Synthesis of nonlinear DNA-binding peptide with binding specificity determinants close to those of 434 Cro-repressor

Design, synthesis and DNA binding activity of a nonlinear 102 residue peptide are reported. The peptide contains four sequence-specific DNA binding domains of 434 Cro protein. These four domains were linked covalently to a symmetrical carboxyterminal crosslinker that contains four arms each ending with an aliphatic aminogroup. From CD studies we have found that in aqueous buffer in the presence of 20% trifluoroethanol the peptide residues assume alpha helical, beta-sheet and random coiled conformations with an alpha helical content of about 16% at room temperature. The alpha helicity is increased up to 40% in the presence of 40% trifluoroethanol. Upon complex formation between the peptide and DNA a change in the peptide conformation takes place which is consistent with an alpha-beta transition in the DNA binding, helix-turn-helix motif of 434 Cro repressor. Evidently residues present in helices alpha(2) and alpha(3) form a beta hairpin which is inserted in the minor DNA groove. The latter inference is supported by our observations that the peptide can displace minor groove binding antibiotic distamycin A from a complex with poly(dA).poly(dT). As revealed from DNase protection studies the peptide exhibits preferences for binding to operator and pseudooperator sites recognized by 434 Cro repressor. It binds strongly to operator sites OR1, OR2 and OR3 and exhibits a greater affinity for pseudooperator site Op1. From analysis of nucleotide sequences in the strong affinity binding sites for the peptide on DNA a conclusion is drawn that it binds to pseudosymmetrical nucleotide sequences 5'-ACAA(W)nCTGT-3', where W is an arbitrary nucleotide. n is equal to six or seven. In the strongest affinity binding site for the peptide on DNA (Op1) motif 5'-ACAA-3' is replaced by sequence 5'-ACCA-3'. A difference in binding specificity shown by the peptide and 434 Cro protein could be attributed to a flexibility of the connecting chains between DNA-binding domains in the peptide molecule as well as to a replacement of Thr - Ala in the alpha 2 helix. Removal of two residues from the N-terminal end of helix alpha 2 in each of the four DNA binding domains of 434 Cro present in the peptide leads to a loss of binding specificity, although the modified peptide binds to DNA unspecifically.     

Cell selectivity and mechanism of action of antimicrobial model peptides containing peptoid residues

To develop a useful method for designing cell-selective antimicrobial peptides and to investigate the effect of incorporating peptoid residues into an alpha-helical model peptide on structure, function, and mode of action, we synthesized a series of model peptides incorporating Nala (Ala-peptoid) into different positions of an amphipathic alpha-helical model peptide (KLW). Incorporation of one or two Nala residues into the hydrophobic helix face of KLW was more effective at disrupting the alpha-helical structure and bacterial cell selectivity than incorporation into the hydrophilic helix face or hydrophobic/hydrophilic interface. Tryptophan fluorescence studies of peptide interaction with model membranes indicated that the cell selectivity of KLW-L9-a and KLW-L9,13-a is closely correlated with their selective interactions with negatively charged phospholipids. KLW-L9,13-a, which has two Nala residues in its hydrophobic helix face, showed a random structure in membrane-mimicking conditions. KLW-L9,13-a exhibited the highest selectivity toward bacterial cells, showing no hemolytic activity and no or less cytotoxicity compared with other peptides against four mammalian cell lines. Unlike other model peptides, KLW-L9,13-a caused no or little membrane depolarization in Staphylococcus aureus or lipid flip-flop in negatively charged vesicles. In addition, KLW-L9,13-a caused very little fluorescent dye leakage from negatively charged vesicles. Furthermore, confocal laser-scanning microscopy and DNA-binding assays showed that KLW-L9,13-a probably exerts its antibacterial action by penetrating the bacterial membrane and binding to cytoplasmic compounds (e.g., DNA), resulting in cell death. Collectively, our results demonstrate that the incorporation of two Nala residues into the central position of the hydrophobic helix face of noncell-selective alpha-helical peptides is a promising strategy for the rational design of intracellular, cell-selective antimicrobial peptides.     

Interactions of the 26-39 fragment of the cro protein from lambda bacteriophage with nucleic acids

A tetradecapeptide with a sequence identical to residues 26-39 of the cro protein from bacteriophage lambda has been synthesized. This peptide has no secondary structure in an aqueous buffer but adopts an alpha-helical conformation in the presence of 20% hexafluoroisopropanol. The fluorescence of the single tyrosyl residue of the cro protein fragment is quenched upon binding to nucleic acids. Proton magnetic resonance has been used to investigate complex formation of the cro protein fragment with a self-complementary decadeoxynucleotide d(AATTGCAATT). Changes in resonance positions and linewidths have been observed for both partners in the 4 complexes which are obtained when either the single-stranded or double-stranded oligonucleotide is mixed with either the random coil or the alpha-helical peptide. These studies are presently extended to the specific complex formed by the cro protein fragment with the OR3 operator sequence.     

Recognition of DNA by single-chain derivatives of the phage 434 repressor: high affinity binding depends on both the contacted and non-contacted base pairs.

Single-chain derivatives of the phage 434 repressor, termed single-chain repressors, contain covalently dimerized DNA-binding domains (DBD) which are connected with a peptide linker in a head-to-tail arrangement. The prototype RR69 contains two wild-type DBDs, while RR*69 contains a wild-type and an engineered DBD. In this latter domain, the DNA- contacting amino acids of thealpha3 helix of the 434 repressor are replaced by the corresponding residues of the related P22 repressor. We have used binding site selection, targeted mutagenesis and binding affinity studies to define the optimum DNA recognition sequence for these single-chain proteins. It is shown that RR69 recognizes DNA sequences containing the consensus boxes of the 434 operators in a palindromic arrangement, and that RR*69 optimally binds to non-palindromic sequences containing a 434 operator box and a TTAA box of which the latter is present in most P22 operators. The spacing of these boxes, as in the 434 operators, is 6 bp. The DNA-binding of both single-chain repressors, similar to that of the 434 repressor, is influenced indirectly by the sequence of the non-contacted, spacer region. Thus, high affinity binding is dependent on both direct and indirect recognition. Nonetheless, the single-chain framework can accommodate certain substitutions to obtain altered DNA-binding specificity and RR*69 represents an example for the combination of altered direct and unchanged indirect readout mechanisms.     

Design and synthesis of peptides capable of specific binding to DNA

In the present communication, design, synthesis and DNA binding activities of the following two peptides are reported: Dns-Gly-Ala-Gln-Lys-Leu-Ala-Cly-Lys-Val-Gly-Thr-Lys-Val-Lys-Val-Gl y-Thr-Lys-Thr - Val-OH (I) and [(H-Ala-Lys-Leu-Ala-Thr-Lys-Ala-Gly-Val-Lys-Gln-Gln-Ser-Ile-Gln-Leu-Ile- Thr- Ala-Aca-Lys-Aca)2Lys-Aca]2Lys-Val-OH (II), where Aca = NH(CH2)5CO--; Dns is a residue of 5-dimethylaminonaphtalene-1-sulfonic acid. Peptide I contains a large fraction (ca.30%) of valyl and threonyl residues, which possess a high potential for beta structure formation. Peptide II contains four repeats of the amino acid sequence present in the presumed DNA binding helix-turn-helix unit of 434 Cro repressor. These four domains are linked in such a way that two domains can interact with two halves a 14 base pair long operator site on DNA. From CD studies we have found that peptide I is in a random coil conformation in the aqueous solution in the presence of 20% trifluoroethanol. By contrast, amino acid residues of peptide II assume alpha helical, beta and random coiled conformations under the same conditions. A change in the secondary structure of the two peptides upon binding to DNA is observed. The difference CD spectra obtained by subtracting the spectra of free DNA from the spectra of peptide I--DNA complexes gives rise to a beta-like pattern. The difference CD spectra obtained for complexes of peptide II with various natural and synthetic DNAs suggest that alpha-beta-transition takes place in the presumed helix-turn-helix repeat units of peptide II upon binding to DNA. Peptide I binds more strongly to poly(dG).poly(dC) than to poly(dA).poly(dT) and poly[d(GC)].poly[d(GC)]. The binding takes place in the minor DNA groove because minor groove binding antibiotic sibiromycin can displace peptide I from a complex with poly(dG).poly(dC). Analysis of footprinting diagramms shows that peptide I specifically protects phosphodiester bonds within operator sites OR1 and OR2 of phage lambda from nuclease cleavage. By contrast, peptide II does not react specifically with operators OR1, OR2 and OR3 of phage 434 although it forms very tight complexes with DNA which are stable in the presence of 1M NH4F.     

A helix-turn motif in the C-terminal domain of histone H1

The structural study of peptides belonging to the terminal domains of histone H1 can be considered as a step toward the understanding of the function of H1 in chromatin. The conformational properties of the peptide Ac-EPKRSVAFKKTKKEVKKVATPKK (CH-1), which belongs to the C-terminal domain of histone H1(o) (residues 99-121) and is adjacent to the central globular domain of the protein, were examined by means of 1H-NMR and circular dichroism. In aqueous solution, CH-1 behaved as a mainly unstructured peptide, although turn-like conformations in rapid equilibrium with the unfolded state could be present. Addition of trifluoroethanol resulted in a substantial increase of the helical content. The helical limits, as indicated by (i,i + 3) nuclear Overhauser effect (NOE) cross correlations and significant up-field conformational shifts of the C(alpha) protons, span from Pro100 to Val116, with Glu99 and Ala117 as N- and C-caps. A structure calculation performed on the basis of distance constraints derived from NOE cross peaks in 90% trifluoroethanol confirmed the helical structure of this region. The helical region has a marked amphipathic character, due to the location of all positively charged residues on one face of the helix and all the hydrophobic residues on the opposite face. The peptide has a TPKK motif at the C-terminus, following the alpha-helical region. The observed NOE connectivities suggest that the TPKK sequence adopts a type (I) beta-turn conformation, a sigma-turn conformation or a combination of both, in fast equilibrium with unfolded states. Sequences of the kind (S/T)P(K/R)(K/R) have been proposed as DNA binding motifs. The CH-1 peptide, thus, combines a positively charged amphipathic helix and a turn as potential DNA-binding motifs.     

Substituting an alpha-helix switches the sequence-specific DNA interactions of a repressor

It has been suggested that many DNA-binding proteins use an alpha-helix for specific sequence recognition. We have used amino acid sequence homologies to identify the presumptive DNA-recognition helices in two related proteins whose structures are unknown--the repressor and cro protein of bacteriophage 434. The 434 repressor and cro protein each bind to three similar sites in the rightward phage 434 operator, OR, and they make different contacts in each binding site, as revealed by the chemical probe dimethyl sulfate. We substituted the putative recognition alpha-helix of 434 repressor with the putative recognition alpha-helix of 434 cro protein to create a hybrid protein named repressor*. The specific DNA contacts made by repressor* are like those of 434 cro protein.     

Two peptide fragments G55-I72 and K97-A109 from staphylococcal nuclease exhibit different behaviors in conformational preferences for helix formation

Two synthetic peptides, SNasealpha1 and SNasealpha2, corresponding to residues G55-I72 and K97-A109, respectively, of staphylococcal nuclease (SNase), are adopted for detecting the role of helix alpha1 (E57-A69) and helix alpha2 (M98-Q106) in the initiation of folding of SNase. The helix-forming tendencies of the two SNase peptide fragments are investigated using circular dichroism (CD) and two-dimensional (2D) nuclear magnetic resonance (NMR) methods in water and 40% trifluoroethanol (TFE) solutions. The coil-helix conformational transitions of the two peptides in the TFE-H2O mixture are different from each other. SNasealpha1 adopts a low population of localized helical conformation in water, and shows a gradual transition to helical conformation with increasing concentrations of TFE. SNasealpha2 is essentially unstructured in water, but undergoes a cooperative transition to a predominantly helical conformation at high TFE concentrations. Using the NMR data obtained in the presence of 40% TFE, an ensemble of alpha-helical structures has been calculated for both peptides in the absence of tertiary interactions. Analysis of all the experimental data available indicates that formation of ordered alpha-helical structures in the segments E57-A69 and M98-Q106 of SNase may require nonlocal interactions through transient contact with hydrophobic residues in other parts of the protein to stabilize the helical conformations in the folding. The folding of helix alpha1 is supposed to be effective in initiating protein folding. The formation of helix alpha2 depends strongly on the hydrophobic environment created in the protein folding, and is more important in the stabilization of the tertiary conformation of SNase.     

The role of helix stabilizing residues in GCN4 basic region folding and DNA binding

Basic region leucine zipper (bZip) proteins contain a bipartite DNA-binding motif consisting of a coiled-coil leucine zipper dimerization domain and a highly charged basic region that directly contacts DNA. The basic region is largely unfolded in the absence of DNA, but adopts a helical conformation upon DNA binding. Although a coil --> helix transition is entropically unfavorable, this conformational change positions the DNA-binding residues appropriately for sequence-specific interactions with DNA. The N-terminal residues of the GCN4 DNA-binding domain, DPAAL, make no DNA contacts and are not part of the conserved basic region, but are nonetheless important for DNA binding. Asp and Pro are often found at the N-termini of alpha-helices, and such N-capping motifs can stabilize alpha-helical structure. In the present study, we investigate whether these two residues serve to stabilize a helical conformation in the GCN4 basic region, lowering the energetic cost for DNA binding. Our results suggest that the presence of these residues contributes significantly to helical structure and to the DNA-binding ability of the basic region in the absence of the leucine zipper. Similar helix-capping motifs are found in approximately half of all bZip domains, and the implications of these findings for in vivo protein function are discussed.     

Jararhagin-derived RKKH peptides induce structural changes in alpha1I domain of human integrin alpha1beta1

Integrin alpha(1)beta(1) is one of four collagen-binding integrins in humans. Collagens bind to the alphaI domain and in the case of alpha(2)I collagen binding is competitively inhibited by peptides containing the RKKH sequence and derived from the metalloproteinase jararhagin of snake venom from Bothrops jararaca. In alpha(2)I, these peptides bind near the metal ion-dependent adhesion site (MIDAS), where a collagen (I)-like peptide is known to bind; magnesium is required for binding. Published structures of the ligand-bound "open" conformation of alpha(2)I differs significantly from the "closed" conformation seen in the structure of apo-alpha(2)I near MIDAS. Here we show that two peptides, CTRKKHDC and CARKKHDC, derived from jararhagin also bind to alpha(1)I and competitively inhibit collagen I binding. Furthermore, calorimetric and fluorimetric measurements show that the structure of the complex of alpha(1)I with Mg(2+) and CTRKKHDC differs from structure in the absence of peptide. A comparison of the x-ray structure of apo-alpha(1)I ("closed" conformation) and a model structure of the alpha(1)I ("open" conformation) based on the closely related structure of alpha(2)I reveals that the binding site is partially blocked to ligands by Glu(255) and Tyr(285) in the "closed" structure, whereas in the "open" structure helix C is unwound and these residues are shifted, and the "RKKH" peptides fit well when docked. The "open" conformation of alpha(2)I resulting from binding a collagen (I)-like peptide leads to exposure of hydrophobic surface, also seen in the model of alpha(1)I and shown experimentally for alpha(1)I using a fluorescent hydrophobic probe.     

An alpha-helical peptide model for electrostatic interactions of proteins with DNA. The N terminus of RecA

A series of synthetic peptides have been studied as models for non-specific protein-DNA interactions. In an alpha-helical conformation, the charged amino acid residues of the N-terminal 24 residues of RecA protein are asymmetrically distributed; at neutral pH there is a +4 charge on one face of the helix and a -3 charge on the other face. Modeling suggests that the positive face of the helix can bind five DNA phosphate groups by electrostatic interactions. Circular dichroism (c.d.) spectra indicate that the analogous peptide, Rec24 (AIDENKQKALAAALGQIEKQFGKG-amide), is largely unstructured in water but becomes highly helical in the presence of DNA. Peptide titrations of fluorescent etheno-DNA confirm that the changes in the c.d. spectrum of the peptide are associated with binding, although a dependence of the c.d. signal on the degree of DNA saturation is observed, indicating that peptide can be bound in more than one conformation. At saturation the peptide binds to 5.0(+/- 0.5) DNA phosphate groups as predicted and the electrostatic nature of the binding is confirmed by a strong dependence on salt concentration. A "mutant" peptide where an acidic glutamate residue replaces an alanine on the basic face of the Rec24 helix exhibits weaker binding to single-stranded DNA, also consistent with the electrostatic nature of the proposed peptide-DNA interaction. Extending Rec24 by ten amino acid residues, where the additional residues do not participate in the helical motif, does not noticeably affect binding. Thus, we show experimentally that an asymmetric charge distribution on an alpha-helix can represent an important element for binding nucleic acids.     

Dual DNA recognition codes of a short peptide derived from the basic leucine zipper protein EmBP1

Sequence-specific DNA binding of short peptide dimers derived from a plant basic leucine zipper protein EmBP1 was studied. A homodimer of the EmBP1 basic region peptide recognized a palindromic DNA sequence, and a heterodimer of EmBP1 and GCN4 basic region peptides targets a non-palindromic DNA sequence when a beta-cyclodextrin/adamantane complex is utilized as a dimerization domain. A homodimer of the EmBP1 basic region peptide binds the native EmBP1 binding 5'-GCCACGTGGC-3' and the native GCN4 binding 5'-ATGACGTCAT-3' sequences with almost equal affinity in the alpha-helical conformation, indicating that the basic region of EmBP1 by itself has a dual recognition codes for the DNA sequences. The GCN4 basic region peptide binds 5'-ATGAC-3' in the alpha-helical conformation, but it neither shows affinity nor helix formation with 5'-GCCAC-3'. Because native EmBP1 forms 100 times more stable complex with 5'-GCCACGTGGC-3' over 5'-ATGACGTCAT-3', our results suggest that the sequence-selectivity of native EmBP1 is dictated by the structure of leucine zipper dimerization domain including the hinge region spanning between the basic region and the leucine zipper.     

The effect of C-terminal helix stabilization on specific DNA binding by monomeric GCN4 peptides.

DNA binding by a 29-residue, monomeric, GCN4 basic region peptide, GCN4br, as well as by peptide br-C, a monomeric basic-region analogue that is helix stabilized at its C-terminal end by a Lys25. Asp29 side-chain lactam-bridged alanine-rich sequence, was studied at 25 C in an aqueous buffer containing 100 mm NaCl. Mixing of both peptides with duplex DNA containing the cAMP-responsive element (CRE) was accompanied by significant helix stabilization in the peptides, whereas mixing of the peptides with duplex DNA containing a scrambled CRE site was not. Peptide NBD-br-C was synthesized as a fluorescent probe to evaluate these peptide-DNA interactions further. Quantitative analysis of the fluorescence quenching of peptide NBD-br-C by CRE half-site DNA indicated the formation of a 1:1 complex with a dissociation constant of 1.41 +/- 0.22 microm. Competitive displacement fluorescence assays of CRE half-site binding gave dissociation constants of 0.65 +/- 0.09 microm for peptide br-C and 3.9 +/- 0.5 microM for GCN4br, which corresponds to a free energy difference of 1.1 kcal/mol that is attributed to the helix stabilization achieved in peptide br-C. This result indicates that helix initiation by the alpha-helical leucine zipper dimerization motif in native bzip proteins, such as GCN4, contributes significantly to the affinity of basic region peptides for their recognition sites on DNA. Our fluorescence assay should also prove useful for determining dissociation constants for CRE binding by other GCN4 basic region analogues under equilibrium conditions and physiological salt concentrations.     

A putative DNA-binding domain in the NUCKS protein

We have studied the DNA-binding properties of a NUCKS-derived, synthetic peptide containing an extended GRP motif. This peptide binds to random-sequence DNA, but did not bind preferentially to poly(dA-dT). A synthetic peptide with the same amino acid composition but with a random sequence did not bind to DNA, suggesting that the structure of the DNA-binding domain plays a pivotal role in the interaction with DNA. NMR and graphic modeling were employed to investigate the structure of the synthetic peptide. It was shown that the DNA-binding peptide constituted an alpha helix in phosphate buffer at pH 5.5. Docking results indicated an almost perfect fit for this small, helical peptide into the major groove of DNA with the possibility of four basic residues interacting with the phosphate backbone of DNA. One consensus site for phosphorylation by Cdk1 is located in the N-terminal end of the DNA-binding peptide. Upon phosphorylation of this site, the binding to DNA was completely prohibited. Immunofluorescence experiments showed that NUCKS was located in the nuclei in proliferating cells in interphase of the cell cycle, but was distributed throughout the cytoplasm in mitotic cells.     

On the importance of van der Waals interaction in the groove binding of DNA with ligands: restrained molecular dynamics study

Comparison of interaction energy between an oligonucleotide and a DNA-binding ligand in the minor and major groove modes was made by use of restrained molecular dynamics. Distortion in DNA was found for the major groove mode whereas less significant changes for both ligand and DNA were detected for the minor groove binding after molecular dynamics simulation. The conformation of the ligand obtained from the major groove mode resembles that computed with the ligand soaked in water. The van der Waals contact energy was found to be as significant as electrostatic energy and more important for difference in binding energy between these two binding modes. The importance of van der Waals force in groove binding was supported by computations on the complex formed by the repressor peptide fragment from the bacteriophage 434 and its operator oligonucleotide.     

Circular dichroism and fluorescence of a tyrosine side-chain residue monitors the concentration-dependent equilibrium between U-shaped and coiled-coil conformations of a peptide derived from the catalytic core of HIV-1 integrase

The peptide denoted K159 (30 residues) derives from the catalytic core (CC) sequence of HIV-1 integrase (IN, residues 147-175). In the crystal structure of CC, the corresponding segment belongs to the alpha4 helix (residues 148-168, including residues Glu 152, Lys 156 and Lys 159, crucial for enzyme activity and DNA recognition), a loop (residues 169-171) and a part of the alpha5 helix (171-175), involved in enzyme dimerization. We used the fluorescence and the circular dichroism (CD) properties in the near-UV of the aromatic side chain of a tyrosine residue added at the C-terminal end of K159 in order to analyze the behavior of the concentrated and diluted peptide in aqueous trifluoroethanol (TFE), in an attempt to connect the information obtainable at high (NMR), medium (CD) and low (fluorescence) concentrations of the peptide. Altogether, the C-terminal tyrosine residue provided indirect information on the global conformation of K159 and on the local orientation and environment of the residue. The propensity of TFE to stabilize alpha-helical conformations in peptides was confirmed in CD and fluorescence experiments at relatively high (20-160 microM) and low (2-16 microM) concentrations, respectively. At relatively high concentration, stabilization of the peptide into alpha-helical conformation favored its auto-association likely in parallel coiled-coil dimers, as pointed out in our previous work [Eur. J. Biochem. 253 (1998) 236]. This was further confirmed by ANS (1-anilinonaphtalene-8-sulfonic acid) analysis and fluorescence temperature coefficient measurement. With diluted K159, a Stern-Volmer analysis with positively and negatively charged quenchers indicated that, when the intermolecular interactions were absent, the tyrosine was in a positively charged environment, as if the peptide folded into a U-shaped conformation similar to that present in the crystal structure of the enzyme.     

Characterization of a synthetic peptide corresponding to a receptor binding domain of mouse interferon gamma.

A receptor binding region of mouse interferon gamma (IFN gamma) has previously been localized to the N-terminal 39 amino acids of the molecule by use of synthetic peptides and monoclonal antibodies. In this report, a detailed analysis of the synthetic peptide corresponding to this region, IFN gamma (1-39), is presented. Circular dichroism (CD) spectroscopy indicated that the peptide has stable secondary structure under aqueous conditions and adopts a combination of alpha-helical and random structure. A peptide lacking two N-terminal amino acids, IFN gamma (3-39), had similar secondary structure and equivalent ability to compete for receptor binding, while peptides lacking four or more N-terminal residues had reduced alpha-helical structure and did not inhibit 125I-IFN gamma binding. Substitution of proline, a helix-destabilizing amino acid, for leucine (residue 8) of a predicted amphipathic alpha-helix (residues 3-12), IFN gamma (1-39) [Pro]8, resulted in a substantial reduction in the helical content of the peptide, supporting the presence of helical structure in this region. However, destabilization of the helix did not reduce the competitive ability of the peptide. A peptide lacking eight C-terminal residues, IFN gamma (1-31), did not block 125I-IFN gamma binding and had no detectable alpha-helical structure, suggesting a requirement of the predicted second alpha-helix (residues 20-34) for receptor interaction and helix stabilization. Substitution of phenylalanine for tyrosine at position 14, IFN gamma (1-39) [Phe]14, a central location of a predicted omega-loop structure, did not affect the secondary structure associated with the region yet resulted in a 30-fold increase in receptor competition.(ABSTRACT TRUNCATED AT 250 WORDS)     

Recognition of DNA sequences by the repressor of bacteriophage 434

The structure of a complex between the DNA-binding domain of phage 434 repressor and a 14 base-pair synthetic DNA operator reveals the molecular interactions important for sequence-specific recognition. A set of contacts with DNA backbone, notably involving hydrogen bonds between peptide-NH groups and DNA phosphates, position the repressor and fix the DNA configuration. Direct interactions between amino acid side chains and DNA bases involve nonpolar van der Waals contacts as well as hydrogen bonds. The structures of the repressor domain and of the 434 cro protein are extremely similar. There appear to be no major conformational changes in the proteins when they bind to DNA.