A New Class of Arylamine Oxygenases: Evidence That P‐aminobenzoate N‐Oxygenase (aurf) is a Di‐iron Enzyme and Further Mechanistic Studies
Michael Simurdiak,Jungkul Lee,Huimin Zhao
DOI: https://doi.org/10.1002/cbic.200600136
IF: 3.2
2006-01-01
ChemBioChem
Abstract:Arylamine oxygenases are enzymes that catalyze the unusual oxidation of an aryl amine group to an aryl nitro group. Although arylamine oxygenases seem to be widespread and used in diverse metabolic reactions, 2] aminopyrrolnitrin oxygenase (PrnD) and p-aminobenzoate oxygenase (AurF) are the only two known examples. Recently, PrnD was biochemically and mechanistically characterized. AurF, however, has had limited characterization, mainly due to the difficulty of obtaining functional enzymes in vitro. 5] In addition, previous database searches (BLAST, PROSITE) failed to identify any known oxidoreductase cofactor binding sites or motifs in the AurF ACHTUNGTRENNUNGsequence; this indicated that AurF could be a novel class of arylamine oxygenases. We performed our own BLAST search on AurF using Biology Workbench (http://workbench.sdsc.edu) and identified four putative proteins that share low sequence homology with AurF (15–17% identity). Further sequence alignment by Clustal W revealed several conserved aspartic acid, glutamic acid, and histidine residues, and the conservation of two copies of an EX28–37DEXXH motif. That this motif is well conserved in several di-iron oxygenases, suggests that AurF might be a dinuclear non-heme iron oxygenase. To test this hypothesis, we cloned the aurF gene from Streptomyces thioluteus (ATCC 12310) into pTrc99A and heterologously expressed it in Escherichia coli BL21 (DE3). In vivo AurF activity in E. coli could not be observed when using the phosphate–saline solution reaction medium, as reported previously in Streptomyces experiments. 14] However, addition of a carbon source such as glucose, glycerol, acetate, or succinate to the reaction media restored in vivo AurF activity in E. coli ; this suggests that a certain cellular process is needed for enzyme function. This result is different from what was observed in experiments with Streptomyces. In our laboratory as well as in other laboratories, p-aminobenzoic acid (pABA) conversion to pnitrobenzoic acid (pNBA) was observed in Streptomyces in a reaction medium lacking a carbon source. Unfortunately, similar to previous studies, the AurF activity was completely abolished after lysis of the E. coli cells. To probe the presence of a di-iron cluster in AurF, each of the eight conserved residues in the EX28–37DEXXH motif was individually mutated to alanine by site-directed mutagenesis. As expected, the activity of each mutant was significantly lowered. No mutant showed greater than 1% of the wild-type AurF activity; this suggests that each residue is important for enzyme function. Note that each mutant enzyme was detected in at least some soluble form by Western blot; this indicates that lack of solubility was not the reason for complete loss of activity. The presence of a di-iron cluster in AurF was further determined by inductively coupled plasma (ICP) atomic emission spectrometry and electron paramagnetic resonance (EPR). For ICP studies, an average iron-to-enzyme ratio was determined to be 2.18, which is in good agreement with the expected value of two iron atoms/enzyme. For EPR studies, according to previous reports, a di-iron cluster is usually stable in three states: oxidized (Fe Fe), mixed-valent (Fe Fe), and reduced (Fe Fe). AurF purified from E. coli cells grown in Luria– Bertani medium generated an oxidized-form EPR spectra (Figure 1A) that resembled a mononuclear manganese(IV) species. This spectrum is also similar to the spectrum generated by a ribonucleotide reductase enzyme containing a Fe Mn center. Based on the ICP data, it is unlikely that AurF naturally contains manganese atom(s). Therefore, the enzyme was expressed in E. coli by using minimal medium with an overabundance of iron. The resulting enzyme generated an EPR-silent oxidized spectrum, and a mixed valence spectrum containing gx, gy, and gz values of 1.94, 1.79, and 1.70, respectively (Figure 1B). These mixed-valence g values are similar to those obtained for methane monooxygenase and ribonucleotide reductase. In addition, the fully reduced form of AurF generated a silent EPR spectrum, which is similar to that which is observed with the di-iron enzyme acyl-carrier protein D desaturase. It is not surprising that the EPR spectrum generated by AurF prepared in E. coli by using Luria–Bertani medium contained a signal from a manganese atom. It has been shown previously that di-iron enzymes can often take up manganese atoms in place of iron atoms. This idea was further tested by producing AurF in E. coli by using minimal medium with an overabundance of manganese. The resulting AurF did not exhibit in vivo activity. Its EPR spectrum was somewhat similar to that of mononuclear manganese; this does not agree with ICP data showing 2.04 manganese atoms/enzyme (see the Supporting Information). A catalytic mechanism was previously proposed for AurF based on in vivo assay data (Scheme 1a). In this mechanism, the first step involves the addition of an oxygen atom to produce p-hydroxylaminobenzoate (pHABA) from pABA; this was confirmed by the isolation of pHABA. The later steps involve the addition of another oxygen atom, followed by a dehydra[a] M. Simurdiak, Dr. J. Lee, Prof. H. Zhao Department of Chemical and Biomolecular Engineering University of Illinois at Urbana–Champaign 600 S. Mathews Avenue, Urbana, IL 61801 (USA) Fax: (+1)217-333-5052 E-mail : zhao5@uiuc.edu [b] Dr. J. Lee Department of Chemical Engineering, Konkuk University 1 Hwayang-dong, Gwangjin-Gu, Seoul, 143-701 (Korea) Supporting information for this article is available on the WWW under http://www.chembiochem.org or from the author.