Inhibition of major groove DNA binding bZIP proteins by positive patch polyamides

https://doi.org/10.1016/S0968-0896(01)00122-5Get rights and content

Abstract

Cell permeable synthetic ligands that bind to predetermined DNA sequences offer a chemical approach to gene regulation, provided inhibition of a broad range of DNA transcription factors can be achieved. DNA minor groove binding polyamides containing aminoalkyl substituents at the N-1 of a single pyrrole residue display inhibitory effects for a bZIP protein which binds exclusively in the DNA major groove. For major groove protein inhibition, specific protein-DNA contacts along the phosphate backbone were targeted with the positively charged dimethylamino substituent on the backbone of a minor groove binding polyamide hairpin. Remarkably, these polyamides bind DNA with enhanced affinity and uncompromised specificity when compared to polyamides with the aminoalkyl moiety at the C-terminus. By adding bZIP transcription factors to the class of protein–DNA complexes that can be disrupted by minor groove binding ligands, these results may increase the functional utility of polyamides as regulators of gene expression.

Introduction

Hairpin polyamides containing pyrrole (Py), imidazole (Im) and hydroxypyrrole (Hp) amino acids are synthetic ligands that bind predetermined DNA sequences in the minor groove with affinities and specificities comparable to many DNA binding proteins.1, 2 Rules have been developed that allow for sequence specific recognition of the DNA minor groove by relating each Watson–Crick base pair with a particular pairing of the aromatic Py, Im and Hp rings.1, 2, 3, 4 The crescent shaped polyamides bind in the minor groove of DNA with pairs of aromatic rings stacked against each other and the walls of the groove, allowing the backbone amide hydrogens and the substituents at the 3-position of the Py, Im and Hp residues to make specific contacts with the edges of the intact base pairs. A γ-aminobutyric acid residue (γ) connects the polyamide subunits in a ‘hairpin’ motif which enhances affinity and unambiguously locks the desired ring pairings in register.

A requirement for the general application of minor groove binding polyamides as regulators of gene transcription is motifs that can effectively inhibit all classes of DNA binding proteins, especially transcription factors which bind regulatory elements in gene promoter regions. Polyamides have been found to interfere with protein-DNA recognition in cases where contacts in the minor groove are important for protein-DNA binding affinity.2 Polyamides targeted to minor groove contacts of transcription factors Ets-1, LEF-1, and TBP inhibited DNA-binding of the each protein in mobility gel shift assays.5 In contrast, polyamides have been shown to bind simultaneously with a some proteins that exclusively occupy the DNA major groove,6, 7, 8 such as the bZIP protein, GCN4. The ubiquity of major groove contacts in protein-DNA recognition provides the impetus to develop approaches for the inhibition of major groove proteins by DNA minor groove binding polyamides.

Hairpin polyamides with the tripeptide Arg-Pro-Arg at the C-terminus have been shown to inhibit GCN4, potentially by competing with the protein side chains for electrostatic contacts to the DNA phosphate backbone.8 However, Arg-Pro-Arg modified polyamides may be limited by the use of α-amino acids with regard to biostability. This lead us to ask the question of whether a N-aminoalkylpyrrole residue could serve as a simple nonpeptide substitute for the Arg-Pro-Arg tripeptide positive patch that can be placed at any position along the polyamide backbone which would allow for discrete targeting of phosphodiester protein contacts.

Analysis of an X-ray crystal structure of a polyamide bound to its cognate DNA site reveals that the N-1 substituent of each pyrrole ring makes its closest phosphate contact two phosphates to the 3′ side of the base the ring pair recognizes, consistent with the minor groove binding nature of the polyamide.9 Polyamides displaying a N-aminoalkyl positive patch could interfere with protein binding to this specific phosphate and consequently inhibit specific contacts with DNA. The increasing availability of protein-DNA crystal structures provide the necessary information to design a polyamide that targets important protein-DNA contacts. Previous attempts to inhibit major groove binding transcription factors with minor groove binding ligands have used polyamine substituents with 10 charges attached to a tripyrrole, likely resulting in a ligand with significantly reduced sequence specificity.8, 10

We report here the synthesis and DNA binding properties of a series of hairpin polyamides containing N-aminoalkyl substitutions at a single pyrrole residue and their ability to inhibit DNA binding by a major groove binding protein, GCN4 (222-281), a member of the bZIP family of transcriptional activators. The X-ray crystal structure of the GCN4–DNA complex reveals several electrostatic contacts between positively charged side chains and the DNA phosphate backbone.11 Of these, ethylation interference experiments suggest that the Lys246-DNA phosphate interaction is particularly important for DNA binding by GCN4.12 Polyamides were designed to place a competing positive charge at the Lys246-DNA phosphate position to disrupt this potentially crucial interaction. A series of polyamides were synthesized that contain multiple amines and selected ether substitutions as controls at the N-1 position of the pyrrole. The DNA binding affinity and specificity of each of these compounds for their target sites was evaluated using quantitative DNase I footprinting and compared to analogues containing the traditional N-methyl substituents. Gel mobility shift assays were employed to investigate the ability of these polyamides to inhibit DNA binding by a major groove binding protein.

Section snippets

Polyamide synthesis

The synthesis of N-aminoalkylpyrrole-containing polyamides Figure 1, Figure 3, Figure 5, Figure 6 (Fig. 1) required the preparation of a new pyrrole monomer suitable for solid phase synthesis and subsequent modification at the N-1 position. (3-Hydroxypropyl)-4-[(tert-butoxycarbonyl)amino]-pyrrole-2-carboxylic acid, (Scheme 1, Figure 2), introduces a 3-hydroxypropyl moiety at the N-1 of pyrrole which can be modified by sulfonylation followed by nucleophilic displacement to afford a variety of

A Py(Dp) residue increases polyamide affinity without compromising specificity

Generally, Py/Im polyamides are prepared using N-methyl aromatic amino acids and a dimethyalaminopropylamide tail on the C-terminus as in Figure 1, Figure 6 and 2. Exchanging the placement of the Dp and N-Me to provide a polyamide with a N-(N′, N′-dimethylaminopropyl)pyrrole residue and a C-terminal N-methyl amide (Figure 1, Figure 5, Figure 6 and Figure 1, Figure 3) affords an isomer of identical molecular weight and composition; however, at least in this case, affinity for the DNA target site

Conclusions

A new class of hairpin polyamides containing a positively charged alkyl amine at the N-1 of a pyrrole ring have been shown to sequence specifically recognize DNA with increased affinity and uncompromised specificity relative to analogues with the amine at the C-terminus. The positive patch provided by a N-diaminoalkylpyrrole residue proved reasonably effective in the inhibition of DNA binding by the major groove binding protein GCN4 (222–281), potentially by competing with the protein side

Experimental

All synthetic reagents were as previously described or obtained from Aldrich.15 Analytical HPLC was performed on a Beckman Gold Nouveau system with a model 126 pump and model 168 diode array detector. A Rainen C18, Microsorb MV, 5 μm, 300×4.6 mm reverse phase column was employed with 0.1% (w/v) TFA:H2O and 1.5% acetonitrile/min. Preparatory HPLC was performed on a Waters DeltaPak 25×100 mm 100 μm C18 column in 0.1% (w/v) TFA, gradient elution 0.25%/min. CH3CN. Resin substitution of synthesized

Acknowledgements

The authors are grateful to the National Institutes of Health for research support, the National Science Foundation for a predoctoral fellowship to R.E.B., Bristol-Myers Squibb and the Ralph M. Parson Foundation for predoctoral fellowships to R.E.B. and N.R.W., and the National Institutes of Health for a research service award to J.W.S. We thank G.M. Hathaway and the Caltech Protein/Peptide Microanalytical Laboratory for MALDI-TOF mass spectrometry.

References (24)

  • P.B. Dervan et al.

    Curr. Opin. Chem. Biol.

    (1999)
  • S. White et al.

    Chem. Biol.

    (1997)
  • L. Neely et al.

    J. Mol. Biol.

    (1997)
  • R.E. Bremer et al.

    Chem. Biol.

    (1998)
  • T.C. Bruice et al.

    Bioorg. Med. Chem.

    (1997)
  • T.E. Ellenberger et al.

    Cell

    (1992)
  • R.E. Bremer et al.

    Bioorg. Med. Chem.

    (2000)
  • I.A. Hope et al.

    Cell

    (1986)
  • P. Konig et al.

    J. Mol. Biol.

    (1993)
  • A.M. Maxam et al.

    Methods Enzymol.

    (1980)
  • J.M. Gottesfeld et al.

    Gene Expression

    (2000)
  • S. White et al.

    Nature

    (1998)
  • Cited by (25)

    • The intercalation of DNA double helices with doxorubicin and nagalomycin

      2007, Journal of Molecular Graphics and Modelling
    • Solid-phase synthesis of DNA binding polyamides on oxime resin

      2002, Bioorganic and Medicinal Chemistry
      Citation Excerpt :

      The change from β-Dp (1) to propanol amide (2) results in similar affinity and specificity. This retention of affinity upon removal of the positively charged Dp moiety has previously been observed.16,17 Presumably, the steric clash of propanol amide versus β-Dp with the exocyclic NH2 of G at N-1 is similar.

    • Principles of Nucleic Acid Structure

      2021, Principles of Nucleic Acid Structure
    View all citing articles on Scopus
    View full text