No abstract

The heme-containing catalase HPII of Escherichia coli consists of a homotetramer in which each subunit contains a core region with the highly conserved catalase tertiary structure, to which are appended N-and C-terminal extensions making it the largest known catalase. HPII does not bind NADPH, a cofactor often found in catalases. In HPII, residues 585-590 of the C-terminal extension protrude into the pocket corresponding to the NADPH binding site in the bovine liver catalase. Despite this difference, residues that define the NADPH pocket in the bovine enzyme appear to be well preserved in HPII. Only two residues that interact ionically with NADPH in the bovine enzyme~Asp212 and His304! differ in HPII~Glu270 and Glu362!, but their mutation to the bovine sequence did not promote nucleotide binding. The active-site heme groups are deeply buried inside the molecular structure requiring the movement of substrate and products through long channels. One potential channel is about 30 Å in length, approaches the heme active site laterally, and is structurally related to the branched channel associated with the NADPH binding pocket in catalases that bind the dinucleotide. In HPII, the upper branch of this channel is interrupted by the presence of Arg260 ionically bound to Glu270. When Arg260 is replaced by alanine, there is a threefold increase in the catalytic activity of the enzyme. Inhibitors of HPII, including azide, cyanide, various sulfhydryl reagents, and alkylhydroxylamine derivatives, are effective at lower concentration on the Ala260 mutant enzyme compared to the wild-type enzyme. The crystal structure of the Ala260 mutant variant of HPII, determined at 2.3 Å resolution, revealed a number of local structural changes resulting in the opening of a second branch in the lateral channel, which appears to be used by inhibitors for access to the active site, either as an inlet channel for substrate or an exhaust channel for reaction products.

The subunit of catalase HPII from Escherichia coli is 753 residues in length and contains a core of approximately 500 residues, with high structural similarity to all other heme catalases. To this core are added extensions of approximately 80 and 180 residues at the N- and C-termini, respectively. The tetrameric structure is made up of a pair of interwoven dimers in which 90 N-terminal residues of each subunit are inserted through a loop formed by the hinge region linking the beta-barrel and alpha-helical domains of the adjacent subunit. A high concentration of proline residues is found in the vicinity of the overlap regions. To study the influence of the extended regions on folding and subunit association of HPII, a diversity of modifications have been introduced. Removal of the complete C-terminal domain or the N-terminal extension, either separately or together, effectively creating a small subunit catalase, resulted in no enzyme accumulation. Systematic truncations showed that only nine C-terminal residues (Ile745 to Ala753) could be removed without significantly affecting the accumulation of active enzyme. Removal or even conservative replacements of the side chain of Arg744 significantly reduced the accumulation of active enzyme despite this residue interacting only with the C-terminal domain. Removal of as few as 18 residues from the N-terminus also reduced accumulation of active enzyme. Changes to other residues in the protein, including residues in the heme binding pocket, also reduced the accumulation of active protein without substantially affecting the enzyme specific activity. Implications of these data for the interdependence of subunit folding and subunit-subunit interactions are discussed.

~Siderophore-negative mutants of Azofobacfer aiwl&i were generated by insertional mutagenesis with a Tn5 construct containing a promoterless l u A B fusion. The use of this construct, delivered on a suicide plasmid by corrjugation, allowed the selection of mutations in iron-repressible genes by virtue of the expression of ironregulated bioluminescence. Although many iron-regulated mutants were selected, only a few could be easily identified as defective in siderophore production. These included a non-fluorescent azotobactin-negative phenotype (strain D27), and strain F1%, which had lost the ability to produce the catechol siderophores azotochelin and aminochelin as well as the lower-afhity chelator 2,3-dihydroxybenzoic acid. Strain D27 had normal production of catechol siderophores, while strain F1% produced 2.5 times as much aotobactin as the wild-type. Two other mutants demonstrated normal catechol levels and either low or relatively unrepressed azotobactin levels. Transformation of the DNA from strain F1% into another spontaneously obtained azotobactin-negative strain (UAI) resulted in strain PIOO, which was unable to produce the known siderophores. Unlike the wild-type and other siderophore-deficient mutants, this strain was unable to grow in the presence of the iron chelator ethylenediamine di-(o-hydroxyphenylacetic acid) (EDDHA; 50 pg ml-l) unless stored iron was carried over in the inoculum. Strain PI00 did grow on iron-limited medium containing EDDHA when the catechol or azotobactin siderophores were provided exogenously. However, strain PlOO gave a positive result in the chrome azurol-S assay . (CAS), a non-specilic assay for siderophores. The CAS activity was iron-repressible and strain PI00 was able to grow and accumulate more iron from the insoluble iron minerals FeS, vivianite and Fe30, than was available by simple diffusion or exchange. Therefore, it appears that iron-limited A. v k h & produces an as yet unidentified low-affinity non-conventional (non-catechol, non-hydroxamate) siderophore.

Site-directed mutagenesis of the katE gene of Escherichia coli was used to change, individually and in combination, Cys438 and Cys669 to serine in catalase HPII. The Cys438-->Ser mutation caused a 30% reduction in the specific activity of the enzyme, whereas the Cys669-->Ser mutation did not affect enzyme activity. The titration of free sulfhydryl groups in HPII revealed that Cys669 was reactive whereas Cys438 was unreactive. Properties of the modification on Cys438 included alkali lability, insensitivity to methylamine, hydroxylamine or reducing agents, and a mass determined by mass spectrometry to be approximately 43 +/- 2 Da. A hemithioacetal structure is consistent with these properties. Although free sulfhydryl groups do not play a significant role in the stability or catalytic mechanism of HPII, the sulfhydryl agent 2-mercaptoethanol caused a 50% inactivation of HPII along with an irreversible change in the absorption spectrum of the protein. Other sulfhydryl agents, including dithiothreitol, cysteine and glutathione, and the organic peroxide, t-butylhydroperoxide, which cannot directly access the active site, do not affect HPII activity, but they do cause a small reversible change in the absorption spectrum, possibly by a mechanism involving superoxide.