-6.H1)

6.H1 and 6.H2 were determined from the following equation22:

6.H = -4.38 / [( I / Tlllax ) - (l / T) ... (7)

where, Till" is the temperature at the peak and T is temperature corresponding to the half height of the peak on the higher temperature side, in the differential melting curve.

Results As reported in our earlier results ll , association of

the (drughMg2+ complexes of CHR and MTR with d(T AGCT AGCT Ah is indicated from changes such as red shift and broadening of the band in the absorption spectra of the free ligand(s) upon addition of the oligonucleotide. Presence of a single isosbestic point supports an analysis in terms of free ligand and a single type of bound species. Increase in the fluorescence quantum yield of the ligands upon association with the oligomer was employed to monitor the binding quantitatively.

Apparent binding constant values were determined using non-linear curve fitting analysis. Representative example from fluorescence titration for CHR is shown in Fig. 2a. Both binding constant and binding stoichiometry were determined from Scatchard plot as well. Representative examples of Scatchard plots are shown in Fig. 2b. Stoichiometry of binding was also calculated directly from the binding isotherms according to our earlier method 17. The binding parameters are summarized in Table I. Comparison of the binding parameters show that there is a difference in binding by the two ligands with the decameric sequence. The binding stoichiometry is significantly higher for CHR as compared to MTR. MTR covers -5 bases per drug molecule whereas CHR covers -10 bases per drug molecule. This indicates that (MTRhMg2+ complex binds to both the available sites in the decamer whereas, (CHRhMg2+ complex binds

1.00 ,.-------------------,

t:;j 0.75 E

u. <1 0.50

u. <1 0.25

a

0.00 CL..-__ '---_----''---_----' __ ---' __ ---'----'

o 1 00 200 300 400 500 [Oligomer], flM -

0.015 .---------------------,

t 0.010

U L:: 0.005

b

o

0.000 L-__ -'--__ --'--__ --1 ___ ..L...-_--.I

0.00 0.05 0.10 ·0.15 0.20 0.25 r" -

10.5,.------------------, c

10.0

to. '::t:,'" 9.5

.EO 9.0

8.5 L-___ ..L...-___ ..l--___ -'--__ --.I

3.30 3.35 3.40 3.45 3.50

1fT X 103 (K'l)

Fig. 2-Binding analysis for the association of the antibiotics wilh d(TAGCTAGCTAh in 20 mM Tris-HCI buffer. plus 10 mM MgCI2• pH 8.0: [(a): SpectroOuoremetric titration of (CHR)2Mg2+ complex (0). at 25°C. 6.F denotes the change in Ouorescence emission intensity (Aex = 470 nm. Ac01=540 nm) after each addition of oligomer w.r.t the free ligand and 6.Fmax denotes the

maximum value of the same as described under Materials and Methods. The connecting curve is generated by non-linear fitting of the experimental points. (b) : Scatchard Riot for interact ion of (CHR)2Mg2+ complex (0) and (MTRhMg-+ complex (0) with the decamer. at 25°C determined from spectroOuoremetric titration. (c) : van't Hoff plot for the interaction of (CHR)2Mg2+ complex (0) and (MTR)2Mg2+ complex (0) with the decamerl

only to a single site which makes its complex with the oligomer asymmetric in nature.

The thermodynamic parameters of association were determined using van't Hoff plot, shown in Fig. 2c. Binding enthalpy was also calculated from the differential DNA melting curves by van't HoW s method as described under Materials and Methods. The calculated parameters are tabulated in Table 2. 6.H values from the two methods are consistent within the limits of experimental error, thereby validating the application of the two methods to understand the mode of association. They are higher for the association of (MTRhMg2+ as compared to

68 INDIAN J BIOCHEM BIOPHYS, VOL. 38, FEBRUARY & APRIL 2001

Table I-Binding parameters for the association of (drug)2Mg2+ complexes with d(TAGCTAGCTA)2, in 20 mM Tris-HCI Buffer, pH 8.0, at 25°C

[The (drug)zMg2+ complexes were prepared by pre-incubating the antibiot ics with 10 mM MgCI2 for an hr, in dark, to ensure complete complex formation. Apparent binding constant Kap was determined from spectronuoremetric titrati on using non-linear curve fitting as described under Materials and Methods . The values within parenthesis denote binding constan t values (Kap) determined from Scatchard eq uation. Intrinsic binding constant (Ko) was determined using Scatchard equation . Binding stoichiometry in terms of number of nuc lcotide bases bound per drug molccule was calcuated from Scatchard plot. Valucs withi n parenthcsis werc calculatcd from thc point of intersection of the two straight lines fitted to a plot of normali zcd increase in nuorescence (t:J.Flt:J.Fmax) against the ratio of molar concentration of the oligomer to drug. t:J.F dcnotes the change in nuorescence emission intensit y (l1.<x=470 nm, A"",=540 nm) aftcr each addition of ol igomcr with respect to the free ligand and t:J.Fmax dcnotes the

maximum value of the same as described under Matcrials and Mcthods. The stoichiometry reponcd ha ve been determincd from fo ur independent sets of experimenLI

Drug K"p x 104 Ko X 105 Stoichiomctry

(M I) (M·I)

CHR 0.8 (1.4) 1.1 8.3 ± 0.5 (9.8 ± 0.2) MTR 0.9 (1.6) 0.7 4.9 ± 0.1 (6.4 ± 0.2)

Table 2-Thcrmodynamic parameters for the association of (drug)2Mg2+ complexcs wi th the oligomer d(TAGCTAGCTAh

[Thc va luc for t:J.G is stated at 25°C. The values for t:J.S and t:J.H werc detcrmined using van't Hoff plots obtained from the variation of binding constant (K"p) , as a function of temperature accordi ng to cq uation (5) as described under Materials and Mcthods. Binding constants have been calculated fro m non-linear curve filling of spectronuoremetric titration results , as described undcr Matcrials and Mcthods. The t:J.H values have been obtained from oligomer mclting profile as described under Materials and Mcthods·1

Druo ~ t:J.G t:J. H t:J.H t:J.S

used (kcallmol) (kcallmol ) (kcallmol) (c.u)

CHR -5.3 - 8.2 - 12.1 -9.7

MTR - .4 - 11.9 -15 .0 - 2 1.8

(CH Rh Mg2+ complex. This is accompanied by a

higher negative value of ~S for MTR as compared to CHR. Higher val ues of ~H and ~S in case of MTR originate from the association of two ligand molecules with t e oligomer.

Additional support regarding the differential interaction of the anti biotics is obtained from an analysis of the melting curves of the ol igomer in absence and presence of saturati ng concentration of the ligands (Fig. 3a). In order to determine the melting curves for the ligand bound oligomers, the amount of

ligand added was so adjusted that the oligomer is completely saturated with the ligand, and there exists only a single species of li gand bound oligomer in the sample. The oligomer was pre-incubated with 10 mM MgCI2 to maintain the same condition in which the binding studies were carried out. Tm values were determined from the corresponding differential melting profiles (Fig. 3b). Since the oligomer sequence is palindromic, its melting proceeds symmetrically, characterized by a sing le sharp melting temperature of 45°C. Upon association with the (MTRhMg2+ complex, the duplex is stabilized further and its T m shifts to 56°C. However, during melting of the oligomer saturated wi th (CHR)2Mg2+ complex the single peak in the differential curve spl its up into two distinct peaks separated by 12°C. Lower peak is at 37°C which is less than the Tm value fo r the free oligomer and the second peak is at 49°C (Fig. 3). Melting of a single DNA molecule with two distinct Tm values may be caused by sign ificant difference in the duplex stability at two different regions in the sequence. Presence of distinct AT and GC rich sequences in different regions of a D. T A molecule can also give rise to more than a single peak in the differential melting curve23. However, in this case with the symmetrical sequence, preferential stabil ization of one part of the duplex originati ng from asymmetric association of CHR ligand gives rise to the double peak.

Discussion We have reported here the in teraction of the

(drughMg2+ complexes of CHR and MTR with the decameric sequence d(T AGCT AGeT A)2. Detailed study on the interaction illustrates the fact th at the two ligands recognize the same sequence differently. This was concluded from the dual observations: (i) binding stoichiometry between the ligand and oligomeric duplex is 2: I for MTR and 1: I for CHR as a result of which (ii ) the two ligands affect the melting of the oligomer different iall y.

A detailed NMR study on the interaction of (MTRhMg2+ complex with the ol igomer has already been reported 18. The overall binding stoichiometry determined in the earlier study for MTR, is 2: I in terms of ligand:oligomeric duplex which is consistent with the present report. It shows that DNA molecule gets ki nked at the intermediate TpA base step in order to facil itate the approach of the two bulky (MTRhMg2+ ligands. The DNA is also distorted locally at the two individual GpC sites where a B~A

CHAKRABARTI & DASGUPT A: INTERACTIONS OF CHROMOMYCIN A3 AND MITHRAMYCIN 69

1.1 1.0

0 0.8 <D N

<l:

"() OJ 0.5

.!::! "6 E L.. 0.2 0 z

0·0 -0.1

295

0.004

0 0.003 -~

r 0.002 "0

-.. .--.

0 <D (\j

~ 0.001 "0

0.000 295

/ /:

~/ /

315

315

./

(a)

335

(b)

335

355

0.008

0.006 --. it

0.004 ---

0 .002

0.000 355

Temperature (K)

Fig. 3-Ultraviolet DNA melting curves for d(TAGCTAGCTA)] in 5 mM Hepes Buffer, plus 10 mM MgCI], pH 8.0 under differcnt conditions: [(a): Normalized valuc of absorbance at 260 nm (AlbO) plOllCd as a fu nction of tempcrature for the following systems: Free

oligomer (100 ~M, - . -); Oligomer saturatcd with (CHR)]Mg2+ complex ([oligomcr]/[CHRI = 2.5, --); Oligomer saturated with (MTRh Mg2+ complex ([oligomer] / IMTR] = 2.5. - - -). (b): Deri va ti ve of absorbance (A2bO) w.rt temperature plolled as a function of tcmperature for the following systems: Frce oligomcr ( I 00 ~M, 1'. ); Oligomer saturated wi th (CHRh Mg2+ complex ([oligomer]/[CHR] = 2.5, 0 ); Oligomer saturated with (MTRhMg2T complex (Ioligomcrl/IMTR I = 2.5, *).

type transition in backbone geometry helps to accommodate the ligand . The two (MTRhMg2+ ligands al ter their conformations in order to fi t into the groove without steric clash between them. Changes in the glycosidic linkage bonds in the C-D-E trisaccharide, in one of the monomers of the dimer­Mg2+ ligand, marks this change.

Com pari on of the bi nding stoichiometries for the associatIOn of (CHRhMg2+ and (MTRhMg2+ complexes with the decamer shows that CHR bi nds to one of the sites in the oligomer and the other site remains unoccupied. In contrast, MTR bi nds to both the sit~s. The (CHRhMg2+-0Iigomer complex is thus asymmetric as compared to the (MTRhMg2+­oligomer complex. The thermodynamic parameters for the association(s) also support a difference in the mode of interaction of the two drugs. As in our earlier reports we have analyzed the observed changes in the thermodynamic parameters as a sum of several contributing terms'7 in the following manner:

where, !1Hbind is the bi nding enthalpy. !1HDNA and !1Hdil1lcr are the enthalpy changes associated with structural al terations in the DNA molecule and the ligand molecule respectively, during their interaction. The !1H values how that binding by MTR involve ' a higher enthalpy change as compared to CHR (Table 2). The value fo r MTR is not double of that fo r CHR, as is expected if only binding enthalpies con tri bute to the total observed !1H. Since association of two (MTRhMg2+ ligands to the decamer involves considerable bending of the DNA molecule around the kink at the TpA base step alongwith local B ~ A transitions at the two binding sites, there is a major contribution from the !1HDNA term. This considerably reduces the negative value of observed !1HIOI for MTR. A higher negative value of !1S for MTR also suggests formation of a more structurally constrained

70 INOIAN J BIOCHEM BIOPHYS, VOL. 38, FEBRUARY & APRIL 2001

Fig. 4-Schematic representation of asymmetric association of (CHRhMg2

+ complex with oligomer.

ligand-DNA complex as compared to CHR. This is consistent with the NMR model of the (MTRhMg2+­oligomer complex (PDB 10 207d I8). !1G values remain comparable for both types of ligands due to enthalpy-entropy compensation.

The observed difference in the binding geometry originates from the nature of the sugars, particularly the substituents present in the Band E sugars. The methoxy group present in the B sugar and acetoxy group in the Band E sugars of CHR impose constraints upon the free rotations about the glycosidic linkages in the disaccharide and trisaccharide chains in the ligand molecu le. This affects the flexibility of (CHRhMg2+ complex which behaves as a rigid molecule. Thus, energetically permissible structural alterations in the ligand molecule like alterations in the glycosidic torsion angles of the C-D-E trisaccharide segment, defined as 'plasticity' of MTR, is absent in case of CHRI8. Earlier reports by Keniry et al., also implies higher flexibility in the (MTRhMg2+ complex during its association with an octameric sequence as compared to CHRI6.

Differences in the nature of the ligand-DNA complexes formed by the two drugs is reflected in the melting profiles of the ligand-oligomer complexes. Asymmetric association of (CHR)2Mg2+ gives rise to two distinct peaks in the melting profile of the decamer. The GpC site, where the (CHRhMg2+ ligand binds, is stabilized. Due to the rigidity of the ligand, considerable extent of DNA distortion is necessary for associat ion at and around the binding site. This leads to partial unwinding and significant destabilization of the second GpC binding site once the first site is occupied. This is indicated from its melting at a lower Tm as compared to the free duplex (Fig. 3). On the other hand, the melting profile for the (MTRhMg2+ bound decamer is characterized by a single peak, because the ligand binds symmetrically (Table 3).

Asymmetric assoclatton of (CHRhMg2+ is schematically shown in Fig. 4.

Differences in extent of DNA distortion induced by MTR and CHR has also been reported earlier for their interaction with polymeric DNA I7. The present report sllcceeds in elucidating the molecular basis of the differential behaviour of the two ligands.

Acknowledgement SC wishes to thank Professor (Dr) B Sinha, Director

of SINP for granting permission to work in the Biophysics Division.

References I Rohr J, Mendez C & Salas J A (1999) Bioorganic Chemistry

27,41-54 2 Gause G F (1975) Olivomycin, Chromomycin and

Mithramycin: Antibiotics !II (Corocan J W & Hahn FE, Eds) pp 197-202, Springer Verlag, Berlin & New York.

3 Chabner B A, Allegra CJ, Curt GA & Calabresi P (1996) Goodman & Gillilan's The Pharmacological Basis of Therapeutics, 9th ed. (Hardman JG, Limbird L E, Molinorr P B, Ruddon R W & Gilman A G, Eds) pp 1233-1 287, McGraw-Hill, New York.

4 Goldberg I H & Friedman P A (1 97 t ) Anl7l1 Rev Biochem 40 , 775-810

5 Waring M J (1981) Annll Rev Biochem 50, 159-192 6 van Oyke M W & Oervan P B (1 983). BiochemistlY 22 ,

2373-2377 7 Stankus A, Goodi sman J & Oabrowiak J C (1992)

Biochemistry 31, 9310-9318 8 Aich P & Oasgupta 0 (1990) Biochem. Biophys Res

COmm11l1 173, 689-696 9 Aich P, Sen R & Oasgupta 0 (1 992) Biochemislly 3 1, 2988-

2997 10 Aich P, Sen R & Oasgupta 0 (1 992) Chem-BiollllIeracl 83 ,

23-33 II Aich P & Oasgupta 0 (1995) Biochemistry 34, 1376- 1385 12 Gao X & Patel 0 J (1989) Biochemisll y 28, 751-762 13 Illarivova R P, Oykhovichnaya 0 E & Bondarenko B N

(1970), AntibiOliki (Moscow) 15, 415-4 i 8 14 Fox K R & Howarth N R ( ]985) Nucleic Acids Res 13, 8695-

8714 15 Sastry M & Patel 0 J (1 993) Biochemistry 32, 6588-6604 16 Keniry M A, Banville 0 L, Simmonds P M & Shafer R H

( 1993) J Mol Bioi 231 , 753-767 17 Majee S, Sen R, Guha S, Bhattacharyya 0 & Oasgupta 0

(1997) Biochemistry 36, 2291-2299 18 Sastry M, Fiala R & Patel 0 J ( 1995) J Mol Bioi 251 , 674-

689 19 Liu R & Chen F (1994) Biochemistry, 33, 1419-1424 20 Wang J L & Edelman G H (1971 ) J Bioi Chem 246, 1185-

1191. 21 Castell an G W (1989) Physical Chemistry (3rd ed.) pp 221-

249, Addi son Wesley/Narosa Publishing House (I ndian student edition), New Oelhi

22 Marky L A & Breslauer K J (1 987) Biopolymers 26, 160 1-1620

23 Leng F, Priebe W & Chaires J (1 998) Biochemistry 37, 1743-1753