molecular modeling study of the norfloxacin-dna complex

Molecular Modeling Study of the Norfloxacin-DNAComplex

http://www.jbsdonline.com

Abstract

Molecular modeling and molecular dynamics were performed to investigate the interactionof norfloxacin with the DNA oligonucleotide 5’-d(ATACGTAT)2. Eight quinolone-DNAbinding structures were built by molecular modeling on the basis of experimental results. A100ps molecular dynamics calculation was carried out on two groove binding models andsix partially intercalating models. The resulting average structures were compared with eachother and to free DNA structure as a reference. The favorable binding mode of norfloxacinto a DNA substrate was pursued by structural assess including steric hindrance, presence ofhydrogen-bonding, non-bonding energies of the complex and presence of abnormal struc-tural distortion. Although two of the intercalative models showed the highest binding ener-gy and the lowest non-bonding interaction energy, they presented structural features whichcontrast with experimental results. On the other hand, one groove binding model demon-strated the most acceptable structure when the experimental observation was accounted. Inthis model, hydrogen bonding of the carbonyl and carboxyl group of the norfloxacin ringswith the DNA bases was present, and norfloxacin binds to the amine group of the guaninebase which protrudes toward the minor groove of B-DNA.

Introduction

Quinolones are a group of low-molecular weight, synthetic, extremely potent anti-bacterial agents (1). The functional target of these drugs is the enzyme DNAgyrase, an essential type II DNA topoisomerase that exists only in bacteria. Somemembers of this class which are currently in clinical use are norfloxacin (2),ofloxacin (3). Norfloxacin (Fig.1), one of the most potent DNA gyrase inhibitorsof the quinolone family, does not directly bind to DNA gyrase but binds to DNAitself (4). Interaction between quinolones including norfloxacin, and supercoiledor relaxed DNA, double stranded and single stranded-DNA, as well as various syn-thetic DNA have been investigated by several spectroscopic methods, biochemicaltechniques and computational methods (4-19). As the results, few binding modesof norfloxacin with various DNA were proposed. Shen and his coworkers pro-posed a cooperative quinolone-DNA binding model for the inhibition of DNAgyrase (7, 8). In the presence of gyrase and non-hydrolyzable ATP analogue,gyrase induces a specific single stranded DNA pocket. Four norfloxacins assem-bled inside the pocket through hydrogen bonds between the carbonyl group of thequinolone rings and the DNA bases of the separated DNA strands. Norfloxacinsare also stabilized by hydrogen bonding, π-π stacking, and tail-to-tail hydrophobicinteractions between assembled norfloxacin molecules. In contrast, Palumbo andhis coworker suggested that the drug was stacked between nucleotides in the pres-ence of Mg2+, preferably in a region of single-stranded DNA (11). Stabilization ofthis interaction was assured by the Mg2+ bridge between the carboxylic and car-bonyl group of quinolone and a phosphate group of the DNA backbone. A theo-retical study using 3-D quantitative structure-activity relationship (QSAR) and

Journal of Biomolecular Structure &Dynamics, ISSN 0739-1102Volume 19, Issue Number 6, (2002)©Adenine Press (2002)

Hyun Mee Lee1

Jong-Ki Kim2

Seog K. Kim1*

1Department of Chemistry

College of Sciences

Yeungnam University

Dae-dong, Kyoungsan City, Kyoung-buk

712-749 Republic of Korea2Department of Biomedical Engineering

School of Medicine

Catholic University of Daegu 3056-6

Daemyung 4dong, Taegu City

705-034 Republic of Korea

1083

Phone: 82 53 810 2362Fax: 82 53 815 5412Email: [email protected]

molecular recognition led Cedergren and coworkers to propose an intercalationmodel for quinolone to the duplex DNA (13), in which the drug and Mg2+ locatedin the major groove. In this model, the structure is stabilized by formation of acomplex between the Mg2+ ion between drug and DNA: The moieties that partici-pate in the complex are the carbonyl and carboxylic oxygen of the drug, a phos-phate group and a purine base of the DNA.

Recently, extensive spectroscopic studies, including circular and linear dichroismspectroscopy on the binding mode of norfloxacin and ofloxacin to various DNAhave been apparent (15-19). The main observations include; (a) an obvious bind-ing of norfloxacin to double-stranded DNA without any mediation from ATP andMg2+, (b) the angle of 65º-80º between the molecular plane of norfloxacin and theDNA helix axis in the complex, (c) a negligible amount of unwinding of supercoiledDNA by norfloxacin which contradicts the classic intercalation model, and (d) inter-action of the norfloxacin molecule with the amine group of the guanine base whichis protruding into the minor groove of DNA. From those observations, it was sug-gested that norfloxacin binds in the minor groove of DNA at the guanine site withthe possibility of partial insertion of aromatic moiety of drug between the base pairs.

Despite the pharmacological interest of quinolone, direct experimental evidence forthe three-dimensional structure of the drug-DNA complex has not been apparent,probably due to both the low association constant of quinolone family to DNA andthe various protonation state of quinolone in aqueous solution, resulting in a het-erogeneous nature of the complex. Indeed, to our knowledge, even the conforma-tion of the isolated drug, either in solution or in the crystalline state has not yet beenreported. Therefore, in this study, we investigated the favorable structure for thenorfloxacin-DNA complex utilizing the molecular modeling and molecular dynam-ics (referred to as MD in this work) method by building up structural features basedon the recent spectroscopic studies in the absence of Mg2+. From the several pos-sible structures, the most probable one, which was chosen by its energy and byknown experimental results, was suggested in this work.

Methods

Structure of the isolated drug and free DNA

The geometry optimization of the drugs was obtained using the Gaussian98 pack-age (20). The geometries of norfloxacin were determined using a 3-21G* basis setat the Hartee-Fork (HF) level by analytical gradient techniques. The drug-DNAsystem was modeled from duplex DNA octamer, d[ATACGTAT]2, containing twoGC base pairs. Double helical octamer structure was built with the biopolymermodules of the SYBYL molecular modeling program (21). The docking of thedrug was subsequently carried out to this DNA.

Modeling of the drug-DNA complex

The starting structures were built by considering some experimental observationson the norfloxacin-DNA complexes, which was summarized in the introduction

1084

Lee et al.

Figure 1: Chemical structure (a) and optimized struc-ture (b) of norfloxacin. The ethyl group of the opti-mized structure was perpendicular to the plane of aro-matic rings and the piperazine ring tilted nearly 45ºwith respect to the molecular plane.

(15-18). Based on the experimental evidence, the norfloxacin molecule conceiv-ably sits in the minor groove and partially intercalated from there. Therefore, thedrug was initially positioned at two main regions, namely, the partially intercalatedposition and the minor groove near the GC base pairs. In the partially intercalatedposition, the orientation of the ethyl group could be either up (toward 5’A) or downwith respect to the molecular plane. The norfloxacin orientation within the plane ofintercalation can be conveniently described by defining an angle ψ swept out fromthe phosphorous-phosphorous axis at the intercalation site, to the C3-C7 axis of thequinolone when viewed down the helical axis (Fig. 2). Initial ψ angles were 0º, 45ºand 90º for each of the ethyl groups in both the up and down positions. At thesethree initial orientations, norfloxacin was inserted between the base pairs with sometilt of the drug’s molecular plane with respect to the DNA base plane, which makesthe ethyl upper or lower initial conformation different. Therefore, the six initialintercalative positions were considered in this work. The second class of startingposition was norfloxacin situated in the minor groove where the carbonyl groupfaced the GC base pairs of DNA with the piperazine group toward either of the cyto-sine bases. Hence the eight starting structures in all, including six intercalativepositions and two minor groove situation binding modes, were considered (Table I).

Molecular dynamics calculation

In all cases, the drug, DNA as well as the drug-DNA complexes were embedded ina cubic box containing 218~359 TIP3P water molecules (22). In order to preventfraying of the terminal base pairs that might occur during the MD process, weakharmonic restraints (force constant of 10kcal/mol) were applied to each Watson-Crick hydrogen bond of the terminal two base pairs at the both ends. The SHAKE

1085Molecular Modeling Study of

the Norfloxacin-DNAComplex

Table IDocking direction of norfloxacin and nomenclature (in this work) in the

model structures.

ψ (0o) ψ (45

o) ψ (90

o) front

(c)

intercalation:upper(a)

Iu1 Iu2 Iu3

down(b)

Id1 Id2 Id3

groove binding:upper(a)

gu

down(b)

gd

The eight starting structures have six intercalative positions depending

on the direction of insertion, the ethyl group conformation and the two

minor groove situations with opposing directions of carbonyl groups.

(see text for explanation)(a)

ethyl group of norfloxacin is in the upper position.(b)

ethyl group of norfloxacin is in the lower position.(c)

carbonyl group directed towards the guanine base in the minor groove.

Figure 2: The orientation of norfloxacin within theplane of intercalation can be conveniently described bydefining an angle ψ swept out from the phosphorous-phosphorous axis at the intercalation site, to the C3-C7axis of quinolone.

algorithm to maintain all the bonds at equilibrium lengths was employed while noother constraints were applied during the simulations (23).

All energy minimization and MD simulations were carried out by the SYBYL pro-gram using the Tripos Force Field on a SUN Sparc workstation. For each startingstructure, the minimization using Powell (24) were carried out until the energy gra-dient reached less than 0.5 kcal/mol. Subsequent MD calculations were performedon those minimized structures under NPT conditions in the two stage: the first 5pssimulation being at 1atm and 400K to overcome minimal local energy and the sec-ond productive simulation for 95ps at 1atm and 300K. The total simulation timewas therefore 100ps using a time step of 1fs, collecting for analysis every 0.1ps.For general analysis purposes, trajectories were saved every 100fs. Final structureswere obtained by averaging trajectories for the last 20ps of MD the simulation, andrefined by additional 100 step energy minimization of the averaged structure usingPowell algorithm where distance dependent constant (ε = 4⋅r) was used withoutexplicit water (25). Structural analysis of the DNA helices was carried out by theCURVES program (26).

Results and Discussion

Structural analysis based on energy calculation

Prior to performing the MD simulation, the norfloxacin structure was optimized bythe ab initio quantum chemical method and was confirmed to be the true energyminimum by observing the absence of any imaginary. In the optimized structureas shown in Figure 1b, the ethyl group was perpendicular to the plane of the aro-matic rings. The piperazine ring tilted nearly 45º with respect to the molecularplane. The plane containing the carbonyl oxygen was parallel to the molecularplane. In reference to the binding mode recently reported by extensive spectro-scopic studies and using the optimized norfloxacin structure, eight docking direc-tions (Table I) were chosen as the starting structures and energy minimization wascarried out on each starting structure prior to the MD calculation. The startingstructures before energy minimization are depicted in Figure 3.

Energy and DNA bending of eight DNA-drug complexes were evaluated and sum-marized in Table II. All the energy for the drug, free DNA and the drug-DNA com-plex were averaged over the last 20ps of the MD. The total energies for the drug-DNA complex were obtained by the Gasteiger-Hücker method excluding the sol-vent from the simulation box. In the same table the binding energy (interactionenergy) was defined by the subtraction of the sum of free DNA and norfloxacinenergy from that of the norfloxacin-DNA complex. Non-bonding interaction ener-gy (van der Waals and electrostatic energy) was calculated for the central two GCbase pairs at the drug-binding site.

1086

Lee et al.

Figure 3: Representative starting structures of the drug-DNA complex for molecular dynamics simulation. (a)iu1 model (b) iu2 model (c) id3 model (d) gu model

As shown in Table II, the non-bonding energies of the intercalation models weregenerally lower compared to those of the groove binding models. This observationmay be understood by the partial stacking of the drug and the DNA bases, whichincreased the van der Waals interaction and electrostatic interaction. It was alsonoticed that the energies including the total energy, binding energy and non-bond-ing interaction energy were relatively variable depending on the direction of thedrug insertion and/or ethyl group in the intercalation models while those energiesin two groove binding models were unaffected. The binding energy of the finalstructure from the starting model iu2 was the lowest and the non-bonding interac-tion energy of the model iu1 was the highest. In the iu2 model (in which the ethylgroup directs upward), the molecular plane of norfloxacin was partially insertedbetween the GC base pair when the angle ψ was 45º (Fig. 3). As it is shown inTable II, the non-bonding energy for the central two bases was lowest for the iu1model. In the final structures which are displayed in Figures 4a and 4b respective-ly, the molecular plane of norfloxacin was inserted between the GC base pair. Inthe intercalation pocket, the carbonyl and carboxylic oxygen nearly poke out to themajor groove. The piperazine ring was not near the intercalation pocket but situ-ated near the phosphate group, and is either rotated by 45º (iu2, Fig. 4b) or remainsparallel (iu1, Fig. 4a) relative to the norfloxacin molecular plane. From the energycalculation, these two final structures were concluded to be the most probable ones.However, these structures did not agree with experimental observation (16-19). Inthese structures, the molecular plane of norfloxacin was perpendicular to the DNAhelix axis, which has never been observed by experiment (16-19). This discrepan-cy suggests that the resulting energy depends highly on the starting structures. Andtherefore, it is difficult to find the most probable structure based on energy calcu-lation alone. Furthermore, in the energy calculation the AT base pair is totally omit-ted assuming that the binding of norfloxacin would not affect adjacent AT basepairs which may not be true.



Table IIEnergy (kcal/mol) and angle of the axis bending (degree) of the norfloxacin-octamer

complex after MD calculation.

gd gu id1 id2 id3 iu1 iu2 iu3

total energy(a)

153.5 153.9 152.2 166.9 161.9 150.8 149.3 152.9

binding energy(b)

29.6 29.2 30.9 16.2 21.2 32.8 33.8 30.2

nonbonded energy(c)

35.4 36.8 23.7 19.6 23.4 16.0 19.0 19.1

axis bend(d)

16.3 16.6 9.8 4.7 6.3 14.1 15.2 9.9

(a)energy from Gasteiger-Hcker method.

(b)interaction energy indicates the energy obtained by the subtraction of the sum of energy of

free DNA and norfloxacin from that of the norfloxacin-DNA complex. The energy of free

DNA is 170.4Kcal and that of the isolated drug is 12.6Kcal.(c)

non-bonding energy is the sum of van der Waals and electrostatic energy in the two

central base pairs(G-C) in the presence of norfloxacin.(d)

The angle of the axis bending obtained from the structure analysis for the drug-DNA

complexes.

Figure 4: The resulting structures from the (a) iu1 model(b) iu2 starting model that exhibits the most stable energy.

Backbone conformation and structural features at binding sites of intercalativemodels

The conformational angles in the 100ps MD simulation for the final structure offree DNA are given in Table III. In the Table, the backbone torsion angles for onlyfour central base pairs which may participate in the drug-DNA interaction, areshown. The backbone torsion angles of octanucleotide were in agreement withthose of energy-refined B-DNA (27). Calculated structures from the MD simula-tion were assessed based on the experimental data and the backbone conformationangle. Backbone analysis allows us to exclude unusual features from the final pos-sible structures of the norfloxacin-DNA complex. The abnormality was judged by;(1) whether the backbone of the center base pair was out of range for normal B-DNA form, (2) whether hydrogen bonding occurred between the sugar and basepair according to structural distortion, and (3) whether the planarity of the base pairremained. When norfloxacin docked into DNA from the intercalation model, thefinal structure exhibited base twisting. These results suggest hydrogen bondingbetween either the amino group of guanine and sugar moiety, or between internalbases, and norfloxacin binding to the amino group of the cytosine base. The finalstructure from the id3 model did not result in any hydrogen bonding between thedrug and DNA. The torsion angles of the calculated structure from the id3 startingmodel are shown in Table IV. The torsion angles α and ξ, in the backbone near thenorfloxacin insertion point were significantly altered, while other torsion anglesremained to be nearly the same compared to those of drug-free B-DNA. The tor-sion angles α, β, γ, ε, ξ calculated in this work are in the gauche-, trans, gauche+,trans, gauche- conformation, respectively. The overall structure of the backbone ofthe calculated structure from the id3 model (Fig. 5) differed little from those of nor-mal B-DNA in spite of the deviation in the α and ξ angles. The torsion angles fromthe id3 model were different from those of the averaged value obtained from thecrystal structures of other classic intercalator-DNA complexes, which were quotedfrom the work done by Shieh and co-workers (Table IV) (28). Although our cal-culated torsion angles cannot be directly compared with those from the crystal data,the difference is large enough to discard our confined model. Furthermore, a sig-nificant twist in base pair was noticed in the complex near the norfloxacin bindingsite, supporting our conclusion.

Backbone conformation and structural features at the binding site of groove bindingmodels

The gu model structures from the minor groove binding model seem to be, in gen-

1088

Lee et al.

Table IIITorsion angles of the drug-free DNA obtained from form MD calculation.

α β γ δ ε ζ χ

Strand 1

A -64.7 167.9 61.9 121.1 167.7 -90.8 -118.9

C -57.1 170.1 57.8 90.2 171.1 -81.7 -154.6

G -65.4 171.0 56.0 123.9 171.3 -87.3 -128.3

T -62.7 173.7 55.5 105.6 174.3 -79.8 -140.4

Strand 2

T -55.8 164.6 50.5 76.0 162.4 -88.9 -145.6

G -80.0 175.8 63.6 145.9 173.9 -83.6 -123.4

C -62.5 168.9 57.3 121.4 179.1 -128.0 -136.8

A -66.1 172.5 56.1 127.5 172.3 -86.4 -126.0

DNA

BDNAa)

-61 180 57 122 173 -91 -119

(a)Torsion angles from energy refined B-DNA from Ref (27).

eral, more favorable compared to those from the intercalative model in terms of thebackbone structure. The torsion angles of the groove binding model obtained fromour work are summarized in Table V. The torsion angles of the gu model structurewere, in general, similar to those of the drug-free DNA, except for little deviationat the δ angle. When a drug is intercalated between DNA bases, elongation,unwinding and stiffen of the DNA stem is expected which results in an increase inLDr value. However, the reported experimental LDr values decreased upon nor-floxacin binding, indicating the bending of DNA (15-18). The DNA bending analy-sis performed in this work showed the larger bending extent for the groove binding



Table IVTorsion angles for the norfloxacin-DNA complex obtained from the inter-

calative starting models.


Strand 1

A -67.0 166.2 54.0 122.5 174.6 -86.2 -118.3

C -60.6 174.3 55.4 83.6 179.8 -81.0 -134.7

G -62.8 178.1 64.9 143.6 170.6 -98.8 -116.4

T -52.4 172.4 54.7 110.8 173.4 -97.5 -140.2

Strand 2

T -70.7 176.5 56.7 137.3 174.0 -91.7 -129.9

G -43.0 172.3 65.5 147.4 170.3 -126.7 -116.6

C -73.1 176.2 59.1 81.2 189.0 -73.1 -144.5

A -63.1 171.6 51.0 128.7 165.8 -94.0 -105.7

id3

Exp.28)

-70 -135 59 -149 -67

Table VTorsion angles for the most favorable norfloxacin-DNA complex obtained from the

groove binding starting model.


Strand 1

A -70.5 175.4 62.5 141.3 171.5 -97.3 -118.6

C -61.1 169.8 58.2 111.4 169.2 -88.8 -139.0

G -62.4 173.7 60.1 111.1 171.1 -88.1 -135.4

T -64.7 169.1 54.5 99.0 176.7 -80.1 -152.6

Strand 2

T -57.1 167.4 61.4 111.7 171.0 -79.7 -137.8

G -62.8 168.7 55.8 85.1 171.5 -80.3 -143.6

C -69.8 168.2 59.4 98.6 176.8 -78.8 -148.5

gu

A -70.1 171.3 67.0 149.7 169.4 -94.1 -105.4

Figure 5: The front (a) and side view (b) of the modelid3 after molecular dynamics simulation.

mode compared to the intercalation mode. The bending is particularly pronouncedfor the gu model structure. The bending angle of the gu structure is 16.6º in com-parison with 6.3º for the id3 model. In the stereo view of the norfloxacin-DNAcomplex (Fig. 6), it could be seen that the angle of the molecular plane of nor-floxacin is 65 - 80º which matches with experiment results (15-18). As it is shownin Figure 6b, the hydrogen bond between the carbonyl as well as carboxylic groupof norfloxacin and amino group of guanine can be seen and was probably respon-sible for the preference of this drug to guanine (18). Another structure, which wasobtained from the gd model structure, seems to be unrealistic because the centerguanines form inter-base hydrogen bonds with adjacent adenines. Furthermore, theplane of the cytosine base deviated too much from planality.

Conclusion

Using molecular modeling methods we concluded that the norfloxacin-DNA com-plex exhibits a non-classical groove binding mode. The molecular plane of the nor-floxacin are at angles of 65 – 70º relative to the DNA helix axis. In the complex,the hydrogen bond between the carbonyl and carboxylic group of the norfloxacinrings and the amine group of guanine base is very conceivable. The formation ofa hydrogen bond may be the reason for the preferentiality of norfloxacin (and ingeneral, quinolone) towards the guanine base.

Acknowledgment

Authors acknowledge the financial support from Korea Research Foundation(Grant no. KRF99-005-D00043).

References and Footnotes

1090

Lee et al.

Figure 6: The stereoview (a) and side view (b) of gumodel after the MD simulation. In figure (b), the possi-ble hydrogen bonds between guanine-NH2 and oxygenof the carbonyl group and carboxyl group of norfloxacinare explicitly shown by arrows.

1.

2.

3.

For a review, Hooper, D. C., and Wolfson, J. S. Quinolone-Antimicrobial Agent, 2nd ed.;American Society for Microbiologists, Washington, DC (1993).H. Koga, A. Itoh, S. Murayama, S. Suzue and T. Irikura, J. Med. Chem. 23, 1358-1363(1980).L. A. Mitscher, P. N. Sharma, D. T. W. Chu, L. L. Shen and A. G. Pernet, J. Med. Chem. 30,2283-2286 (1987).



4.5.

6.

7.

8.9.10.

11.

12.

13.14.15.

16.

17.

18.

19.

20.

21.22.

23.24.25.26.

27.28.

L. L. Shen, and A. G. Pernet, Proc. Natl. Acad. Sci. USA 82, 307-311 (1985).G. Palù, S. Valisena, M. Perachi and M. Palumbo, Biochem. Pharmacol. 37, 1887-1888(1988).L. L. Shen, W. E. Kohlbrenner, D. Weigi and J. Baranowski, J. Biol. Chem. 264, 2973-2978(1989).L. L. Shen, L. A. Mitscher, P. N. Sharma, T. J. O’Donnel, D. W. T. Chu, C. SS. Cooper, T.Rosen and A. G. Pernet, Biochemistry 28, 3886-3894 (1989).L. L. Shen, J. Baranowski and A. G. Pernet, Biochemistry, 28, 3879-3894 (1989).L. L. Shen, Biochem Pharmarcol. 38, 2042-2044 (1989).J. Y. Fan, D. Sun, H. T. Yu, S. M. Kerwin and L. H. Hurley, J. Med. Chem. 38, 408-424(1995).G. Palù, S. Valisena, M. Peracchi and M. Palumbo, Proc. Natl. Acad. Sci. USA89, 9671-9675(1992).C. Sissi, M. Andreolli, V. Cecchetti, A. Fravolini, B. Gatto and M. Palumbo, Bioorg. & Med.Chem. 6, 1555-1561 (1998).B. Llorente, F. Leclerc and R. Cedergren, Bioorg. & Med. Chem. 4, 61-71 (1996).C. Bailly, P. Colson and C. Houssier, Biochem. Biophys. Res. Commun. 243, 844-848 (1998).G.-W. Son, J.–A. Yeo, M.-S. Kim, S. K. Kim, A. Holmén, B. Åkerman and B. Nordén, J.Am. Chem. Soc. 120, 6451-6457 (1998).J.-A. Yeo, T.-S. Cho, S.K. Kim, H.R. Moon, G.-J. jhon and W. Nam, Bull. Kor. Chem. Soc.19, 449-457 (1998).G.-W. Son, J.–A. Yeo, J.–M. Kim, S. K. Kim, H. R. Moon and W. Nam, Biophys. Chem. 74,225-236 (1998).E.-J. Lee, J.-A. Yeo, G.-J. Lee, S. W, Han and S. K. Kim, Eur. J. Biochem., 267, 6018-6024(2000).E.-J. Lee, J.-A. Yeo, K. Jung, H. J. Hwangbo, G.-J. Lee, S. K. Kim, Arch. Biochem. Biophys.,395, 21-24. (2001).M. J. Frich, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, V.G. Zakzewski, Jr. J. A. Montgomery, R. E. Stratman, J. C. Burant, S. Dapprich, J. M. Millam,A. D. Daniels, K. N. Kudin, M. C. Strain, O. Farkas, J. Tomasi, V. Barone, M. Cossi, R.Cammi, B. Mennucci, C. Pomelli, C. Adamo, S. Clifford, J. Ochterski, G. A. Petersson, P. Y.Ayala, Q. Cui, K. Morokuma, D. K. Malick, A. D. Rabuck, K. Raghavachari, J. B. Foresman,J. Cioslowski, J. V. Ortiz, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R.Gomperts, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng, A. Nanayakkara,C. Gonzalez, M. Challacombe, P. M. W. Gill, B. Johnson, W. Chen. M. W. Wong, J. L.Andres, M. Head-Gordon, E. S. Replogle, J. A. Pople, Gaussian 98, Revision A. 3., Gaussian,Inc., Pittsburgh, PA, 1998.SYBYL 6.2, Molecular Modeling Program; Tripos Assoc. Inc.: St Louis.W. L. Jorgensen, J. Chandrasekhar, J. D. Madura, R. Impey and M. Klein, J. Chem. Phys.79, 926-935 (1983).J. P. Ryckaert, G. Cicotti and H. J. C. Berendsen, J. Comput. Phys. 23, 327-341 (1977).M. Clark, R. D. Cramer III and N. Van Opdenbosch, J. Comp. Chem. 10, 982-1012 (1989).M. Orozco, C. A. Laughton, P. Herzyk and S. Neidle, J. Biomol. Str. Dyn. 8, 359-373 (1990).R. Laver and H. Sklenar, CURVES 5.3, Laboratorie de Biochimie Theorique CNRS, Institude Biologie Physico-Chimique, Paris, France. (1990).M. Levitt, Proc. Nat. Acad. Sci. USA 75, 640-644 (1978).H. S. Shieh, H. M. Berman, M. Dabrow, and S. Neidle, Nucl. Acids Res. 8, 85-97 (1980).

Date Received: December 14, 2001

Communicated by the Editor Ramaswamy H Sarma

molecular modeling study of the norfloxacin-dna complex

Documents