Recognition of DNA by designed ligands at subnanomolar concentrations

Selection of peptides that functionally replace a zinc finger in the Sp1 transcription factor by using a yeast combinatorial library

We have developed a strategy for the identification of peptides able to functionally replace a zinc finger domain in a transcription factor. This strategy could have important ramifications for basic research on gene regulation and for the development of therapeutic agents. In this study in yeast, we expressed chimeric proteins that included a random peptide combinatorial library in association with two zinc finger domains and a transactivating domain. The library was screened for chimeric proteins capable of activating transcription from a target sequence in the upstream regulatory regions of selectable or reporter genes. In a screen of approximately 1.5 x 10(7) transformants we identified 30 chimeric proteins that exhibited transcriptional activation, some of which were able to discriminate between wild-type and mutant DNA targets. Chimeric library proteins expressed as glutathione S-transferase fusions bound to double-stranded oligonucleotides containing the target sequence, suggesting that the chimeras bind directly to DNA. Surprisingly, none of the peptides identified resembled a zinc finger or other well-known transcription factor DNA binding domain.

Can drugs be designed?

Many examples are now emerging of the successful use of rational, structure-based methods in drug discovery. Of particular note is the development of imaginative NMR-based methods for rapid routes to ligand design. Our understanding of the chemistry underlying protein-ligand interactions, however, remains relatively poor and a major limitation in our ability to truly design drugs.

New DNA minor-groove binding molecules with high sequence-selectivities and binding affinities

New DNA minor-groove binding molecules with high binding affinities and sequence-selectivities are described. The effects of structural changes in ligands with a three-ring backbone on their DNA-binding properties have been studied. Of a pool of eight potential ligands, two showed binding affinities and sequence-selectivities rivaling distamycin. The two best ligands share a common structural motif that involves a para-di-substituted benzene ring flanked by two meta-di-substituted benzene rings.

Defining GC-specificity in the minor groove: side-by-side binding of the di-imidazole lexitropsin to C-A-T-G-G-C-C-A-T-G

BACKGROUND

Polyamide drugs, such as netropsin, distamycin and their lexitropsin derivatives, can be inserted into a narrow B-DNA minor groove to form 1:1 complexes that can distinguish AT base pairs from GC, but cannot detect end-for-end base-pair reversals such as TA for AT. In contrast, 2:1 side-by-side polyamide drug complexes potentially are capable of such discrimination. Imidazole (Im) and pyrrole (Py) rings side-by-side read a GC base pair with the Im ring recognizing the guanine side. But the reason for this specific G-Im association is unclear because the guanine NH2 group sits in the center of the groove. A 2:1 drug:DNA complex that presents Im at both ends of a GC base pair should help unscramble the issue of imidazole reading specificity.

RESULTS

We have determined the crystal structure of a 2:1 complex of a di-imidazole lexitropsin (DIM), an analogue of distamycin, and a DNA decamer with the sequence C-A-T-G-G-C-C-A-T-G. The two DIM molecules sit antiparallel to one another in a broad minor groove, with their cationic tails widely separated. Im rings of one drug molecule stack against amide groups of the other. DIM1 rests against nucleotides C7A8T9G10 of strand 1 of the helix, whereas DIM2 rests against G14G15C16C17 on strand 2. All DIM amide nitrogens donate hydrogen bonds to N and O atoms on the floor of the DNA groove and, in addition, the two Im rings on DIM2 accept hydrogen bonds from guanine N2 amines, thereby providing specific reading. The guanine N2 amine can bond to Im on its own side of the groove, but not on the cytosine side, because of limits on close approach of the two Im rings and the geometry of sp2 hybridization about the amide nitrogen.

CONCLUSIONS

Im and Py rings distinguish AT from GC base pairs because of steric factors involving the bulk of the guanine amine, and the ability of Im to form a hydrogen bond with the amine. Side-by-side Im and Py rings differentiate GC from CG base pairs because of tight steric contacts and sp2 hybridization at the amine nitrogen atom, with the favored conformations being G/Im,Py/C and C/Py,Im/G. Discrimination between AT and TA base pairs may be possible using bulkier rings, such as thiazole to select the A end of the base pair.

On the pairing rules for recognition in the minor groove of DNA by pyrrole-imidazole polyamides

BACKGROUND

Cell-permeable small molecules that target predetermined DNA sequences with high affinity and specificity have the potential to control gene expression. A binary code has been developed to correlate DNA sequence with side-by-side pairings between N-methylpyrrole (Py) and N-methylimidazole (Im) carboxamides in the DNA minor groove. We set out to determine the relative energetics of pairings of Im/Py, Py/Im, Im/Im, and Py/Py for targeting G.C and A.T base pairs. A key specificity issue, which has not been previously addressed, is whether an Im/Im pair is energetically equivalent to an Im/Py pair for targeting G.C base pairs.

RESULTS

Equilibrium association constants were determined at two five-base-pair sites for a series of four six-ring hairpin polyamides, in order to test the relative energetics of the four aromatic amino-acid pairings opposite G.C and A.T base pairs in the central position. We observed that a G.C base pair was effectively targeted with Im/Py but not Py/Im, Py/Py, or Im/Im. The A.T base pair was effectively targeted with Py/Py but not Im/Py, Py/Im, or Im/Im.

CONCLUSIONS

An Im/Im pairing is energetically disfavored for the recognition of both A.T and G.C. This specificity will create important limitations on undesirable slipped motifs that are available for unlinked dimers in the minor groove. Baseline energetic parameters will thus be created which, using the predictability of the current pairing rules for specific molecular recognition of double-helical DNA, will guide further second-generation polyamide design for DNA recognition.

Targeting the minor groove of DNA

Small molecules that specifically bind with high affinity to any predetermined DNA sequence in the human genome will be useful tools in molecular biology and, potentially, in human medicine. Pairing rules have been developed to control rationally the sequence specificity of minor groove binding polyamides containing N-methylimidazole and N-methylpyrrole amino acids. Using simple molecular shapes and a two-letter aromatic amino acid code, pyrrole-imidazole polyamides achieve affinities and specificities comparable to DNA-binding proteins.

Estimation of the DNA sequence discriminatory ability of hairpin-linked lexitropsins

Three- and four-ring polyamides containing N-methylimidazole and N-methylpyrrole, and their hairpin-linked derivatives, bind side-by-side in the minor groove of DNA in a sequence-specific manner. The sequences recognized by side-by-side molecules are dependent on the pairings of the polyamide rings to the bases. In this study we report a mathematical model for estimating the free energies of binding for gamma-aminobutyric acid-linked polyamides to 5- and 6-bp DNA sequences. The model parameters are calibrated by a least-squares fit to 35 experimental binding constants. The model performs well in cross-validation experiments and the parameters are consistent with previously proposed empirical rules of polyamide-DNA binding. We apply the model to the design of targeted polyamides, evaluating the ability of the proposed polyamides to bind to a DNA sequence of interest while minimizing binding to the remaining DNA sequences.

Design of polydactyl zinc-finger proteins for unique addressing within complex genomes

Zinc-finger proteins of the Cys2-His2 type represent a class of malleable DNA-binding proteins that may be selected to bind diverse sequences. Typically, zinc-finger proteins containing three zinc-finger domains, like the murine transcription factor Zif268 and the human transcription factor Sp1, bind nine contiguous base pairs. To create a class of proteins that would be generally applicable to target unique sites within complex genomes, we have utilized structure-based modeling to design a polypeptide linker that fuses two three-finger proteins. Two six-fingered proteins were created and demonstrated to bind 18 contiguous bp of DNA in a sequence-specific fashion. Expression of these proteins as fusions to activation or repression domains allows transcription to be specifically up- or down-modulated within human cells. Polydactyl zinc-finger proteins should be broadly applicable as genome-specific transcriptional switches in gene therapy strategies and the development of novel transgenic plants and animals.

Coralyne has a preference for intercalation between TA.T triples in intramolecular DNA triple helices

Intercalating ligands may improve both the stability and sequence specificity of triple helices. Numerous intercalating drugs have been described, including coralyne, which preferentially binds triple helices, though its sequence specificity has been reported to be low [Lee,J.S., Latimer,L.J.P. and Hampel,K.J. (1993) Biochemistry , 32, 5591-5597]. In order to analyse the sequence preferences of coralyne we have used a combination of DNase I footprinting, UV melting, UV-visible spectrophotometry, circular dichroism and NMR spectroscopy to examine defined intermolecular triplexes and intramolecular triplexes linked either by hexaethylene glycol chains or by octandiol chains. DNase I footprinting demonstrated that coralyne has a moderate preference for triplexes over duplexes, but a substantial preference for TA.T triplets compared with CG. C+triplets. The drug was found to have essentially no effect on the melting temperatures of duplexes of the kind d(A)n.d(T)n or d(GA)n.d(TC)n. In contrast, it increased the T m for triplexes of the kind d(T)nd(A)n.dTn, but had little effect on the stability of d(TC)nd(GA).d(CT)n at either low or high pH. On binding to DNA triplexes, there is a large change in the absorption spectrum of coralyne and also a substantial fluorescence quenching that can be attributed to intercalation. The changes in the optical spectra have been used for direct titration with DNA. For triplexes d(T)6d(A)6.d(T)6, the Kd at 298 K was 0.5-0.8 microM. In contrast, the affinity for d(TC) nd(GA)n.d(CT)n triplexes was 6- to 10-fold lower and was characterized by smaller changes in the absorption and CD spectra. This indicates a preference for intercalation between TAT triples over CG.C+/TA.T triples. NMR studies confirmed interaction by intercalation. However, a single, secondary binding was observed at high concentrations of ligand to the triplex d(AGAAGA-L-TCTTCT-L-TCTTCT), presumably owing to the relatively low difference in affinity between the TA.T site and the competing, neighbouring sites.

Structure-based design of a new bisintercalating anthracycline antibiotic

A new bisintercalating anthracycline antibiotic, WP631, has been designed and synthesized. The rational design of the new compound was based upon the geometry of monomeric anthracyclines bound to DNA oligonucleotides observed in high-resolution crystal structures. Monomeric units of daunorubicin have been linked through their reactive 3' NH2 substituents on the daunosamine moieties to form the new bisanthracycline WP631. Viscosity studies confirmed that WP631 binds to DNA by bisintercalation. Differential scanning calorimetry and UV melting experiments were used to measure the ultratight binding of WP631 to DNA. The binding constant for the interaction of WP631 with herring sperm DNA was determined to be 2.7 x 10(11) M-1 at 20 degrees C. The large, favorable binding free energy of -15.3 kcal mol-1 was found to result from a large, negative enthalpic contribution of -30.2 kcal mol-1. A molecular model was generated that shows the favorable stereochemical fit of the linker in the DNA minor groove. The cytotoxicity of WP631 was compared to that of doxorubicin using MCF-7-sensitive and MCF-7/VP-16 MRP-mediated multidrug-resistant cell lines. These initial studies showed that while WP631 is slightly less cytotoxic than doxorubicin in the sensitive cell line, it appears to overcome MRP-mediated multidrug resistance and was much more cytotoxic against the MCF-7/VP-16 cell line than was doxorubicin. The design of new potential anticancer agents based on known structural principles was found to produce a compound with significantly increased DNA binding affinity and with interesting biological activity.

Effects of the A.T/T.A degeneracy of pyrrole--imidazole polyamide recognition in the minor groove of DNA

Pairing rules have been developed to predict the sequence specificity of minor groove binding polyamides containing pyrrole (Py) and imidazole (Im) amino acids. An Im/Py pair distinguishes G.C from C.G and both of these from A.T/T.A base pairs. A Py/Py pair appears not to distinguish A.T from T.A base pairs. To test the extent of this degeneracy, the affinity and binding orientation of the hairpin polyamide ImPyPy-gamma-PyPyPy-beta-Dp were measured for eight possible five base pair 5'-TG(A,T)(3)-3' match sites. Affinity cleavage experiments using a polyamide with an EDTA.Fe(II) moiety at the carboxy terminus, ImPyPy-gamma-PyPyPy-beta-Dp-EDTA.Fe(II), are consistent with formation of an oriented 1:1 hairpin polyamide complex at all eight 5'-TG(A,T)(3)-3' binding sites [20 mM HEPES, 200 mM NaCl, 50 mg/ml glycogen, pH 7.0, 22 degrees C, 5 mM DTT, 1 mM Fe(II)]. Quantitative DNase I footprint titration experiments reveal that ImPyPy-gamma-PyPyPy-beta-Dp binds all eight 5'-TG(A,T)(3)-3' target sites with only a 12-fold difference in the equilibrium association constants between the strongest site, 5'-TGTTT-3' (Ka = 2.1 x 10(8) M-1), and the weakest site, 5'-TGAAT-3' (Ka = 1.8 x 10(7) M-1) (10 mM Tris.HCl, 10 mM KCl, 10 mM MgCl2, 5 mM CaCl2, pH 7.0, 22 degrees C). This relatively small range indicates that the Py/Py pair is approximately degenerate for recognition of A,T base pairs, affording generality with regard to targeting sequences of mixed A.T/T.A composition.