Search

The structure of the cyclic AMP receptor-promoter complex in solution was studied in the range of 0.2-50 microM cAMP by measurements of the electric birefringence at 0.1 M salt using a lac promoter DNA with 121 bp and with the CAP binding site at its center. An excess of protein required for complete conversion of the promoter DNA into the specific complex seems to be partly due to nonspecific binding. The specific complex is associated with a decay time constant of 1.36 micros at 3 degrees C, a positive birefringence, and a permanent dipole moment demonstrated by pulse reversal. These attributes were observed at cAMP concentrations between 3 and 50 muM and are characteristic of the specific complex. Model calculations demonstrate that the DNA bending angle under these conditions is 92 degrees . The observed positive birefringence does not result from the combination of the calculated quasi-permanent dipole and the orientation of the helix axes alone but is due to coupling of translational and rotational diffusion. When the cAMP concentration is decreased below 3 microM, the positive birefringence turns to a negative one with a transition center at 1.5 microM. The transition is too narrow for a model with induction of the specific cyclic AMP receptor-promoter complex after binding of a single cAMP to the cyclic AMP receptor dimer but is consistent with induction of this complex after binding of two cAMP molecules. The cyclic AMP receptor-promoter complex is driven into its specific bent form in vitro in the range of cAMP concentrations corresponding to that required for gene regulation in vivo.

Many biological processes, such as gene transcription and replication, involve opening and closing of short regions of double-stranded DNA (dsDNA). Few techniques, however, can study these processes in real time or at the single-molecule level. Here, we present a Förster resonance energy transfer (FRET) assay that monitors the state of DNA (double- vs single-stranded) at a specific region within a DNA fragment, at both the ensemble level and the single-molecule level. The assay utilizes two closely spaced fluorophores: a FRET donor fluorophore (Cy3B) on the first DNA strand and a FRET acceptor fluorophore (ATTO647N) on the complementary strand. Because our assay is based on quenching and dequenching FRET processes, i.e., the presence or absence of contact-induced fluorescence quenching, we have named it a "quenchable FRET" assay or "quFRET". Using lac promoter DNA fragments, quFRET allowed us to sense transcription bubble expansion and compaction during abortive initiation by bacterial RNA polymerase. We also used quFRET to confirm the mode of action of gp2 (a phage-encoded protein that acts as a potent inhibitor of Escherichia coli transcription) and rifampicin (an antibiotic that blocks transcription initiation). Our results demonstrate that quFRET should find numerous applications in many processes involving DNA opening and closing, as well as in the development of new antibacterial therapies involving transcription.

Lactose repressor protein (LacI) utilizes an allosteric mechanism to regulate transcription in Escherichia coli, and the transition between inducer- and operator-bound states has been simulated by targeted molecular dynamics (TMD). The side chains of amino acids 149 and 193 interact and were predicted by TMD simulation to play a critical role in the early stages of the LacI conformational change. D149 contacts IPTG directly, and variations at this site provide the opportunity to dissect its role in inducer binding and signal transduction. Single mutants at D149 or S193 exhibit a minimal change in operator binding, and alterations in inducer binding parallel changes in operator release, indicating normal allosteric response. The observation that the double mutant D149A/S193A exhibits wild-type properties excludes the requirement for inter-residue hydrogen bond formation in the allosteric response. The double mutant D149C/S193C purified from cell extracts shows decreased sensitivity to inducer binding while retaining wild-type binding affinities and kinetic constants for both operator and inducer. By manipulating cysteine oxidation, we show that the more reduced state of D149C/S193C responds to inducer more like the wild-type protein, whereas the more oxidized state displays diminished inducer sensitivity. These features of D149C/S193C indicate that the novel disulfide bond formed in this mutant impedes the allosteric transition, consistent with the role of this region predicted by TMD simulation. Together, these results establish the requirement for flexibility in the spatial relationship between D149 and S193 rather than a specific D149-S193 interaction in the LacI allosteric response to inducer.

HMGB proteins are abundant non-histone components of eukaryotic chromatin. The biological function of DNA sequence-nonspecific HMGB proteins is obscure. These proteins are composed of one or two conserved HMG box domains, each forming three alpha-helices that fold into a sequence-nonspecific DNA-binding module recognizing the DNA minor groove. Box A and box B homology domains have subtle sequence differences such that box B domains bend DNA strongly while DNA bending by isolated box A domains is weaker. Both box A and box B domains preferentially bind to distorted DNA structures. Here we show using DNA cyclization kinetics assays in vitro and Escherichia coli DNA looping assays in vivo that an isolated HMG box A domain derived from human HMGB2 folds poorly and does not enhance apparent DNA flexibility. Surprisingly, substitution of a small number of cationic residues from the N-terminal leader of a functional yeast box B protein, Nhp6Ap, confers the ability to enhance DNA flexibility. These results demonstrate important roles for cationic leader amino acids in HMGB folding, DNA interaction, and DNA bending.

Fapy.dG is produced in DNA as a result of oxidative stress from a precursor that also forms OxodG. Bypass of Fapy.dG in a shuttle vector in COS-7 cells produces G --T transversions slightly more frequently than does OxodG (Kalam, M. A., et al. (2006) Nucleic Acids Res. 34, 2305). The effect of Fapy.dG on replication in Escherichia coli was studied by transfecting M13mp7(L2) bacteriophage DNA containing the lesion within the lacZ gene in 4 local sequence contexts. For comparison, experiments were carried out side-by-side on OxodG. The efficiency of lesion bypass was determined relative to that of a genome containing native nucleotides. Fapy.dG was bypassed less efficiently than OxodG. Bypass efficiency of Fapy.dG and OxodG increased modestly in SOS-induced cells. Mutation frequencies at the site of the lesions in the originally transfected genomes were determined using the REAP assay (Delaney, J. C., Essigmann, J. M. (2006) Methods Enzymol. 408, 1). G --T transversions were the only mutations observed above background when either Fapy.dG or OxodG was bypassed. OxodG mutation frequencies ranged from 3.1% to 9.8%, whereas the G --T transversion frequencies observed upon Fapy.dG bypass wereor=1.9% in wild-type E. coli. In contrast to OxodG bypass, Fapy.dG mutation frequencies were unaffected by carrying out experiments in mutM/mutY cells. Overall, these experiments suggest that Fapy.dG is at most weakly mutagenic in E. coli. Steady-state kinetic experiments using the Klenow fragment of DNA polymerase I from E. coli suggest that a low dA misincorporation frequency opposite Fapy.dG and inefficient extension of a Fapy.dG:dA base pair work synergistically to minimize the levels of G --T transversions.

Protein structural change underlies many signal transduction processes. Although end-state structures are known for various allosteric proteins, intermediates are difficult to observe. Recently, targeted molecular dynamics simulation (TMD) was used to examine the conformational transition and predict relevant intermediates for wild-type lactose repressor (LacI). A catalog of involved residues suggests that the transition of this homodimer is asymmetric and that K84 is a prominent participant in the dynamic N-subdomain interface. Previous experiments indicated that hydrophobic substitutions at position 84 engender slowed, biphasic inducer binding kinetics, which might reflect the same phenomena observed in TMD. Here, we report biochemical confirmation that DNA and inducer binding remain allosterically linked in K84A and K84L, albeit with a differential smaller than that found in wild-type LacI. Other features of these mutant proteins are consistent with an allosteric conformational shift that approximates that of the wild type. As a consequence, these repressors can be utilized to explore an unanswered question about LacI function: How many inducers (one or two per dimer) are required to diminish operator affinity? The biphasic natures of the K84L and K84A inducer association rates allow direct correlation between the two distinct inducer binding events and operator release. Indeed, the kinetics of operator release for the K84A and K84L closely parallel those for the second inducer binding event. Together with implications from previous equilibrium results for wild-type and mutant proteins, these kinetic data demonstrate that binding of two inducers per dimeric DNA binding unit is required to release the operator in these variant LacI proteins.

Cytochrome c(550) of the photosystem II complex of cyanobacteria is an unusual member of the large protein family of monoheme c-type cytochromes. Despite the fact that it shares considerable amino acid sequence similarity and has a protein fold similar to the other members of the family, Cyt.c(550) has a midpoint potential (E(m7) = -250 mV) that is much lower than the positive midpoint potentials characteristic (E(m7) = 100-300mV) of this cytochrome family. An E. coli heterologous expression system involving secretion of the recombinant protein from Synechocystis PCC6803 to the periplasm was utilized to allow production of wild-type and mutant forms of the cytochrome. For most of the variants studied, the yield of protein was significantly enhanced by growth at 28 degrees C and inclusion of sucrose and betaine, in addition to isopropyl-beta-d-thiogalactoside (IPTG), to the growth medium of the E. coli expression host. Analysis of the protein products revealed that the wild-type protein maintained the redox and visible spectroscopic characteristics of the authentic protein. Mutations in the residues engaging in hydrogen bond interactions with the heme propionate (Asn49) and the axial 6th ligand His92 (Pro93) resulted in small (12-20 mV), but reproducible, upshifts in midpoint redox potential. Substitution of the axial ligand His92 with Met produced no discernible changes in the optical spectrum relative to the wild-type despite the fact that in this mutant, unlike the others studied here, the thioether linkage either was not formed or was highly labile as evidenced by loss of the heme during SDS-PAGE. On the other hand, the midpoint potential of the C550-H92M mutant was upshifted by approximately 70 mV. This value is significantly less of a perturbation than that observed in a similar mutant that is natively expressed in Thermosynechoccocus but appears to have an intact thioether linkage between the heme and the polypeptide moiety.

Transporter ProP of Escherichia coli, a member of the major facilitator superfamily (MFS), acts as an osmosensor and an osmoregulator in cells and after purification and reconstitution in proteoliposomes. H(+)-osmoprotectant symport via ProP is activated when medium osmolality is elevated with membrane impermeant osmolytes. The three-dimensional structure of ProP was modeled with the crystal structure of MFS member GlpT as a template. This GlpT structure represents the inward (or cytoplasm)-facing conformation predicted by the alternating access model for transport. LacZ-PhoA fusion analysis and site-directed fluorescence labeling substantiated the membrane topology and orientation predicted by this model and most hydropathy analyses. The model predicts the presence of a proton pathway within the N-terminal six-helix bundle of ProP (as opposed to the corresponding pathway found within the C-terminal helix bundle of its paralogue, LacY). Replacement of residues within the N-terminal helix bundle impaired the osmotic activation of ProP, providing the first indication that residues outside the C-terminal domain are involved in osmosensing. Some residues that were accessible from the periplasmic side, as predicted by the structural model, were more susceptible to covalent labeling in permeabilized membrane fractions than in intact bacteria. These residues may be accessible from the cytoplasmic side in structures not represented by our current model, or their limited exposure in vivo may reflect constraints on transporter structure that are related to its osmosensory mechanism.

Thyroid transcription factor 1 (TTF-1) is a 42 kDa homeodomain (HD) containing the tissue-specific transcription factor of Nkx2 family members (also termed TEBP and Nkx2.1). TTF-1 is an essential transcription factor required for lung development and lung-specific gene expression. Transgenic mice carrying TTF-1 DNA-binding site mutations completely abolish expression of the human surfactant protein B (hSP-B) 1.5 kb lacZ reporter gene in the lung in vivo. Acetylation of transcription factors by nuclear receptor coactivators is an important mechanism for gene regulation. TTF-1 is acetylated by nuclear receptor coactivators including the activator of the thyroid and retinoic acid receptor, CREB-binding protein, and steroid receptor coactivator 1 (SRC-1) in cell transfection and immunoprecipitation studies. A glutathionine transferase pull-down assay shows TTF-1 direct interaction with the SRC-1 histone acetyltransferase domain. Site-specific mutagenesis identifies that the lysine residue at position 182 in the TTF-1 HD is acetylated in respiratory epithelial cells. Mutation at this acetylation site shows a dominant negative effect on SP-B gene transcription. The study supports a concept that acetylation is an important mechanism for TTF-1 activity.

The vast increase in available data from the "-omics" revolution has enabled the fields of structural proteomics and structure prediction to make great progress in assigning realistic three-dimensional structures to each protein molecule. The challenge now lies in determining the fine structural details that endow unique functions to sequences that assume a common fold. Similar problems are encountered in understanding how distinct conformations contribute to different phases of a single protein's dynamic function. However, efforts are hampered by the complexity of these large, three-dimensional molecules. To overcome this limitation, structural data have been recast as two-dimensional networks. This analysis greatly reduces visual complexity but retains information about individual residues. Such diagrams are very useful for comparing multiple structures, including (1) homologous proteins, (2) time points throughout a dynamics simulation, and (3) functionally different conformations of a given protein. Enhanced structural examination results in new functional hypotheses to test experimentally. Here, network representations were key to discerning a difference between unliganded and inducer-bound lactose repressor protein (LacI), which were previously presumed to be identical structures. Further, the interface of unliganded LacI was surprisingly similar to that of the K84L variant and various structures generated by molecular dynamics simulations. Apo-LacI appears to be poised to adopt the conformation of either the DNA- or inducer-bound structures, and the K84L mutation appears to freeze the structure partway through the conformational transition. Additional examination of the effector binding pocket results in specific hypotheses about how inducer, anti-inducer, and neutral sugars exert their effects on repressor function.

HIV-1 reverse transcriptase (RT) is a highly error prone DNA polymerase. We assessed whether the ability of RT to bind nucleotide substrates affects viral mutagenesis. Structural modeling predicts that the V148 and Q151 residues influence the interaction between RT and the incoming dNTP. When we introduce either a V148I or Q151N mutation, RT fidelity increases 8.7- or 13-fold, respectively, as measured by the M13 lacZalpha forward mutation assay. Interestingly, pre-steady state kinetic studies demonstrated that these mutations do not alter polymerase fidelity during the first step of mutation synthesis, misincorporation. Rather, the V148I and Q151N mutations alter RT fidelity by weakening the ability of the polymerase to complete mismatch extension, the second step of mutation synthesis. While both these mutations minimally affect the binding of RT (K(D)) to a mismatched template-primer complex (T/P), these mutant RTs are significantly impaired in their ability to bind (K(d)) and chemically incorporate (k(pol)) nucleotide substrate onto a mismatched T/P. These differences in binding and catalysis translate into 24- and 15.9-fold increase in mismatch extension fidelity for the V148I and Q151N RT mutants, respectively. Finally, we employed a cell-based pseudotyped HIV-1 mutation assay to determine whether changes in these dNTP binding residues alter RT fidelity in vivo. We found that the V148I and Q151N mutant viruses had 3.8- and 5.7-fold higher fidelities than wild-type viruses, respectively, indicating that the molecular interaction between HIV-1 RT and the dNTP substrate contributes to viral mutagenesis.

High-throughput assays for enzyme catalysis that can be applied to a broad range of chemical reactions are key to advances in directed evolution and proteomics. Recently, we reported such a general assay, chemical complementation, which links enzyme catalysis to reporter gene transcription in vivo using the yeast three-hybrid assay. In this proof-of-principle experiment, it was shown that a wild-type beta-lactamase enzyme could be isolated from a pool of inactive mutants using a lacZ screen. Ideally, however, such an assay should be able to distinguish enzymes based on their catalytic activity. Thus, here, we set out to determine if the catalytic efficiency of an enzyme variant does in fact correlate with its level of transcription activation in the chemical complementation assay. First, the reaction mechanism for the cleavage of the beta-lactam substrate used in the chemical complementation proof-of-principle experiment was determined. Then a series of beta-lactamase variants was designed to span several orders of magnitude in k(cat)/K(m). The activity of each variant was determined both in vitro using purified enzyme and in vivo in the chemical complementation transcription assay. Beta-lactamase variants spanning three-orders of magnitude in k(cat)/K(m) could be distinguished in the assay, and the catalytic efficiency of each variant correlated with its level of transcription activation in vivo. These results establish the suitability of chemical complementation for the directed evolution of enzymes with improvements in catalytic activity and for profiling the relative substrate specificities of a group of enzymes in proteomics applications.

The effect of superhelicity of DNA templates on transcription is well documented in several cases. However, the amount of supercoiling that is needed to bring about any changes and the steps at which such effects are exerted were not systematically studied. We investigated the effect of DNA supercoiling on transcription from a set of promoters present on a plasmid by using a series of topoisomers with different superhelical densities ranging from totally relaxed to more than physiological. In vitro transcription assays with these topoisomers in the absence and presence of gene regulatory proteins showed that the effect of negative supercoiling on intrinsic transcription varies from promoter to promoter. Some of those promoters, in which DNA superhelicity stimulated transcription, displayed specific optima of superhelical density while others did not. The results also showed that the amounts of abortive RNA synthesis from two of the promoters decreased and full-length RNA increased with increasing supercoiling, indicating for the first time an inverse relationship between full-length and abortive RNA synthesis and supporting a role of DNA superhelicity in promoter clearance. DNA supercoiling might also influence the point of RNA chain termination. Furthermore, the effect of varying the amount of supercoiling on the action of gene regulatory proteins suggested the mode of action, which is consistent with previous results. Our results underscore the importance of DNA supercoiling in fine-tuning promoter activities, which should be relevant in cell physiology given that local changes in chromosomal supercoiling must occur in different environments.

The activated Escherichia coli cAMP receptor protein, CRP, is capable of regulating the expression of more than 20 genes by binding to specific DNA sites. DNA bending is an important structural feature that has been observed in the regulatory mechanism of gene expression by CRP. On the basis of the results of the fluorescence energy transfer study of the gal P1 promoter, gal bends asymmetrically upon binding to CRP, although DNA bends symmetrically in the CRP-lac complex. The flanking sequence proximal to the TGTGA motif is involved in a sharper bend than the other side with an overall bending angle of approximately 90-125 degrees, without wrapping around the CRP molecule. To understand the factors that control the symmetry in DNA bending, a series of DNA sequences was tested to dissect the contribution of half-sites and flanking sequences, using the natural gal P1 and lac P1 sequences as initial targets. The extent of DNA bending induced by CRP was monitored by the difference in fluorescence anisotropy between free DNA and the DNA-CRP complex. The extent of bending was sequence-dependent, and most importantly, the symmetry of bending was a function of the symmetry of the DNA sequence. For example, in the lac promoter the two binding half-sites (TGTGA and TCACT) were almost symmetric as an inverted repeat. The recognition F-helices of the two CRP subunits would bind to these half-sites with a 2-fold symmetry. The flanking sequences (ATAAA and CATTA) were almost identical mirror images. Thus, they are expected to bend in a similar manner. Finally, the sequence symmetry properties of a series of natural CRP promoters were analyzed. A strong tendency for symmetry sequence was encoded in class I promoter sites but not in class II promoter sites. Results from this analysis support the conclusion that the geometry of the CRP-DNA complex plays a major role in determining the molecular mechanism in gene transcription.

Trp-999 is a key residue for the action of beta-galactosidases (Escherichia coli). Several site specific substitutions (Phe, Gly, Tyr, Leu) for Trp-999 were made. Each substitution caused greatly decreased affinities for substrates and inhibitors that bind in the "shallow" mode, while the affinities of inhibitors that bind in the "deep" mode were not decreased nearly as much. This shows that Trp-999 is important for binding in the shallow mode. The residue is also very important for binding glucose to galactosyl-beta-galactosidase (as a transgalactosidic acceptor). Substitution greatly diminished the affinity for glucose. Substitutions also changed the activation thermodynamics and, subsequently, the rates of the catalytic reactions. The enthalpies of activation of the glycolytic bond cleavage step (galactosylation, k(2)) became less favorable while the entropies of activation of that step became more favorable as a result of the substitutions. Differing magnitudes of these enthalpic and entropic effects with ONPG as compared to PNPG caused the k(2) values for ONPG to decrease but to increase for PNPG. The enthalpies of activation for the common hydrolytic step (degalactosylation, k(3)) increased while the entropies of activation for this step did not change much. As a result, k(3) became small and rate determining for each substituted enzyme. The substitutions caused the rate constant (k(4)) of the transgalactosidic acceptor reactions with glucose (for the formation of allolactose) to become much larger and of the same order of magnitude as the normally large rate constants for transgalactosidic acceptor reactions with small alcohols. This is probably because glucose can approach with less restriction in the absence of Trp-999. However, since glucose binds very poorly to the galactosyl-beta-galactosidases with substitutions for Trp-999, the proportion of lactose molecules converted to allolactose is small. Thus, Trp-999 is also important for ensuring that an appropriate proportion of lactose is converted to allolactose.

The structures of a series of complexes designed to mimic intermediates along the reaction coordinate for beta-galactosidase are presented. These complexes clarify and enhance previous proposals regarding the catalytic mechanism. The nucleophile, Glu537, is seen to covalently bind to the galactosyl moiety. Of the two potential acids, Mg(2+) and Glu461, the latter is in better position to directly assist in leaving group departure, suggesting that the metal ion acts in a secondary role. A sodium ion plays a part in substrate binding by directly ligating the galactosyl 6-hydroxyl. The proposed reaction coordinate involves the movement of the galactosyl moiety deep into the active site pocket. For those ligands that do bind deeply there is an associated conformational change in which residues within loop 794-804 move up to 10 A closer to the site of binding. In some cases this can be inhibited by the binding of additional ligands. The resulting restricted access to the intermediate helps to explain why allolactose, the natural inducer for the lac operon, is the preferred product of transglycosylation.

Human immunodeficiency virus type 1 (HIV-1) reverse transcriptase (RT) is a putative source of the genomic hypermutation that promotes rapid evolution of HIV-1. To understand the molecular strategies that create a highly mutagenic DNA polymerase active site in HIV-1 RT, we investigated the roles of four residues in the beta 8-alpha E loop. Gln151, which interacts with the sugar of the incoming dNTP, and Lys154, which interacts with the template, yielded site-directed mutants with increased fidelity, suggesting that these residues are directly involved in the mutagenic architecture of the active site. Mutations at Gln151 and Lys154 also reduced processivity. Q151N RT showed enhanced ability to discriminate between TTP and AZT triphosphate, consistent with the observation that the Q151M mutation confers AZT resistance in vivo. Mutations at Gly152 greatly decreased RT activity; molecular modeling suggests that Gly152 is critical for the proper geometric alignment that permits base-pairing of the incoming dNTP with the template. Mutations at Trp153 reduced the expression level, and presumably the stability, of RT proteins in bacteria. These observations support the conclusion that interactions of active site residues in the beta 8-alpha E loop with incoming dNTPs and the template are determinants of the accuracy, processivity, and substrate selectivity of HIV-1 RT.

We have used random sequence mutagenesis and complementation in a bacterial selection system to establish a large library of immunodeficiency virus type 1 (HIV-1) reverse transcriptase (RT) mutants with amino acid substitutions in the beta3-beta4 region of the fingers subdomain [Kim, B., Hathaway, T. R., and Loeb, L. A. (1996) J. Biol. Chem. 271, 4872-4878]. We show here that one of these mutants, D76V, exhibits increased accuracy of copying both DNA and RNA templates in a primer extension assay with biased dNTP pools. More detailed analysis of DNA-dependent polymerization showed that the D76V mutation conferred an up to 14-fold increase in fidelity of nucleotide insertion and a 9-fold reduced mutation rate in an M13mp2 lacZalpha forward mutation assay. Substitution at D76 with positively charged (D76R) and nonpolar (D76V and D76I) residues increased replicational accuracy, while substitutions with negatively charged (D76E) and polar residues (D76S and D76C) had little effect on fidelity. We propose that D76 affects replicational accuracy by mediating interaction between the fingers subdomain and the single-stranded template. Our work shows that the Escherichia coli complementation system can yield HIV RT mutants with increased fidelity that have not been isolated from the natural host and that are valuable in understanding the molecular bases of replicational accuracy.

This study describes the first complete in vitro error specificity analysis of a mutator DNA polymerase that is altered in a residue not predicted to contact either the DNA or dNTP substrate. We examined this mutator form of polymerase beta (Y265C) in order to elucidate the critical role tyrosine 265 plays in the accuracy of DNA synthesis. Our results demonstrate that an increase in both frame shift errors in homonucleotide repeat sequences and base substitution errors contribute nearly equally to the Y265C mutator phenotype. The models described for production of these errors, primer/template misalignment and base misincorporation, respectively, are distinctly different, suggesting the Y265C alteration affects discrimination against both types of error production pathways. In addition, Y265C displays a 530-fold increase in multiple errors within the 203-base pair target region examined, relative to that of wild type. Processivity studies revealed that Y265C retains the near distributive nature of DNA synthesis characteristic of the wild type polymerase beta. Therefore, multiple errors exhibited by Y265C most likely result from independent polymerase binding events. Localization of tyrosine 265 in the X-ray crystallographic structure suggests this residue may play a role in mediating a conformational change of the polymerase [Pelletier, H., et al. (1996) Biochemistry 35, 12742-12761]. A conformational change is predicted to enhance the accuracy of DNA synthesis by imposing an induced fit selection against premutational intermediates. The observed loss of discrimination against both misalignment-mediated and misincorporation-mediated errors produced by polymerase Y265C is consistent with such a model.

Annexin II tetramer (AIIt) is a Ca2+-dependent, phosphatidylserine-binding, and F-actin-bundling phosphoprotein which is localized to both the extracellular and cytoplasmic surfaces of the plasma membrane. The tetramer is composed of two p36 heavy chains and two p11 light chains. We have produced prokaryotic cDNA expression constructs for both p36 and p11. Both proteins were expressed in large amounts in Escherichia coli upon induction with IPTG. Electrospray ionization mass spectrometry and amino acid sequence analysis of purified recombinant p36 (rp36) and recombinant p11 (rp11) suggested that the recombinant proteins were identical to their native counterparts except for the lack of N-terminal acetylation of rp36. Furthermore, the non-acetylated rp36 bound rp11 and formed AIIt. The circular dichroism spectra and urea denaturation profiles of acetylated AIIt and non-acetylated rAIIt were identical. In addition, both the acetylated AIIt and non-acetylated rAIIt were similar in their Ca2+ dependence and concentration dependence of phospholipid liposome aggregation, chromaffin granule aggregation, and F-actin bundling. These results suggest that N-terminal acetylation of p36 is not in fact necessary for binding of the protein to p11 and that N-terminal acetylation does not affect the conformational stability of AIIt or the in vitro activities of AIIt. The availability of large amounts of rAIIt will facilitate further characterization of the structure-function relationships of the protein.

Escherichia coli cAMP receptor protein (CRP) controls more than 20 genes. There are significant differences in the promoter regions in these genes. Thus, an elucidation of the mechanism of CRP action requires knowledge about the mode of selectivity in these promoters. An earlier study [Heyduk, T., & Lee, J. C. (1990) Proc. Natl. Acad. Sci. U.S.A. 81, 1744-8] indicates that the CRP(cAMP)1 conformer exhibits the highest affinity for the lac PI site in the lac operon. It is conceivable that the CRP conformer that binds with the highest affinity to these other sites may not be CRP(cAMP)1. To investigate this possibility, the binding of CRP to nine CRP binding sites was studied as a function of cAMP concentration. The CRP binding sites employed in this investigation were chosen to represent the primary promoter sites from class I (lac site PI) and class II (sites PI of gal and crp) as well as secondary CRP binding sites (crp site PII and cat PII) to further understand the molecular mechanism of CRP in controlling the transcription of these bacterial genes. The affinity of CRP for three synthetic CRP binding sites was also examined to explore the contribution of the inverted repeat region and sequences surrounding the recognition motifs. The synthetic sequences are gallac which contains the lac recognition motifs in the background of gal, modified cat PII which contains an 8-base pair spacer between the recognition motifs rather than the 7-base pair sequence naturally found in cat PII, and a random sequence that has no known similarity to any CRP binding site found in nature. The apparent affinities of these sequences for CRP were quantitatively determined to be biphasic in their cAMP dependence. The CRP(cAMP)1 conformer was found to have the highest affinity for all of the DNA sequences examined. No specific affinity was observed for these sequences with free CRP and CRP(cAMP)2. The affinity of CRP for DNA was sequence-dependent and increased in the following order: random < cat site PII, modified cat site PII, crp sites PI and PII < gal site PI < lac site PI < gallac. These results indicate that the entire CRP binding site sequence and its natural variability provide information to CRP. These promoter sites which appear to have different mechanisms at the molecular level are transcriptionally controlled by the same CRP conformer, CRP(cAMP)1. Thus, the regulation of transcription by CRP is more subtle than choosing different conformational forms of CRP. Using "physiological" concentrations of various components, a computer simulation study was conducted to illustrate the possible consequences of the thermodynamic parameters determined in this study. It is evident that the promoters of protein systems regulating the transport and metabolism of carbohydrates are responsive to low cAMP concentrations. However, the promoter for controlling the expression of CRP is highly regulated by the fluctuation of cAMP concentration.

Discrimination between allowed and disallowed amino acid substitutions provides a powerful method for analysis of protein structure and function. Site-directed mutagenesis allows specific hypotheses to be tested, but its systematic application to entire structural motifs is inefficient. This limitation may be overcome by genetic selection, which allows rapid scoring of thousands of randomly (or pseudorandomly) generated mutants. To facilitate structural dissection of DNA-binding proteins, we have designed two generally applicable bacterial selection systems: pPLUS selects for the ability of a protein to bind to a user-defined DNA sequence, whereas pMINUS selects against such binding. Complementary positive and negative selections allow rapid mapping of critical residues. To test and calibrate the systems, we have investigated the bipartite POU domain of the human B-cell-specific transcription factor Oct-2. (i) An invariant residue (Asn347) in the DNA-recognition helix of the POU-specific homeodomain (POUHD) was substituted by each of the 19 other possible amino acids. The mutant proteins each exhibited decreased specific DNA binding as defined in vivo by genetic selection and in vitro by gel retardation; relative affinities correlate with phenotypes in the positive and negative selection systems. An essential role for Asn347 in wild-type POUHD-DNA recognition is consistent with homologous Asn-adenine interactions in cocrystal structures of canonical homeodomains. (ii) Extension of pPLUS/pMINUS selection to the POU-specific subdomain (POUs) is demonstrated by analysis of mutations in its putative helix-turn-helix (HTH) element.(ABSTRACT TRUNCATED AT 250 WORDS)

The transcriptional control protein MerR is a metalloregulatory switch, activating transcription of a mercury resistance operon in the presence of mercuric ions and repressing transcription in their absence. We report here the construction and utilization of a synthetic merR gene and a single-copy merT'-lacZ fusion reporter for mutagenic analysis of the MerR protein's function. Site-directed mutagenesis of clustered acidic residues within the central region of the MerR protein indicated that these residues are important to the protein's ability to repress transcription. Quadruple or sextuple mutations involving residues E83 and E84 and other nearby acidic residues result in a repression-deficient (RD) phenotype. One of the mutant proteins was purified and shown by gel shift assay to retain binding to its operator DNA with an affinity similar to wild-type protein, suggesting that transcriptional repression does not correlate with MerR binding affinity. A small region of merR corresponding to residues 81-92 also was mutagenized in a search for other RD mutants and for mutants displaying sufficient transcriptional activation in the absence of mercuric ion to be classified as constitutive activation (CA) mutants. In this case, oligonucleotide-directed randomization of the target region and a screening/selection protocol were employed. Sixteen different mutants with an RD phenotype were identified, as well as eight different mutants with a CA phenotype. A high frequency of S87C mutations is evident in the RD set of mutants. The CA mutants have a high incidence of S86C and A89V mutations. The CA double mutant S86C/A89V was purified and found to bind to its DNA site with an affinity similar to that of the wild-type protein. Chemical nuclease activity assays indicate that the nonmercurated S86C/A89V CA mutant has a DNA distortion activity identical to that of mercurated wild-type MerR. A unique disulfide bond bridging this CA mutant's dimer interface was found and is proposed to constrain protein conformation in a manner analogous to mercuric ion binding in the wild-type protein.

Lactose repressor protein (LacI), a negative transcriptional regulator in Escherichia coli, relies on an allosteric conformational change for its function. The LacI effector isopropyl-β,D-thiogalactoside (IPTG) promotes this allosteric response and engages the side chains of residues N125 and D149 based on the crystallographic structure of LacI·IPTG. Targeted molecular dynamics (TMD) simulations have indicated involvement of these side chains during the protein structural changes in response to inducer binding. To examine this region further, we applied stochastic boundary molecular dynamics (SBMD) simulation and identified a transient interaction between residues N125 and D149. On the basis of these data, we introduced substitutions for either/both residues and analyzed their impact on protein function. The substitutions utilized were alanine to preclude hydrogen bonding or cysteine to allow disulfide bond formation, which was not observed for N125C/D149C. Minimal impacts were observed on operator affinity for all substitutions, but D149C, N125A/D149A, and N125C/D149C bound to IPTG with 5-8-fold lower affinity than wild-type LacI, and exhibited decreased allosteric amplitude (K(RI/O)/K(R/O)). Of interest, the double mutants did not exhibit an allosteric response to an alternate inducer, 2-phenylethyl-β,D-galactoside (PhEG), despite demonstration of PhEG binding. Further, the presence of the anti-inducer, o-nitrophenyl-β,D-fucoside (ONPF), enhanced operator affinity for wild-type LacI and all other mutant proteins examined, but behaved as an inducer for N125A/D149A, decreasing operator binding affinity. These results confirm the role of residues 125 and 149 in ligand binding and allosteric response and illustrate how readily the function of a regulatory protein can be altered.

C(4)-Dicarboxylate uptake transporter B (DcuB) of Escherichia coli is a bifunctional transporter that catalyzes fumarate/succinate antiport and serves as a cosensor of the sensor kinase DcuS. Sites and domains of DcuB were analyzed for their topology relative to the cytoplasmic or periplasmic side of the membrane and their accessibility to the water space. For the topology studies, DcuB was fused at 33 sites to the reporter enzymes PhoA and LacZ that are only active when located in the periplasm or the cytoplasm, respectively. The ratios of the PhoA and LacZ activities suggested the presence of 10 or 11 hydrophilic loops, and 11 or 12 α-helical transmembrane domains (TMDs). The central part of DcuB allowed no clear topology prediction with LacZ/PhoA fusions. The sites of DcuB accessible to the hydrophilic thiol reagent 4-acetamido-4'-maleimidylstilbene-2,2'-disulfonate (AMS) were determined with variants of DcuB that carried single Cys residues. After intact cells were labeled with the membrane-impermeable AMS, denatured cells were differentially labeled with the thiol reagent polyethylene-glycol-maleimide (PEGmal) and analyzed for a mass shift. From 35 positions 17 were accessible to AMS in intact bacteria. The model derived from topology and accessibility suggests 12 TMDs for DcuB and a waterfilled cavity in its central part. The cavity ends with a cytoplasmic lid accessible to AMS from the periplasmic side. The sensory domain of DcuB is composed of cytoplasmic loop XI/XII and a membrane integral region with the regulatory residues Thr396/Asp398 and Lys353.

Guanine residues in the lac operator were substituted with the isosteric analogue of 7-deazaguanine. The observed equilibrium dissociation constants for lac repressor binding to substituted operators were measured in 10 mM Tris, 150 mM KCl, 0.1 mM EDTA, and 0.1 mM DTE, pH 7.6, at 25 degrees C, using either a standard equilibrium assay or a competition assay. Of the seven individual sites tested, only three significantly altered the complex affinity. Two of these sites are symmetrically related and are positioned in the major grooves that are known to interact directly with repressor, and the third site is located in the central core region of the operator. The alkylation interference assay, which identifies essential phosphate sites, was used to define the extent of perturbation on the protein-DNA complex by the modified nucleotide. Chemical footprinting data for the singly substituted operator done at a single concentration of protein reflected the alignment of sterically incompatible groups or disruption in the local conformation, but did not appear to alter general phosphate backbone interactions. Similar experiments with a doubly substituted operator revealed that the free energy of binding was an additive combination of each of the individual sites, and the alkylation interference data were similar to the singly substituted and wild-type operators. A constitutive mutant which also had a higher binding constant had a similar alkylation interference pattern.

Regulation of the initiation of gene transcription from the gal operon of Escherichia coli is activated by the binding of CAP (catabolite activator protein) to a site centered at base pair -41.5 relative to the S1 start site of transcription. This operon is repressed by the specific binding of Gal repressor (GalR) to two operators, OE and OI, centered at -60.5 and +53.5, respectively. It has been proposed that this negative regulation results from the interaction of GalR dimers bound to OE and OI to form a protein-mediated "looped complex" [cf. Adhya, S. (1989) Annu. Rev. Genet. 23, 207-230]. In order to test whether DNA-bound CAP would facilitate or inhibit the binding of GalR, the simultaneous binding of these proteins was studied by quantitative DNase I footprint titration analysis. These studies demonstrate that GalR binding is noncooperative in the presence and in the absence of CAP and that GalR and CAP bind to the gal operon independently. No evidence was found that CAP stabilizes a putative Gal repressor-mediated protein-DNA looped complex. It has been shown that the gal operon can be negatively regulated by the binding of Lac repressor (LacI) to a gal operon in which OE and OI were both modified to be recognized by LacI [Haber, R., & Adhya, S. (1988) Proc. Natl. Acad. Sci. U.S.A. 85, 9683-9687]. In contrast to GalR, LacI binds to the chimeric gal operon with moderate cooperativity via the formation of a stable protein-DNA looped complex.(ABSTRACT TRUNCATED AT 250 WORDS)

The adk gene from Bacillus stearothermophilus was cloned and overexpressed in Escherichia coli under the control of the lac promoter. The primary structure of B. stearothermophilus adenylate kinase exhibited 76% identity with the enzyme from Bacillus subtilis, 60% identity with the enzyme from Lactococcus lactis, and 42% identity with the enzyme from E. coli. The most striking property of the adenylate kinase from B. stearothermophilus is the presence of a structural zinc atom bound to four cysteines in a zinc finger-like fashion. The ability to coordinate zinc is predicted also for a number of other isoforms of bacterial adenylate kinases. Furthermore, the tightly bound metal ion contributes to the high thermodynamic stability of adenylate kinase from B. stearothermophilus.

The 31P NMR spectra of various 14-base-pair lac operators bound to both wild-type and mutant lac repressor headpiece proteins were analyzed to provide information on the backbone conformation in the complexes. The 31P NMR spectrum of a wild-type symmetrical operator, d(TGTGAGCGCTCACA)2, bound to the N-terminal 56-residue headpiece fragment of a Y7I mutant repressor was nearly identical to the spectrum of the same operator bound to the wild-type repressor headpiece. In contrast, the 31P NMR spectrum of the mutant operator, d(TATAGAGCGCTCATA)2, wild-type headpiece complex was significantly perturbed relative to the wild-type repressor-operator complex. The 31P chemical shifts of the phosphates of a second mutant operator, d(TGTGTGCGCACACA)2, showed small but specific changes upon complexation with either the wild-type or mutant headpiece. The 31P chemical shifts of the phosphates of a third mutant operator, d(TCTGAGCGCTCAGA)2, showed no perturbations upon addition of the wild-type headpiece. The 31P NMR results provide further evidence for predominant recognition of the 5'-strand of the 5'-TGTGA/3'-ACACT binding site in a 2:1 protein to headpiece complex. It is proposed that specific, strong-binding operator-protein complexes retain the inherent phosphate ester conformational flexibility of the operator itself, whereas the phosphate esters are conformationally restricted in the weak-binding operator-protein complexes. This retention of backbone torsional freedom in strong complexes is entropically favorable and provides a new (and speculative) mechanism for protein discrimination of different operator binding sites. It demonstrates the potential importance of phosphate geometry and flexibility on protein recognition and binding.

The secondary structures of the cAMP receptor protein (CRP), a complex of CRP and cAMP, and a cAMP-independent receptor protein mutant (CRP*141 gln) were examined by using Raman spectroscopy. Spectra were obtained from CRP and CRP*141 gln dissolved in 0.3 M NaCl and 30 mM sodium phosphate at protein concentrations of 30-40 mg/mL. CRP and CRP.cAMP1 were compared at lower protein concentrations (10-12 mg/mL) in a solvent of 0.35 M NaCl and 20 mM sodium phosphate. Raman analysis indicates that CRP structural changes induced by one bound cAMP or by the Gly to Gln mutation at residue 141 are small. Spectra of the three CRP samples are essentially identical from 400 to 1900 cm-1. This result differs from the Raman spectroscopy study of CRP and CRP.cAMP2 cocrystals [DeGrazia et al. (1990) Biochemistry 29, 3557]. The latter work showed spectral differences between CRP and CRP.cAMP2 consistent with alterations in the protein conformation. These studies indicate that CRP and CRP.cAMP1 in solution are similar in structure and differ from CRP.cAMP2 cocrystals. Protease digestion and a DNA binding assay were also employed to characterize the wild-type and mutant proteins. CRP*141 gln exhibited the same conformational characteristics of previously reported cAMP-independent mutant proteins. It was sensitive to proteolytic attack in the absence of cAMP, or upon addition of cGMP. In the absence of cAMP, both wild-type and mutant CRPs bound noncooperatively to a 62 bp lac promoter DNA. The equilibrium constants were approximately 10(6) M-1 in 0.1 M Na+. CRP*141 gln had a 2-4-fold higher affinity for the 62 bp DNA than CRP.(ABSTRACT TRUNCATED AT 250 WORDS)

Mutants in putative helix VIII of lactose permease that retain the ability to accumulate lactose were created by cassette mutagenesis. A mutagenic insert encoding amino acid residues 259-278 was synthesized chemically by using reagents contaminated with 1% each of the other three bases and ligated into a KpnI/BclI site in the lacY gene in plasmid pGEM-4. Mutants that retain transport activity were selected by transforming a strain of Escherichia coli containing a wild-type lacZ gene, but deleted in lacY, with the mutant library and identifying colonies that transport lactose on indicator plates. Sequencing of the mutated region in lacY in 129 positive colonies reveals 43 single amino acid mutations at 26 sites and 26 multiple mutations. The variable amino acid positions are largely on one side of the putative alpha-helix, a stripe opposite Glu269. This mutable stripe of low information content is probably in contact with the membrane phospholipids.

Escherichia coli ribosomal protein S4 is a translational repressor regulating the expression of four ribosomal genes in the alpha operon. In vitro studies have shown that the protein specifically recognizes an unusual mRNA pseudoknot secondary structure which links sequences upstream and downstream of the ribosome binding site for rpsM (S13) [Tang, C. K., & Draper, D. E. (1989) Cell 57, 531]. We have prepared fusions of the rpsM translational initiation site and lacZ that allows us to detect repression in cells in which overproduction of S4 repressor can be induced. Twenty-five mRNA sequence variants have been introduced into the S13-lacZ fusions and the levels of translational repression measured. Sets of compensating base changes confirm the importance of the pseudoknot secondary structure for translational repression. An A residue in a looped, single-stranded sequence is also required for S4 recognition and may contact S4 directly. Comparison of translational repression levels and S4 binding constants for the set of mRNA mutations show that nine mutants are repressed much more weakly than predicted from their affinity for S4; in extreme cases no repression can be detected for variants with unchanged S4 binding. We suggest that the mRNA contains functionally distinct ribosome and repressor binding sites that are allosterically coupled. Mutations can relieve translational repression by disrupting the linkage between the two sites without altering S4 binding. This proposal assigns to the mRNA a more active role in mediating translational repression than found in other translational repression systems.

The H1 protein is a likely candidate for structuring DNA in the bacterial nucleoid. We have studied determinants leading to its binding to DNA (and in particular to Escherichia coli lac and gal promoters) in vitro through the pattern of attack of both DNaseI and the copper-o-phenanthroline complex [(OP)2Cu+]. The binding of H1 depends on the primary sequence of DNA. H1 also associates with recognition sites for specific proteins, in particular with the Pribnow box and the CRP binding site. Binding of H1 to the Pribnow box of the wild-type lac promoter does not change the pattern of nucleolytic digestion with (OP)2Cu+. In contrast, binding of H1 to the strong lac promoter mutants Ps and UV5 appears to change the conformational state of this DNA. Similar changes in accessibility of the minor groove surrounding the respective binding sites were observed for both H1-DNA and CRP-DNA complexes.

Activation of G protein coupled receptors (GPCRs) by binding of ligand is the initial event in diverse cellular signaling pathways. To examine the frequency and diversity of mutations that cause constitutive activation of one particular GPCR, the yeast alpha-factor receptor, we screened libraries of random mutations for constitutive alleles. In initial screens for mutant receptor alleles that exhibit signaling in the absence of added ligand, 14 different point mutations were isolated. All of these 14 mutants could be further activated by alpha-factor. Ten of the mutants also acquired the ability to signal in response to binding of desTrp(1)¿Ala(3)ălpha-factor, a peptide that acts as an antagonist toward normal alpha-factor receptors. Of these 10 mutants, at least eight alleles residing in the third, fifth, sixth, and seventh transmembrane segments exhibit bona fide constitutive signaling. The remaining alleles are hypersensitive to alpha-factor rather than constitutive. They can be activated by low concentrations of endogenous alpha-factor present in MATa cells. The strongest constitutively active receptor alleles were recovered multiple times from the mutational libraries, and extensive mutagenesis of certain regions of the alpha-factor receptor did not lead to recovery of any additional constitutive alleles. Thus, only a limited number of mutations is capable of causing constitutive activation of this receptor. Constitutive and hypersensitive signaling by the mutant receptors is partially suppressed by coexpression of normal receptors, consistent with preferential association of the G protein with unactivated receptors.

The carboxy-terminal domain of the alpha subunit of Escherichia coli RNA polymerase, which is connected with the core part of RNA polymerase through a long flexible linker, plays decisive roles in transcription activation by directly interacting with a large number of transcription factors and upstream (UP) element DNA. Here we constructed a set of mutant RNA polymerases, each containing a mutant alpha subunit with an altered interdomain linker. Deletion of three amino acids from the linker exhibited 50% inhibition of cAMP receptor protein- (CRP-) dependent lac P1 transcription. Deletion of six amino acids completely knocked out the activity. Insertion of three amino acids did not affect the activity, whereas 40-60% inhibition was observed after insertion of one, two, or four amino acids. Substitution of 10 consecutive glycine residues resulted in nearly 90% reduction of the CRP-dependent activity, whereas 50% activity was retained after substitution of 10 proline residues or a sequence expected to form a strong alpha-helix. Essentially the same results were obtained with UP element-dependent rrnB P1 transcription. These observations altogether suggest that (i) sufficient length of the interdomain linker is required for transcription activation mediated by the alpha carboxy-terminal domain, (ii) the linker is not totally unstructured but has structural and torsional preferences to facilitate positioning of the carboxy-terminal domain to a proper location for the interaction with CRP and UP element, and (iii) CRP-dependent activation and UP element-dependent activation share a common intermediary state in which the positioning of the alpha carboxy-terminal domain is of primary importance.

The topology of the beta-subunit of the oxaloacetate Na+ pump (OadB) was probed with the alkaline phosphatase (PhoA) and beta-galactosidase (lacZ) fusion technique. Additional evidence for the topology was derived from amino acid alignments and comparative hydropathy profiles of OadB with related proteins. Consistent results were obtained for the three N-terminal and the six C-terminal membrane-spanning alpha-helices. However, the two additional helices that were predicted by hydropathy analyses between the N-terminal and C-terminal blocks did not conform with the fusion results. The analyses were therefore extended by probing the sideness of various engineered cysteine residues with the membrane-impermeant reagent 4-acetamido-4'-maleimidylstilbene-2, 2'-disulfonate. The results were in accord with those of the fusion analyses, suggesting that the protein folds within the membrane by a block of three N-terminal transmembrane segments and another one with six C-terminal transmembrane segments. The mainly hydrophobic connecting segment is predicted not to traverse the membrane fully, but to insert in an undefined manner from the periplasmic face. According to our model, the N-terminus is at the cytoplasmic face and the C-terminus is at the periplasmic face of the membrane.

We have mapped sites of tamoxifen adduct formation, in the lacI gene using the polymerase STOP assay, following reaction in vitro with alpha-acetoxytamoxifen and horseradish peroxidase (HRP)/H(2)O(2) activated 4-hydroxytamoxifen. For both compounds, most adduct formation occurred on guanines. However, one adenine, within a run of guanines, generated a strong polymerase STOP site with activated 4-hydroxytamoxifen, and a weaker STOP site with alpha-acetoxytamoxifen at the same location. In Escherichia coli the lac I gene reacted with 4-hydroxytamoxifen was more likely to be mutated (2 orders of magnitude) than when reacted with alpha-acetoxytamoxifen, despite the greater DNA adduct formation by alpha-acetoxytamoxifen. This correlates with the greater predicted ability of activated 4-hydroxytamoxifen adducts to disrupt DNA structure than alpha-acetoxytamoxifen adducts. For lac I reacted with activated 4-hydroxytamoxifen, a hot spot of base mutation was located in the region of the only adenosine adduct. No mutational hot spots were observed with alpha-acetoxytamoxifen. Our data clearly shows a lack of correlation between gross adduct number, as assayed by (32)P-postlabeling and mutagenic potential. These data indicate the importance of minor adduct formation in mutagenic potential and further that conclusions regarding the mutagenicity of a chemical may not be reliably derived from the gross determination of adduct formation.

Using mouse Ltk(-) cells (L13-17 cells) that had been transfected with a plasmid in which the lacZ gene had been ligated downstream of 1.4 kbp of the sequence of the promoter of the mouse gene for metallothionein-I (MT-I) as a reporter gene, we examined 268 organic compounds for the ability to activate this promoter. We found that PF1070A, an antibiotic produced by Humicola sp., efficiently activated the MT promoter and caused marked enhancement of beta-galactosidase activity in L13-17 cells. The extent of activation by PF1070A was almost equivalent to that by of zinc ions, the most effective known inducer of the synthesis of MT. PF1070A also caused marked elevation of the levels of the mRNA for MT and of MT itself in L13-17 cells. A similar result was obtained in human HeLa-S3 cells. When PF1070A was added to the culture medium simultaneously with cadmium ion or dexamethasone, the level of expression of the reporter gene was markedly elevated, compared to the level of expression induced by each agent independently. The effect of PF1070A was reduced considerably by deletion of nucleotides at positions -150 and -149 from the site of initiation of transcription in the promoter region of the MT gene and also by deletion of the seven bases located at positions -49 to -43. Since no known cis element was found in these two regions, PF1070A might be a new type of inducer of MT synthesis that promotes expression of the gene for MT via a mechanism completely different from those exploited by other known agents. These results also suggest the presence of a system for control of transcription of the gene for MT that has not previously been recognized. Both cadmium ions and bismuth ions induce the synthesis of MT by acting on the metal response element (MRE). Bismuth ions had no significant effect on the promoter activity that had already reached a maximum level in response to treatment with the optimal concentration of cadmium ion. By contrast, PF1070A further and markedly increased the promoter activity. This result suggests that it is possible to increase the concentration of MT in tissue using PF1070A as an inducer even in cases where the MRE-mediated activation of the MT promoter has already been induced by the accumulation of cadmium, as is the case in a clinical setting. PF1070A may prove to be an excellent inducer of MT synthesis that is effective and clinically applicable. Moreover, use of PF1070A in combination with salts of heavy metals might be useful in controlling expression of a transfected gene that is regulated by the MT promoter since PF1070A can activate the MT promoter to an extent that cannot be achieved with heavy metal ions alone, when PF1070A is used in combination with zinc ions at a concentration of the latter considerably below the toxic level.

Base-specific interactions between promoter DNA and Escherichia coli RNA polymerase are regulated by a sigma (sigma) protein during transcription initiation. To map spatial relations between evolutionarily conserved regions of the primary sigma (sigma 70) and each DNA strand along the lacUV5 promoter in the transcriptionally active "open" complex, we have used a cysteine-tethered cutting reagent to cleave DNA strands. The chemical nuclease FeBABE [iron (S)-1-(p-bromoacetamidobenzyl)ethylenediaminetetraacetate] was conjugated to single-cysteine mutants of sigma 70 at sites 132C, 376C, 396C, 422C, 496C, 517C, or 581C. After formation of open promoter complexes between lacUV5 DNA and RNA polymerase holoenzymes carrying conjugated sigma 70 subunits, we observed promoter DNA cleavage spanning at least 60 bases, between positions -48 and +12. The results show that sigma 70 region 2.1, otherwise implicated in core enzyme binding, is proximal to the nontemplate strand of lacUV5 DNA between the -10 promoter element and positions as far downstream of the transcription start site as +12. Conserved region 3.2 of sigma 70 is proximal to the template strand near the +1 transcription start site, and region 3.1 is positioned between the lacUV5-10 and -35 promoter elements. We propose a model for the orientation of sigma 70 and DNA in the open complex.

The rate of UV-induced deamination of cytosine to uracil at a specific site in double-stranded (ds) DNA was monitored using a genetic reversion assay. M13mp2C141 ds DNA was exposed to 160 J/m2 UV (254 nm), incubated at 37 degrees C, pH 7.4, for various time intervals to allow for deamination, and treated with Escherichia coli photolyase in the presence of 365 nm light to reverse cyclobutane-type pyrimidine dimers. Upon transfection into uracil-glycosylase deficient (ung-) E. coli cells, the mutation (i.e., reversion) frequencies in the CCCC target sequence increased greatly with post-UV time of incubation at 37 degrees C, nearly doubling every day that the DNA had been held at 37 degrees C. After 8 days, the reversion frequencies had increased by two orders of magnitude upon transfection into ung- cells, relative to isogenic ung+ cells, indicating that most of the mutations arising in UV/photolyase-treated ds DNA were C--T mutations mediated by a uracil intermediate. Sequencing of the revertants revealed that all mutations were single C--T or tandem double CC--TT mutations. An increasing percentage of tandem double CC--TT mutations was found with longer post-UV incubation times, yet none occurred if the post-UV delay time step was omitted before photoreversal. After a 4-day delay between UV and photoreversal at 37 degrees C, greater than 84% of the total revertants had tandem double CC--TT mutations. Thus, the generation of a tandem double mutation is a time-dependent process that arises in DNA after the initial UV exposure. The rate of appearance (with a pseudo-first-order rate constant ca. 10(-6) s-1) of tandem double mutations during incubation of UV-irradiated DNA is inconsistent with two random, independently occurring mutational events and suggests a concerted deamination of both residues in a tandem cytosine pyrimidine (CC) dimer. Considering that deamination in a CC dimer occurred here with a half-life of ca. 5 days, in contrast to the measured half-life of ca. 20,000 years for spontaneous (non-UV-treated) cytosine deamination for the same target, these studies show that the formation of pyrimidine dimers in DNA increases the rate of deamination by six orders of magnitude, leading to the accelerated formation of single C--T and tandem double CC--TT mutations.

Single-stranded DNA of the lacUV-5 promoter formed at the active site of Escherichia coli RNA polymerase during transcription is specifically cleaved by the redox active tetrahedral cuprous chelates of 1,10-phenanthroline and its derivatives. The cleavage sites are observed in the open, initiating, and elongating complexes. Redox-inert, tetrahedral cuprous chelates of neocuproine (2,9-dimethyl-1,10- phenanthroline) and its 5-phenyl and 4-phenyl derivatives protect the template strand of DNA from scission within these steady state intermediates and inhibit transcription. Although these cuprous chelates of neocuproine bind at multiple sites within three distinct enzyme intermediates, the highest affinity site is within the elongation complex. The I50 of 5 microM for the 2:1 5-phenylneocuproine cuprous complex ((5 phi NC)2Cu+) in runoff transcription therefore primarily reflects its intermediate. The neocuproine cuprous chelates are novel transcription inhibitors because they bind to single-stranded DNA sites generated during the course of catalysis by RNA polymerase.

The chimeric nuclease Fis-OP has been used to identify novel Fis binding sites. Tethering the chemical nuclease OP-Cu+ to position 73 of the protein with a newly developed longer acetyl-beta-alanylamino spacer has facilitated the localization of two high-affinity Fis binding sequences in a 3 kb pUC19 plasmid. The shorter acetamido linker has allowed the chimeric nuclease to locate two strong Fis binding sites in the 50 kb phage lambda genome. All four sites reside in biologically interesting loci and have been confirmed by gel-retardation and DNase I footprint analyses. A newly discovered site resides in the lac operon of Escherichia coli. The binding of Fis to this site may antagonize repression by the LacI repressor. These studies demonstrate the feasibility of applying chimeric chemical nucleases to the task of identifying functional protein binding sites of biological interest within genomes without any assumption about their sequence preference.

The role of Asp274 in inducer binding of lac repressor has been explored by spectroscopic measurements, fluorescence quenching, in vitro induction assays, and chemical modification of mutants with conservative substitutions at this site. Although no fluorescence emission shift or characteristic UV difference spectrum was observed at high inducer concentration, fluorescence quenching, effects on operator binding, and chemical modification results indicate indirectly that the mutants Asp274-->Asn and Asp274-->Glu bind sugar, albeit with very low affinity (> 0.1 M). Consistent with very weak inducer binding indicated by protection from fluorescence quenching by iodide, operator binding activity of these two mutant proteins is altered at very high IPTG concentration, although in opposite directions. The distinct effects of inducer on operator binding in these two mutant proteins as well as substantial differences in the effect of sugar ligand on chemical modification of Cys107 and Cys140 by 2-(bromoacetamido)-4-nitrophenol suggest that the conformation of the protein before and after association with sugar may differ in these mutant proteins. Fluorescence quenching assays of lac mutant proteins at Asp274 indicate the proximity of Trp220 to the side chain at position 274, consistent with the location of this residue in the structural model of lac repressor and in the crystallographic structure of the homologous purine repressor. From these results, we conclude that Asp274 is in the inducer binding site, that the character of this residue is crucial to inducer binding, and that interaction of sugar with the side chain at this position may be associated with the conformational change necessary for generating high affinity ligand binding.

An ribooligonucleotide, UGGAA, complementary to the template strand of the lacUV-5 promoter can hybridize to the transcription "bubble" of the open complex formed by Escherichia coli RNA polymerase. Its site-specific binding, measured by gel retardation, enzyme inhibition, and chemical nuclease footprinting, is dependent on catalysis by RNA polymerase and the sequence of the hybridizing ribooligonucleotide. When UGGAA is linked to the chemical nuclease 1,10-phenanthroline copper, site-specific scission of the template strand of the transcriptionally active gene is observed. The formation of single-stranded DNA at transcription start sites by RNA polymerases provides a target for antigene strategies.

DNA containing the lac UV5 promoter was alkylated using bifunctional sulfur and nitrogen mustards and a monofunctional sulfur mustard. The alkylation sites were mapped using Taq polymerase, and the effect of alkylation on the formation of the DNA-RNA polymerase complex was determined using gel retardation. Alkylation was observed at all G residues in the template strand. Exposure of the alkylated DNA to Escherichia coli RNA polymerase resulted in the formation of a DNA-enzyme complex that was more stable, prior to initiation, than the complex formed with nonalkylated DNA. The DNA-RNA polymerase complex formed with the alkylated DNA also demonstrated decreased ability to progress along the full length of the DNA template. These observations show that, in addition to inducing transcriptional blockages, mustards also influence the interaction between RNA polymerase and its promoter. The ability to interfere with protein-DNA interactions may contribute significantly to the effects of these compounds in eukaryotic systems with their complex array of transcription factors.

Ionizing radiations induce numerous damages in DNA, especially strand breaks. The hydroxyl radical OH., produced by the radiolysis of water, is mainly responsible for this effect. The fact that strand breakage occurs at all nucleotides and that bound proteins may locally radioprotect DNA at the binding site lead us to develop a radiolytic footprinting method to study DNA-protein interactions. Three different radiations were used: beta rays, gamma photons, and fast neutrons. In order to validate this technique, three well-known interaction systems were tested: the lac repressor-lac operator of Escherichia coli, the cyclic AMP receptor protein (CRP) of E. coli and its specific site in the lac regulation region, and the core nucleosome. Radiolytic footprinting gives results similar to those obtained by more classical probes: DNase I, complexes of orthophenanthroline (OP) and copper, complexes of ethylenediaminetetraacetate ion (EDTA) and iron, and UV light. For the same system (lac repressor), irradiation with either gamma photons or fast neutrons gives identical results.

The overexpression of a nonfusion product of human beta-globin in Escherichia coli from its cDNA sequence has been accomplished for the first time. Expression of beta-globin from its native cDNA required the use of the strong bacteriophage T7 promoter. In this system, beta-globin accumulated to approximately 10% of total E. coli proteins. alpha-Globin was not expressed in the T7 system using the native cDNA. For the expression of alpha-globin, synthetic genes containing optimal E. coli codons were constructed. Neither synthetic alpha- nor beta-globin gene alone was expressed from the lac or tac promoter. Globin expression was achieved when the two synthetic alpha- and beta-globin genes were combined as an operon downstream of the lac promoter. The two proteins combined intracellularly with endogenous heme, which was concomitantly overproduced to yield tetrameric hemoglobin as roughly 5-10% of total E. coli protein. Cloning the alpha- and beta-globin cDNAs in a construct identical with the lac promoter did not yield globin production, establishing the requirement for optimal codon usage. The recombinant beta-globin from the T7 expression system was purified and reconstituted in vitro with heme and native alpha chains. N-terminal analyses showed that the beta-globin produced in the T7 system and the tetrameric hemoglobin produced from the synthetic genes contained an additional beta 1 methionine residue. Two additional mutants, beta 1 Val----Met and beta 1 Val----Ala were produced using the T7 system. Functional and structural properties of the purified hemoglobins will be discussed in the following papers.