Role of glutamic acid-181 in DNA-sequence recognition by the catabolite gene activator protein (CAP) of Escherichia coli: altered DNA-sequence-recognition properties of [Val181]CAP and [Leu181]CAP

cAMP Activation of the cAMP Receptor Protein, a Model Bacterial Transcription Factor

The active and inactive structures of the Escherichia coli cAMP receptor protein (CRP), a model bacterial transcription factor, are compared to generate a paradigm in the cAMP-induced activation of CRP. The resulting paradigm is shown to be consistent with numerous biochemical studies of CRP and CRP*, a group of CRP mutants displaying cAMP-free activity. The cAMP affinity of CRP is dictated by two factors: (i) the effectiveness of the cAMP pocket and (ii) the protein equilibrium of apo-CRP. How these two factors interplay in determining the cAMP affinity and cAMP specificity of CRP and CRP* mutants are discussed. Both the current understanding and knowledge gaps of CRP-DNA interactions are also described. This review ends with a list of several important CRP issues that need to be addressed in the future.

Studies of IscR reveal a unique mechanism for metal-dependent regulation of DNA binding specificity

IscR from Escherichia coli is an unusual metalloregulator in that it globally regulates transcription by recognizing two different DNA motifs in a Fe-S dependent manner. Here, we report structural and biochemical studies of IscR, which suggest remodeling of the protein-DNA interface upon Fe-S ligation broadens the DNA binding specificity from binding a type 2 motif to both type 1 and 2 motifs. Analysis of an apo-IscR variant with relaxed target-site discrimination identified a key residue in wild-type apo-IscR that we propose makes unfavorable interactions with a type 1 motif. Upon Fe-S binding, these interactions are apparently removed, thereby allowing holo-IscR to bind both type 1 and 2 motifs. These data suggest a novel mechanism of ligand-mediated DNA site recognition, whereby metallocluster ligation relocates a protein specificity determinant to expand DNA target site selection, allowing a broader transcriptomic response by holo-IscR.

Repression by cyclic AMP receptor protein at a distance

In a previous study of promoters dependent on the Escherichia coli cyclic AMP receptor protein (CRP), carrying tandem DNA sites for CRP, we found that the upstream-bound CRP could either enhance or repress transcription, depending on its location. Here, we have analyzed the interactions between CRP and the C-terminal domains of the RNA polymerase α subunits at some of these promoters. We report that the upstream-bound CRP interacts with these domains irrespective of whether it up- or downregulates promoter activity. Hence, disruption of this interaction can lead to either down- or upregulation, depending on its location.

IMPORTANCE

Many bacterial promoters carry multiple DNA sites for transcription factors. While most factors that downregulate promoter activity bind to targets that overlap or are downstream of the transcription start and -10 element, very few cases of repression from upstream locations have been reported. Since more Escherichia coli promoters are regulated by cyclic AMP receptor protein (CRP) than by any other transcription factor, and since multiple DNA sites for CRP are commonplace at promoters, our results suggest that promoter downregulation by transcription factors may be more prevalent than hitherto thought, and this will have implications for the annotation of promoters from new bacterial genome sequences.

Protein-protein interactions between sigma(70) region 4 of RNA polymerase and Escherichia coli SoxS, a transcription activator that functions by the prerecruitment mechanism: evidence for "off-DNA" and "on-DNA" interactions

According to the prerecruitment hypothesis, Escherichia coli SoxS activates the transcription of the genes of the SoxRS regulon by forming binary complexes with RNA polymerase (RNAP) that scan the chromosome for class I and class II SoxS-dependent promoters. We showed previously that the alpha subunit's C-terminal domain plays a role in activating both classes of promoter by making protein-protein contacts with SoxS; some of these contacts are made in solution in the absence of promoter DNA, a critical prediction of the prerecruitment hypothesis. Here, we identified seven single-alanine substitutions of the region 4 of sigma(70) (sigma(70) R4) of RNAP that reduce SoxS activation of class II promoters. With genetic epistasis tests between these sigma(70) R4 mutants and positive control mutants of SoxS, we identified 10 pairs of amino acids that interact with each other in E. coli. Using the yeast two-hybrid system and affinity immobilization assays, we showed that SoxS and sigma(70) R4 can interact in solution (i.e., "off-DNA"). The interaction requires amino acids of the class I/II (but not the class II) positive control surface of SoxS, and five amino acids of sigma(70) R4 that reduce activation in E. coli also reduce the SoxS-sigma(70) R4 interaction in yeast. One of the epistatic interactions that occur in E. coli also occurs in the yeast two-hybrid system (i.e., off-DNA). Importantly, we infer that the five epistatic interactions occurring in E. coli that require an amino acid of the class II surface occur "on-DNA" at class II promoters. Finding that SoxS contacts sigma(70) R4 both off-DNA and on-DNA is consistent with the prerecruitment hypothesis. Moreover, SoxS is now the first example of an E. coli transcriptional activator that uses a single positive control surface to make specific protein-protein contacts with two different subunits of RNAP.

Engineering transcription factors with novel DNA-binding specificity using comparative genomics

The transcriptional program for a gene consists of the promoter necessary for recruiting RNA polymerase along with neighboring operator sites that bind different activators and repressors. From a synthetic biology perspective, if the DNA-binding specificity of these proteins can be changed, then they can be used to reprogram gene expression in cells. While many experimental methods exist for generating such specificity-altering mutations, few computational approaches are available, particularly in the case of bacterial transcription factors. In a previously published computational study of nitrogen oxide metabolism in bacteria, a small number of amino-acid residues were found to determine the specificity within the CRP (cAMP receptor protein)/FNR (fumarate and nitrate reductase regulatory protein) family of transcription factors. By analyzing how these amino acids vary in different regulators, a simple relationship between the identity of these residues and their target DNA-binding sequence was constructed. In this article, we experimentally tested whether this relationship could be used to engineer novel DNA–protein interactions. Using Escherichia coli CRP as a template, we tested eight designs based on this relationship and found that four worked as predicted. Collectively, these results in this work demonstrate that comparative genomics can inform the design of bacterial transcription factors.

Characterization of DNA-binding specificity and analysis of binding sites of the Pseudomonas aeruginosa global regulator, Vfr, a homologue of the Escherichia coli cAMP receptor protein

Vfr, a global regulator of Pseudomonas aeruginosa virulence factors, is a homologue of the Escherichia coli cAMP receptor protein, CRP. Vfr is 91% similar to CRP and maintains many residues important for CRP to bind cAMP, bind DNA, and interact with RNA polymerase at target promoters. While vfr can complement an E. coli crp mutant in beta-galactosidase production, tryptophanase production and catabolite repression, crp can only complement a subset of Vfr-dependent phenotypes in P. aeruginosa. Using specific CRP binding site mutations, it is shown that Vfr requires the same nucleotides as CRP for optimal transcriptional activity from the E. coli lac promoter. In contrast, CRP did not bind Vfr target sequences in the promoters of the toxA and regA genes. Footprinting analysis revealed Vfr protected sequences upstream of toxA, regA, and the quorum sensing regulator lasR, that are similar to but significantly divergent from the CRP consensus binding sequence, and Vfr causes similar DNA bending to CRP in bound target sequences. Using a preliminary Vfr consensus binding sequence deduced from the Vfr-protected sites, Vfr target sequences were identified upstream of the virulence-associated genes plcN, plcHR, pbpG, prpL and algD, and in the vfr/orfX, argH/fimS, pilM/ponA intergenic regions. From these sequences the Vfr consensus binding sequence, 5'-ANWWTGNGAWNY : AGWTCACAT-3', was formulated. This study suggests that Vfr shares many of the same functions as CRP, but has specialized functions, at least in terms of DNA target sequence binding, required for regulation of a subset of genes in its regulon.

Characterization of CprK1, a CRP/FNR-type transcriptional regulator of halorespiration from Desulfitobacterium hafniense

The recently identified CprK branch of the CRP (cyclic AMP receptor protein)-FNR (fumarate and nitrate reduction regulator) family of transcriptional regulators includes proteins that activate the transcription of genes encoding proteins involved in reductive dehalogenation of chlorinated aromatic compounds. Here we report the characterization of the CprK1 protein from Desulfitobacterium hafniense, an anaerobic low-G+C gram-positive bacterium that is capable of reductive dechlorination of 3-chloro-4-hydroxyphenylacetic acid (Cl-OHPA). The gene encoding CprK1 was cloned and functionally overexpressed in Escherichia coli, and the protein was subsequently purified to homogeneity. To investigate the interaction of CprK1 with three of its predicted binding sequences (dehaloboxes), we performed in vitro DNA-binding assays (electrophoretic mobility shift assays) as well as in vivo promoter probe assays. Our results show that CprK1 binds its target dehaloboxes with high affinity (dissociation constant, 90 nM) in the presence of Cl-OHPA and that transcriptional initiation by CprK1 is influenced by deviations in the dehaloboxes from the consensus TTAAT----ATTAA sequence. A mutant CprK1 protein was created by a Val-->Glu substitution at a conserved position in the recognition alpha-helix that gained FNR-type DNA-binding specificity, recognizing the TTGAT----ATCAA sequence (FNR box) instead of the dehaloboxes. CprK1 was subject to oxidative inactivation in vitro, most likely caused by the formation of an intermolecular disulfide bridge between Cys11 and Cys200. The possibility of redox regulation of CprK1 by a thiol-disulfide exchange reaction was investigated by using two Cys-->Ser mutants. Our results indicate that a Cys11-Cys200 disulfide bridge does not appear to play a physiological role in the regulation of CprK1.

Regulation of sexual dimorphism: mutational and chemogenetic analysis of the doublesex DM domain

Doublesex (dsx) is a transcription factor in Drosophila that regulates somatic sexual differentiation. Male- and female-specific splicing isoforms of DSX share a novel DNA-binding domain, designated the DM motif. Broadly conserved among metazoan sex-determining factors, the DM domain contains a nonclassical zinc module and binds in the DNA minor groove. Here, we characterize the DM motif by site-directed and random mutagenesis using a yeast one-hybrid (Y1H) system and extend this analysis by chemogenetic complementation in vitro. The Y1H system is based on a sex-specific Drosophila enhancer element and validated through studies of intersexual dsx mutations. We demonstrate that the eight motif-specific histidines and cysteines engaged in zinc coordination are each critical and cannot be interchanged; folding also requires conserved aliphatic side chains in the hydrophobic core. Mutations that impair DNA binding tend to occur at conserved positions, whereas neutral substitutions occur at nonconserved sites. Evidence for a specific salt bridge between a conserved lysine and the DNA backbone is obtained through the synthesis of nonstandard protein and DNA analogs. Together, these results provide molecular links between the structure of the DM domain and its function in the regulation of sexual dimorphism.

Catabolite activator protein: DNA binding and transcription activation

Recently determined structures of the Escherichia coli catabolite activator protein (CAP) in complex with DNA, and in complex with the RNA polymerase alpha subunit C-terminal domain (alphaCTD) and DNA, have yielded insights into how CAP binds DNA and activates transcription. Comparison of multiple structures of CAP-DNA complexes has revealed the contributions of direct and indirect readout to DNA binding by CAP. The structure of the CAP-alphaCTD-DNA complex has provided the first structural description of interactions between a transcription activator and its functional target within the general transcription machinery. Using the structure of the CAP-alphaCTD-DNA complex, the structure of an RNA polymerase-DNA complex, and restraints from biophysical, biochemical and genetic experiments, it has been possible to construct detailed three-dimensional models of intact class I and class II transcription activation complexes.

Two cAMP receptor proteins with different biochemical properties in the filamentous cyanobacterium Anabaena sp. PCC 7120

Two open reading frames (ORFs), alr0295 and alr2325, are found to encode putative cAMP receptor proteins (CRPs) in the genome of the filamentous cyanobacterium Anabaena sp. PCC 7120. These ORFs were named cAMP receptor protein-like gene A in Anabaena sp. PCC 7120 (ancrpA) and cAMP receptor protein-like gene B in Anabaena sp. PCC 7120 (ancrpB), respectively, and those translated products were investigated. The equilibrium dialysis measurements revealed that AnCrpA bound with cAMP specifically, while AnCrpB bound with both cAMP and cGMP, though the affinity for cGMP was weak. The binding affinity for cAMP of AnCrpA showed the lowest dissociation constant, approximately 0.8 microM, among bacterial CRPs. A gel mobility shift assay elucidated that AnCrpA and AnCrpB formed a complex with the consensus DNA sequence in the presence of cAMP, although AnCrpB did not have ordinary DNA-binding motifs.

Architectural requirements for optimal activation by tandem CRP molecules at a class I CRP-dependent promoter

The Escherichia coli cyclic AMP receptor protein (CRP) activates transcription at target promoters by interacting with the C-terminal domain of the RNA polymerase alpha subunit. We have constructed a set of promoters carrying tandem DNA sites for CRP with one site centred at position -61.5 and the other site located at different upstream positions. Optimal CRP-dependent activation of transcription is observed when the upstream DNA site for CRP is located at position -93.5 or at position -103.5. Evidence is presented to suggest that activation by the upstream-bound CRP molecule is due to interaction with the C-terminal domain of the RNA polymerase alpha subunit.

Indirect readout of DNA sequence at the primary-kink site in the CAP-DNA complex: alteration of DNA binding specificity through alteration of DNA kinking

The catabolite activator protein (CAP) sharply bends DNA in the CAP-DNA complex, introducing a DNA kink, with a roll angle of approximately 40 degrees and a twist angle of approximately 20 degrees, between positions 6 and 7 of the DNA half-site, 5'-A(1)A(2)A(3)T(4)G(5)T(6)G(7)A(8)T(9)C(10)T(11)-3' ("primary kink"). CAP recognizes the base-pair immediately 5' to the primary-kink site, T:A(6), through an "indirect-readout" mechanism involving sequence effects on the energetics of primary-kink formation. CAP recognizes the base-pair immediately 3' to the primary-kink site, G:C(7), through a "direct-readout" mechanism involving formation of a hydrogen bond between Glu181 of CAP and G:C(7). Here, we report that substitution of the carboxylate side-chain of Glu181 of CAP by the one-methylene-group-shorter carboxylate side-chain of Asp changes DNA binding specificity at position 6 of the DNA half site, changing specificity for T:A(6) to specificity for C:G(6), and we report a crystallographic analysis defining the structural basis of the change in specificity. The Glu181-->Asp substitution eliminates the primary kink and thus eliminates indirect-readout-based specificity for T:A(6). The Glu181-->Asp substitution does not eliminate hydrogen-bond formation with G:C(7), and thus does not eliminate direct-readout-based specificity for G:C(7).

Superimposed levels of regulation of the 4-hydroxyphenylacetate catabolic pathway in Escherichia coli

The regulation of the Pg promoter, which controls the expression of the meta operon of the 4-hydroxyphenylacetic acid (4-HPA) catabolic pathway of Escherichia coli W, has been examined through in vivo and in vitro experiments. By using Pg-lacZ fusions we have demonstrated that Pg is a promoter only inducible in the stationary phase when cells are grown on glucose as the sole carbon and energy source. This strict catabolite repression control is mediated by the cAMP receptor protein (CRP). This event does not require the presence of the specific HpaR repressor or the 4-HPA permease (HpaX), excluding the involvement of a typical inducer exclusion mechanism. However, the acetic acid excreted in the stationary phase by the cells growing in glucose acts as an overflow metabolite, which can provide the energy to produce cAMP and to adapt the cells rapidly to the utilization of a new less preferred carbon source such as the aromatic compounds. Although Pg is not a final sigma(38)-dependent promoter, it is activated by the global regulator integration host factor (IHF) in the stationary phase of growth. Gel retardation assays have demonstrated that both CRP and IHF simultaneously bind to the Pg upstream region. DNase I footprint experiments showed that cAMP-CRP and IHF binding sites are centered at -61.5 and -103, respectively, with respect to the transcription start site +1 of the Pg promoter.

Mutational analysis of conserved residues in the putative DNA-binding domain of the response regulator Spo0A of Bacillus subtilis

The Spo0A protein of Bacillus subtilis is a DNA-binding protein that is required for the expression of genes involved in the initiation of sporulation. Spo0A binds directly to and both activates and represses transcription from the promoters of several genes required during the onset of endospore formation. The C-terminal 113 residues are known to contain the DNA-binding activity of Spo0A. Previous studies identified a region of the C-terminal half of Spo0A that is highly conserved among species of endospore-forming Bacillus and Clostridium and which encodes a putative helix-turn-helix DNA-binding domain. To test the functional significance of this region and determine if this motif is involved in DNA binding, we changed three conserved residues, S210, E213, and R214, to Gly and/or Ala by site-directed mutagenesis. We then isolated and analyzed the five substitution-containing Spo0A proteins for DNA binding and sporulation-specific gene activation. The S210A Spo0A mutant exhibited no change from wild-type binding, although it was defective in spoIIA and spoIIE promoter activation. In contrast, both the E213G and E213A Spo0A variants showed decreased binding and completely abolished transcriptional activation of spoIIA and spoIIE, while the R214G and R214A variants completely abolished both DNA binding and transcriptional activation. These data suggest that these conserved residues are important for transcriptional activation and that the E213 residue is involved in DNA binding.

A study of the CopF repressor of plasmid pAMbeta1 by phage display

We studied DNA binding of a transcriptional repressor, CopF, displayed on a filamentous phage. Mutagenesis of a putative helix-turn-helix motif of CopF and of certain bases of the operator abolished the protein-DNA interaction, establishing the elements involved in CopF function and showing that phage display can be used to study repressor proteins.

Identification of the subunit of cAMP receptor protein (CRP) that functionally interacts with CytR in CRP-CytR-mediated transcriptional repression

At promoters of the Escherichia coli CytR regulon, the cAMP receptor protein (CRP) interacts with the repressor CytR to form transcriptionally inactive CRP-CytR-promoter or (CRP)(2)-CytR-promoter complexes. Here, using "oriented heterodimer" analysis, we show that only one subunit of the CRP dimer, the subunit proximal to CytR, functionally interacts with CytR in CRP-CytR-promoter and (CRP)(2)-CytR-promoter complexes. Our results provide information about the architecture of CRP-CytR-promoter and (CRP)(2)-CytR-promoter complexes and rule out the proposal that masking of activating region 2 of CRP is responsible for the transcriptional inactivity of the complexes.

Broad-host-range shuttle vectors for screening of regulated promoter activity in viridans group streptococci: isolation of a pH-regulated promoter

Viridans group streptococci are major constituents of the normal human oral flora and are also identified as the predominant pathogenic bacteria in native valve infective endocarditis. Little information is available regarding the regulation of gene expression in viridans group streptococci, either in response to changes in the oral environment or during development of endocarditis. We therefore constructed a set of broad-host-range vectors for the isolation of promoters from viridans group streptococci that are activated by specific environmental stimuli in vitro or in vivo. A genomic library of Streptococcus gordonii strain CH1 was constructed in one of the new vectors, and this library was introduced into a homologous bacterium by using an optimized electroporation protocol for viridans group streptococci. Because viridans group streptococci entering the bloodstream from the oral cavity encounter an increase in pH, we selected promoters upregulated by this specific stimulus. One of the selected promoter sequences showed homology to the promoter region of the hydA gene from Clostridium acetobutylicum, the expression of which is known to be regulated by the environmental pH. The isolation of this pH-regulated promoter shows that S. gordonii can sense an increase in the environmental pH, which serves as a signal for bacterial gene activation. Furthermore, this demonstrates the usefulness of these new selection vectors in research on adaptive gene expression of viridans group streptococci and possibly also of other gram-positive bacteria.

A genetic method for dissecting the mechanism of transcriptional activator synergy by identical activators

Pairs of transcriptional activators in prokaryotes have been shown to activate transcription synergistically from promoters with two activator binding sites. In some cases, such synergistic effects result from cooperative binding, but in other cases each DNA-bound activator plays a direct role in the activation process by interacting simultaneously with separate surfaces of RNA polymerase. In such cases, each DNA-bound activator must possess a functional activating region, the surface that mediates the interaction with RNA polymerase. When transcriptional activation depends on two or more identical activators, it is not straightforward to test the requirement of each activator for a functional activating region. Here we describe a method for directing a mutationally altered activator to either one or the other binding site, and we demonstrate the use of this method to examine the mechanism of transcriptional activator synergy by the Escherichia coli cyclic AMP receptor protein (CRP) working at an artificial promoter bearing two CRP-binding sites.

The role of DNA-protein salt bridges in molecular recognition: a model study

A theoretical study is presented of the influence of salt bridges between cationic side chains and DNA phosphates on DNA conformation and flexibility. The DNA sequence studied is that of the catabolite activator protein binding oligomer from the crystallized complex. The effect of salt bridges is modeled by neutralization of net phosphate charges for the groups involved in such interactions in the crystallized complex. Energy-optimized conformations are obtained by molecular mechanics using the JUMNA program. Base sequence dependence is studied by moving the phosphate neutralization pattern along the sequence and also by point mutations. Normal mode analysis is used to evaluate DNA flexibility. The results obtained show that the free oligomer is already precurved in the direction favored by the protein, and the effect of phosphate neutralization is principally to increase this curvature. This effect is, however, strongly sequence dependent. In addition, it is shown that oligomer flexibility cannot be explained by a simple superposition of the properties of successive dinucleotide steps, strong long-range coupling effects are observed. In all the cases examined, phosphate neutralization, however, leads to a reduction in oligomer flexibility.

Altering the DNA-binding specificity of Mu transposase in vitro

We describe the isolation of a variant of Mu transposase (MuA protein) which can recognize altered att sites at the ends of Mu DNA. No prior knowledge of the structure of the DNA binding domain or its mode of interaction with att DNA was necessary to obtain this variant. Protein secondary structure programs initially helped target mutations to predicted helical regions within a subdomain of MuA demonstrated to harbor att DNA binding activity. Of the 54 mutant positions examined, only two showed decreased affinity for att DNA, while eight others affected assembly of the Mu transpososome. A variant impaired in DNA binding [MuA(R146V)], and predicted to be in the recognition helix of an HTH motif, was challenged with altered att sites created from degenerate oligonucleotides to select for novel DNA binding specificity. DNA sequences bound to MuA(R146V) were detected by gel-retardation, and following several steps of PCR amplification/enrichment, were identified by cloning and sequencing. The strategy allowed recovery of an altered att site for which MuA(R146V) showed higher affinity than for the wild-type site, although this site was bound by wild-type MuA as well. The altered association between MuA(R146V) and an altered att site target was competent in transposition. We discuss the strengths and limitations of this methodology, which has applications in dissecting the functional role of specific protein-DNA associations.

Transcription activation at promoters carrying tandem DNA sites for the Escherichia coli cyclic AMP receptor protein: organisation of the RNA polymerase alpha subunits

We have constructed a family of promoters carrying tandem DNA sites for the Escherichia coli cyclic AMP receptor protein (CRP), with one of the sites centred between base-pairs 41 and 42 upstream from the transcription start site, and the second site located further upstream. In vivo activity measurements show that the activity of these promoters is completely dependent on CRP and that, depending on the precise location, CRP bound at the upstream site increases transcription activation. Hydroxyl radical footprinting was exploited to investigate the binding of CRP and RNA polymerase holoenzyme (RNAP) to these promoters. The study shows that the C-terminal domains of the RNAP alpha subunits bind adjacent to the upstream CRP and that their precise positioning depends on the location of upstream-bound CRP. The C-terminal domains of the RNAP alpha subunits interact with both the upstream and downstream-bound CRP via activating region 1 of CRP.

A lethal mutant of the catabolite gene activator protein CAP of Escherichia coli

The dimeric catabolite gene activator protein (CAP) of Escherichia coli uses its recognition helix to bind with each subunit the DNA sequence motif 5' G-7T-6G-5A-4 3'. It makes a direct amino acid-base contact with E181 and cytosine in position-5' on the reverse strand. While testing mutants of CAP in position 181 for specificity changes, we found that CAP E181Q is lethal in high amounts for the E. coli strains we used for cloning. We cloned this CAP mutant successfully in cya- strains, where CAP is inactive. Examination of the in vitro binding activities of CAP E181Q, and of in vivo activity when present in low, non-lethal amounts, revealed loss of specificity but not of binding capacity for its DNA targets. It binds well to CAP consensus with G or T in position-5, better to CAP consensus with A, C in position-5, quite well to lambda consensus operator with G in position-7 and rather weakly to lambda consensus.

The structure of a CAP-DNA complex having two cAMP molecules bound to each monomer

The 2.2 A resolution crystal structure of the Escherichia coli catabolite gene activator protein (CAP) complexed with cAMP and a 46-bp DNA fragment reveals a second cAMP molecule bound to each protein monomer. The second cAMP is in the syn conformation and is located on the DNA binding domain interacting with the helix-turn-helix, a beta-hairpin from the regulatory domain and the DNA (via water molecules). The presence of this second cAMP site resolves the apparent discrepancy between the NMR and x-ray data on the conformation of cAMP, and explains the cAMP concentration-dependent behaviors of the protein. In addition, this site's close proximity to mutations affecting transcriptional activation and its water-mediated interactions with a DNA recognition residue (E181) and DNA raise the possibility that this site has biological relevance.

Effect of cAMP binding site mutations on the interaction of cAMP receptor protein with cyclic nucleoside monophosphate ligands and DNA

Although cAMP binding to wild type cAMP receptor protein (CRP) induces specific DNA binding and activates transcription, cyclic nucleoside monophosphate (cNMP) binding to the CRP mutant Ser128 --> Ala does not, whereas the double CRP mutant Thr127 --> Leu/Ser128 --> Ala activates transcription even in the absence of cNMP. Isothermal titration calorimetry measurements on the cNMP binding reactions to the S128A and T127L/S128A mutants show that the reactions are mainly entropically driven as is cAMP binding to CRP. In contrast to cAMP binding to CRP, the binding reactions are noncooperative and exothermic with binding enthalpies (DeltaHb) ranging from -23.4 +/- 0.9 kJ mol-1 for cAMP binding to S128A at 39 degrees C to -4.1 +/- 0.6 kJ mol-1 for cAMP binding to T127L/S128A at 24 degrees C and exhibit enthalpy-entropy compensation. To account for the inactivity of the S128A mutant, in vitro and in vivo DNA binding experiments were performed on the cAMP-ligated S128A mutant. The cAMP-ligated S128A mutant binds to the consensus DNA binding site with approximately the same affinity as that of cAMP-ligated CRP but forms a different type of complex, which may account for loss of transcriptional activity by the mutant. Energy minimization computations on the cAMP-ligated S128A mutant show that amino acid conformational differences between S128A and CRP occur at Ser179, Glu181, and Thr182 in the center of the DNA binding site, implying that these conformational changes may account for the difference in DNA binding.

A single substitution in the putative helix-turn-helix motif of the pleiotropic activator PrfA attenuates Listeria monocytogenes virulence

PrfA, the regulator of virulence-gene expression in the pathogenic bacterium Listeria monocytogenes, displays sequence similarity to members of the CAP-FNR family of transcriptional regulators. To test the functional significance of this similarity, we constructed and analysed substitutions of two amino acids of PrfA predicted to contact DNA, i.e. Ser-184 and Ser-183. Substitution of Ser-184 by Ala reduced DNA binding and virulence-gene activation, and attenuated the virulence in a mouse model of infection. In contrast, substitution of Ser-183 by Ala had the opposite effect in these functional assays. A 17bp DNA sequence, which includes a putative PrfA site, was shown to be sufficient for target-site recognition by PrfA and PrfA-S183A. Our results strongly support the hypothesis that PrfA is a structural and functional homologue of CAP. In addition, they establish a clear correlation between DNA binding by PrfA, virulence-gene activation, and virulence.

Molecular characterization and transcriptional analysis of the putative hydrogenase gene of Clostridium acetobutylicum ATCC 824

A 2.8-kbp DNA region of Clostridium acetobutylicum ATCC 824 containing the putative hydrogenase gene (hydA) was cloned and sequenced. The 1,745-bp hydA encodes a 64,415-Da protein and presents strong identity with the [Fe] hydrogenase genes of Desulfovibrio and Clostridium species. The level of the putative hydA mRNA was high in cells from an acidogenic or an alcohologenic phosphate-limited continuous culture, while it was comparatively very low in cells from a solventogenic phosphate-limited continuous culture. These results were in agreement with the hydrogenase protein level, indicating that expression of hydA is regulated at the transcriptional level. Primer extension analysis identified a major transcriptional start site 90 bp upstream of the hydA start codon. The position of a putative rho-independent transcription terminator immediately downstream of the termination codon is in agreement with the size of the hydA transcript (1.9 kb) determined by Northern (RNA) blot experiments and confirms that the gene is transcribed as a monocistronic operon. Two truncated open reading frames (ORFs) were identified downstream and upstream of hydA and in opposite directions. The amino acid sequence deduced from ORF2 presents strong identity with ortho phosphoribosyl transferases involved in pyrimidine synthesis. The amino acid sequence deduced from ORF3 presents no significant similarity to any sequence in various available databases.

CUS1, a suppressor of cold-sensitive U2 snRNA mutations, is a novel yeast splicing factor homologous to human SAP 145

The function of U2 snRNA in splicing is mediated by the proteins of the U2 small nuclear ribonucleoprotein. To identify proteins that influence the function of U2 snRNA we carried out a screen for mutations in Saccharomyces cerevisiae that suppress the cold-sensitive growth defect of a mutation in U2 stem loop IIa, a structure important for the stable association of the U2 snRNP with pre-mRNA. The screen identified three dominant suppressor genes, one of which, CUS1-54, encodes an essential splicing protein required for U2 snRNP addition to the spliceosome. The suppressor protein rescues the spliceosome assembly defect of the mutant U2 in vitro, indicating that suppression is direct. Allele specificity tests show that the suppressor does not simply bypass the requirement for U2 stem loop IIa. Extra copies of wild-type CUS1, but not CUS1-54, suppress the temperature-sensitive prp11 and prp5 mutations, linking CUS1 protein to a subset of other factors required at the same step of spliceosome assembly. CUS1 is homologous to SAP 145, a component of the mammalian U2 snRNP that interacts with pre-mRNA. The yeast genome also encodes a homolog of human SAP 49, a protein that interacts strongly with both SAP 145 and pre-mRNA, underscoring the conservation of U2 snRNP proteins that function in spliceosome assembly.

Acetate utilization is inhibited by benzoate in Alcaligenes eutrophus: evidence for transcriptional control of the expression of acoE coding for acetyl coenzyme A synthetase

During batch growth of Alcaligenes eutrophus on benzoate-acetate mixtures, benzoate was the preferred substrate, with acetate consumption being delayed until the rate of benzoate consumption had diminished. This effect was attributed to a transcriptional control of the synthesis of acetyl coenzyme A (acetyl-CoA) synthetase, an enzyme necessary for the entry of acetate into the central metabolic pathways, rather than to a biochemical modulation of the activity of this enzyme. Analysis of a 2.4-kb mRNA transcript hybridizing with the A. eutrophus acoE gene confirmed this repression effect. In a benzoate-limited chemostat culture, derepression was observed, with no increase in the level of expression following an acetate pulse. Benzoate itself was not the signal triggering the repression of acetyl-CoA synthetase. This role was played by catechol, which transiently accumulated in the medium when high specific rates of benzoate consumption were reached. The lack of rapid inactivation of the functional acetyl-CoA synthetase after synthesis has been stopped enables A. eutrophus to retain the capacity to metabolize acetate for prolonged periods while conserving minimal protein expenditure.

Sequence analysis of the phs operon in Salmonella typhimurium and the contribution of thiosulfate reduction to anaerobic energy metabolism

The phs chromosomal locus of Salmonella typhimurium is essential for the dissimilatory anaerobic reduction of thiosulfate to hydrogen sulfide. Sequence analysis of the phs region revealed a functional operon with three open reading frames, designated phsA, phsB, and phsC, which encode peptides of 82.7, 21.3, and 28.5 kDa, respectively. The predicted products of phsA and phsB exhibited significant homology with the catalytic and electron transfer subunits of several other anaerobic molybdoprotein oxidoreductases, including Escherichia coli dimethyl sulfoxide reductase, nitrate reductase, and formate dehydrogenase. Simultaneous comparison of PhsA to seven homologous molybdoproteins revealed numerous similarities among all eight throughout the entire frame, hence, significant amino acid conservation among molybdoprotein oxidoreductases. Comparison of PhsB to six other homologous sequences revealed four highly conserved iron-sulfur clusters. The predicted phsC product was highly hydrophobic and similar in size to the hydrophobic subunits of the molybdoprotein oxidoreductases containing subunits homologous to phsA and phsB. Thus, phsABC appears to encode thiosulfate reductase. Single-copy phs-lac translational fusions required both anaerobiosis and thiosulfate for full expression, whereas multicopy phs-lac translational fusions responded to either thiosulfate or anaerobiosis, suggesting that oxygen and thiosulfate control of phs involves negative regulation. A possible role for thiosulfate reduction in anaerobic respiration was examined. Thiosulfate did not significantly augment the final densities of anaerobic cultures grown on any of the 18 carbon sources tested. on the other hand, washed stationary-phase cells depleted of ATP were shown to synthesize small amounts of ATP on the addition of the formate and thiosulfate, suggesting that the thiosulfate reduction plays a unique role in anaerobic energy conservation by S typhimurium.

Modulation of DNA-binding specificity within the nuclear receptor family by substitutions at a single amino acid position

Regulation of gene expression involves a large number of transcription factors with unique DNA-binding properties. Many transcription factors belong to families of related proteins that bind to similar but distinct sequences. In this study we have analyzed how amino acid substitutions at a single position in the DNA-binding domain modulate the DNA-binding specificity within the nuclear receptor family of transcription factors. All possible amino acids were introduced at the first position in the DNA recognition helix, and the specificities of the mutants were analyzed using response elements containing all combinations of bases at two variable base pair positions. All mutant proteins were functional in DNA binding, and could be divided into classes of mutants with different response element specificities. By combining functional data with analysis of the structural effects of the mutations by molecular modeling, we could identify both prohibitive steric interactions as well as positive interactions, such as hydrogen bonds, that function as important determinants for specificity. Only the residues found naturally in the glucocorticoid and estrogen receptors, glycine and glutamate, produce unique binding specificities. The specificities of the other mutants overlap with each other somewhat but the substitutions clearly have potential to contribute to diversity within the nuclear receptor family.

The vfr gene product, required for Pseudomonas aeruginosa exotoxin A and protease production, belongs to the cyclic AMP receptor protein family

The synthesis of exotoxin A (ETA) by Pseudomonas aeruginosa is a complex, regulated event. Several ETA putative regulatory mutants of P. aeruginosa PA103 have previously been characterized (S. E. H. West, S. A. Kaye, A. N. Hamood, and B. H. Iglewski, Infect. Immun. 62:897-903, 1994). In addition to ETA production, these mutants, PA103-15, PA103-16, and PA103-19, were also deficient in the production of protease and in regA P1 promoter activity. RegA is a positive regulator of ETA transcription. We cloned a gene, designated vfr for virulence factor regulator, that restored ETA and protease production to parental levels in these mutants. In addition, transcription from the regA P1 promoter was restored. In Escherichia coli, when vfr was overexpressed from a phage T7 promoter, a protein with an apparent molecular mass of 28.5 kDa was produced. Analysis of the deduced amino acid sequence of vfr revealed that the expected protein is 67% identical and 91% similar over a 202-amino-acid overlap to the E. coli cyclic AMP receptor protein (CAP or Crp). The cloned vfr gene complemented the beta-galactosidase- and tryptophanase-deficient phenotypes of E. coli RZ1331, a crp deletion mutant. However, the E. coli crp gene under the control of the tac promoter did not complement the ETA-deficient or protease-deficient phenotype of PA103-15 or PA103-16. The ability of vfr to restore both ETA and protease production to these mutants suggests that vfr is a global regulator of virulence factor expression in P. aeruginosa.

High-specificity DNA cleavage agent: design and application to kilobase and megabase DNA substrates

Strategies to cleave double-stranded DNA at specific DNA sites longer than those of restriction endonucleases (longer than 8 base pairs) have applications in chromosome mapping, chromosome cloning, and chromosome sequencing--provided that the strategies yield high DNA-cleavage efficiency and high DNA-cleavage specificity. In this report, the DNA-cleaving moiety copper:o-phenanthroline was attached to the sequence-specific DNA binding protein catabolite activator protein (CAP) at an amino acid that, because of a difference in DNA bending, is close to DNA in the specific CAP-DNA complex but is not close to DNA in the nonspecific CAP-DNA complex. The resulting CAP derivative, OP26CAP, cleaved kilobase and megabase DNA substrates at a 22-base pair consensus DNA site with high efficiency and exhibited no detectable nonspecific DNA-cleavage activity.

Location and orientation of an activating region in the Escherichia coli transcription factor, FNR

We have characterized a number of mutations in fnr that interfere with FNR-dependent transcription activation at two promoters where the FNR-binding site is centred around 41 1/2 bp upstream from the transcription start site. The substituted residues in all but one of these FNR mutants are clustered around a presumed surface-exposed beta-turn containing G85 which, we suggest, forms an activating region that contacts RNA polymerase at these promoters. Using the 'oriented heterodimers' method described elsewhere, we show that this activating region on the promoter-proximal subunit of the FNR dimer is sufficient to activate transcription initiation. In contrast, this region is not essential for activation of a third FNR-dependent promoter where the FNR-binding site is centred at 61 1/2 bp upstream from the transcription start site. However, a substitution at S73 interferes with FNR-dependent activation at both this promoter and promoters in which the FNR site is located at 41 1/2 bp from the transcript start, suggesting that FNR may contain a second activating region.

Transcription activation at Class I CAP-dependent promoters

Catabolite gene activator protein (CAP)-dependent promoters can be grouped into three classes, based on the requirement for transcription activation and the position of the DNA site for CAP. Class I CAP-dependent promoters require only CAP for transcription activation and have the DNA site for CAP located upstream of the DNA site for RNA polymerase. Amino acids 156 to 162 of the promoter-proximal subunit of CAP are essential for transcription activation at Class I CAP-dependent promoters, but are not essential for DNA binding, and are not essential for DNA bending. In the structure of the CAP-DNA complex, these amino acids are located in a surface loop and form a cluster on the surface of the CAP-DNA complex. Amino acids 261, 265, and 270 of the alpha subunit of RNA polymerase are essential for response to transcription activation by CAP at Class I CAP-dependent promoters. Several lines of evidence indicate that transcription activation at Class I CAP-dependent promoters requires a direct protein-protein contact between amino acids 156 to 162 of the promoter-proximal subunit of CAP and a molecule of RNA polymerase bound adjacent to CAP on the same face of the DNA helix. It is a strong possibility that this direct protein-protein contact involves amino acids 261 and 265 of the alpha subunit of RNA polymerase.

The Escherichia coli cAMP receptor protein (CRP) represses the Rhizobium meliloti dctA promoter in a cAMP-dependent fashion

The expression of the Rhizobium meliloti C4-dicarboxylic acid permease gene (dctA) is controlled by the sensor DctB and the transcriptional regulator, DctD. The R. meliloti Dct system has been reconstituted in Escherichia coli. Expression of the dctA promoter is DctBD dependent and is induced in the presence of C4-dicarboxylic acids (dCA). Other carbon sources also influence dctA expression. We demonstrate that the cAMP receptor protein (CRP) has a repressive effect on the dctA promoter. A mutated CRP molecule (CRP-H159L), unable to activate catabolic promoters (but still proficient in DNA binding), gives similar results. This suggests that the CRP-cAMP complex represses the dctA promoter activity by direct interaction with the DNA. Direct binding of the CRP-cAMP complex to the dctA promoter was confirmed in vitro by gel mobility-shift assays. Sequence analysis of the dctA promoter indicates that the most likely binding sites for CRP are the two confirmed DctD-binding sites. It is proposed that the CRP-cAMP complex competes with DctD for occupancy of these sites. Since in the presence of CRP-cAMP complex the uninduced levels of dctA expression are reduced, whereas induced levels are largely unaffected, such competition appears to be an essential regulatory feature of dctA expression.

Escherichia coli cyclic AMP receptor protein mutants provide evidence for ligand contacts important in activation

The three-dimensional model of the Escherichia coli cyclic AMP (cAMP) receptor protein (CRP) shows that several amino acids are involved as chemical contacts for binding cAMP. We have constructed and characterized mutants at four of these positions, E72, R82, S83, and R123. The mutations were made in wild-type crp as well as a cAMP-independent crp, crp*. The activities of the mutant proteins were characterized in vivo for their ability to activate the lac operon. These results provide genetic evidence to support that E72 and R82 are essential and S83 and R123 are important in the activation of CRP by cAMP.

Incorporation of an EDTA-metal complex at a rationally selected site within a protein: application to EDTA-iron DNA affinity cleaving with catabolite gene activator protein (CAP) and Cro

We have developed a simple procedure to incorporate an EDTA-metal complex at a rationally selected site within a full-length protein. Our procedure has two steps: In step 1, we use site-directed mutagenesis to introduce a unique solvent-accessible cysteine residue at the site of interest. In step 2, we derivatize the resulting protein with S-(2-pyridylthio)cysteaminyl-EDTA-metal, a novel aromatic disulfide derivative of EDTA-metal. We have used this procedure to incorporate an EDTA-iron complex at amino acid 2 of the helix-turn-helix motif of each of two helix-turn-helix motif sequence-specific DNA binding proteins, catabolite gene activator protein (CAP) and Cro, and we have analyzed EDTA-iron-mediated DNA affinity cleavage by the resulting protein derivatives. The CAP derivative cleaves DNA at base pair 2 of the DNA half-site in the protein-DNA complex, and the Cro derivative cleaves DNA at base pairs -3 to 5 of the DNA half-site in the protein-DNA complex. We infer that amino acid 2 of the helix-turn-helix motif of CAP is close to base pair 2 of the DNA half-site in the CAP-DNA complex in solution and that amino acid 2 of the helix-turn-helix motif of Cro is close to base pairs -3 to 5 of the DNA half-site in the Cro-DNA complex in solution.(ABSTRACT TRUNCATED AT 250 WORDS)

Determination of the orientation of a DNA binding motif in a protein-DNA complex by photocrosslinking

We have developed a straightforward biochemical method to determine the orientation of the DNA binding motif of a sequence-specific DNA binding protein relative to the DNA site in the protein-DNA complex. The method involves incorporation of a photoactivatable crosslinking agent at a single site within the DNA binding motif of the sequence-specific DNA binding protein, formation of the derivatized protein-DNA complex, UV-irradiation of the derivatized protein-DNA complex, and determination of the nucleotide(s) at which crosslinking occurs. We have applied the method to catabolite gene activator protein (CAP). We have constructed and analyzed two derivatives of CAP: one having a phenyl azide photoactivatable crosslinking agent at amino acid 2 of the helix-turn-helix motif of CAP, and one having a phenyl azide photoactivatable crosslinking agent at amino acid 10 of the helix-turn-helix motif of CAP. The results indicate that amino acid 2 of the helix-turn-helix motif is close to the top-strand nucleotides of base pairs 3 and 4 of the DNA half site in the CAP-DNA complex, and that amino acid 10 of the helix-turn-helix motif is close to the bottom-strand nucleotide of base pair 10 of the DNA half site in the CAP-DNA complex. The results define unambiguously the orientation of the helix-turn-helix motif relative to the DNA half site in the CAP-DNA complex. Comparison of the results to the crystallographic structure of the CAP-DNA complex [Schultz, S., Shields, S. & Steitz, T. (1991) Science 253, 1001-1007] indicates that the method provides accurate, high-resolution proximity and orientation information.

DNA binding specificity and sequence of Xanthomonas campestris catabolite gene activator protein-like protein

The Xanthomonas campestris catabolite gene activator protein-like protein (CLP) can substitute for the Escherichia coli catabolite gene activator protein (CAP) in transcription activation at the lac promoter (V. de Crecy-Lagard, P. Glaser, P. Lejeune, O. Sismeiro, C. Barber, M. Daniels, and A. Danchin, J. Bacteriol. 172:5877-5883, 1990). We show that CLP has the same DNA binding specificity as CAP at positions 5, 6, and 7 of the DNA half site. In addition, we show that the amino acids at positions 1 and 2 of the recognition helix of CLP are identical to the amino acids at positions 1 and 2 of the recognition helix of CAP:i.e., Arg at position 1 and Glu at position 2.

How the EcoRI endonuclease recognizes and cleaves DNA

One popular recombinant DNA tool is the EcoRI endonuclease, which cleaves DNA at GAATTC sites and serves as a paradigm for sequence specific DNA-enzyme interactions. The recently revised X-ray crystal structure of an EcoRI-DNA complex reveals EcoRI employs novel DNA recognition motifs, a four alpha-helix bundle and two extended chains, which project into the major groove to contact substrate purines and pyrimidines. Interestingly, pyrimidine contacts had been predicted based on genetic and biochemical studies. Current work focuses on the EcoRI active site structure, enzyme and substrate conformational changes during catalysis, and host-restriction system interactions.

Molecular analysis of the lac operon encoding the binding-protein-dependent lactose transport system and beta-galactosidase in Agrobacterium radiobacter

The genes coding for the binding-protein-dependent lactose transport system and beta-galactosidase in Agrobacterium radiobacter strain AR50 were cloned and partially sequenced. A novel lac operon was identified which contains genes coding for a lactose-binding protein (lacE), two integral membrane proteins (lacF and lacG), an ATP-binding protein (lacK) and beta-galactosidase (lacZ). The operon is transcribed in the order lacEFGZK. The operon is controlled by an upstream regulatory region containing putative -35 and -10 promoter sites, an operator site, a CRP-binding site probably mediating catabolite repression by glucose and galactose, and a regulatory gene (lacl) encoding a repressor protein which mediates induction by lactose and other galactosides in wild-type A. radiobacter (but not in strain AR50, thus allowing constitutive expression of the lac operon). The derived amino acid sequences of the gene products indicate marked similarities with other binding-protein-dependent transport systems in bacteria.

Genetic identification of the DNA binding domain of Escherichia coli LexA protein

Two genetic approaches were taken to define the DNA binding domain of LexA protein, the repressor of the Escherichia coli SOS regulon. First, several dominant negative lexA mutants defective in DNA binding were isolated. The mutations altered amino acids in a region similar to the helix-turn-helix, a DNA binding domain of other repressors and DNA binding proteins. Second, the region encoding the predicted DNA recognition helix was subjected to oligonucleotide-directed mutagenesis and mutant LexA proteins with altered or relaxed specificity for several recA operator positions were isolated. By examining the effects of a series of amino acid substitutions on repressor specificity, it was shown that a glutamic acid residue at position 45 in LexA protein is important for recognition of the first base pair (G.C) in the recA operator.

A RAP1-interacting protein involved in transcriptional silencing and telomere length regulation

The yeast RAP1 protein is a sequence-specific DNA-binding protein that functions as both a repressor and an activator of transcription. RAP1 is also involved in the regulation of telomere structure, where its binding sites are found within the terminal poly(C1-3A) sequences. Previous studies have indicated that the regulatory function of RAP1 is determined by the context of its binding site and, presumably, its interactions with other factors. Using the two-hybrid system, a genetic screen for the identification of protein-protein interactions, we have isolated a gene encoding a RAP1-interacting factor (RIF1). Strains carrying gene disruptions of RIF1 grow normally but are defective in transcriptional silencing and telomere length regulation, two phenotypes strikingly similar to those of silencing-defective rap1s mutants. Furthermore, hybrid proteins containing rap1s missense mutations are defective in an interaction with RIF1 in the two-hybrid system. Taken together, these data support the idea that the rap1s phenotypes are attributable to a failure to recruit RIF1 to silencers and telomeres and suggest that RIF1 is a cofactor or mediator for RAP1 in the establishment of a repressed chromatin state at these loci. By use of the two-hybrid system, we have isolated a mutation in RIF1 that partially restores the interaction with rap1s mutant proteins.

EcoRI DNA methyltransferase-DNA interactions

We present a novel strategy with synthetic hemimethylated DNA substrates containing uracil for thymine and inosine for guanosine replacements and EcoRI DNA methyltransferase to characterize the importance of major groove hydrophobic groups to the sequence-specific modification of DNA. The bacterial Mtase uses S-adenosyl-L-methionine to methylate the double-stranded DNA site 5'GAATTC3' at the N6 position of the central adenosine of each strand. Uracil substitution in either strand at the outer thymine (5'GAATUC3') causes 2.2- and 1.7-fold improvements in specificity (kcat/KmDNA). The fact that the specificity constant for the substrate containing uracil in both strands is identical to the value expected for noninteracting substitutions suggests that no significant methyltransferase-DNA interactions are altered beyond the site of either substitution. Similar analysis of the internal thymine (5'GAAUTC3') also shows these methyl groups to make a negative contribution to specificity, although the observed nonadditivity with the doubly modified substrate clearly shows methyltransferase-DNA interactions beyond the site of substitution to be affected in this case. To further probe the effect of analogue incorporation on methyltransferase-DNA interactions beyond the site of substitution, the relatively "silent" and additive uracil changes (5'GAATUC3') were combined with inosine for guanosine substitutions (e.g., 5'IAATTC3') known to have significant negative effects on specificity. In contrast to the additivity observed with the outer thymines, these studies show significant changes in methyltransferase-DNA interactions caused by the removal of the thymine methyls. Our results implicate a complex and flexible methyltransferase-DNA interface in which subtle structural changes in the substrate are transmitted over the entire canonical site.(ABSTRACT TRUNCATED AT 250 WORDS)

Non-additivity of sequence-specific enzyme-DNA interactions in the EcoRI DNA methyltransferase

We describe a novel strategy to characterize protein-DNA interactions involving monomeric enzymes such as DNA methyltransferases (Mtases). This strategy is applied to our investigation of the EcoRI DNA Mtase, which binds its double stranded recognition site 5'-G-AATTC-3' and methylates the central adenosine of each strand using S-adenosyl-L-methionine as the methyl donor. We show that prior methylation of adenosine in either strand does not perturb catalysis. In contrast, substrates substituted with deoxyinosine at either guanosine position (T-BMI5 and TI5-BM) show the minor groove residing N2 amino group of both guanosines contribute to DNA recognition since specificity constants for the modified substrates are reduced 13 and 39 fold. Similar analysis of a substrate containing deoxyinosine at both positions (TI5-BMI5) clearly shows that some communication occurs between the sites. To determine the extent to which structural changes in the DNA alone contribute to this lack of additivity, we performed DNA melting analysis of the singly and doubly substituted substrates, and also found non-additivity. Although our functional and structural analyses suggest that deoxyinosine incorporation causes long range conformational effects, the similarity of KmAdoMet for all substrates suggests that no large-scale structural changes occur in the Mtase-DNA-AdoMet complex. Our results support the following conclusions: 1) The non-additivity shown in this system contrasts with the widespread demonstration of additivity involving repressors [Lehming et al., 1990; Takeda et al., 1989; Ebright et al., 1987], suggesting that sequence discrimination by enzymes may involve more complex mechanisms. Further, this non-additivity precludes quantitative assignment of individual interactions and we suggest that future analyses of this and related enzyme systems with base analogs include detailed information about the long range structural consequences of individual substitutions. 2) Although TI5-BM and T-BMI5 are shown to be radically different by thermodynamic analysis, the similar specificity constants with the Mtase suggest that the underlying structural differences (e.g., altered helical parameters of the DNA) are not critical for sequence-recognition. 3) The significance of minor groove Mtase-DNA interactions to specificity is confirmed.