High-resolution transcription assay for probing drug-DNA interactions at individual drug sites

New anthracenedione derivatives with improved biological activity by virtue of stable drug-DNA adduct formation

Mitoxantrone is an anticancer agent that acts as a topoisomerase II poison, however, it can also be activated by formaldehyde to form DNA adducts. Pixantrone, a 2-aza-anthracenedione with terminal primary amino groups in its side chains, forms formaldehyde-mediated adducts with DNA more efficiently than mitoxantrone. Molecular modeling studies indicated that extension of the "linker" region of anthracenedione side arms would allow the terminal primary amino greater flexibility and thus access to the guanine residues on the opposite DNA strand. New derivatives based on the pixantrone and mitoxantrone backbones were synthesized, and these incorporated primary amino groups as well as extended side chains. The stability of DNA adducts increased with increasing side chain length of the derivatives. A mitoxantrone derivative bearing extended side chains (7) formed the most stable adducts with ∼100-fold enhanced stability compared to mitoxantrone. This finding is of great interest because long-lived drug-DNA adducts are expected to perturb DNA-dependent functions at all stages of the cell cycle.

In vitro transcription assay for resolution of drug-DNA interactions at defined DNA sequences

A major class of anticancer agents in current clinical use exerts its anticancer effects by binding covalently or non-covalently to DNA. A detailed understanding of the nature of these drug-DNA complexes would be expected to lead to better uses of these drugs, and also assist with the design of improved drug derivatives. Here, we present a transcriptional footprinting assay that can be implemented to define the DNA sequence specificity and kinetics associated with drug-DNA complexes. The basic steps involve the formation of drug-DNA complexes, the formation of synchronised initiated transcripts, and finally transcriptional elongation to reveal drug blockage sites that impede the progression of RNA polymerase. We have used the "in vitro transcription assay" to investigate many covalent drug-DNA interactions; most notably those obtained using anthracycline anticancer agents such as doxorubicin and anthracenedione-based anticancer agents, including mitoxantrone and pixantrone.

Formaldehyde-activated Pixantrone is a monofunctional DNA alkylator that binds selectively to CpG and CpA doublets

The topoisomerase II poison mitoxantrone is important in the clinical management of human malignancies. Pixantrone, a novel aza-anthracenedione developed to improve the therapeutic profile of mitoxantrone, can efficiently alkylate DNA after formaldehyde activation. In vitro transcriptional analysis has now established that formaldehyde-activated pixantrone generates covalent adducts selectively at discrete CpG or CpA dinucleotides, suggesting that the activated complex binds to guanine or cytosine (or both) bases. The stability of pixantrone adduct-induced transcriptional blockages varied considerably, reflecting a mixture of distinct pixantrone adduct types that may include relatively labile monoadducts and more stable interstrand cross-links. 6,9-Bis-[[2-(dimethylamino)ethyl]amino]benzo[g]isoquinoline-5,10-dione (BBR 2378), the dimethyl N-substituted analog of pixantrone, could not form adducts, suggesting that pixantrone alkylates DNA through the primary amino functions located in each side chain of the drug. Pixantrone generated DNA adducts only when guanine was present in substrates and exhibited a lack of adduct formation with inosine-containing polynucleotides, confirming that the N2 amino group of guanine is the site for covalent attachment of the drug. Mass spectrometric analysis of oligonucleotide-drug complexes confirmed that formation of covalent pixantrone-DNA adducts is mediated by a single methylene linkage provided by formaldehyde and that this occurs only with guanine-containing double stranded oligonucleotide substrates. CpG methylation, an epigenetic modification of the mammalian genome, significantly enhanced the generation of pixantrone-DNA adducts within a methylated DNA substrate, indicating that the methylated dinucleotide may be a favored target in a cellular environment.

Quantitative microarray profiling of DNA-binding molecules

A high-throughput Cognate Site Identity (CSI) microarray platform interrogating all 524 800 10-base pair variable sites is correlated to quantitative DNase I footprinting data of DNA binding pyrrole-imidazole polyamides. An eight-ring hairpin polyamide programmed to target the 5 bp sequence 5'-TACGT-3' within the hypoxia response element (HRE) yielded a CSI microarray-derived sequence motif of 5'-WWACGT-3' (W = A,T). A linear beta-linked polyamide programmed to target a (GAA)3 repeat yielded a CSI microarray-derived sequence motif of 5'-AARAARWWG-3' (R = G,A). Quantitative DNase I footprinting of selected sequences from each microarray experiment enabled quantitative prediction of Ka values across the microarray intensity spectrum.

Programmable oligomers for minor groove DNA recognition

The four Watson-Crick base pairs of DNA can be distinguished in the minor groove by pairing side-by-side three five-membered aromatic carboxamides, imidazole (Im), pyrrole (Py), and hydroxypyrrole (Hp), four different ways. On the basis of the paradigm of unsymmetrical paired edges of aromatic rings for minor groove recognition, a second generation set of heterocycle pairs, imidazopyridine/pyrrole (Ip/Py) and hydroxybenzimidazole/pyrrole (Hz/Py), revealed that recognition elements not based on analogues of distamycin could be realized. A new set of end-cap heterocycle dimers, oxazole-hydroxybenzimidazole (No-Hz) and chlorothiophene-hydroxybenzimidazole (Ct-Hz), paired with Py-Py are shown to bind contiguous base pairs of DNA in the minor groove, specifically 5'-GT-3' and 5'-TT-3', with high affinity and selectivity. Utilizing this technology, we have developed a new class of oligomers for sequence-specific DNA minor groove recognition no longer based on the N-methyl pyrrole carboxamides of distamycin.

Minimization of a protein-DNA dimerizer

A protein-DNA dimerizer constructed from a DNA-binding polyamide and the peptide FYPWMKG facilitates the binding of a natural transcription factor Exd to an adjacent DNA site. The Exd binding domain can be reduced to a dipeptide WM attached to the polyamide through an epsilon-aminohexanoic acid linker with retention of protein-DNA dimerizer activity. Screening a library of analogues indicated that the tryptophan indole moiety is more important than methionine's side chain or the N-terminal acetamide. Remarkably, switching the stereochemistry of the tryptophan residue (l to d) stabilizes the dimerizer*Exd*DNA ternary complex at 37 degrees C. These observations provide design principles for artificial transcription factors that may function in concert with the cellular regulatory circuitry.

Sequence-specific fluorescence detection of DNA by polyamide-thiazole orange conjugates

Fluorescent methods to detect specific double-stranded DNA sequences without the need for denaturation may be useful in the field of genetics. Three hairpin pyrrole-imidazole polyamides 2-4 that target their respective sequences 5'-WGGGWW-3', 5'-WGGCCW-3', and 5'-WGWWCW-3' (W = A or T) were conjugated to thiazole orange dye at the C-termini to examine their fluorescence properties in the presence and absence of match duplex DNA. The conjugates fluoresce weakly in the absence of DNA but showed significant enhancement (>1000-fold) upon the addition of 1 equiv of match DNA and only slight enhancement with the addition of mismatch DNA. The polyamide-dye conjugates bound specific DNA sequences with high affinity (Ka > 10(8) M(-1)) and unwound the DNA duplex through intercalation (unwinding angle, phi, approximately 8 degrees). This new class of polyamides provides a method to specifically detect DNA sequences without denaturation.

DNA Cleavage by the Photolysis of Cyclopentadienyl Metal Complexes: Mechanistic Studies and Sequence Selectivity of Strand Scission by CpW(CO)3CH3

[Reaction: see text]. The photolysis of CpW(CO)3Me has been shown to produce methyl radicals and to cleave DNA in a single-stranded manner, and preliminary evidence implicated a carbon-centered radical in this process. In this work, the mechanism of strand scission in this reaction was determined to occur by hydrogen atom abstraction from the 4'- and 5'-positions of the deoxyribose moiety of the backbone of DNA. Additionally, in a side reaction that does not lead to frank strand scission, all four bases of DNA are methylated under these conditions; however, none of these base or backbone modifications lead to the formation of abasic sites.

The speed of RNA transcription and metabolite binding kinetics operate an FMN riboswitch

Riboswitches are genetic control elements that usually reside in untranslated regions of messenger RNAs. These folded RNAs directly bind metabolites and undergo allosteric changes that modulate gene expression. A flavin mononucleotide (FMN)-dependent riboswitch from the ribDEAHT operon of Bacillus subtilis uses a transcription termination mechanism wherein formation of an RNA-FMN complex causes formation of an intrinsic terminator stem. We assessed the importance of RNA transcription speed and the kinetics of FMN binding to the nascent mRNA for riboswitch function. The riboswitch does not attain thermodynamic equilibrium with FMN before RNA polymerase needs to make a choice between continued transcription and transcription termination. Therefore, this riboswitch is kinetically driven, and functions more like a "molecular fuse." This reliance on the kinetics of ligand association and RNA polymerization speed might be common for riboswitches that utilize transcription termination mechanisms.

Visualising DNA: Footprinting and 1-2D Gels

The study of molecular recognition of DNA by natural and synthetic ligands has made enormous progress due in large part to the discovery and development of methods for separating DNA fragments by gel electrophoresis in one and two dimensions, and for characterizing DNA-ligand complexes by footprinting techniques.

Expanding the repertoire of heterocycle ring pairs for programmable minor groove DNA recognition

The discrimination of the four Watson-Crick base pairs by minor groove DNA-binding polyamides have been attributed to the specificity of three five-membered aromatic amino acid subunits, 1-methyl-1H-imidazole (Im), 1-methyl-1H-pyrrole (Py), and 3-hydroxy-1H-pyrrole (Hp) paired four different ways. The search for additional ring pairs that demonstrate DNA-sequence specificity has led us to a new class of 6-5 fused bicycle rings as minor groove recognition elements. The affinities and specificities of the hydroxybenzimidazole/pyrrole (Hz/Py) and hydroxybenzimidazole/benzimidazole (Hz/Bi) pairs for each of the respective Watson-Crick base pairs within the sequence context 5'-TGGXCA-3' (X = A, T, G, C) were measured by quantitative DNaseI footprinting titrations. The Hz/Py and Hz/Bi distinguish T.A from A.T. Hairpin polyamides containing multiple Hz/Py pairs were examined and were shown to mimic the Hp/Py pair with regard to affinity and specificity. Therefore, the Hz/Py pair may be considered a second-generation replacement for the Hp/Py pair.

Toward artificial developmental regulators

A polyamide-peptide conjugate is designed which recruits sequence specifically the developmental regulator Exd to a cognate DNA site. In particular, an eight-ring hairpin polyamide (Im-Im-Py(C3H6NHR)-Py-gamma-Im-Py-Py-Py-beta-Dp) with a heptapeptide (R = Ac-Phe-Tyr-Pro-Trp-Met-Lys-Gly-) attached on a central ring was shown to induce cooperative binding of the Drosophila Hox protein cofactor Exd with a Kd of 4.4 nM in vitro, an order of magnitude more efficient than the natural Hox protein partner Ubx. The conjugate joins two sequence specific domains, one for DNA and one for the protein. This small molecule thus serves as a cooperative protein-DNA dimerizer, which mimics the natural Hox family of developmental regulators.

DNA binding ligands targeting drug-resistant bacteria: structure, activity, and pharmacology

We describe the lead optimization and structure-activity relationship of DNA minor-groove binding ligands, a novel class of antibacterial molecules. These compounds have been shown to target A/T-rich sites within the bacterial genome and, as a result, inhibit DNA replication and RNA transcription. The optimization was focused on N-terminal aromatic heterocycles and C-terminal amines and resulted in compounds with improved in vivo tolerability and excellent in vitro antibacterial potency (MIC >/= 0.031 microg/mL) against a broad range of Gram-positive pathogens, including drug-resistant strains such as methicillin-resistant Stapylococcus aureus (MRSA), penicillin-resistant Streptococcus pneumoniae (PRSP), and vancomycin-resistant Enterococcus faecalis (VRE). In a first proof-of-concept study, a selected compound (35) showed in vivo efficacy in a mouse peritonitis model against methicillin-sensitive S. aureus infection with an ED(50) value of 30 mg/kg.

Allosteric inhibition of protein--DNA complexes by polyamide--intercalator conjugates

The sequence-specific inhibition of essential protein-DNA contacts in the promoter of a gene is a central issue for the regulation of gene expression by chemical methods. Hairpin polyamides have been shown to inhibit protein-DNA complexes in some but not all cases. For example, polyamides co-occupy the same DNA sequence in the minor groove in the presence of major-groove binding bZip proteins. Four hairpin polyamide-acridine conjugates were synthesized and shown to bind the minor groove of DNA with high affinity in a sequence-specific manner. The polyamide-acridine conjugates were shown to unwind DNA (phi = 14-15 degrees), evidence for intercalation by the acridine moiety. Importantly, the polyamide-intercalator conjugates, which combine sequence-specific groove binding with proximal local unwinding, inhibit major-groove DNA binding by the GCN4 bZip protein. This class of DNA binding molecules creates a sequence-specific allosteric change in DNA structure and has the potential to be a general inhibitor of transcription factor binding independent of the specific protein-DNA structure.

Sequence-selective DNA cleavage by a chimeric metallopeptide

A chimeric metallopeptide derived from the sequences of two structurally superimposable motifs was designed as an artificial nuclease. Both DNA recognition and nuclease activity have been incorporated into a small peptide sequence. P3W, a 33-mer peptide comprising helices alpha2 and alpha3 from the engrailed homeodomain and the consensus EF-hand Ca-binding loop binds one equivalent of lanthanides or calcium and folds upon metal binding. The conditional formation constants (in the presence of 50 mM Tris) of P3W for Eu(III) (K(a) = (2.1 +/- 0.1) x 10(5) M(-1)) and Ce(IV) (K(a) = (2.6 +/- 0.1) x 10(5) M(-1)) are typical of isolated EF-hand peptides. Circular dichroism studies show that 1:1 CeP3W is 26% alpha-helical and EuP3W is up to 40% alpha-helical in the presence of excess metal. The predicted helicity of the folded peptide based on helix length and end effects is about 50%, showing the metallopeptides are significantly folded. EuP3W has considerably more secondary structure than our previously reported chimeras (Welch, J. T.; Sirish, M.; Lindstrom, K. M.; Franklin, S. J. Inorg. Chem. 2001, 40, 1982-1984). Eu(III)P3W and Ce(IV)P3W nick supercoiled DNA at pH 6.9, although EuP3W is more active at pH 8. CeP3W cleaves linearized, duplex DNA as well as supercoiled plasmid. The cleavage of a 5'-(32)P-labeled 121-mer DNA fragment was followed by polyacrylamide gel electrophoresis. The cleavage products are 3'-OPO(3) termini exclusively, suggesting a regioselective or multistep mechanism. In contrast, uncomplexed Ce(IV) and Eu(III) ions produce both 3'-OPO(3) and 3'-OH, and no evidence of 4'-oxidative cleavage termini with either metal. The complementary 3'-(32)P-labeled oligonucleotide experiment also showed both 5'-OPO(3) and 5'-OH termini were produced by the free ions, whereas CeP3W produces only 5'-OPO(3) termini. In addition to apparent regioselectivity, the metallopeptides cut DNA with modest sequence discrimination, which suggests that the HTH motif binds DNA as a folded domain and thus cleaves selected sequences. The de novo artificial nuclease LnP3W represents the first small, underivatized peptide that is both active as a nuclease and sequence selective.

Imidazopyridine/Pyrrole and hydroxybenzimidazole/pyrrole pairs for DNA minor groove recognition

The DNA binding properties of fused heterocycles imidazo[4,5-b]pyridine (Ip) and hydroxybenzimidazole (Hz) paired with pyrrole (Py) in eight-ring hairpin polyamides are reported. The recognition profile of Ip/Py and Hz/Py pairs were compared to the five-membered ring pairs Im/Py and Hp/Py on a DNA restriction fragment at four 6-base pair recognition sites which vary at a single position 5'-TGTNTA-3', where N = G, C, T, A. The Ip/Py pair distinguishes G.C from C.G, T.A, and A.T, and the Hz/Py pair distinguishes T.A from A.T, G.C, and C.G, affording a new set of heterocycle pairs to target the four Watson-Crick base pairs in the minor groove of DNA.

Pausing of DNA polymerases on duplex DNA templates due to ligand binding in vitro

Using the recently developed peptide nucleic acid (PNA)-assisted assay, which makes it possible to extend a primer on duplex DNA, we study the sequence-specific inhibition of the DNA polymerase movement along double-stranded DNA templates imposed by DNA-binding ligands. To this end, a plasmid vector has been prepared featuring the polylinker with two flanking priming sites to bi-directionally initiate the primer-extension reactions towards each other. Within this plasmid, we have cloned a set of random DNA sequences and analyzed the products of these reactions with several phage and bacterial DNA polymerases capable of strand-displacement synthesis. Two of them, ø29 and modified T7 (Sequenase 2.0) enzymes, were found to be most potent for primer extension in the presence of DNA-binding ligands. We used these enzymes for a detailed study of ligand-induced pausing effects with four ligands differing in modes of binding to the DNA double-helix. GC-specific intercalator actinomycin D and three minor groove-binders, chromomycin A(3) (GC-specific), distamycin A and netropsin (both AT-specific), have been chosen. In the presence of each ligand both selected DNA polymerases experienced multiple clear-cut pauses. Each ligand yielded its own characteristic pausing pattern for a particular DNA sequence. The majority of pausing sites could be located with a single-nucleotide resolution and corresponded to the preferred binding sites known from the literature for the ligands under study. Besides, DNA polymerases stalled exactly at the positions occupied by PNA oligomers that were employed to initiate the primer extension. These findings provide an important insight into the DNA polymerase performance. In addition, the high-resolution ligand-induced pausing patterns we obtained for the first time for DNA polymerase elongation on duplex DNA may become a valuable addition to the existing arsenal of methods used to monitor duplex DNA interactions with various DNA-binding ligands, including drugs.

Sequence specific fluorescence detection of double strand DNA

Methods for the fluorescent detection of specific sequences of double strand DNA in homogeneous solution may be useful in the field of human genetics. A series of hairpin polyamides with tetramethyl rhodamine (TMR) attached to an internal pyrrole ring were synthesized, and the fluorescence properties of the polyamide-fluorophore conjugates in the presence and absence of duplex DNA were examined. We observe weak TMR fluorescence in the absence of DNA. Addition of >/=1:1 match DNA affords a significant fluorescence increase over equimolar mismatch DNA for each polyamide-TMR conjugate. Polyamide-fluorophore conjugates offer a new class of sensors for the detection of specific DNA sequences without the need for denaturation. The polyamide-dye fluorescence-based method can be used to screen in parallel the interactions between aromatic ring pairs and the minor groove of DNA even when the binding site contains a non-Watson-Crick DNA base pair. A ranking of the specificity of three polyamide ring pairs-Py/Py, Im/Py, and Im/Im-was established for all 16 possible base pairs of A, T, G, and C in the minor groove. We find that Im/Im is an energetically favorable ring pair for minor groove recognition of the T.G base pair.