Exhaustive Review (Last Two Years) on Biotin Function and Biosynthesis

• Loading metrics

Open up Access

Peer-reviewed

Inquiry Commodity

Remarkable Diversity in the Enzymes Catalyzing the Final Stride in Synthesis of the Pimelate Moiety of Biotin

• Madelyn M. Shapiro,
• Vandana Chakravartty,
• John E. Cronan

ten

• Published: November 9, 2012
• https://doi.org/10.1371/journal.pone.0049440

Figures

Abstract

Biotin synthesis in Escherichia coli requires the functions of the bioH and bioC genes to synthesize the precursor pimelate moiety by use of a modified fatty acid biosynthesis pathway. However, it was previously noted that bioH has been replaced with bioG or bioK inside the biotin synthetic gene clusters of other bacteria. We report that each of four BioG proteins from diverse bacteria and 2 cyanobacterial BioK proteins functionally supercede E. coli BioH in vivo. Moreover, purified BioG proteins accept esterase action confronting pimeloyl-ACP methyl ester, the physiological substrate of BioH. Two of the BioG proteins cake biotin synthesis when highly expressed and these toxic proteins were shown to have more promiscuous substrate specificities than the not-toxic BioG proteins. A postulated BioG-BioC fusion protein was shown to functionally replace both the BioH and BioC functions of E. coli. Although the BioH, BioG and BioK esterases catalyze a common reaction, the proteins are evolutionarily singled-out.

Citation: Shapiro MM, Chakravartty V, Cronan JE (2012) Remarkable Variety in the Enzymes Catalyzing the Last Step in Synthesis of the Pimelate Moiety of Biotin. PLoS I 7(11): e49440. https://doi.org/10.1371/journal.pone.0049440

Editor: Christophe Herman, Baylor College of Medicine, United States of America

Received: July 23, 2012; Accepted: October ix, 2012; Published: November 9, 2012

Copyright: © 2012 Shapiro et al. This is an open-access article distributed under the terms of the Creative Eatables Attribution License, which permits unrestricted use, distribution, and reproduction in whatsoever medium, provided the original author and source are credited.

Funding: Funding was provided past NIH NIAID AI15650. The funders had no part in study design, data drove and analysis, decision to publish, or preparation of the manuscript.

Competing interests: The authors take declared that no competing interests be.

Introduction

Biotin (vitamin H) is an essential enzyme cofactor required by all three domains of life. It functions as a covalently-bound prosthetic group which mediates the transport of CO₂ in many vital metabolic carboxylation, decarboxylation and transcarboxylation reactions [1], [two]. Although biotin is an essential cofactor, our knowledge of its biosynthesis remains fragmentary. Labeling studies in Escherichia coli suggested that most of the carbon atoms of biotin are derived from pimelic acrid, a seven carbon α,ω-dicarboxylic acid [3], [4]. The pathway whereby the pimelate moiety is synthesized was a long-standing puzzle until contempo work in E. coli showed that it is fabricated past a modification of the fat acid synthesis pathway that allows synthesis of dicarboxylic fatty acids past a machinery reminiscent of that proposed in 1963 [5]. 2 enzymes, BioC and BioH, hijack a fraction of the fat acid biosynthetic capacity to make the pimelate moiety. In this, the kickoff complete biotin synthetic pathway, BioC converts the free carboxyl grouping of a malonyl thioester to its methyl ester [6]. Methylation cancels the accuse of the carboxyl group and provides a methyl carbon to mimic the methyl ends of normal acyl bondage to give a species approximating the substrates ordinarily accepted by the fat acrid synthetic enzymes (Figure 1). Ii cycles of the standard elongation-reduction-dehydration-reduction cycle of fatty acid synthesis results in the acyl carrier protein (ACP) thioester of monomethyl pimelate. The methyl ester of this product is then cleaved past BioH to requite pimeloyl-ACP which reacts with alanine in the BioF reaction to requite the start intermediate of biotin ring associates. Thus, the methyl ester disguises the biotin synthetic intermediates such that they are accustomed as substrates by the fatty acid constructed pathway [6]. Although carbon concatenation elongation requires that the carboxyl group of the primer cease of the acyl concatenation be neutralized by a methyl grouping [six], information technology must be freed subsequently in the pathway because the carboxyl is required for biotin protein ligase-catalyzed attachment of biotin to its cognate enzyme proteins.

Figure 1. The Eastward. coli biotin constructed pathway.

The biotin synthetic pathway is initiated (a) by BioC-catalyzed and S-adenosyl-L-methionine (SAM) mediated methylation of malonyl-ACP. The methyl group is red. The malonyl ACP methyl ester enters the fat acid synthetic cycle as the primer. (b) Following for 2 rounds of the fatty acid concatenation elongation bicycle the resulting pimeloyl-ACP methyl ester is then (c) hydrolyzed past BioH to course pimeloyl-ACP which is a substrate for BioF to begin assembly of the biotin rings (d). Abbreviations: SAH, S-adenosyl-homocysteine; AON, 8-amino-7-oxononanoate.

https://doi.org/x.1371/journal.pone.0049440.g001

In E. coli the biotin synthetic genes are located in two distant genome locations. The bioA, bioB, bioF, bioC and bioD genes are clustered and transcribed past two face-to-face promoters in a bidirectional operon [7]. However bioH, the remaining biotin factor, is located far from the bio operon (Figure 2) and unlike the other bio genes its transcription is not regulated by BirA, the E. coli bifunctional repressor-biotin poly peptide ligase [8]–[ten]. This cistron system is in contrast to those of many other bacteria (e.grand., the Pseudomonadaceae, Bacillus cereus) where bioH is located within the biotin operon immediately upstream of bioC [xi] and is well integrated into the operon (the coding sequences of biotin operon genes often overlap). Thus, the E. coli bioH gene may have been more recently acquired than the more "domesticated" bioH genes located in bio operons. Eastward. coli is not the but bacterium in which bioH is removed from the bio operon. Yersinia sp., Shewnella sp., and Serratia proteamaculans share this property, although just the final of these has been shown to functionally replace E. coli bioH [12].

Effigy 2. The differing configurations of the biotin genes of diverse leaner.

Eastward. coli and P. aeruginosa both contain bioH just P. aeruginosa has bioH within its bio operon upstream of bioC whereas the E. coli bioH is located elsewhere on the chromosome. H. influenzae and C. jejuni accept bioG within their bio operons upstream of bioC. N. meningitidis has both bioH and bioG upstream of divide copies of bioC. B. fragilis encodes a fusion of BioC and BioG. In place of BioG or BioH, most BioC-containing blue-green alga carrying, such as Synechococcus spp. and P. marinus, take bioK upstream of bioC.

https://doi.org/10.1371/journal.pone.0049440.g002

Based on bioinformatics analyses Rodionov and coworkers [11] reported that BioH is something of a "wild card" among biotin synthesis enzymes because in some bacteria the cistron has been displaced from the biotin operon by other genes (bioG bioK and bioZ) (Figure 2). These workers proposed that their ascertainment can be explained "either past utilization of different sources for biotin biosynthesis or by nonorthologous displacements of the BioC-linked proteins" [11]. It should likewise be noted that like BioH, BioG and BioK are upstream of and overlapping with BioC in all of the organisms examined (Effigy ii). In our piece of work we furthered and tested the hypotheses of Rodionov and coworkers [eleven] and establish that BioG and BioK are members of the α,β-hydrolase family like BioH and therefore seemed likely to be esterases able to cleave the methyl ester of pimeloyl-ACP whereas BioZ, which is confined to the α-proteobacteria, probably plays a very unlike biosynthetic role. We report that BioG and BioK proteins of diverse bacteria tin can replace the BioH part in the E. coli biotin synthetic pathway and that several purified BioG proteins cleave the methyl ester of pimeloyl-ACP in vitro with varying degrees of specificity.

Materials and Methods

Growth media

Genetic manipulations were done in LB broth or agar [xiii]. Strains grown on M9 minimal medium or agar contained 0.2% arabinose or 0.two% glycerol plus avidin (0.1 U/ml). When supplemented with biotin, iv nM was the final concentration. The antibiotics used were (µg/ml) sodium ampicillin. 100: kanamycin sulfate; spectinomycin sulfate, 50 and chloramphenicol, 25. The genomic DNAs were obtained from the ATCC.

Plasmid Constructs

The strains, plasmids and primers used are listed in Tabular array one. To gather constructs for the complementation analysis, the bioG coding sequences were PCR amplified from the genomic DNAs of H. influenzae Rd KW20, North. meningitidis MC58, and C. jejuni 81–176 [14] were using primers P2 and P7, P3 and P4 and P29 and P30, respectively. The bioGC factor was PCR amplified from B. fragilis ATCC 25285 genomic DNA using primers P6 and P8. Custom, codon optimized bioK genes of P. marinus MIT 9211 and Synechococcus sp. CC9902 were synthesized with brake sites added on the vector by IDT, Inc. such both could be amplified using the aforementioned primers, P31 and P32. The PCR products of H. influenzae and B. fragilis were digested with KpnI and HindIII and ligated into pBAD322 digested with the same enzymes to class pMad9 and pMad6, respectively. The PCR products from North. meningitidis, C. jejuni, P. marinus and Synechococcus sp. were similarly digested and ligated into pBAD322 [15] to form pMad12, pMad76, pMad77, and pMad78, respectively, except that the enzymes used were NcoI and HindIII for North. meningitidis and XbaI and SalI for C. jejuni bioG plus both bioK genes.

To assemble constructs for overexpression and purification, the bioGs of H. influenzae, Northward. meningitidis, and C. jejuni were amplified using primers P15 and P16, P17 and P18, and P46 and P30, respectively, whereas B. fragilis bioGC was amplified using primers P19 and P20. The bioG PCR products of H. influenzae and N. meningitidis and bioGC from B. fragilis were digested with XbaI and XhoI and ligated into pET28b+ digested with the same enzymes to give pMad23, pMad27 and pMad40, respectively, in which the putative esterase genes encode C-terminal hexahistidine-tagged proteins. Similarly the bioG PCR product of C. jejuni was digested with NdeI and SalI and ligated into pET28b+ cut with the same enzymes to requite pMad97 which encodes an North-terminal hexahistidine-tagged protein.

Site directed mutagenesis was done using the QuickChange site-directed mutagenesis method (Stratagene) with primers P52 and P53 to alter the putative active site residues of H. influenzae BioG (pMad23) (Table 1). Pfu polymerase was the PCR polymerase. The PCR reaction product (pMad70) was ethanol precipitated and and so introduced into strain DH5α by chemic transformation. The mutations were verified by sequencing (ACGT, Inc). The plasmids pMad9, pMad12, pMad76, pMad77, and pMAD78 were transformed into Eastward. coli strain STL24 (ΔbioH) whereas pMad6 was transformed into strain STL25 (ΔbioC ΔbioH). The overexpression constructs, pMad23, pMad27, pMad40, pMad97, and pMad70 were each transformed into BL21(DE3) and Tuner (Novagen).

High level expression and purification of BioH and BioG

The protocol to purify BioH and BioG was adapted from that used to purify E. coli BioH [6]. Strains BL21(DE3) or Tuner (Novagen) carrying a pET28b+ plasmid encoding a BioG or E. coli BioH were grown to OD₆₀₀ of 1 in LB-kanamycin medium at 37°C and overexpression was induced past addition of one mM IPTG. The cistron products encoded by pMad23 and pMad40 were soluble when transformed into BL21(DE3) and grown at 37°C for 3 h or 21°C for 16 h, respectively, whereas the gene products encoded by pMad27, pMad97, and pMad70 were soluble in Tuner later on incubation at 21°C for xvi h. The cells were harvested by centrifugation and the prison cell pellets were done with M9 salts and stored at −20°C.

All poly peptide manipulations were done at 4°C or on ice. The cell pellets were resuspended in lysis buffer containing 50 mM 3-(North-morpholino)propanesulfonic acid (MOPS), 10% glycerol, 5 mM 2-mercaptoethanol, 0.five M NaCl and 20 mM imidazole (pH 7.five). The suspension was passed twice through a French pressure cell and then centrifuged 15,000 RPM for ane h to isolate the soluble excerpt which was mixed for 30 min with Ni-NTA resin (Qiagen) that had been previously equilibrated in lysis buffer. The resin was and then washed twice with lysis buffer and twice with wash buffer (lysis buffer containing 40 mM imidazole). After resuspension in wash buffer the resin was loaded into a column. After flow through of the wash buffer was consummate the cavalcade was eluted with lysis buffer containing 180 mM imidazole and fractions were collected. Following purity estimation by SDS-PAGE, the fractions were pooled and dialyzed overnight using Slide-A-Lyzer cassettes (Pierce Chemical) against a buffer of 25 mM MOPS, ten% glycerol, 1 mM tris(2-carboxyethyl)phosphine (TCEP) and 0.2 M NaCl (pH 7.five). E. coli BioH, H. influenzae BioG, and B. fragilis BioGC were full-bodied using Millipore centrifugal concentrators (10,000 MWCO). The proteins were so flash frozen and stored at −80°C.

The hexahistidine-tagged BioG and BioGC proteins were stale under vacuum, and the mass was analyzed past MALDI-TOF/ESI mass spectrometry at University of Illinois, Schoolhouse of Chemical Sciences Mass Spectrometry Laboratory. Size exclusion chromatography was washed on a Superdex 200 analytical size exclusion column calibrated with poly peptide standards from Bio-Rad. The BioG proteins eluted betwixt the chicken ovalbumin (44 kDa) and equine myoglobin (17 kDa) protein standards indicating monomeric proteins. Given that both partners in the BioGC fusion protein are monomeric, BioGC was expected be monomeric and this was the case.

Esterase activity assays

Each reaction contained 50 mM Tris-HCl (pH 7.0), 5% glycerol, 40 µM pimeloyl-ACP methyl ester (or a shorter or longer homologue) and 5 µg/ml of a putative esterase. The mixtures were incubated for i h at 37°C and the products were run on a 20% polyacrylamide gel with 2.v Chiliad urea at 130 V for 3 h. ACP was expressed and purified as previously described [16]. The mono-methyl esters of the dicarboxylic acids were obtained and converted to ACP thioesters using acyl-ACP synthetase as previously described [vi].

Results

BioG, BioK and BioH share conserved residues characteristic of esterase activity

We performed bioinformatics analyses of sixteen BioGs and seventeen BioKs using Muscle [17], [18] and constitute that these proteins had the hallmarks of α,β-hydrolases, about notably all contain the aspartic acrid, histidine and putative catalytic serine residues feature of esterases [19] that aligned with those of BioH proteins (Figure three). Conservation of the catalytic triad regions among a number of BioG, BioK and BioH sequences suggested that BioG and BioK proteins could be capable of operation in place of E. coli BioH. Annotation that an E. coli BioH crystal structure was obtained several years agone [20] that demonstrated the catalytic triad and identified serine-82 as the nucleophile. More recently, the structure of a BioH-methyl-pimeloyl-ACP circuitous was determined that allowed demonstration that BioH action prevents elongation of the pimeloyl moiety to a physiologically useless product [21].

Figure three. Sequence alignments putative biotin constructed esterases evidence conserved catalytic triad residues.

Homologues of BioH, BioG and another putative isozymes, BioK, were obtained from the SEED database (http://theseed.uchicago.edu/FIG/alphabetize.cgi). Shown are some of the sequences from a MUSCLE [18] alignment with a −one extension gap penalty. The putative catalytic sites are shaded in yellow. The residue numbers (given in Italics) are those of East. coli BioH, H. influenzae BioG and P. marinus BioK.

https://doi.org/x.1371/periodical.pone.0049440.g003

BioG and BioK both replace Eastward. coli BioH in vivo

All known bioG and bioK genes are establish immediately upstream of bioC where the operon-sited bioH genes are plant [xi]. Moreover, Bacteriodes fragilis encodes a protein that appears to have BioG fused to BioC [11]. Given these genomic contexts plus our identification of the proteins as putative α,β-hydrolases we tested if expression the bioG genes of several diverse bacteria and the bioK genes of two blue-green alga could complement the biotin auxotrophy of an E. coli strain carrying a deletion of bioH (ΔbioH) and thereby allow growth in the absence of biotin. The genes tested were based on several criteria. H. influenzae bioG was chosen due to the fairly close evolutionary relationship of this bacterium with E. coli whereas Northward. meningitidis bioG was chosen considering its genome also contains a bioH bioC cluster in add-on to the bioG bioC cluster. B. fragilis bioG was called because its coding sequence is fused to that of bioC. The C. jejuni bioG was chosen because the protein shares only 27% and 24% sequence identity with the BioGs of H. influenzae and North. meningitides (which are 84% identical to ane another). The two cyanobacterial bioK genes, those of P. marinus MIT-9211 and Synechococcus sp. CC9902 were called because cyanobacterial proteins tend to have niggling sequence similarity and these two proteins share only 35% sequence identity.

To test the part of these genes in the Due east. coli biotin constructed pathway, pBAD322 plasmid derivatives carrying either bioG or bioK were transformed into a ΔbioH derivative of E. coli strain MG1655 and the transformants were streaked onto M9 minimal media lacking biotin that contained either 0.two% arabinose (the inducer of the araBAD promoter) or 0.2% glycerol (which gives basal expression) as sole carbon source (Figure 4).

Effigy iv. Expression of the bioG and bioK genes in E. coli replaces bioH part in vivo.

East. coli strain STL24 (ΔbioH) was transformed with derivative of pBAD322 carrying various bioG or bioK genes. The transformants were streaked on M9 agar plates in the design shown on the plate diagram containing the carbon source shown in either the presence or absence of biotin (bio). All plates were incubated at 37°C except those expressing P. marinus bioK which were incubated at 25°C. To prevent cross-feeding plates divided into three zones by plastic walls were used. Console A. Arabinose every bit carbon source, STL24 ΔbioH transformed with pBAD322 carrying no insert (lower left third), expressing bioG or bioK (pinnacle third of each plate) and the wild type strain transformed with pBAD322 (lower right of each plate). Panel B. The inoculation design was the same as Console A and glycerol was the carbon source in place of arabinose. Console C. The streaking pattern is given by the plate diagram. Arabinose was the carbon source and the test strain was Eastward. coli strain STL25 (ΔbioCΔbioH) transformed with pBAD322 carrying no insert (lower left third), bioGC (top third) or the wild type strain transformed with the vector pBAD322 (lower right third).

https://doi.org/x.1371/periodical.pone.0049440.g004

In the absence of biotin and the presence of arabinose the E. coli ΔbioH strain expressing H. influenzae bioG grew similarly to the wild type strain whereas the ΔbioH strain carrying the empty vector showed no growth (Effigy 4A). On glycerol, which allows only basal expression from the araBAD promoter, growth as well proceeded but more slowly. In contrast the N. meningitidis and C. jejuni bioG genes, respectively, showed poor and no growth of the East. coli ΔbioH strain in the presence of arabinose simply robust growth occurred when biotin was added (Figure 4A). When glycerol was the carbon source both bioG genes supported growth in the absence of biotin (Figure 4B). These growth data indicate that the toxicity of arabinose induction is limited to the biotin synthetic pathway.

The bioK genes of Synechococcus and P. marinus also allowed growth of the E. coli ΔbioH strain in the absence of biotin, but merely upon arabinose induction (Figure 4A). However, the growth supported by P. marinus bioK occurred merely at a low growth temperature (25°C) and required four days of incubation (Figure 4B).

To test role of the putative bifunctional enzyme encoded by the B. fragilis bioGC factor the plasmid was transformed into a ΔbioH ΔbioC doubly deleted derivative of E. coli strain MG1655 and streaked on M9 agar containing 0.2% arabinose and lacking biotin. This strain showed potent growth whereas the strain transformed with the empty pBAD322 vector failed to abound (Effigy 4C). Thus the B. fragilis cistron replaced the functions of two E. coli bio genes, bioC and bioH, indicating that both the bioG and bioC domains of the protein are functional.

Although the growth requirements vary, these data all betoken that BioG and BioK proteins of diverse sequence functionally supervene upon E. coli BioH. The complementation of bioH by bioG and bioK also indicates that like BioH, their functions are likely interrelated with that of BioC equally implied past the functional fusion of the B. fragilis BioC and BioG domains. In addition, the in vivo experiments suggest that non-specific hydrolysis of biotin intermediates can occur.

BioG and BioGC recognize a biotin precursor in vitro

Although the complementation assay and bioinformatics studies indicated that, similar BioH, BioG and BioK part every bit esterases, further studies to characterize these enzymes required in vitro studies. Constructs encoding hexahistidine-tagged versions of the BioG and BioGC proteins were expressed and the purified proteins were readily obtained by Ni^two+-chelate chromatography (Figure 5). The purified proteins were analyzed by both MALDI mass spectroscopy and size exclusion chromatography (Tabular array ii). Unfortunately, this was non the example for the BioK proteins. Both BioK proteins invariably formed insoluble inclusion bodies under a wide diverseness of expression conditions and thus no active proteins were obtained. The activities of the BioG proteins were adamant using a gel electrophoretic mobility shift assay [sixteen]. ACP is a dynamic protein, which attains a large effective radius in this partially denaturing gel system. The protein construction is stabilized against denaturation by acyl chains fastened to the ACP prosthetic group with the degree of stabilization depending on the length and polarity of the acyl chain [sixteen].

Figure 5. Purification of His₆-tagged BioG proteins.

Samples (10 µl) of each eluted fraction were analyzed by electrophoresis on 10% SDS-polyacrylamide gels. The lysate and soluble fractions are given in lanes L and Southward, respectively. The protein was eluted from the Ni-NTA column with a buffer containing 200 mM imidazole. The fractions shown were pooled and dialyzed as described in Experimental Procedures. Low range molecular weight markers are shown in lanes marked Chiliad.

https://doi.org/10.1371/journal.pone.0049440.g005

The BioG and BioGC proteins were assayed for conversion of pimeloyl-ACP methyl ester, the physiological BioH substrate, to pimeloyl-ACP. As reported previously the reaction mixture containing BioH gave a ring of lower mobility indicating hydrolysis of the ester moiety (Effigy half-dozen) This is because the new charged ω-carboxyl group plus loss of the hydrophobic methyl ester destabilized the hydrophobic ACP acyl chain bounden cleft causing the ACP moiety to expand. The BioG proteins of H. influenzae, N. meningitidis, and C. jejuni besides as the B. fragilis BioGC fusion protein hydrolyzed pimeloyl-ACP methyl ester (Figure 5, Lanes 4–half-dozen), which shows that under these conditions, the BioGs recognize and hydrolyze the same substrate every bit E. coli BioH. Hence, these data are in fantabulous accord with the in vivo complementation data. Notation that this assay has been validated by mass spectroscopy [6].

Figure 6. The BioG proteins cleave the ester grouping of pimeloyl-ACP methyl ester.

The reaction mixtures containing pimeloyl-ACP methyl ester were mixed with purified BioH equally a positive control or a purified BioG from one of four different leaner as shown. Lane i lacked enzyme added. Post-obit incubation for 1 h 10 µl of the reaction mixture was analyzed by electrophoresis on a 20% polyacrylamide gel containing ii.5 Chiliad urea.

https://doi.org/x.1371/periodical.pone.0049440.g006

H. influenzae BioG is a serine esterase

Prior work had shown that substitution of alanine for the putative serine nucleophile of Due east. coli BioH abolished both the in vivo and in vitro action of the protein [6]. To test if H. influenzae BioG functioned similarly we constructed an S65A derivative of the protein in the pET28b+ expression vector. Based on prior data nosotros expected that the repressed level of expression of wild type H. influenzae BioG from the pET28b+ promoter would be sufficient to permit growth of the E. coli ΔbioH strain in the absence of biotin. Indeed robust growth was observed (Figure 7A). In contrast the plasmid that encoded the BioG S65A protein failed to allow growth under these weather. Upon biotin supplementation all strains grew. All the same, the lack of complementation by BioG S65A could accept been the upshot of inclusion torso formation by the mutant protein. This was not the case. The mutant protein was duplicate from the wild type BioG poly peptide in that it was readily expressed in soluble form and purified (Figure 7B). However, in contrast to the wild blazon BioG protein, the BioG S65A protein had no detectable activity in the pimeloyl-ACP methyl ester cleavage analysis (Figure 7C). Taken together with the in vivo information, it is clear that H. influenzae BioG, like E. coli BioH, is a serine esterase.

Figure 7. Loss of BioG function upon exchange of the putative active site serine with alanine.

Panel A. East. coli strain STL24 (ΔbioH) was transformed with plasmids encoding H. influenzae BioG S65A (right), wild type H. influenzae BioG domain (pinnacle) or the empty pET28b+ vector (left). The transformants were streaked onto M9 plates containing 0.2% glucose. Panel B Purification of the S65A BioG. Eluted fractions (10 µl) were analyzed by electrophoresis on a 10% SDS-polyacrylamide gel. The lysate and soluble fractions are shown in lanes L and Southward, respectively. Panel C. The H. influenzae BioG S65A poly peptide was assayed for esterase activity as in Figure vi.

https://doi.org/ten.1371/journal.pone.0049440.g007

The BioG proteins accept differing degrees of substrate promiscuity

In the complementation experiments presented above some of the bioG genes required induction of expression past arabinose for robust growth in the absence of biotin whereas others grew well without induction (glycerol every bit carbon source), but grew very poorly in the presence of arabinose. Moreover, addition of biotin to the arabinose plates allowed normal growth. These results indicated that toxicity was specific to the biotin synthetic pathway. The virtually straightforward interpretation is that the toxic BioGs are either less specific or are produced at higher levels than the non-toxic proteins and the excesss activity aborts biotin synthesis. Nosotros favor that onetime possibility because toxic and nontoxic BioG proteins showed similar levels of expression in extracts prepared for protein purification (Figure v). The virtually plausible biotin pathway target for the toxic BioG would be the short intermediates of pimeloyl moiety synthesis. We assayed hydrolysis of ii such intermediates, the ACP thioesters of malonate methyl ester and glutarate methyl ester. Gel mobility shift assays with the four BioG proteins indicate that the two toxic proteins were more promiscuous in their substrate cleavage (Effigy 8). When assayed on the physiological C7 substrate and the non-physiological C9 substrate, the four enzymes had comparable activities. However, both of the toxic BioGs, those of C. jejuni and N. meningitides, broken both the C3 and C5 substrates whereas neither of the non-toxic BioGs H. influenzae and B. fragilis cleaved the C3 substrate. H. influenzae BioG cleaved the C5 substrate whereas B. fragilis BioG was the least promiscuous of the four enzymes in that information technology had only trace activity on the C5 substrate (annotation that the effective BioG concentration was one-half that of the other enzymes due to the greater size of the fusion protein). These results testify that each of these BioGs are capable of hydrolyzing substrates other than the C7 substrate required in biotin synthesis every bit is the instance of E. coli BioH which slowly cleaves the C5 acyl-ACP methyl ester moiety [6]. Information technology should be noted that assay by gel mobility shift is non suited for kinetic determinations considering in order to observe shifted bands at low substrate concentrations, hydrolysis of a major fraction of the substrate is required and thus the reaction kinetics would exist progressively altered during the assays.

Effigy 8. Assay of the purified BioG on shorter and longer analogues of pimeloyl-ACP methyl ester.

Reaction mixtures each containing an acyl-ACP methyl ester were either left untreated (−) or treated (+) with 1 of the BioG proteins as described in Materials and Methods.

https://doi.org/10.1371/journal.pone.0049440.g008

The BioH, BioG and BioK esterases fall into distinct clades

Although BioH, BioG, BioK recognize the same substrate and share the same catalytic residues, their overall sequence identity is low. To examine the phylogenetic relationships of these proteins we first entered a few BioH, BioG and BioK sequences into the Pfam database [22] which placed the esterases into different poly peptide families within the same clan (CL0028). BioHs were members of the α,β -hydrolase 6 family, BioGs the DUF452 family unit and BioKs are members of the α,β-hydrolase 5 family.

To examine the evolutionary distances of BioG, BioH, and BioK a minimum evolution phylogenetic tree was constructed from five BioGs, four BioHs and 5 BioKs both relative to one other and to outlier sequences from two other families of the same association, two bacterial S-formylglutathione hydrolases (esterase family unit) and iii eukaryotic lipases (lipase family unit) (Figure ix). The proteins grouped into 5 clades, every bit expected, with relatively loftier bootstrap values for the nodes linking all of the genes inside each clade (96% for BioG, 97% for BioK, 89% for BioH, 87% for the lipases, and 100% for the esterases (Effigy vii). This shows that despite the biotin synthesis proteins sharing the same biological role, each had followed its own evolutionary path as seen for the outlier esterases of unlike biological functions. Even so, conclusions regarding the relative evolutionary distances of each of the clades from i some other cannot exist drawn because the node positions betwixt the clades bear witness little bootstrap support.

Figure 9. BioH, BioG and BioK are evolutionarily singled-out.

The evolutionary human relationship betwixt sequences from several α,β-hydrolase families was inferred using the Mega5 [32]. Sequences from other families α,β-hydrolases were obtained from the Pfam database [22]. The bootstrap percentage values for 1000 replicates are shown adjacent to the branches. The optimal tree is drawn to scale, with branch lengths in the same units equally those of the evolutionary distances (the number of amino acid residue substitutions per site). The scale represents a l% difference in compared residues per length. The analysis involved 23 amino acid sequences. All positions containing gaps and missing information were eliminated. The final dataset had a full of 148 positions. Bootstrap values lower than 80% are not shown.

https://doi.org/x.1371/periodical.pone.0049440.g009

The ability to determine evolutionary distances between each of the sequences within a biotin synthetic esterase clade vary, however, some distances can be inferred. In the case of BioK, the 97% bootstrap value suggest that the P. marinus and Synechococcus sp BioKs diverged from a different ancestor than did the other BioKs, however, there is niggling bootstrap support for the position of the nodes linking each of the BioK sequences to one another. However this shows that fifty-fifty a protein sequence within a clade can follow an evolutionary path singled-out from some other protein within that clade. In the case of BioH, there is strong evidence that E. coli and Y. enterocolitica BioHs are more than closely related to one another than to P. aeruginosa BioH (100% and 89% bootstrap support, respectively), however, there was insufficient bootstrap support to assess the altitude of N. meningitidis BioH to any of the other BioHs. In dissimilarity, there is strong bootstrap support for the relative distances of all of the BioGs. The distance between the BioGs from N. meningitidis and H. influenzae are close (100% bootstrap support) compared to the other BioGs. The evolutionary distance from B. fragilis BioG to H. influenzae and N. meningitidis BioG is longer (96% bootstrap support) whereas C. jejuni BioG is nearly distant from any of the other BioG genes (96% bootstrap support). These results show that inside both the BioG and BioH clades, the evolutionary distances between sequences can vary profoundly even if those sequences share an evolutionary path. Despite the wide variation in evolutionary distances between all of the BioG, BioK, and BioH proteins, the in vivo results show that each poly peptide is capable of performing the same hydrolytic function on the C7 acyl-ACP methyl ester moiety. All the same, considering that supposedly related BioGs (from H. influenzae and N. meningitidis) take different enzymatic activities under the same weather condition, it shows that that even BioGs having similar sequences can accept enzyme activities that differ more than two BioGs separated past a greater evolutionary distance (North. meningitidis and C. jejuni).

Discussion

Given the stiff conservation of the other biotin synthetic enzymes across biology, the diversity of proteins that catalyze cleavage of pimeloyl-ACP methyl ester is hitting. None of these enzymes appear to accept been newly evolved or acquired because their coding sequences frequently overlap with both the downstream bioC coding sequence and the upstream bioF coding sequence (Figure 2, Table 3). Therefore, the esterase genes seem well integrated into their respective operons. Indeed, the translational coupling imparted by overlapping genes should result in a set up ratio of esterase activity to that of BioC and BioF which seems important because high level expression of certain BioG proteins results in loss of the power to supercede East. coli BioH seen upon moderate expression (Figure four). Other workers have reported that overproduction of Due east. coli BioH compromises E. coli biotin synthesis [10]. Equally noted in a higher place some BioH proteins, notably that of E. coli, are not encoded inside a biotin constructed operon, but elsewhere on the genome. Such freestanding genes are not readily identified considering bacterial genomes encode many esterases (Due east. coli has at least xv) [23]. Indeed, the Due east. coli bioH gene was discovered only when deletion assay of a neighboring cistron cluster engendered a biotin requirement [24]. Hence, there may well be unrecognized examples of BioG and BioK genes located outside biotin gene clusters. Freestanding bioK genes would seem particularly difficult to recognize considering cyanobacterial proteins take little sequence similarity even among bacteria idea to exist closely related. I possible caption for the diversity of the esterases of biotin synthesis relative to the rather strict conservation seen in the biotin ring formation enzymes is that ester hydrolysis is a simple reaction whereas ring formation requires much more complex chemical science. Indeed, collaborative work from this laboratory has shown that a P. aeruginosa PAO1 esterase of unknown role can accomplish BioH action past simple amino acid substitutions [25]. Note that like East. coli BioH [26] and virtually all other α,β-hydrolases [27], [28], the BioG proteins behave equally monomeric proteins in solution (Table 2). Given these data and the finding that Bacillus cereus BioC is monomeric [29] the BioGC poly peptide seemed likely to be monomeric and this was the instance (Table ii).

A possible caveat to our in vitro data is that we have used E. coli ACP rather than the cognate ACP of each of the bacteria. Even so, since each of the BioGs (besides every bit the BioKs) replaced BioH function in E. coli the BioGs conspicuously recognize the substrate when attached to E. coli ACP. Based on the structure of the complex of BioH with methyl pimeloyl-ACP [21] this may exist due to the highly conserved helix II of ACP. The interactions of BioH with the substrate ACP moiety are exclusively with helix II and all of the BioG-containing organisms we tested have ACP helix Ii sequences very like to that of E. coli.

The B. fragilis BioG-BioC fusion poly peptide postulated by Rodionov and coworkers [eleven] has been expressed and has both of the postulated activities, the protein simultaneously replaces both BioH and BioC in E. coli and its BioG activity has been demonstrated in vitro. It is interesting that of the Bacteroides species of known genome sequence merely the three B. fragilis genomes encode the fusion protein. The other Bacteroides genomes (B. thetaiotaomicron, B. xylanisolvens, B. vulgatus, B. helcogenes and B. salanitronis) each encode discrete BioG and BioC proteins. The B. thetaiotaomicron proteins can be readily aligned (58–60% amino acid residuum identity) with the B. fragilis fusion protein and the alignments leave only a gap of ten residues betwixt the sequences that align with BioG and BioC. The fusion protein sequence opposite the alignment gap is NLAPAAAASS, a sequence that closely resembles the flexible linker regions that allow inter-subunit communication in enzymes such as pyruvate dehydrogenase and acetyl-CoA carboxylase [thirty], [31]. Given the gene club of biotin operons formation of the fusion protein can readily be envisioned. However, bifunctional fusion proteins generally catalyze sequent steps in a pathway (e.one thousand., the bifunctional Due east. coli TrpC and TrpD proteins) whereas the BioC-BioG poly peptide does non; the two reactions are separated by two cycles of fat acid synthesis. However, if we consider fatty acid synthesis to be an essential cell procedure that must always be performed (because cells must make membranes), then BioC-catalyzed methylation and BioG catalyzed ester cleavage tin can be considered consecutive steps.

In conclusion the enzymes that remove the methyl group of methyl-pimeloyl-ACP prove a diversity that appears of long standing. Each of the leaner nosotros studied announced to accept acquired a gene that encodes an α,β-hydrolase that performs the required function without disruption of other cellular processes. The gene became integrated into the biotin constructed operon where it remains a stable entity. Whether the gene encodes BioH, BioG or BioK seems of no consequence and there appears to be little or no selective pressure to favor one gene over another.

Acknowledgments

We thank Dr. Steven Lin for bacterial strains and advice.

Author Contributions

Conceived and designed the experiments: MMS JEC. Performed the experiments: MMS VC. Analyzed the data: MMS JEC VC. Contributed reagents/materials/analysis tools: MMS. Wrote the paper: MMS JEC.

References

1. ane. Attwood PV, Wallace JC (2002) Chemical and catalytic mechanisms of carboxyl transfer reactions in biotin-dependent enzymes. Acc Chem Res 35: 113–120.
   • View Article
   • Google Scholar
2. two. Knowles JR (1989) The mechanism of biotin-dependent enzymes. Annu Rev Biochem 58: 195–221.
   • View Article
   • Google Scholar
3. iii. Ifuku O, Miyaoka H, Koga N, Kishimoto J, Haze South, et al. (1994) Origin of the carbon atoms of biotin. Eur J Biochem 220: 585–591.
   • View Commodity
   • Google Scholar
4. 4. Sanyal I, Lee S, Flint DH (1994) Biosynthesis of pimeloyl-CoA, a biotin forerunner in Escherichia coli, follows a modified fatty acid synthesis pathway: ¹³C-labeling studies. J Am Chem Soc 116: 2637–2638.
   • View Article
   • Google Scholar
5. 5. Lezius A, Ringelmann E, Lynen F (1963) Zur biochemischen Funktion des Biotins. 4. Die Biosynthese des Biotins. Biochem Z 336: 510–525.
   • View Commodity
   • Google Scholar
6. 6. Lin S, Hanson RE, Cronan JE (2010) Biotin synthesis begins by hijacking the fatty acid synthetic pathway. Nat Chem Biol 6: 682–688.
   • View Article
   • Google Scholar
7. 7. Cleary PP, Campbell A (1972) Deletion and complementation analysis of the biotin cistron cluster of Escherichia coli. J Bacteriol 112: 830–839.
   • View Article
   • Google Scholar
8. 8. Barker DF, Campbell AM (1980) Use of bio-lac fusion strains to written report regulation of biotin biosynthesis in Escherichia coli. J Bacteriol 143: 789–800.
   • View Commodity
   • Google Scholar
9. ix. Cronan JE Jr (1989) The Eastward. coli bio operon: transcriptional repression by an essential protein modification enzyme. Cell 58: 427–429.
   • View Article
   • Google Scholar
10. 10. Koga N, Kishimoto J, Haze S, Ifuku O (1996) Assay of the bioH gene of Escherichia coli and its issue on biotin productivity. J Ferment Bioeng 81: 482–487.
    • View Article
    • Google Scholar
11. eleven. Rodionov DA, Mironov AA, Gelfand MS (2002) Conservation of the biotin regulon and the BirA regulatory signal in Eubacteria and Archaea. Genome Res 12: 1507–1516.
    • View Article
    • Google Scholar
12. 12. Akatsuka H, Kawai E, Sakurai Northward, Omori 1000 (2003) The Serratia marcescens bioH gene encodes an esterase. Factor 302: 185–192.
    • View Article
    • Google Scholar
13. xiii. Miller J (1992) A short course in bacterial genetics: a laboratory manual and handbook for Escherichia coli and related bacteria: Cold Jump Harbor Laboratory Press.
14. 14. Balaban Grand, Hendrixson DR (2011) Polar flagellar biosynthesis and a regulator of flagellar number influence spatial parameters of cell division in Campylobacter jejuni. PLoS Pathog 7: e1002420.
    • View Article
    • Google Scholar
15. 15. Cronan JE (2006) A family of arabinose-inducible Escherichia coli expression vectors having pBR322 copy control. Plasmid 55: 152–157.
    • View Article
    • Google Scholar
16. 16. Cronan JE, Thomas J (2009) Bacterial fatty acid synthesis and its relationships with polyketide synthetic pathways. Methods Enzymol 459: 395–433.
    • View Article
    • Google Scholar
17. 17. Edgar RC (2004) MUSCLE: a multiple sequence alignment method with reduced time and infinite complexity. BMC Bioinformatics 5: 113.
    • View Article
    • Google Scholar
18. 18. Edgar RC (2004) Musculus: multiple sequence alignment with high accuracy and high throughput. Nucleic Acids Res 32: 1792–1797.
    • View Article
    • Google Scholar
19. nineteen. Polgar L (2005) The catalytic triad of serine peptidases. Cell Mol Life Sci 62: 2161–2172.
    • View Commodity
    • Google Scholar
20. 20. Sanishvili R, Yakunin AF, Laskowski RA, Skarina T, Evdokimova E, et al. (2003) Integrating structure, bioinformatics, and enzymology to discover office: BioH, a new carboxylesterase from Escherichia coli. J Biol Chem 278: 26039–26045.
    • View Commodity
    • Google Scholar
21. 21. Agarwal V, Lin Due south, Lukk T, Nair S, Cronan J (2012) Structure of the Enzyme-ACP Substrate Gatekeeper Complex Required for Biotin Synthesis. Proc Natl Acad Sci U S A 109 in press.
    • View Article
    • Google Scholar
22. 22. Finn RD, Mistry J, Tate J, Coggill P, Heger A, et al. (2010) The Pfam protein families database. Nucleic Acids Res 38: D211–222.
    • View Commodity
    • Google Scholar
23. 23. Xie X, Wong WW, Tang Y (2007) Improving simvastatin bioconversion in Escherichia coli by deletion of bioH. Metab Eng nine: 379–386.
    • View Article
    • Google Scholar
24. 24. Schwartz One thousand (1966) Location of the maltose A and B loci on the genetic map of Escherichia coli. J Bacteriol 92: 1083–1089.
    • View Article
    • Google Scholar
25. 25. Flores H, Lin S, Contreras-Ferrat One thousand, Cronan JE, Morett Due east (2012) Evolution of a new function in an esterase: simple amino acid substitutions enable the activeness present in the larger paralog, BioH. Protein Eng Des Sel 25: 387–395.
    • View Article
    • Google Scholar
26. 26. Tomczyk NH, Nettleship JE, Baxter RL, Crichton HJ, Webster SP, et al. (2002) Purification and characterisation of the BIOH protein from the biotin biosynthetic pathway. FEBS Lett 513: 299–304.
    • View Commodity
    • Google Scholar
27. 27. Heikinheimo P, Goldman A, Jeffries C, Ollis DL (1999) Of barn owls and bankers: a lush variety of alpha/beta hydrolases. Structure 7: R141–146.
    • View Commodity
    • Google Scholar
28. 28. Jochens H, Hesseler Grand, Stiba K, Padhi SK, Kazlauskas RJ, et al. (2011) Protein engineering of alpha/beta-hydrolase fold enzymes. Chembiochem 12: 1508–1517.
    • View Commodity
    • Google Scholar
29. 29. Lin S, Cronan JE (2012) The BioC O-Methyltransferase Catalyzes Methyl Esterification of Malonyl-Acyl Carrier Protein, an Essential Step in Biotin Synthesis. J Biol Chem in press.
    • View Article
    • Google Scholar
30. 30. Cronan JE Jr (2002) Interchangeable enzyme modules. Functional replacement of the essential linker of the biotinylated subunit of acetyl-CoA carboxylase with a linker from the lipoylated subunit of pyruvate dehydrogenase. J Biol Chem 277: 22520–22527.
    • View Article
    • Google Scholar
31. 31. Perham RN (2000) Swinging arms and swinging domains in multifunctional enzymes: catalytic machines for multistep reactions. Annu Rev Biochem 69: 961–1004.
    • View Commodity
    • Google Scholar
32. 32. Tamura Thou, Peterson D, Peterson Due north, Stecher Yard, Nei M, et al. (2011) MEGA5: molecular evolutionary genetics assay using maximum likelihood, evolutionary altitude, and maximum parsimony methods. Mol Biol Evol 28: 2731–2739.
    • View Article
    • Google Scholar

dunnhatomentand.blogspot.com

Source: https://journals.plos.org/plosone/article?id=10.1371%2Fjournal.pone.0049440