Crystal structure of HlyU, the hemolysin gene transcription activator, from Vibrio cholerae N16961 and functional implications.

HlyU in Vibrio cholerae is known to be the transcriptional activator of the hemolysin gene, HlyA and possibly a regulator of other virulence factors influencing growth, colonization and pathogenicity of this infective agent. Here we report the crystal structure of HlyU from V. cholerae N16961 (HlyU_Vc) at 1.8Å. The protein, with five α-helices and three ß-strands in the topology of α1-α2-ß1-α3-α4-ß2-ß3-α5, forms a homodimer. Helices α3-α4 and a ß sheet form the winged helix-turn-helix (wHTH) DNA-binding motif common to the transcription regulators of the SmtB/ArsR family. In spite of an overall fold similar to SmtB/ArsR family, it lacks any metal binding site seen in SmtB. A comparison of the dimeric interfaces showed that the one in SmtB is much larger and have salt bridges that can be disrupted to accommodate metal ions. A model of HlyU-DNA complex suggests bending of the DNA. Cys38 in the structure was found to be modified as sulfenic acid; the oxidized form was not seen in another structure solved under reducing condition. Although devoid of any metal binding site, the presence of a Cys residue exhibiting oxidation-reduction suggests the possibility of the existence of a redox switch in transcription regulation. A structure-based phylogenetic analysis of wHTH proteins revealed the segregation of metal and non-metal binding proteins as well as those in the latter group that are under redox control.

Escherichia coli HflK and HflC can individually inhibit the HflB (FtsH)-mediated proteolysis of lambdaCII in vitro.

LambdaCII is the key protein that influences the lysis/lysogeny decision of lambda by activating several phage promoters. The effect of CII is modulated by a number of phage and host proteins including Escherichia coli HflK and HflC. These membrane proteins copurify as a tightly bound complex 'HflKC' that inhibits the HflB (FtsH)-mediated proteolysis of CII both in vitro and in vivo. Individual purification of HflK and HflC has not been possible so far, since each requires the presence of the other for proper folding. We report the first purification of HflK and HflC separately as active and functional proteins and show that each can interact with HflB on its own and each inhibits the proteolysis of CII. They also inhibit the proteolysis of E. coli sigma(32) by HflB. We show that at low concentrations each protein is dimeric, based on which we propose a scheme for the mutual interactions of HflB, HflK and HflC in a supramolecular HflBKC protease complex.

Properties of HflX, an enigmatic protein from Escherichia coli.

The Escherichia coli gene hflX was first identified as part of the hflA operon, mutations in which led to an increased frequency of lysogenization upon infection of the bacterium by the temperate coliphage lambda. Independent mutational studies have also indicated that the HflX protein has a role in transposition. Based on the sequence of its gene, HflX is predicted to be a GTP-binding protein, very likely a GTPase. We report here purification and characterization of the HflX protein. We also specifically examined its suggested functional roles mentioned above. Our results show that HflX is a monomeric protein with a high (30% to 40%) content of helices. It exhibits GTPase as well as ATPase activities, but it has no role in lambda lysogeny or in transposition.

Probing the antiprotease activity of lambdaCIII, an inhibitor of the Escherichia coli metalloprotease HflB (FtsH).

The CIII protein encoded by the temperate coliphage lambda acts as an inhibitor of the ubiquitous Escherichia coli metalloprotease HflB (FtsH). This inhibition results in the stabilization of transcription factor lambdaCII, thereby helping the phage to lysogenize the host bacterium. LambdaCIII, a small (54-residue) protein of unknown structure, also protects sigma(32), another specific substrate of HflB. In order to understand the details of the inhibitory mechanism of CIII, we cloned and expressed the protein with an N-terminal six-histidine tag. We also synthesized and studied a 28-amino-acid peptide, CIIIC, encompassing the central 14 to 41 residues of CIII that exhibited antiproteolytic activity. Our studies show that CIII exists as a dimer under native conditions, aided by an intersubunit disulfide bond, which is dispensable for dimerization. Unlike CIII, CIIIC resists digestion by HflB. While CIII binds to HflB, it does not bind to CII. On the basis of these results, we discuss various mechanisms for the antiproteolytic activity of CIII.

Role of C-terminal residues in oligomerization and stability of lambda CII: implications for lysis-lysogeny decision of the phage.

A crucial element in the lysis-lysogeny decision of the temperate coliphage lambda is the phage protein CII, which has several interesting properties. It promotes lysogeny through activation of three phage promoters p(E), p(I) and p(aQ), recognizing a direct repeat sequence TTGCN6TTGC at each. The three-dimensional structure of CII, a homo-tetramer of 97 residue subunits, is unknown. It is an unstable protein in vivo, being rapidly degraded by the host protease HflB (FtsH). This instability is essential for the function of CII in the lysis-lysogeny switch. From NMR and limited proteolysis we show that about 15 C-terminal residues of CII are highly flexible, and may act as a target for proteolysis in vivo. From in vitro transcription, isothermal calorimetry and gel chromatography of CII (1-97) and its truncated fragments CIIA (4-81/82) and CIIB (4-69), we find that residues 70-81/82 are essential for (a) tetramer formation, (b) operator binding and (c) transcription activation. Presumably, tetramerization is necessary for the latter functions. Based on these results, we propose a model for CII structure, in which protein-protein contacts for dimer and tetramer formation are different. The implications of tetrameric organization, essential for CII activity, on the recognition of the direct repeat sequence is discussed.

Crystallization and X-ray analysis of the transcription-activator protein C1 of bacteriophage P22 in complex with the PRE promoter element.

The transcription-activator protein C1 of the temperate phage P22 of Salmonella typhimurium plays a key role in the lytic versus lysogenic switch of the phage. A homotetramer of 92-residue polypeptides, C1 binds to an approximate direct repeat similar to the transcription activator CII of coliphage λ. Despite this and several other similarities, including 57% sequence identity to coliphage CII, many biochemical observations on P22 C1 cannot be explained based on the structure of CII. To understand the molecular basis of these differences, C1 was overexpressed and purified and subjected to crystallization trials. Although no successful hits were obtained for the apoprotein, crystals could be obtained when the protein was subjected to crystallization trials in complex with a 23-mer promoter DNA fragment (PRE). These crystals diffracted very well at the home source, allowing the collection of a 2.2 Šresolution data set. The C1-DNA crystals belonged to space group P21, with unit-cell parameters a = 87.27, b = 93.58, c = 111.16 Å, ß = 94.51°. Solvent-content analysis suggests that the asymmetric unit contains three tetramer-DNA complexes. The three-dimensional structure is expected to shed light on the mechanism of activation by C1 and the molecular basis of its specificity.

Revisiting the mechanism of activation of cyclic AMP receptor protein (CRP) by cAMP in Escherichia coli: lessons from a subunit-crosslinked form of CRP.

Cyclic AMP receptor protein (CRP), the global transcription regulator in prokaryotes, is active only as a cAMP-CRP complex. Binding of cAMP changes the conformation of CRP, transforming it from a transcriptionally 'inactive' to an 'active' molecule. These conformers are also characterized by distinct biochemical properties including the ability to form an S-S crosslink between the C178 residues of its two monomeric subunits. We studied a CRP variant (CRP(cl)), in which the subunits are crosslinked. We demonstrate that CRP(cl) can activate transcription even in the absence of cAMP. Implications of these results for the crystallographically-determined structure of cAMP-CRP are discussed.

Dispensability of zinc and the putative zinc-binding domain in bacterial glutamyl-tRNA synthetase.

The putative zinc-binding domain (pZBD) in Escherichia coli glutamyl-tRNA synthetase (GluRS) is known to correctly position the tRNA acceptor arm and modulate the amino acid-binding site. However, its functional role in other bacterial species is not clear since many bacterial GluRSs lack a zinc-binding motif in the pZBD. From experimental studies on pZBD-swapped E. coli GluRS, with Thermosynechoccus elongatus GluRS, Burkholderia thailandensis GluRS and E. coli glutamyl-queuosine-tRNA(Asp) synthetase (Glu-Q-RS), we show that E. coli GluRS, containing the zinc-free pZBD of B. thailandensis, is as functional as the zinc-bound wild-type E. coli GluRS, whereas the other constructs, all zinc-bound, show impaired function. A pZBD-tinkered version of E. coli GluRS that still retained Zn-binding capacity, also showed reduced activity. This suggests that zinc is not essential for the pZBD to be functional. From extensive structural and sequence analyses from whole genome database of bacterial GluRS, we further show that in addition to many bacterial GluRS lacking a zinc-binding motif, the pZBD is actually deleted in some bacteria, all containing either glutaminyl-tRNA synthetase (GlnRS) or a second copy of GluRS (GluRS2). Correlation between the absence of pZBD and the occurrence of glutamine amidotransferase CAB (GatCAB) in the genome suggests that the primordial role of the pZBD was to facilitate transamidation of misacylated Glu-tRNA(Gln) via interaction with GatCAB, whereas its role in tRNA(Glu) interaction may be a consequence of the presence of pZBD.

Preliminary X-ray crystallographic analysis of an engineered glutamyl-tRNA synthetase from Escherichia coli.

The nature of interaction between glutamyl-tRNA synthetase (GluRS) and its tRNA substrate is unique in bacteria in that many bacterial GluRS are capable of recognizing two tRNA substrates: tRNAGlu and tRNAGln. To properly understand this distinctive GluRS-tRNA interaction it is important to pursue detailed structure-function studies; however, because of the fact that tRNA-GluRS interaction in bacteria is also associated with phylum-specific idiosyncrasies, the structure-function correlation studies must also be phylum-specific. GluRS from Thermus thermophilus and Escherichia coli, which belong to evolutionarily distant phyla, are the biochemically best characterized. Of these, only the structure of T. thermophilus GluRS is available. To fully unravel the subtleties of tRNAGlu-GluRS interaction in E. coli, a model bacterium that can also be pathogenic, determination of the E. coli GluRS structure is essential. However, previous attempts have failed to crystallize E. coli GluRS. By mapping crystal contacts of a homologous GluRS onto the E. coli GluRS sequence, two surface residues were identified that might have been hindering crystallization attempts. Accordingly, these two residues were mutated and crystallization of the double mutant was attempted. Here, the design, expression, purification and crystallization of an engineered E. coli GluRS in which two surface residues were mutated to optimize crystal contacts are reported.