Different conformations and site selectivity of HO-2-Co(III)-bleomycin A2 and Co (III)-bleomycin A2 bound to DNA oligomers

Kinetic studies of the interaction between antitumor antibiotics and DNA using quartz crystal microbalance

OBJECTIVE

Kinetic studies of the interaction process between antitumor antibiotics, Mitomycin C (MMC) and Bleomycin (BLM), and DNA were performed with a novel analytical method, piezoelectric quartz crystal (PQC) impedance analysis.

DESIGN AND METHODS

DNA was directly immobilized on the Au-electrode surface of a piezoelectric quartz crystal by adsorption. The DNA-modified piezoelectric sensor was in contact with Mitomycin C and Bleomycin solution, respectively.

RESULTS

The experimental results demonstrated that antitumor antibiotics concentration had an effect on the interaction. A pseudo-first-order kinetic model for the interaction between antitumor antibiotics and DNA was derived to describe the process. All fitted results were well in agreement with the corresponding experimental results. The kinetic parameters, the interaction rate constants (k(MMC) and k(BLM)), were determined by fitting experimental data to the model. At 37 +/- 0.5 degrees C, the k(MMC) and k(BLM) values obtained were 4.56 (+/-0.02) x 10(-3) and 9.11 (+/-0.02) mM(-1) s(-1), respectively.

CONCLUSION

Piezoelectric quartz crystal impedance (PQCI) analysis is a very powerful method to study the kinetic process of antitumor drugs and DNA interaction.

The solution structure of the Ga(III)-bleomycin A2 complex resolved by NMR and molecular modeling; interaction with d(CCAGGCCTGG)

The solution structure of the Ga(III)-bleomycin A2 complex (GaBLM) has been determined using 2D NMR methods in combination with molecular dynamics calculations. Complete assignment of the amide and amine protons, observation of 80 NOEs and measurement of 15 (3)JH(-H) coupling constants provided us with a well-defined structure using a restrained simulated annealing protocol. On the basis of distance and dihedral angle constraints agreement, along with potential energy considerations, the favored model is a five-coordinate complex with the primary amine of beta-aminoalanine holding the axial position of a distorted tetragonal pyramid. The disaccharide moiety of GaBLM is not a ligand, sharing the same side of the equatorial plane with the axial amine ligand. Titration of the self-complementary oligonucleotide d(CCAGGCCTGG) with GaBLM results in the formation of only one 1:1 complex in slow exchange on the NMR time scale. Our data indicate that the bithiazole moiety intercalates between the C6*G15 and C7*G14 base pairs, in a similar mode to that reported by earlier studies. Structural implications and comparisons to other metallo-bleomycins are discussed.

Structural studies on metallobleomycins: The interaction of Pt(II) and Pd(II) with bleomycin

Two of the most successful chemotherapeutic agents used in the treatment of several neoplasias are bleomycin and cisplatin. Both drugs attack the DNA leading to the cancer cells death via different mechanisms. In view of the fact that the combination with each other leads to enhanced activity with less sever side effects, we have undertaken NMR studies on the complexes formed between bleomycin and PtII, PdII, cisplatin and transplatin. Herein we present a brief review of the studies on metallobleomycins which were carried out by our lab and others, as an outline of the results obtained using NMR in combination to circular dichroism spectroscopy. Our data indicate that in most cases and under several conditions studied, both metal ions form similar complexes with BLM while more than one species are present in the solution. Structural implications and comparisons with other metallobleomycins are being discussed. .

Structures of HO(2)-Co(III)bleomycin A(2) bound to d(GAGCTC)(2) and d(GGAAGCTTCC)(2): structure-reactivity relationships of Co and Fe bleomycins

HO(2)-Co(III)bleomycin is a model for HO(2)-Fe(III)bleomycin, which initiates single and double strand cleavage of DNA. In order to enlarge the understanding of its structure and reactivity, three-dimensional structures of HO(2)-Co(III)bleomycin bound to two DNA oligomers, d(GAGCTC)(2) (I) and d(GGAAGCTTCC)(2) (II), that have 5'-GC-3' binding sites, have been determined by nuclear magnetic resonance (NMR) methods. Besides previously recognized determinants of binding selectivity, a probable hydrogen bond was detected between the pyrimidinyl acetamido NH(2) and the carbonyl of cytosine base paired to G at the recognition site. Another hydrogen bond between the NH of the dimethylsulfonium R group and N7 of guanine opposite cytosine at the GC site may contribute to specification of the pyrimidine. Substitution of G with inosine shifted HO(2)-Co(III)Blm A(2)[bond]I and Fe(III)Blm[bond]I into fast exchange on the NMR time scale, supporting the role of the 2-amino group in site specification for each molecule. The conformationally stable metal-domain linker established a close-packed adduct with the minor groove in which the hydroperoxide ligand occupies a sterically constrained pocket that is isolated from the solvent. The hydroperoxide group is directed toward one of the two cytosine H4' hydrogens but is sterically blocked from access to the other by the drug. These findings enlarge the structural understanding of selective binding of Co(III)/Fe(III)Blm species at G-pyrimidine sites. They also rationalize the instability of a number of ligands bound to Co(III)/Fe(III)Blm at specific binding sequences and the relative unreactivity of Fe(III)Blm[bond]I with ascorbate as well as its lack of interaction with spin labels.

Raman spectroscopy of an O(2)-Co(II)bleomycin-calf thymus DNA adduct: alternate polymer conformations

Bleomycin (Blm) is an antitumor agent which binds to specific sequences of DNA and as HO(2)-Fe(III)Blm causes single and double strand cleavage. In the present investigation, binding of O(2)-Co(II)Blm to a native DNA polymer, calf thymus DNA, was examined using conventional Raman spectroscopy. O(2)-Co(II)Blm is a model for O(2)-Fe(II)Blm, the direct precursor of HO(2)-Fe(III)Blm. Although the DNA polymer retained a predominant B-form structure, Raman spectral evidence was obtained for localized structural changes to A, C and Z-DNA forms. The presence of these alternate DNA forms within B-DNA implied the presence of B/A, B/C and B/Z junctions. The observed changes in DNA secondary structure were attributed to perturbation of structural water resulting from binding of O(2)-Co(II)Blm within the minor groove.

Conformational/configurational analysis of all the binding geometries of cobalt(III) bleomycin

No crystal structure of metallobleomycin (BLM) exists, and the exact coordination mode of the ligand is unknown. To date, spectroscopic investigations of BLM complexes and crystal structures of BLM models have been used to propose its metal coordination sites. This has led to contradictory interpretations of the metal coordination sphere in BLM. Inorganic molecular mechanics and configurational/conformational searches were used to analyze HOO-CoBLM A2, H2O-CoBLM A2, and HOO-CoPEP with commonly proposed binding geometries. The lowest energy binding geometry found was one with the mannose carbamoyl bound to the cobalt ion. The Monte Carlo dihedral and translational variational searches were able to find most of the configurations available to cobalt(III) bleomycin in the three binding geometries examined.

A model of the structure of HOO-Co.bleomycin bound to d(CCAGTACTGG): recognition at the d(GpT) site and implications for double-stranded DNA cleavage

BACKGROUND

The bleomycins (BLMs) are a family of natural products used clinically as antitumor agents. In the presence of their required cofactors, iron and oxygen, BLMs bind to and mediate single-stranded and double-stranded DNA cleavage. Recently, two dimensional nuclear magnetic resonance (2D NMR) spectroscopic studies and molecular modeling have provided a picture of how the hydroperoxide form of cobalt BLM A2 (HOO-CoBLM), an analog of 'activated' iron BLM (HOO-FeBLM), binds to a d(GpC) motif and of the basis for both sequence specificity and chemical specificity of DNA cleavage.

RESULTS

The solution structure of HOO-CoBLM bound to d(CCAGTACTGG) containing a 'hot spot' for double-stranded DNA cleavage at T5 and T15 is reported using constraints from 2D NMR spectroscopy. The mode of binding and basis for sequence specificity and chemical specificity of cleavage is almost identical to that of a d(GpC) motif. This structure has allowed formulation of a structural model for how a single molecule of FeBLM can mediate a double-stranded DNA cleavage event without dissociation from the DNA.

CONCLUSIONS

The structural similarity of HOO-CoBLM bound to d(GpT) in d(CCAGTACTGG) compared to a d(GpC) motif suggests a general paradigm for the binding of HOO-CoBLM to DNA and, by analogy, for the binding of the biological significant entity HOO-FeBLM.

Mapping of the cleavage-associated bleomycin binding site on DNA with a new method based on site-specific blockage of the minor groove with N2-isobutyrylguanine

Although the binding of various forms of bleomycin to DNA has been studied extensively, the transient nature of the activated bleomycin species which ultimately attacks DNA has largely precluded direct examination of its physical interactions with DNA. In an attempt to map the minimum binding site required for this species to effect DNA cleavage, several oligonucleotide duplexes were synthesized, each of which contained a single N2-isobutyrylguanine moiety at a specific position in the sequence. These duplexes were end-labeled, and sequence-specific bleomycin-induced cleavage was assessed in each strand of each duplex. Isobutyrylguanine substitution immediately 5' to a primary bleomycin target site suppressed bleomycin-induced cleavage by more than 10-fold. Substitution two bases 5' to a target site suppressed cleavage by about 4-fold, and substitution directly opposite the target site suppressed cleavage by 7-fold. Substitution immediately 3' to the target site, or at other more distant positions 3' or 5', had little or no effect. In cases where cleavage at a primary site was strongly suppressed, cleavage at the corresponding secondary site (the putative site of the second break in a bleomycin-induced double-strand break) was also inhibited, even when the secondary site was several bases away from the isobutyrylguanine substitution. The results suggest that the binding site required for bleomycin-induced DNA cleavage spans a region of approximately 2 or 3 bp in the minor groove, including the base associated with the sugar attacked and one or two bases to its 5' side. Computer-based molecular modeling indicated that these results are consistent with the predictions of recently proposed models in which the bithiazole is intercalated immediately 3' to the cleavage site, and the iron coordination site binds in the minor groove immediately 5' to the cleavage site. Both the empirical data and the modeling studies suggest that N2-isobutyrylguanine substitution effectively blocks the minor groove, but without significantly disturbing DNA secondary structure. Thus, it is proposed that site-specific incorporation of N2-isobutyrylguanine may provide a general method for mapping binding sites of minor groove-binding ligands on DNA.