Detecting the DNA kinks in a DNA-CRP complex in solution with iodine-125 radioprobing

Auger-electron-emitting radioisotopes such as 125I produce DNA strand breaks within nanometer range of the decay site. Here we analyze these breaks in order to study changes in DNA conformation upon binding with cyclic AMP receptor protein (CRP) in solution. The clear difference we found in break frequency in the CRP-DNA complex, as compared to the naked DNA duplex, correlates with the increased distances between the deoxyriboses and the radioiodine atom caused by the CRP-induced kink observed in the cocrystal. Thus, we demonstrate that 125I radioprobing can be used to study fine conformational changes of DNA within DNA-protein complexes.

Analysis of genetic instability during mammary tumor progression using a novel selection-based assay for in vivo mutations in a bacteriophage lambda transgene target

Genetic instability is thought to be responsible for the numerous genotypic changes that occur during neoplastic transformation and metastatic progression. To explore the role of genetic instability at the level of point mutations during mammary tumor development and malignant progression, we combined transgenic mouse models of mutagenesis detection and oncogenesis. Bitransgenic mice were generated that carried both a bacteriophage lambda transgene to assay mutagenesis and a polyomavirus middle T oncogene, mammary gland-targeted expression of which led to metastatic mammary adenocarcinomas. We developed a novel assay for the detection of mutations in the lambda transgene that selects for phage containing forward mutations only in the lambda cII gene, using an hfl- bacterial host. In addition to the relative ease of direct selection, the sensitivity of this assay for both spontaneous and chemically induced mutations was comparable to the widely used mutational target gene, lambda lacI, making the cII assay an attractive alternative for mutant phage recovery for any lambda-based mouse mutagenesis assay system. The frequencies of lambda cII- mutants were not significantly different in normal mammary epithelium, primary mammary adenocarcinomas, and pulmonary metastases. The cII mutational spectra in these tissues consisted mostly of G/C-->A/T transitions, a large fraction of which occurred at CpG dinucleotides. These data suggest that, in this middle T oncogene model of mammary tumor progression, a significant increase in mutagenesis is not required for tumor development or for metastatic progression.

Analysis of the spacer DNA between the cyclic AMP receptor protein binding site and the lac promoter

The role of the spacer region DNA between the cyclic AMP receptor protein (CRP) site and the RNA polymerase in the lac promoter was examined. We wanted to determine whether the wild-type DNA sequence of this region was an absolute requirement for CRP activation of lac transcription. The sequence of a 9-bp stretch of the spacer, from -41 to -49 relative to the start of transcription, was randomized, and the effect of randomization on lac expression was investigated in vitro and in vivo. We found that the spacer contains no specific sequence determinants for CRP activation of lac transcription; fewer than 1% of the mutants displayed greater than a 50% decrease in CRP activation of lac transcription.

Effect of the FruR regulator on transcription of the pts operon in Escherichia coli

The promoters of the pts operon of Escherichia coli are controlled by the cyclic AMP receptor protein (CRP) complexed with cAMP (CRP.cAMP). In addition, glucose stimulates pts operon expression in vivo. The pts promoter region has a fructose repressor (FruR)-binding site (the FruR box) that partially overlaps with one of the CRP.cAMP-binding sites. The effects of the pleiotropic transcriptional regulator FruR on pts operon expression were studied to determine whether the in vivo glucose effect on pts operon expression is mediated by FruR. In vitro, FruR can repress P1b transcription, which is activated by CRP.cAMP, and restore P1a transcription, which is repressed by CRP.cAMP. FruR can displace CRP.cAMP from its binding site in the presence of RNA polymerase even though FruR and CRP.cAMP can bind simultaneously to their partially overlapping binding sites in the absence of RNA polymerase. FruR had very little effect on the transcription of the P0 promoter, which is most important for regulation by glucose. Consistent with the in vitro results, pts P0 transcription did not increase as much in cells grown in the presence of fructose or in fruR- mutant cells as in cells grown in the presence of glucose. These results suggest that FruR alone does not mediate the in vivo glucose effect on pts operon expression.

CRP-DNA complexes: inducing the A-like form in the binding sites with an extended central spacer

The consensus DNA sequence for binding of the Escherichia coli cyclic AMP receptor protein (CRP) has two symmetrically related inverted recognition elements TGTGA:TCACA, separated by a variable spacer, normally 6 bp long. We have shown that the CRP-cAMP complex, when bound to synthetic binding sites with an extended 8 bp spacer segment, induces an increase in the DNA circular dichroism (CD). The CD change at lambda > 275 nm agrees with the shift of approximately one helical turn of DNA into A-like form. The B-conformation is preserved for CRP binding sites similar to that in the lac and uxaCA promoters with 6 bp spacers. Another effect accompanying DNA binding is a dramatic increase of the negative CD magnitude in the spectral region of the ligand cAMP, at lambda < 272 nm. This effect is observed when CRP binds to specific sites with 6 or 8 bp spacers as well as to non-specific DNA. We reason that the A-like form arises by compressing and unwinding the DNA in CRP-DNA complexes having 8 bp central spacers. This serves to maintain a fixed length and twisting angle and is controlled by the protein's relatively rigid frame. This model is consistent with the observation that some binding sites with 6 bp spacers may also show the CD increase inherent to the sites with the extended 8 bp spacers. These 6 bp spacers are characterized by an increased twisting angle that requires their unwinding to bind to CRP. We propose that a mutual adaptation between CRP and binding sites by local untwisting and a B-->A-like transition in the DNA is of general importance and may occur in other protein-DNA complexes, such as the complex of RNA polymerase with promoter DNA.

An arcane role of DNA in transcription activation

The mechanism by which the cAMP receptor protein (CRP) activates transcription has been investigated using the lac promoter of Escherichia coli. For transcription activation, an interaction between DNA-bound CRP and RNA polymerase is not sufficient. CRP must bind to a site in the same DNA and close to the promoter. CRP action requires an intact spacer DNA to provide a rigid support in building a CRP-RNA polymerase protein bridge or to allow a conformational change in the DNA to be transmitted to the lac promoter using the protein bridge as a structural support.

Promoter switch in the Escherichia coli pts operon

The ptsH operon of Escherichia coli is controlled by two promoters P0 and P1, each of which is regulated by cyclic AMP receptor protein (CRP) complexed with cAMP (CRP.cAMP). We have studied the in vitro as well as in vivo transcriptional regulation of these two promoters. Each promoter exhibits a switching mechanism in vitro, where, depending upon the presence or absence of CRP.cAMP, transcription is initiated from different start sites termed a and b. P0 (P0a) is affected by supercoiling: when the template is linear, transcription initiation is switched to a site 3 base pairs upstream (P0b) and becomes more CRP.cAMP dependent. Transcription from the P1 promoter (P1a) switches initiation sites to 7 base pairs downstream (P1b) in the presence of CRP.cAMP. Most transcription in vivo was from P1a, and P0b could not be detected in vivo. Glucose has independent positive effects on pts expression in vivo. The results indicate that the two different regulatory mechanisms (one through CRP.cAMP, the other through glucose) are working together for fine control of pts expression.

Bacterial expression of Chinese hamster regulatory type-I and catalytic subunits of cyclic AMP-dependent protein kinase and mutational analysis of the type-I regulatory subunit

The type-I regulatory subunit (RI) of the cyclic AMP-dependent protein kinase (PKA) from Chinese hamster ovary (CHO) cells has been cloned and expressed in a strain of BL21(DE3) Escherichia coli lacking adenylate cyclase [BL21(DE3)/delta cya]. RI expressed in this bacterial system free of cyclic AMP is soluble and can reconstitute functional PKA. Recombinant CHO C alpha is predominantly insoluble with some active soluble protein. C beta is entirely insoluble and inactive. Soluble recombinant RI and soluble recombinant C alpha can associate in vitro and be activated by cyclic AMP. Six site-directed mutations of RI were generated to study the interaction of cyclic AMP with RI and RI-C alpha subunit interactions. Four cyclic AMP-binding-site point mutants were generated [W261R (tryptophan to arginine at position 261), a novel mutation in site A; V376G, a novel mutation in site B; G200E (site A), and Y370F (site B), previously described in bovine RI were introduced into the CHO RI for comparison purposes]. Mutants W261R, Y370F, and G200E demonstrated decreased 8-N3-[3H]cyclic AMP binding as well as 5-fold reduced affinity for [3H]cyclic AMP, with threefold increased EC50 values for cyclic AMP activation of kinase activity from reconstituted mutant holoenzymes. The mutation at V376G did not alter cyclic AMP binding or activation by cyclic AMP of mutant holoenzyme. A truncation mutant, G200Stop, which lacks both cyclic AMP-binding sites, did not bind cyclic AMP but can inhibit C alpha subunit activity. A novel mutation outside the cyclic AMP-binding regions of RI (V89A) weakened the interaction with C alpha indicated by a 7-fold lower EC50 for mutant holoenzyme activation by cyclic AMP.

Promoter resurrection by activators--a minireview

Frequently, in nature, defective promoters can be resurrected by activator proteins in response to cellular demands. The activators bind to nearby DNA sites for action. Various protein-protein and DNA-protein contacts involving activators, RNA polymerase, and different segments of DNA in and around a defective promoter form a DNA-multiprotein complex (cage) which enhances transcription.

Pivotal role of amino acid at position 138 in the allosteric hinge reorientation of cAMP receptor protein

The cAMP receptor protein (CRP) of Escherichia coli needs cAMP for an allosteric change to regulate gene expression by binding to specific DNA sites. The hinge region connecting the DNA-binding domain to the cAMP-binding domain has been proposed to participate in the cAMP-induced allosteric change necessary to adjust C and D alpha-helices for movement of the DNA-binding F alpha-helix away from the protein surface. The role of the hinge region for a conformation change in CRP was tested by studying the effects of single amino acid substitutions at residue 138 located within the hinge. Physiological studies of wild-type and mutant cells and biochemical analysis of purified wild-type and mutant CRP revealed at least three groups of altered CRPs: (i) CRP that behaves like wild type (CRP+); (ii) CRP that binds cAMP but does not complete the structural changes required for specific DNA binding, proteolytic cleavage, and transcription activation (CRPallo); and (iii) CRP that shows some or all of these conformational changes without cAMP (CRP*). These results show a pivotal role of position 138 from which change emanates and provide further evidence that a hinge reorientation involving residue 138 is involved in the interhelical adjustments.

Allosteric changes in the cAMP receptor protein of Escherichia coli: hinge reorientation

The cAMP receptor protein (CRP) of Escherichia coli is a dimer of a two-domain subunit. It requires binding of cAMP for a conformational change in order to function as a site-specific DNA-binding protein that regulates gene activity. The hinge region connecting the cAMP-binding domain to the DNA-binding domain is involved in the cAMP-induced allosteric change. We studied the structural changes in CRP that are required for gene regulation by making a large number of single and double amino acid substitutions at four different positions in or near the hinge. To achieve cAMP-independent transcription by CRP, amino acid residues 138 (located within the hinge region) and 141 (located in the D alpha-helix adjacent to the hinge) must be polar. This need for polar residues at positions 138 and 141 suggests an interaction that causes the C and D alpha-helices to come together. As a consequence, the F alpha-helix is released from the D alpha-helix and can interact with DNA. At position 144 in the D alpha-helix and within interacting distances of the F alpha-helix, replacement of alanine by an amino acid with a larger side chain, regardless of its nature, allows cAMP independence. This result indicates that pushing against the F alpha-helix may be a way of making the helix available for DNA binding. We believe that the cAMP-induced allosteric change involves similar hinge reorientation to adjust the C and D alpha-helices, allowing outward movement of the F alpha-helix.

Evidence for two promoters upstream of the pts operon: regulation by the cAMP receptor protein regulatory complex

Several potential target sites for multiple regulatory mechanisms were previously identified in the 5' flanking region of the pts operon. We have investigated the in vitro interactions of the cAMP receptor protein (CRP).cAMP regulatory complex with two DNA binding sites, by gel mobility-shift assays, and report the analysis of the functional role of each of the binding sites in vivo. Promoter-reporter gene fusion studies identified two CRP.cAMP-dependent promoters (the previously identified P1 and another promoter, P0) upstream of ptsH. The crr promoters (P2) within ptsI may be negatively regulated by CRP.cAMP.

A mutation defining ultrainduction of the Escherichia coli gal operon

Tn10 insertion in the galS (ultrainduction factor) gene of Escherichia coli allows the gal operon to be constitutively expressed at a very high level, equal to that seen in a delta galR strain in the presence of an inducer. The insertion has been mapped by criss-cross Hfr matings and by marker rescue into Kohara phages at 46 min on the E. coli chromosome.

Further inducibility of a constitutive system: ultrainduction of the gal operon

In wild-type Escherichia coli, expression of the gal operon is negatively regulated by the Gal repressor and is induced 10- to 15-fold when the repressor is inactivated by an inducer. In strains completely deleted for galR, the gene which encodes the Gal repressor, the operon is derepressed by only 10-fold without an inducer. But this derepression is increased further by threefold during cell growth in the presence of an inducer, D-galactose or D-fucose. This phenomenon of extreme induction in the absence of Gal repressor is termed ultrainduction--a manifestation of further inducibility in a constitutive setup. Construction and characterization of gene and operon fusion strains between galE and lacZ, encoding beta-galactosidase as a reporter gene, show that ultrainduction occurs at the level of transcription and not translation. Transcription of the operon, from both the cyclic AMP-dependent P1 and the cyclic nucleotide-independent P2 promoters, is subject to ultrainduction. The wild-type galR+ gene has an epistatic effect on ultrainducibility: ultrainduction is observed only in cells devoid of Gal repressor protein. Titration experiments show the existence of an ultrainducibility factor that acts like a repressor and functions by binding to DNA segments (operators) to which Gal repressor also binds to repress the operon.

Characterization of the binding of cAMP and cGMP to the CRP*598 mutant of the E. coli cAMP receptor protein

Wild type cAMP receptor protein (CRP) activates in vitro lac transcription only in the presence of cAMP. In contrast the mutant CRP*598 (Arg-142 to His, Ala-144 to Thr) can activate lac transcription in the absence of cyclic nucleotide or at concentrations of cAMP below that required by CRP. To further characterize the properties of CRP*598, the binding of cAMP and cGMP to CRP and CRP*598 has been determined. The intrinsic binding constant (K) values obtained for cAMP binding are: CRP, 1.9 x 10(4) M-1; CRP*598, 3.8 x 10(5) M-1. The K values obtained for cGMP binding are: CRP, 2.9 x 10(4) M-1; CRP*598, 2.7 x 10(4) M-1. The results indicate that the affinity of CRP and CRP*598 for cGMP is relatively unchanged while the affinity of CRP*598 for cAMP is approximately twenty times greater than that shown by CRP. Binding of cAMP by CRP and cGMP by CRP or CRP*598 exhibits slight negative cooperativity. The major difference seen is that CRP*598 binds cAMP with strong positive cooperativity. The importance of the unsubstituted N6 position of the adenine moiety is also shown by the similar affinity of both forms of CRP for N6-butyryl cAMP. The cAMP binding properties evinced by CRP*598 suggest that its intrinsically altered conformation may be related to that assumed by CRP in a CRP-DNA or a cAMP-CRP-DNA complex.

Cooperative DNA binding of heterologous proteins: evidence for contact between the cyclic AMP receptor protein and RNA polymerase

Four cAMP-independent receptor protein mutants (designated CRP* mutants) isolated previously are able to activate in vivo gene transcription in the absence of cAMP and their activity can be enhanced by cAMP or cGMP. One of the four mutant proteins, CRP*598 (Arg-142 to His, Ala-144 to Thr), has been characterized with regard to its conformational properties and ability to bind to and support abortive initiation from the lac promoter. In the absence of cGMP, CRP*598 shows a more open conformation than CRP, as indicated by its sensitivity to proteolytic attack and 5,5'-dithiobis(2-nitrobenzoic acid)-mediated subunit crosslinking. Binding of wild-type CRP to its site on the lac promoter and activation of abortive initiation by RNA polymerase on this promoter are effected by cAMP but not by cGMP. CRP*598 can activate lacP+-directed abortive initiation in the presence of cAMP and less efficiently in the presence of cGMP or in the absence of cyclic nucleotide. DNase I protection ("foot-printing") indicates that cAMP-CRP* binds to its site on the lac promoter whereas unliganded CRP* and cGMP-CRP* form a stable complex with the [32P]lacP+ fragment only in the presence of RNA polymerase, showing cooperative binding of two heterologous proteins. This cooperative binding provides strong evidence for a contact between CRP and RNA polymerase for activation of transcription. Although cGMP binds to CRP, it cannot replace cAMP in effecting the requisite conformational transition necessary for site-specific promoter binding. In contrast, the weakly active unliganded CRP*598 can be shifted to a functional state not only by cAMP but also by cGMP and RNA polymerase.

Cyclic AMP-induced conformational change of cyclic AMP receptor protein (CRP): intragenic suppressors of cyclic AMP-independent CRP mutations

We isolated and characterized crp mutations in Escherichia coli that allow cyclic AMP (cAMP) receptor protein to function without cAMP. These mutants defined a region involved in the cAMP-induced allosteric change of cAMP receptor protein that is necessary for activation of the protein. Currently, we have isolated intragenic suppressors of the crp mutations. These crp (Sup) mutants require cAMP for activity. The crp (Sup) mutations map in regions which define new sites of changes involved in cAMP receptor protein activation. From these results, we suggest that to activate cAMP receptor protein cAMP brings about (i) a hinge reorientation to eject the DNA-binding F alpha-helices, (ii) proper alignment between the two subunits, and (iii) an adjustment between the position of the two domains. Cyclic GMP fails to effect the last step.

Sites of allosteric shift in the structure of the cyclic AMP receptor protein

We have characterized crp mutations in E. coli that allow CRP to function without cAMP. crp* mutants carrying a deletion of the gene encoding adenylate cyclase (cya) show significant lac expression. Cyclic GMP, normally an ineffective activator of CRP+, can stimulate these mutant CRP*s to permit greater lac expression in vivo. Cyclic AMP binding to the amino-terminal domain of CRP+ induces an allosteric transition that changes the DNA-binding property of the carboxy domain. The CRP* phenotype is caused by substitution of amino acids with bulkier side chains in the D alpha-helix of the protein's carboxy domain, near the hinge connecting the two domains. These results are consistent with a model in which the mutant CRP*s assume, in part, a conformation normally evoked only by cAMP binding: one in which the relative orientation of the C, D, and F alpha-helices is altered. We define precisely the amino acids of these alpha-helices that interact to cause the allosteric shift.

Occurrence of Vibrio cholerae serotype O1 in Maryland and Louisiana estuaries

Vibrio cholerae serotype O1 has been isolated from Chesapeake Bay in Maryland and estuaries and sewers in Louisiana. The occurrence of V. cholerae O1 in the aquatic environment in the absence of human disease suggests that this organism survives and multiples in the natural environment.

Comparative study of the aerobic, heterotrophic bacterial flora of Chesapeake Bay and Tokyo Bay

A comparative study of the bacterial flora of the water of Chesapeake Bay and Tokyo Bay was undertaken to assess similarities and differences between the autochthonous flora of the two geographical sites and to test the hypothesis that, given similarities in environmental parameters, similar bacterial populations will be found, despite extreme geographic distance between locations. A total of 195 aerobic, heterotrophic bacterial strains isolated from Chesapeake Bay and Tokyo Bay water were examined for 115 biochemical, cultural, morphological, nutritional, and physiological characters. The data were analyzed by the methods of numerical taxonomy. From sorted similarity matrices, 77% of the isolates could be grouped into 30 phena and presumptively identified as Acinetobacter-Moraxella, Caulobacter, coryneforms, Pseudomonas, and Vibrio spp. Vibrio and Acinetobacter species were found to be common in the estuarine waters of Chesapeake Bay, whereas Acinetobacter-Moraxella and Caulobacter predominated in Tokyo Bay waters, at the sites sampled in the study.