Cooperative protein-DNA interactions: effects of KCl on lambda cI binding to OR.

The effects of monovalent salt activity on the site-specific and cooperative interactions of cI repressor with its three operator sites OR were studied by using quantitative DNase I footprint titration methods. Individual-site binding isotherms were obtained for binding repressor dimers to each site of wild-type OR and to mutant operator templates in which binding to one or two sites has been eliminated. The standard Gibbs energies for intrinsic binding, delta G1, delta G2, and delta G3, and cooperative interactions, delta G12 and delta G23, were determined at each condition (range 50-200 mM KCl). It is found that the dimer affinity for each of the three sites increases as [KCl] decreases, a striking result given that the monomer-dimer equilibrium shifts toward monomer formation under identical solution conditions [Koblan, K. S., & Ackers, G. K. (1991) Biochemistry (preceding paper in this issue)]. The magnitudes of ion-linked effects are found to differ at the three operator sites, while the intrinsic interaction binding free energies for sites OR1 and OR3 change in parallel over the entire range of [KCl]. The KCl dependencies at OR1 and OR3 represent the average release of 3.7 +/- 0.6 and 3.8 +/- 0.6 apparent ions, respectively. By contrast, the KCl dependency of OR2 binding corresponds to the displacement of 5.2 +/- 0.7 apparent ions. The ability of cI repressor to discriminate between the three operator sites thus appears linked to ion binding/release reactions.     

How lambda repressor and lambda Cro distinguish between OR1 and OR3

Although lambda repressor and lambda Cro bind to the same six operators on the phage chromosome, the fine specificities of the two proteins differ: repressor binds more tightly to OR1 than to OR3, and vice versa for Cro. In this paper, we change base pairs in the operators and amino acids in the proteins to analyze the basis for these preferences. We find that these preferences are determined by residues 5 and 6 of the recognition helices of the two proteins and by the amino-terminal arm, in the case of repressor. We also find that the most important base pairs in the operator which enable repressor and Cro to discriminate between OR1 and OR3 are position 3 (for Cro) and positions 5 and 8 (for repressor). These and previous results show how repressor and Cro recognize and distinguish between two related operator sequences.     

F''-coded, temperature-sensitive lambda cI857 repressor gene for easy construction and regulation of lambda promoter-dependent expression systems.

We describe the construction and properties of an F' factor which carries the temperature-sensitive cI857 allele of the repressor gene of coliphage lambda and which lacks the lambda cro function. This episome can easily be transferred to any F- and F' Escherichia coli strain, thus facilitating the construction and regulation of lambda promoter-dependent expression systems without the use of defective prophages.     

Coupled energetics of lambda cro repressor self-assembly and site-specific DNA operator binding II: cooperative interactions of cro dimers

The bacteriophage lambda relies on interactions of the cI and cro repressors which self assemble and bind the two operators (O(R) and O(L)) of the phage genome to control the lysogenic to lytic switch. While the self assembly and O(R) binding of cI have been investigated in detail, a more complete understanding of gene regulation by phage lambda also requires detailed knowledge of the role of cro repressor as it dimerizes and binds at O(R) sites. Since dimerization and operator binding are coupled processes, a full elucidation of the regulatory energetics in this system requires that the equilibrium constants for dimerization and cooperative binding be determined. The dimerization constant for cro has been measured as a prelude to these binding studies. Here, the energetics of cro binding to O(R) are evaluated using quantitative DNaseI footprint titration techniques. Binding data for wild-type and modified O(R) site combinations have been simultaneously analyzed in concert with the dimerization energetics to obtain both the intrinsic and cooperative DNA binding energies for cro with the three O(R) sites. Binding of cro dimers is strongest to O(R)3, then O(R)1 and lastly, O(R)2. Adjacently bound repressors exhibit positive cooperativity ranging from -0.6 to -1.0 kcal/mol. Implications of these, newly resolved, energetics are discussed in the framework of a dynamic model for gene regulation. This characterization of the DNA-binding properties of cro repressor establishes the foundation on which the system can be explored for other, more complex, regulatory elements such as cI-cro cooperativity.     

Energetics of subunit dimerization in bacteriophage lambda cI repressor: linkage to protons, temperature, and KCl.

A common feature of gene regulatory systems is the linkage between reversible protein oligomerization and DNA binding. Experimental dissection using temperature dependence of the subunit-subunit energetics and their linkage to processes such as ion binding and release is necessary for characterization of the chemical forces that contribute to cooperativity and site specificity. We have therefore studied the effects of temperature, proton activity, and monovalent salt on monomer-dimer assembly of the lambda cI repressor using a recently developed gel chromatographic procedure. This technique has made possible studies in the previously inaccessible picomolar concentration ranges where the assembly reactions occur. Upon formation of the dimer interface in the range pH 5-9, we find an overall absorption of protons which is temperature-dependent. The dimerization reaction displays a large negative enthalpy of association at all conditions studied (pH 5, 7, and 9). The reaction is also dependent on monovalent salt concentration: subunit association is weaker at low-salt conditions. The results suggest that a repulsive interaction between negatively charged side chains (i.e., aspartates and glutamates) on each monomer surface is attenuated by increasing concentrations of KCl. Formation of the dimer interface may be mediated by absorption of cations which stabilize the complex.     

Digestion of the lambda cI repressor with various serine proteases and correlation with its three dimensional structure

Partial proteolysis of the lambda cI repressor has been carried out systematically with trypsin, chymotrypsin, elastase, endoproteinase Glu-C, kallikrein, and thrombin. The cleavage sites have been determined by (i) comparison of fragments produced and observed in SDS-polyacrylamide gel with known fragments and plots of distance migrated versus log (molecular weight of fragment), (ii) partial Edman sequencing of the stable C-terminal fragments to identify cleavage points, and (iii) electrospray mass spectrometry of fragments produced. Most cleavage points are found to occur in the region 86-137, saving some in the N-terminal domain observed for trypsin and Glu-C. Region 86-137 can be further subdivided into three regions 86-91, 114-121, and 128-137 prone to cleavage, with intermediate regions resistant to cleavage to all six proteases. These resistant regions show that much of the region 93-131 previously called a 'linker' is actually part of the C-domain as first proposed in all models from our laboratory. Region 92-114 includes the cleavage site Ala-Gly, which must be buried in the intact repressor. The observed cleavage points in region 114-137 can be used to judge the best among three previously proposed models since they differ from each other in the structure of region 93-131. Model 1j5g is adjudged to be better than model 1lwq (which is based on 1kca, a crystal structure) as susceptible residues are more exposed in the former and lack of cleavages at six sites is better explained. Likewise, the models 1j5g and 1lwq are compared with a recent crystal structure of fragment 101-229 in 2ho0 and another low resolution crystal structure in 3bdn.     

Anatomy of specific interactions between lambda repressor and operator DNA

Recognition of specific DNA sequences by proteins is essential for regulation of gene expression. To fully understand the recognition mechanism, it is necessary to understand not only the structure of the specific protein-DNA interactions but also the energetics. We therefore performed a computer analysis in which a phage DNA-binding protein, lambda repressor, was used to examine the changes in binding free energy (DeltaDeltaG) and its energy components caused by single base mutations. We then determined which of the calculated energy components best correlated with the experimental data. The experimental DeltaDeltaG values were well reproduced by the calculations. Component analysis revealed that the electrostatic and hydrogen bond energies were most strongly correlated with the experimental data. Among the 51 single base-substitution mutants examined, positive DeltaDeltaG values, corresponding to weakened binding, were caused by the loss of favorable electrostatic interactions and hydrogen bonds, the introduction of steric collisions and electrostatic repulsion, the loss of favorable interactions with a thymine methyl group, and the increase of unfavorable hydration energy from isolated DNA. This analysis also showed distinct patterns of recognition at A-T and G-C positions, as different combinations of energy components were involved in DeltaDeltaG caused by the two substitution types. We have thus been able to identify the energy components that most strongly correlate with sequence-dependent DeltaDeltaG and determine their contribution to the specificity of DNA sequence recognition by the lambda repressor. Application of this method to other systems should provide additional insight into the molecular mechanism of protein-DNA recognition.     

Amino acid changes in the repressor of bacteriophage lambda due to temperature-sensitive mutations in its cI gene and the structure of a highly temperature-sensitive mutant repressor

The mutant cIts genes from seven different lambdacIts phages carrying tsU50, tsU9, tsU46, ts1, tsU51, tsI-22 and ts2 mutations were cloned in plasmid. The positions of these mutations and the resulting changes of amino acids in the repressor were determined by DNA sequencing. The first four mutations mapping in the N-terminal domain show the following changes: I21S, G53S, A62T and V73A, respectively. Of the three remaining mutations mapping in the C-terminal domain, cItsI-22 and cIts2 show N207T and K224E substitutions respectively, while the mutant cItsU51 gene carries F141I and P153L substitutions. Among these ts repressors, CIts2 having the charge-reversal change K224E was overexpressed from tac promoter in a plasmid and purified, and its structure and function were studied. Operator-binding studies suggest that the ts2 repressor is somewhat defective in monomer-dimer equilibrium and/or cooperativity even at permissive temperatures and loses its operator-binding ability very rapidly above 25 degrees C. Comparative studies of fluorescence and CD spectra, sulfhydryl group reactivity and elution behaviour in size-exclusion HPLC of both wild-type and ts2-mutant repressors at permissive and non-permissive temperatures suggest that the C-terminal domain of the ts2 repressor carrying a K224E substitution has a structure that does not favor tetramer formation at non-permissive temperatures.     

Differential regulation of lambda pL and pR promoters by a cI repressor in a broad-host-range thermoregulated plasmid marker system

Plasmid systems with unique markers were constructed to assess the fate of recombinant DNA and genetically manipulated bacteria in soil and freshwater model environments. On such constructs the marker gene, xylE (for catechol 2,3-dioxygenase), is expressed from the lambda promoter pL or pR, each of which is controlled by the temperature-sensitive lambda repressor c1857. Combinations of these elements were cloned into the broad-host-range plasmid pKT230 to form pLV1010 (pL-xylE), pLV1011 (pL-xylE-c1857), and pLV1013 (pR-xylE-c1857). The recombinant plasmids were introduced into different gram-negative bacteria. The thermoregulated system of pLV1013 functioned well in a range of species, with xylE induction being readily achieved by elevation of the temperature from 28 to 37 degrees C. There was a difference in the induction of catechol 2,3-dioxygenase activity, depending on whether xylE was expressed from pL (pLV1011) or pR (pLV1013). Our observations on testing the different systems in a number of hosts suggest that genes carried by the DNA of genetically engineered microorganisms may not be expressed in a predictable manner following transfer from the release host to other species.     

Comparison of operator-specific and nonspecific DNA binding of the lambda cI repressor: [KCl] and pH effects.

The effects of proton and KCl activity on the nonspecific lambda cI repressor-DNA interactions and on the site-specific repressor-O(R) interactions were compared, in order to assess their roles in site specificity. The repressor-O(R) interactions were studied by using DNase I footprint titration. The Gibbs free energy changes for binding and for cooperativity were determined between 25 and 300 mM KCl, from individual-site isotherms for the binding of repressor to O(R) and to reduced-valency mutants. The proton-linked effects on repressor-O(R) interactions have been published [Senear, D. F., & Ackers, G. K. (1990) Biochemistry 29, 6568-6577; Senear, D. F., & Bolen, D. W. (1991) Methods Enzymol. (in press)]. Nonspecific binding was studied by using a nitrocellulose filter binding assay, which proved advantageous in this case, due to the relatively weak nonspecific binding, and precipitation of repressor-DNA complexes. Filter binding provided measurements at low binding density where precipitation did not occur. The data provide estimates of the Gibbs free energy changes for nonspecific, intrinsic binding, but not for cooperativity. The KCl concentration dependencies of the intrinsic binding constants indicate that ion release plays similar roles in distinguishing between the operators and in discriminating operator from nonoperator DNA. Binding to DNA is accompanied by net proton absorption. Near neutral pH, proton linkages to operator and nonoperator binding are the same. Differences at acid and at basic pH implicate the same ionizable repressor groups in distinguishing between the operators and in discriminating operator from nonoperator DNA. The results indicate similar overall modes of operator and nonoperator binding of repressor, but implicate indirect effects of DNA sequence as important contributors to sequence recognition.     

Protein-protein interactions in a higher-order structure direct lambda site-specific recombination

The highly directional site-specific recombination of bacteriophage lambda is tightly regulated by the binding of three different proteins to a complex array of sites. The manner in which these reactions are both stimulated and inhibited by co-operative binding of proteins to specific sites on the P arm of attP and AttR has been elucidated by correlation of nuclease protection with recombination studies of both wild-type and mutant DNAs. In addition to co-operative forces, there is a specific competitive interaction that allows the protein-DNA complex to serve as a "biological switch". This switch does not depend upon the simple occlusion of DNA binding sites by neighboring proteins; but, rather, the outcome of this competition is dependent on long-range interactions that vary between the higher-order structures of attP and attR. These higher-order structures are dependent on cooperative interactions involving three proteins binding to five or more sites.     

Differential regulation of lambda pL and pR promoters by a cI repressor in a broad-host-range thermoregulated plasmid marker system.

Plasmid systems with unique markers were constructed to assess the fate of recombinant DNA and genetically manipulated bacteria in soil and freshwater model environments. On such constructs the marker gene, xylE (for catechol 2,3-dioxygenase), is expressed from the lambda promoter pL or pR, each of which is controlled by the temperature-sensitive lambda repressor c1857. Combinations of these elements were cloned into the broad-host-range plasmid pKT230 to form pLV1010 (pL-xylE), pLV1011 (pL-xylE-c1857), and pLV1013 (pR-xylE-c1857). The recombinant plasmids were introduced into different gram-negative bacteria. The thermoregulated system of pLV1013 functioned well in a range of species, with xylE induction being readily achieved by elevation of the temperature from 28 to 37 degrees C. There was a difference in the induction of catechol 2,3-dioxygenase activity, depending on whether xylE was expressed from pL (pLV1011) or pR (pLV1013). Our observations on testing the different systems in a number of hosts suggest that genes carried by the DNA of genetically engineered microorganisms may not be expressed in a predictable manner following transfer from the release host to other species.     

Translational frameshift sites within bacteriophage lambda genes rexA and cI

Phage lambda's cI-rexA-rexB operon displays an intriguing example of regulation by an unexplained mechanism of polarity. We have identified three potential -1 translational frameshift sites and present a model for translational frameshift suppression by lambda's CI repressor as a mechanism of regulating operon polarity, implying an additional role for CI self-regulation.     

Sites of contact between lambda operators and lambda repressor.

DNA bearing lambda operator sequences was methylated by dimethyl sulfate (DMS) in the presence or absence of lambda repressor. Under the experimental conditions, DMS methylates only the purine residues. The presence of lambda repressor affects only the methylation of certain G residues in the operators. Repressor blocks the methylation of certain G's and enhances the methylation of other G's. Since the reactive ring-nitrogen of G lies in the major groove of double-stranded DNA, and the reactive ring-nitrogen of A lies in the minor groove, the above results imply that the repressor makes contacts in the major groove of the helix. The repressor effect on G-methylation is sharply confined to the three 17 base pair units within each lambda operator previously proposed as the repressor-binding sites.