This article was downloaded by: [York University Libraries] On: 05 January 2015, At: 10:31 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
Journal of Biomolecular Structure and Dynamics Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/tbsd20
The Binding of Prototype Lexitropsins to the Minor Groove of DNA: Quantum Chemical Studies a
a
c
Paul Mazurek , Waldo Feng , Kenhai Shukla , a
Anne-Marie Sapse & J. William Lown
b
a
Department of Biophysics and Physiology , Mount Sinai School of Medicine , 1 Gustav Levy Place, New York , N.Y. b
Department of Chemistry , University of Alberta Edmonton , Alberta , Canada , T6G 2G2 c
New Jersey Institute of Technology , New Jersey Published online: 21 May 2012.
To cite this article: Paul Mazurek , Waldo Feng , Kenhai Shukla , Anne-Marie Sapse & J. William Lown (1991) The Binding of Prototype Lexitropsins to the Minor Groove of DNA: Quantum Chemical Studies, Journal of Biomolecular Structure and Dynamics, 9:2, 299-313, DOI: 10.1080/07391102.1991.10507914 To link to this article: http://dx.doi.org/10.1080/07391102.1991.10507914
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any
losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content.
Downloaded by [York University Libraries] at 10:31 05 January 2015
This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions
Journal of Biomolecular Structure & Dynamics, ISSN 0739-1102 Volume 9, Issue Number 2 (1991). ©Adenine Press (1991).
The Binding of Prototype Lexitropsins to the Minor Groove of DNA: Quantum Chemical Studies
Downloaded by [York University Libraries] at 10:31 05 January 2015
Paul Mazurek 1, Waldo Fengl, Kenhai Shukla4, . 1 2* • • 3* Anne-Mane Sapse' , and J. Wilham Lown 1
Department of Biophysics and Physiology Mount Sinai School of Medicine 1 Gustav Levy Place New York, N.Y. 2
Permanent address City University of New York and Rockefeller University New York, N.Y. 10021 3
Department of Chemistry University of Alberta Edmonton, Alberta, Canada, T6G 2G2 ~ew Jersey Institute of Technology
New Jersey Abstract Ab initio calculations (Hartree-Fock) using the 6-31G basis set have been performed on two prototype lexitropsins or information-reading molecules. The latter are DNA minor groove binding agents related to the A· T recognizing netropsin in which each of the two N-methylpyrrole moieties is replaced in turn by 1-methylimidazole and which thereby confers the property of recognizing G · C sites.Ab initio treatment was possible by examining composites of separate non-conjugated segments of the molecules. Geometry optimized conformations, energies and distribution of electrostatic charges within the molecules were derived. The ab initio derived parameters of the geometry optimized conformations of these lexitropsins were used to interpret their interaction with different sequences within the minor groove of B-DNA.
Introduction The regulation of gene expression by control proteins (promoters, repressors) in both procaryotes and eucaryotes requires the specific recognition of both singlestranded and double-stranded nucleic acid base sequences (2). Control proteins and xenobiotics appear to utilize two different channels of information in reading the base sequence in double helical DNA (2k). In general the major groove is *Authors to whom correspondence should be addressed.
299
300
Mazurek et a/.
Downloaded by [York University Libraries] at 10:31 05 January 2015
employed by control proteins [although in some cases interactions extend into the minor groove (2h,i,j)], while the minor groove is utilized by certain polymerases and many xenobiotics including antibiotics (2f,k). The reason for the selection of the major groove by control proteins is presumably because of the higher information content (defined by the number of and discriminatory capacity of hydrogen bond accepting and donating sites). Nevertheless, from the viewpoint of artificial control of gene expression, the minor groove offers advantages in that, unlike the major groove which may often be occupied, it is relatively accessible to attack. This may be the reason for the evolutionary development of minor groove selective antibiotics by microorganisms to combat competitors. Nature provides examples of sequence selective minor groove binding antibiotics including netropsin (3), distamycin (4,5), anthelvencin (6), and kikumycin (7). Evidence from biochemical pharmacology indicates that these oligopeptide antibiotics act to block the template function ofDNA by binding selectivity to (AT)n sequences in the minor groove (5). Analysis of the structural requirements for the molecular recognition of netropsin and distamycin for DNA (2k,8) suggested appropriate structural modification of the parent antibiotics could lead to a predictable alteration in DNA sequence recognition. The resulting compounds have been termed "lexitropsins" or information reading molecules (8). Examples of such lexitropsins bearing hydrogen bond accepting heterocycles including imidazole, furan, 1,2,4-triazole, and thiazole have been shown, by MPE complementary strand footprinting, high field NMR analysis and additional spectroscopic techniques, to exhibit the anticipated altered sequence recognition (9). The latter phenomenon together with the efficiency of DNA binding in tum can be related to their biological properties (10) including, in certain cases, selective suppression of gene expression (lOd). The components of molecular recognition that determine the sequence selectivity include electrostatic attraction between ligand and DNA(ll ), hydrogen bonds from the amide NH groups which bridge the strands to the exposed adenine N(3) and thymine 0(2) on adjacent sites and between inward directed hetero atoms and guanine-2-NH 2 sites(2k,8,9f). The semantophoric, or information-reading process, is evidently carried out largely by van der Waals nonbonded contacts between the ligand and surface of the minor groove (2k,8,9f). Such hydrophobic interactions have been invoked in the recognition of the lac repressor (12). Theoretical studies on minor groove binding ligands to date have made use of the methods of molecular mechanics including the AMBER program as well as molecular modeling through the use of computer graphics (13). The important contributions of Pullman and coworkers (11), pioneers in the application of theoretical methods to lexitropsins, indicate that the binding of the drug to the DNA is not only due to hydrogen bonding but also to electrostatic effects, i.e. the interaction between the electrostatic field in the grooves of the DNA and that of the drug. Pullman proposes that the lower affinity of netropsin and distamycin for (GC)n sequences compared with their affinity for AT stretches is not only due to the steric hindrance afforded by the guanine-2-NH 2 but also because GC sequences feature a less negative potential than AT sequences in the minor groove and consequently interact less efficiently
Lexitropsin DNA Binding
301
Downloaded by [York University Libraries] at 10:31 05 January 2015
1
2
3
Figure 1: Schematic representation showing the numbering of the atoms of the parent antibiotic netropsin I, and the two prototype imidazole-bearing lexitropsins in the present study 2 and 3.
with the biscationic oligopeptide antibiotics (lid). Accumulating experimental evidence from complementary strand footprinting (9b,d-h), high-field NMR analysis of ligand-oligonucleotide complexes (9c,d,f,g,h) and microcalorimetry (9g,h) have provided strict tests for such theoretical predictions. To date, however, no ab initio quantum chemical treatment of DNA groove selective ligands has been reported. Presumably the reason is that the relatively large size of such compounds precludes ab initio calculations at high computational level, such as that provided by a split-valence Gaussian basis set. This study addresses itself to this question and applies ab initio methods to the study of certain prototype lexitropsins in which each of the two N-methylpyrrole moieties of netropsin is successively replaced by 1-methylimidazole. Included are calculations of preferred conformations,
302
Mazurek et a/.
Downloaded by [York University Libraries] at 10:31 05 January 2015
a
b
H
c
H
N
¥~ N C I
CH3
H
II
0
N
I
'
CH3
Figure 2: The three fragments a, band c of the lexitropsins used for the 6-31G optimization.
Lexitropsin DNA Binding
303
energies and electrostatic distributions. The ab initio derived parameters of the geometry optimized conformations of these lexitropsins are then used to interpret their interaction with different sequences within the minor groove of B-DNA Methods
Downloaded by [York University Libraries] at 10:31 05 January 2015
Drugs
The agents selected for study are prototype lexitropsins, or information-reading DNA minor groove binding ligands, 2 and 3 (Figure 1). The molecules bear an imidazole moiety in place of one or other of the N-methylpyrrole groups in the parent natural antitumor antibiotic netropsin 1 (Figure 1). The synthesis and characterization of compounds 2 and 3 (9a) as well as their DNA binding constants (3,9b) and sequence selective binding (3,9b) and NMR analysis of their DNA complexes (9c,d,i) have been reported. Lexitropsins bearing imidazole moieties were designed and synthesized to invoke hydrogen bonding between the inward-directed imidazole nitrogen and the G-2-NH2 group in the minor groove (9f). The present ab initio calculations were designed to assist in the interpretation ofthe observed properties of such drugs themselves and of their interaction with their cell receptors, principally duplex DNA 0/igodeoxyn ucleotides
The double stranded DNA structures were generated in the B conformation with a program that constructs a computerized model restricting all bond lengths and bond angles to those established experimentally for fiber DNA This study used a B-DNA segment [GCGAATTGCGh. The crystal structure of the complete dodecamer, from which this segment was derived [CGCGAATTCGCGh was determined by Dickerson et al. (17) and is stored in the Brookhaven Protein Data Bank, from which it was retrieved via the Quanta program. Ab initio Calculations Lexitropsins
The model used for ab initio geometry optimization of an imidazole-containing lexitropsin is shown in Figure 1 (2). The ab initio calculations were performed using 631 G and ST0-3G basis sets, as implemented by the Gaussian-86 computer program (14). Since the 6-31G basis set, which is a split valence set, is more reliable insofar as predicting accurate geometries than the minimal ST0-3G basis set an attempt was made to calculate all the bond lengths and angles via 6-31 G geometry optimization. Indeed it is known that minimum basis set calculations predict, for instance, an exaggerated length for the CN bond in formamide, while the 6-31G basis set predicts a CN bond length closer to experimental values (15d). However it is not feasible to treat the lexitropsin molecules in their entirety with the 6-31G basis set owing to their size. Accordingly, we examined separate fragments of the lexitropsins. As pointed out by Gago et al. (13b ), this procedure may present difficulties owing to the fact that the structure is conjugated. However this potential source of error may be
Mazurek et a/.
Downloaded by [York University Libraries] at 10:31 05 January 2015
304
Figure 3: Computer graphic representation of the three conformations considered of structure 2. (A) optimized conformation a; (B) the fully coplanar conformation ~; (C) conformation y with dihedral angle C 10C9N 7C8 at 63 o and all other dihedral angles set at 180°. A', B' and C' show the corresponding space filling versions of the three conformations.
Downloaded by [York University Libraries] at 10:31 05 January 2015
Lexitropsin DNA Binding
305
Figure 3 continued
306
Mazurek et a/. Table I Optimized Bond Lengths, Angles and Dihedral Angles (A, Degrees) for Conformation a for Lexitropsins 2 and 3. Parameter
NsC6 c6c4 C4N6
Downloaded by [York University Libraries] at 10:31 05 January 2015
N6Cs N4C4 C 3N 4 0 1C 3 clcl N 3C 2 CINJ NICI=NlCI CsCs N7Cs C9N7 CIOC9 CIICIO NsCII c12Ns cl4cll N9C14 02Cs OJCI4 c1sN9 cl6cls cl7cl6 NIOCI7=NIICI7
Value
1.396 1.368 1.379 1.300 1.399 1.344 1.241 1.514 1.486 1.344 1.330 1.447 1.371 1.433 1.368 1.368 1.396 1.396 1.447 1.344 1.211 1.211 1.399 1.564 1.507 1.318
Parameter C4C6Ns N6C4C6 C 5N 6C 4 N4C4N6 C 3N 4C 4 C 2C 3N 4 N 3C 2C 3 CINJCl CsCsNs N7CsCs CwC~7
N9Cl4Cll cl6clsN9 cl7cl6cls Nwcl7cl6=Nllcl7cl6 N 3C 2C 3N 4 Nwcl7cl6cls CIOC9N7C8 cl6clsN9cl4 cl7cl6clsN9
Value
105° lW
105° ur 125° ll5° 108° 125° 125° 109° 109° 120° ll0° ur 120° 180° 99° 63° 75° 19r
avoided by a judicious choice offragments. Since there is no conjugation through a CH2 group, using the parameters of the guanidinium ion and amidinium ion moieties should not introduce any error. For the other molecular components the use of overlapping segments a, b and c (Figure 2) provides reliable values for the intervening segment bond lengths and angles. The segments examined are shown in Figure 2. Thus it can be seen that fragment C provides overlap between segments a and b in such a way as to contain the C 5-C8 bond, present in 2. The bond length value thus obtained was also used for the C 11 -C 14 bond in 2 and 3. The guanidinium ion has been extensively studied at the 631G level by one of us (AMS) and coworkers (15a,b,c), so the geometry is known. Moreover, it has been demonstrated that replacing one of the hydrogens in guanidinium by a methyl group does not alter the geometry significantly (16). Therefore, through optimization, the guanidinium parameters have been maintained at their established values. Once the bond lengths and bond angles were obtained for the fragments they were used for the description of the whole molecule. The dihedral angles between the fragments and within the fragments were obtained by optimizing the complete molecules at the ST0-3G level, while all the bond lengths and angles were maintained at the 6-31G
Lexitropsin DNA Binding
307
Table II Energies in Atomic Units and Energy Differences in kcal/mole.
~
-1386.4598
D. E (1·2or3) 0 4.5
Conformation y
-1386.4543
7.9
Conformation a
Downloaded by [York University Libraries] at 10:31 05 January 2015
Conformation
-1386.4669
values obtained by fragment optimization. These dihedral angles are: N 3C 2C 3N 4, CIOC9N7C8(a), c16ci5N9CI4• c17c16ci5N9, and NIOC17C16cl5• As mentioned above, the angles were optimized for the whole molecule at the ST0-3G level. In order to assess differences between the STO-3G results and those using 6-31 Gin this respect, the dihedral angle between the two rings in fragment C has been optimized at both STO-3G and 6-31 G levels. It was found that the differences are small, both methods predicting a dihedral angle between 60 and 70°. The structure shown in Figure 1 (3) has not been optimized. Its parameters have been obtained by using the results for Figure I (2) and reversing the positions of the pyrrole and imidazole rings for 3.
Drug-DNA Interactions It has been established experimentally that lexitropsins interact with DNA by insertion into the minor groove (2k,8,9). The parent natural products netropsin and distamycin bind to (AT)n sequences (8) whereas when rationally altered structurally by e.g. replacing the pyrrole units with imidazole or other appropriate heterocycle, the resulting lexitropsins are capable of recognizing GC sites (9f).
This study focused on the structural components of the molecular recognition process influenced by the energetics of drug binding. The study used a B-DNA segment [GCGAATTCGCh. The site targeted for the insertion of the drug is the AATT sequence The crystal structure of the complete dodecamer, from which this segment was derived, [CGCGAATTCGCGh was determined by Dickerson eta!. (17) and is stored in the Brookhaven Protein Data Bank, from which it was retrieved via the Quanta program. Account was taken in the calculations of the sheath of water molecules within the grooves (17b) and the necessity to displace an appropriate number of them during incorporation of the drug molecules (18). The Quanta program (version 3.2) was used to insert conformations~ andy of each lexitropsin into the minor groove of the segment, with the aid of the rotational-translational matrix, aiming for a minimum in energy. The energy calculated by the Quanta program represents a sum of electrostatic and van der Waals energies. This energy is expressed in kcal/mole however it is not calibrated in a way which would afford significant values. Therefore it was used to provide only qualitative results for comparison purposes. During the insertion of individual drugs into the minor groove via Quanta energy optimization, the different conformations of the drugs were maintained and only their interaction with the DNA was considered. No CHARMm calculations have been performed.
Mazurek et a/.
Downloaded by [York University Libraries] at 10:31 05 January 2015
308
180.00 160.00 1-40.00 120.00 100.00
so.oo 60.00
Journal of Biomolecular Structure and Dynamics Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/tbsd20
The Binding of Prototype Lexitropsins to the Minor Groove of DNA: Quantum Chemical Studies a
a
c
Paul Mazurek , Waldo Feng , Kenhai Shukla , a
Anne-Marie Sapse & J. William Lown
b
a
Department of Biophysics and Physiology , Mount Sinai School of Medicine , 1 Gustav Levy Place, New York , N.Y. b
Department of Chemistry , University of Alberta Edmonton , Alberta , Canada , T6G 2G2 c
New Jersey Institute of Technology , New Jersey Published online: 21 May 2012.
To cite this article: Paul Mazurek , Waldo Feng , Kenhai Shukla , Anne-Marie Sapse & J. William Lown (1991) The Binding of Prototype Lexitropsins to the Minor Groove of DNA: Quantum Chemical Studies, Journal of Biomolecular Structure and Dynamics, 9:2, 299-313, DOI: 10.1080/07391102.1991.10507914 To link to this article: http://dx.doi.org/10.1080/07391102.1991.10507914
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any
losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content.
Downloaded by [York University Libraries] at 10:31 05 January 2015
This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions
Journal of Biomolecular Structure & Dynamics, ISSN 0739-1102 Volume 9, Issue Number 2 (1991). ©Adenine Press (1991).
The Binding of Prototype Lexitropsins to the Minor Groove of DNA: Quantum Chemical Studies
Downloaded by [York University Libraries] at 10:31 05 January 2015
Paul Mazurek 1, Waldo Fengl, Kenhai Shukla4, . 1 2* • • 3* Anne-Mane Sapse' , and J. Wilham Lown 1
Department of Biophysics and Physiology Mount Sinai School of Medicine 1 Gustav Levy Place New York, N.Y. 2
Permanent address City University of New York and Rockefeller University New York, N.Y. 10021 3
Department of Chemistry University of Alberta Edmonton, Alberta, Canada, T6G 2G2 ~ew Jersey Institute of Technology
New Jersey Abstract Ab initio calculations (Hartree-Fock) using the 6-31G basis set have been performed on two prototype lexitropsins or information-reading molecules. The latter are DNA minor groove binding agents related to the A· T recognizing netropsin in which each of the two N-methylpyrrole moieties is replaced in turn by 1-methylimidazole and which thereby confers the property of recognizing G · C sites.Ab initio treatment was possible by examining composites of separate non-conjugated segments of the molecules. Geometry optimized conformations, energies and distribution of electrostatic charges within the molecules were derived. The ab initio derived parameters of the geometry optimized conformations of these lexitropsins were used to interpret their interaction with different sequences within the minor groove of B-DNA.
Introduction The regulation of gene expression by control proteins (promoters, repressors) in both procaryotes and eucaryotes requires the specific recognition of both singlestranded and double-stranded nucleic acid base sequences (2). Control proteins and xenobiotics appear to utilize two different channels of information in reading the base sequence in double helical DNA (2k). In general the major groove is *Authors to whom correspondence should be addressed.
299
300
Mazurek et a/.
Downloaded by [York University Libraries] at 10:31 05 January 2015
employed by control proteins [although in some cases interactions extend into the minor groove (2h,i,j)], while the minor groove is utilized by certain polymerases and many xenobiotics including antibiotics (2f,k). The reason for the selection of the major groove by control proteins is presumably because of the higher information content (defined by the number of and discriminatory capacity of hydrogen bond accepting and donating sites). Nevertheless, from the viewpoint of artificial control of gene expression, the minor groove offers advantages in that, unlike the major groove which may often be occupied, it is relatively accessible to attack. This may be the reason for the evolutionary development of minor groove selective antibiotics by microorganisms to combat competitors. Nature provides examples of sequence selective minor groove binding antibiotics including netropsin (3), distamycin (4,5), anthelvencin (6), and kikumycin (7). Evidence from biochemical pharmacology indicates that these oligopeptide antibiotics act to block the template function ofDNA by binding selectivity to (AT)n sequences in the minor groove (5). Analysis of the structural requirements for the molecular recognition of netropsin and distamycin for DNA (2k,8) suggested appropriate structural modification of the parent antibiotics could lead to a predictable alteration in DNA sequence recognition. The resulting compounds have been termed "lexitropsins" or information reading molecules (8). Examples of such lexitropsins bearing hydrogen bond accepting heterocycles including imidazole, furan, 1,2,4-triazole, and thiazole have been shown, by MPE complementary strand footprinting, high field NMR analysis and additional spectroscopic techniques, to exhibit the anticipated altered sequence recognition (9). The latter phenomenon together with the efficiency of DNA binding in tum can be related to their biological properties (10) including, in certain cases, selective suppression of gene expression (lOd). The components of molecular recognition that determine the sequence selectivity include electrostatic attraction between ligand and DNA(ll ), hydrogen bonds from the amide NH groups which bridge the strands to the exposed adenine N(3) and thymine 0(2) on adjacent sites and between inward directed hetero atoms and guanine-2-NH 2 sites(2k,8,9f). The semantophoric, or information-reading process, is evidently carried out largely by van der Waals nonbonded contacts between the ligand and surface of the minor groove (2k,8,9f). Such hydrophobic interactions have been invoked in the recognition of the lac repressor (12). Theoretical studies on minor groove binding ligands to date have made use of the methods of molecular mechanics including the AMBER program as well as molecular modeling through the use of computer graphics (13). The important contributions of Pullman and coworkers (11), pioneers in the application of theoretical methods to lexitropsins, indicate that the binding of the drug to the DNA is not only due to hydrogen bonding but also to electrostatic effects, i.e. the interaction between the electrostatic field in the grooves of the DNA and that of the drug. Pullman proposes that the lower affinity of netropsin and distamycin for (GC)n sequences compared with their affinity for AT stretches is not only due to the steric hindrance afforded by the guanine-2-NH 2 but also because GC sequences feature a less negative potential than AT sequences in the minor groove and consequently interact less efficiently
Lexitropsin DNA Binding
301
Downloaded by [York University Libraries] at 10:31 05 January 2015
1
2
3
Figure 1: Schematic representation showing the numbering of the atoms of the parent antibiotic netropsin I, and the two prototype imidazole-bearing lexitropsins in the present study 2 and 3.
with the biscationic oligopeptide antibiotics (lid). Accumulating experimental evidence from complementary strand footprinting (9b,d-h), high-field NMR analysis of ligand-oligonucleotide complexes (9c,d,f,g,h) and microcalorimetry (9g,h) have provided strict tests for such theoretical predictions. To date, however, no ab initio quantum chemical treatment of DNA groove selective ligands has been reported. Presumably the reason is that the relatively large size of such compounds precludes ab initio calculations at high computational level, such as that provided by a split-valence Gaussian basis set. This study addresses itself to this question and applies ab initio methods to the study of certain prototype lexitropsins in which each of the two N-methylpyrrole moieties of netropsin is successively replaced by 1-methylimidazole. Included are calculations of preferred conformations,
302
Mazurek et a/.
Downloaded by [York University Libraries] at 10:31 05 January 2015
a
b
H
c
H
N
¥~ N C I
CH3
H
II
0
N
I
'
CH3
Figure 2: The three fragments a, band c of the lexitropsins used for the 6-31G optimization.
Lexitropsin DNA Binding
303
energies and electrostatic distributions. The ab initio derived parameters of the geometry optimized conformations of these lexitropsins are then used to interpret their interaction with different sequences within the minor groove of B-DNA Methods
Downloaded by [York University Libraries] at 10:31 05 January 2015
Drugs
The agents selected for study are prototype lexitropsins, or information-reading DNA minor groove binding ligands, 2 and 3 (Figure 1). The molecules bear an imidazole moiety in place of one or other of the N-methylpyrrole groups in the parent natural antitumor antibiotic netropsin 1 (Figure 1). The synthesis and characterization of compounds 2 and 3 (9a) as well as their DNA binding constants (3,9b) and sequence selective binding (3,9b) and NMR analysis of their DNA complexes (9c,d,i) have been reported. Lexitropsins bearing imidazole moieties were designed and synthesized to invoke hydrogen bonding between the inward-directed imidazole nitrogen and the G-2-NH2 group in the minor groove (9f). The present ab initio calculations were designed to assist in the interpretation ofthe observed properties of such drugs themselves and of their interaction with their cell receptors, principally duplex DNA 0/igodeoxyn ucleotides
The double stranded DNA structures were generated in the B conformation with a program that constructs a computerized model restricting all bond lengths and bond angles to those established experimentally for fiber DNA This study used a B-DNA segment [GCGAATTGCGh. The crystal structure of the complete dodecamer, from which this segment was derived [CGCGAATTCGCGh was determined by Dickerson et al. (17) and is stored in the Brookhaven Protein Data Bank, from which it was retrieved via the Quanta program. Ab initio Calculations Lexitropsins
The model used for ab initio geometry optimization of an imidazole-containing lexitropsin is shown in Figure 1 (2). The ab initio calculations were performed using 631 G and ST0-3G basis sets, as implemented by the Gaussian-86 computer program (14). Since the 6-31G basis set, which is a split valence set, is more reliable insofar as predicting accurate geometries than the minimal ST0-3G basis set an attempt was made to calculate all the bond lengths and angles via 6-31 G geometry optimization. Indeed it is known that minimum basis set calculations predict, for instance, an exaggerated length for the CN bond in formamide, while the 6-31G basis set predicts a CN bond length closer to experimental values (15d). However it is not feasible to treat the lexitropsin molecules in their entirety with the 6-31G basis set owing to their size. Accordingly, we examined separate fragments of the lexitropsins. As pointed out by Gago et al. (13b ), this procedure may present difficulties owing to the fact that the structure is conjugated. However this potential source of error may be
Mazurek et a/.
Downloaded by [York University Libraries] at 10:31 05 January 2015
304
Figure 3: Computer graphic representation of the three conformations considered of structure 2. (A) optimized conformation a; (B) the fully coplanar conformation ~; (C) conformation y with dihedral angle C 10C9N 7C8 at 63 o and all other dihedral angles set at 180°. A', B' and C' show the corresponding space filling versions of the three conformations.
Downloaded by [York University Libraries] at 10:31 05 January 2015
Lexitropsin DNA Binding
305
Figure 3 continued
306
Mazurek et a/. Table I Optimized Bond Lengths, Angles and Dihedral Angles (A, Degrees) for Conformation a for Lexitropsins 2 and 3. Parameter
NsC6 c6c4 C4N6
Downloaded by [York University Libraries] at 10:31 05 January 2015
N6Cs N4C4 C 3N 4 0 1C 3 clcl N 3C 2 CINJ NICI=NlCI CsCs N7Cs C9N7 CIOC9 CIICIO NsCII c12Ns cl4cll N9C14 02Cs OJCI4 c1sN9 cl6cls cl7cl6 NIOCI7=NIICI7
Value
1.396 1.368 1.379 1.300 1.399 1.344 1.241 1.514 1.486 1.344 1.330 1.447 1.371 1.433 1.368 1.368 1.396 1.396 1.447 1.344 1.211 1.211 1.399 1.564 1.507 1.318
Parameter C4C6Ns N6C4C6 C 5N 6C 4 N4C4N6 C 3N 4C 4 C 2C 3N 4 N 3C 2C 3 CINJCl CsCsNs N7CsCs CwC~7
N9Cl4Cll cl6clsN9 cl7cl6cls Nwcl7cl6=Nllcl7cl6 N 3C 2C 3N 4 Nwcl7cl6cls CIOC9N7C8 cl6clsN9cl4 cl7cl6clsN9
Value
105° lW
105° ur 125° ll5° 108° 125° 125° 109° 109° 120° ll0° ur 120° 180° 99° 63° 75° 19r
avoided by a judicious choice offragments. Since there is no conjugation through a CH2 group, using the parameters of the guanidinium ion and amidinium ion moieties should not introduce any error. For the other molecular components the use of overlapping segments a, b and c (Figure 2) provides reliable values for the intervening segment bond lengths and angles. The segments examined are shown in Figure 2. Thus it can be seen that fragment C provides overlap between segments a and b in such a way as to contain the C 5-C8 bond, present in 2. The bond length value thus obtained was also used for the C 11 -C 14 bond in 2 and 3. The guanidinium ion has been extensively studied at the 631G level by one of us (AMS) and coworkers (15a,b,c), so the geometry is known. Moreover, it has been demonstrated that replacing one of the hydrogens in guanidinium by a methyl group does not alter the geometry significantly (16). Therefore, through optimization, the guanidinium parameters have been maintained at their established values. Once the bond lengths and bond angles were obtained for the fragments they were used for the description of the whole molecule. The dihedral angles between the fragments and within the fragments were obtained by optimizing the complete molecules at the ST0-3G level, while all the bond lengths and angles were maintained at the 6-31G
Lexitropsin DNA Binding
307
Table II Energies in Atomic Units and Energy Differences in kcal/mole.
~
-1386.4598
D. E (1·2or3) 0 4.5
Conformation y
-1386.4543
7.9
Conformation a
Downloaded by [York University Libraries] at 10:31 05 January 2015
Conformation
-1386.4669
values obtained by fragment optimization. These dihedral angles are: N 3C 2C 3N 4, CIOC9N7C8(a), c16ci5N9CI4• c17c16ci5N9, and NIOC17C16cl5• As mentioned above, the angles were optimized for the whole molecule at the ST0-3G level. In order to assess differences between the STO-3G results and those using 6-31 Gin this respect, the dihedral angle between the two rings in fragment C has been optimized at both STO-3G and 6-31 G levels. It was found that the differences are small, both methods predicting a dihedral angle between 60 and 70°. The structure shown in Figure 1 (3) has not been optimized. Its parameters have been obtained by using the results for Figure I (2) and reversing the positions of the pyrrole and imidazole rings for 3.
Drug-DNA Interactions It has been established experimentally that lexitropsins interact with DNA by insertion into the minor groove (2k,8,9). The parent natural products netropsin and distamycin bind to (AT)n sequences (8) whereas when rationally altered structurally by e.g. replacing the pyrrole units with imidazole or other appropriate heterocycle, the resulting lexitropsins are capable of recognizing GC sites (9f).
This study focused on the structural components of the molecular recognition process influenced by the energetics of drug binding. The study used a B-DNA segment [GCGAATTCGCh. The site targeted for the insertion of the drug is the AATT sequence The crystal structure of the complete dodecamer, from which this segment was derived, [CGCGAATTCGCGh was determined by Dickerson eta!. (17) and is stored in the Brookhaven Protein Data Bank, from which it was retrieved via the Quanta program. Account was taken in the calculations of the sheath of water molecules within the grooves (17b) and the necessity to displace an appropriate number of them during incorporation of the drug molecules (18). The Quanta program (version 3.2) was used to insert conformations~ andy of each lexitropsin into the minor groove of the segment, with the aid of the rotational-translational matrix, aiming for a minimum in energy. The energy calculated by the Quanta program represents a sum of electrostatic and van der Waals energies. This energy is expressed in kcal/mole however it is not calibrated in a way which would afford significant values. Therefore it was used to provide only qualitative results for comparison purposes. During the insertion of individual drugs into the minor groove via Quanta energy optimization, the different conformations of the drugs were maintained and only their interaction with the DNA was considered. No CHARMm calculations have been performed.
Mazurek et a/.
Downloaded by [York University Libraries] at 10:31 05 January 2015
308
180.00 160.00 1-40.00 120.00 100.00
so.oo 60.00
