ABSTRACT
BioF (8-amino-7-oxononanoate synthase) is a strictly conserved enzyme that catalyzes the first step in assembly of the fused heterocyclic rings of biotin. The BioF acyl chain donor has long been thought to be pimeloyl-CoA. Indeed, in vitro the Escherichia coli and Bacillus sphaericus enzymes have been shown to condense pimeloyl-CoA with l-alanine in a pyridoxal 5′-phosphate-dependent reaction with concomitant CoA release and decarboxylation of l-alanine. However, recent in vivo studies of E. coli and Bacillus subtilis suggested that the BioF proteins of the two bacteria could have different specificities for pimelate thioesters in that E. coli BioF may utilize either pimeloyl coenzyme A (CoA) or the pimelate thioester of the acyl carrier protein (ACP) of fatty acid synthesis. In contrast, B. subtilis BioF seemed likely to be specific for pimeloyl-CoA and unable to utilize pimeloyl-ACP. We now report genetic and in vitro data demonstrating that B. subtilis BioF specifically utilizes pimeloyl-CoA.
IMPORTANCE Biotin is an essential vitamin required by mammals and birds because, unlike bacteria, plants, and some fungi, these organisms cannot make biotin. Currently, the biotin included in vitamin tablets and animal feeds is made by chemical synthesis. This is partly because the biosynthetic pathways in bacteria are incompletely understood. This paper defines an enzyme of the Bacillus subtilis pathway and shows that it differs from that of Escherichia coli in the ability to utilize specific precursors. These bacteria have been used in biotin production and these data may aid in making biotin produced by biotechnology commercially competitive with that produced by chemical synthesis.
INTRODUCTION
Biotin, previously known as vitamin H and vitamin B7, is an essential enzyme cofactor found in all three domains of life. This cofactor is necessary for essential steps of central metabolism, including fatty acid synthesis and amino acid degradation (1). De novo synthesis of biotin is restricted to archaea, bacteria, plants, and some fungi (2). Animals, including humans, cannot synthesize biotin and must rely on exogenous sources for the vitamin and thus the enzymes of biotin biosynthesis have become popular drug targets for development of novel antimicrobial agents (3–5). Biotin consists of two fused heterocyclic rings plus a valeric acid acyl chain (Fig. 1). Assembly of the first heterocyclic ring (the tetrahydrothiophene ring) plus the valeric acid chain begins with the BioF reaction, catalyzed by 8-amino-7-oxononanoate (AON) synthase (enzyme EC 2.3.1.47), which is often called 7-keto-8-aminopelargonate synthase in the literature; AON is often abbreviated as KAPA. The mechanism of BioF, a pyridoxal 5′-phosphate-dependent enzyme which catalyzes the decarboxylative condensation of l-alanine with a monothioester of pimelic acid (heptanedioic acid) to form 8(S)-amino-7-oxononanoate is well studied (6–12). However, the identity of the thiol moiety of the thioester substrate used in vivo remained unclear and it seemed possible that it could vary among different biotin-producing organisms. In Eisenberg's pioneering studies of biotin synthesis (8) his insight was that synthesis of AON could proceed by a mechanism analogous to that of synthesis of a heme precursor, δ-aminolevulinic acid, in mammals in which glycine is condensed with succinyl-CoA. Based on this hypothesis, pimeloyl-CoA and l-alanine were the substrates tested for AON synthesis in Escherichia coli crude extracts. However, later studies with purified E. coli BioF showed that the enzyme required 25-μM pimeloyl-CoA for half-maximal activity (11). Since biotin synthesis is a very low-demand pathway and E. coli requires only a few hundred biotin molecules per cell (13), this substrate concentration seemed excessive. Moreover, the BioF of Bacillus sphaericus, which (unlike E. coli) encodes a pimeloyl-CoA synthetase that converts pimelic acid to pimeloyl-CoA, has a reported 1-μM Michaelis constant for pimeloyl-CoA (9). Thus, it seemed possible that pimeloyl-CoA was acting as a model substrate in the E. coli BioF reaction and would have only a limited in vivo relevance. This would not be unprecedented. Several fatty acid synthetic enzymes are known to function with acyl-CoA substrates. The usual pattern is that acyl-CoAs give good maximal activities, but only at much higher concentrations than those for the cognate acyl-ACPs (14).
The canonical pathway of assembly of the heterocyclic rings of biotin (A) and the B. subtilis enzymes of pimeloyl-thioester synthesis (B and C). Note that although the BioW reaction is well characterized (17, 46), the BioI cleavage of long-chain acyl-ACP substrates has not been shown to be catalytic in vitro (22). Abbreviations: AON, 8-amino-7-oxononanoate; DAN, 2,8-amino-7-oxononanoate; SAM, S-adenosyl-l-methionine; AMTOD, S-adenosyl-2-oxo-4-thiomethylbutyrate; 5′-DOA, 5′-deoxyadenosine. In the literature 2,8-amino-7-oxononanoate is often informally called 7,8-diaminopelargonic acid (DAPA) and AON is often informally called KAPA.
Recent studies have shown that E. coli uses a modified fatty acid synthetic pathway to produce pimelic acid in thioester linkage to the acyl carrier protein (ACP) of fatty acid synthesis, which functions in the BioF reaction (15, 16). In contrast, although Bacillus subtilis also produces the pimelate moiety by fatty acid synthesis, the pimeloyl-CoA synthetase encoded by bioW (17) is required for biotin synthesis (18). This finding plus the ability of the combination of BioW plus exogenous pimelic acid to bypass E. coli mutants blocked in pimelate synthesis (16, 19) suggested that in E. coli either pimeloyl-CoA or pimeloyl-ACP could support BioF activity, whereas in B. subtilis BioF might be unable to utilize pimeloyl-ACP. We report genetic and in vitro data indicating that B. subtilis BioF utilizes only pimeloyl-CoA for AON synthesis. We also report a bacterial BioWF fusion protein that has both BioW and BioF activities.
RESULTS
Expression of B. subtilis BioF failed to allow growth of an E. coli ΔbioF strain unless BioW was also expressed.Although to date all biotin-producing organisms studied encode a recognized BioF based on the active site residues known from the crystal structures and mutagenesis studies of the E. coli (6) and Mycobacterium smegmatis BioF proteins (E. coli numbering positions Lys236, His133, Glu175, Asp204, and His207) (7, 20), the overall sequences generally do not show high percentages of conserved residues. Most BioF proteins show only about 30% sequence identity with one another, with little dependence on the biological source. For example, alignment of the B. sphaericus BioF with that of the plant Arabidopsis thaliana shows 32% sequence identity, a value identical to that seen in alignments of the BioF proteins of B. sphaericus and E. coli. Moreover, although the two proteins have essentially the same crystal structure, the M. smegmatis and E. coli BioF proteins have only 37% sequence identity (20). The E. coli and B. subtilis proteins reflect this trend in that the two proteins show only 34% sequence identity.
Given that expression of the A. thaliana enzyme was reported to allow growth of an E. coli bioF mutant strain in the absence of biotin (21), we expected that B. subtilis BioF expression would complement an E. coli ΔbioF strain. However, this was not the case. In the absence of biotin, growth of the E. coli ΔbioF strain was observed only upon expression of both B. subtilis BioF and BioW proteins in the presence of exogenous pimelic acid (Fig. 2). These results indicated that B. subtilis BioF was unable to accept pimeloyl-ACP produced by E. coli but became functional when pimelic acid was converted to pimeloyl-CoA by BioW.
BioW is required for complementation of an E. coli ΔbioF strain expressing B. subtilis BioF. The E. coli ΔbioF strain STL115 was transformed with a plasmid that expressed B. subtilis bioF (upper right sector) or with the B. subtilis bioF plasmid plus a second plasmid expressing B. subtilis bioW (lower right sector). The two plasmids express the B. subtilis genes from the E. coli araBAD promoter and have compatible replication origins and different antibiotic selections. Empty vector controls were plated in the left hand sectors. The carbon sources were l-arabinose and glycerol, each at 0.2%.
We have also tested whether in its native context B. subtilis BioF is unable to use pimeloyl-ACP derived from its native ACP. Our prior work showed that BioW is an essential biotin synthetic enzyme whereas BioI is not (18). BioW produces pimeloyl-CoA whereas BioI is a cytochrome P450 enzyme that binds and cleaves long-chain acyl-ACPs to pimeloyl-ACP (22, 23). These phenotypes suggested that B. subtilis BioF would be unable to utilize pimeloyl-ACP. However, the B. subtilis and E. coli ACPs are only 60% identical and although B. subtilis ACP functionally replaces E. coli ACP when expressed in E. coli (24), it remained possible that E. coli ACP is incompatible with B. subtilis BioF. To test this possibility, we introduced an isopropyl-β-d-thiogalactopyranoside (IPTG)-inducible E. coli bioF gene into the chromosomes of the bioW and ΔbioW ΔbioI strains to provide an assay for pimeloyl-ACP synthesis. Note that the bioW disruption allele contains a promoter that expresses the downstream genes (18). Expression of E. coli bioF allowed growth of the strain lacking BioW but not growth of the strain that lacked both BioW and BioI (Fig. 3). Hence, in contrast to E. coli BioF, B. subtilis BioF was unable to utilize the BioI-generated pimelate thioester of its cognate ACP. As noted above, we and others (16, 19) have reported that E. coli BioF can use pimeloyl-CoA synthesized from exogenous pimelate by B. subtilis BioW, as shown by the robust growth of E. coli strains (ΔbioC or ΔbioH) blocked in pimeloyl-ACP synthesis.
BioI is expressed in functional form in B. subtilis and provides a substrate for E. coli BioF but fails to provide a substrate for B. subtilis BioF. The genetic constructs are given below the plates. The upper construct is that used in the left-hand plates and the lower construct is that used in the right-hand plates. E. coli BioF expression was induced with 1 mM IPTG. The orange boxes represent bioW or bioI sequences, whereas the blue boxes represent the E. coli bioF sequences. The disruption plasmid pMUTIN4 contains a promoter to transcribe the downstream genes (18). The bioW::pMUTIN4 strain was MM43 and the bioW::pMUTIN4 ΔbioI strain was MM61.
B. subtilis BioF weakly utilizes pimeloyl-ACP in vitro.Unlike the E. coli (6, 11), B. sphaericus (9), and mycobacterial (7, 20) proteins, B. subtilis BioF invariably formed insoluble aggregates (inclusion bodies) when expressed in E. coli, despite testing of a variety of expression protocols. A fusion protein composed of B. subtilis BioF coupled to the maltose-binding protein solubility tag was also of no avail. The BioF protein aggregated upon removal of the solubility tag and the intact fusion protein was inactive. Therefore, the inclusion bodies were solubilized with guanidine HCl and the protein was refolded by slow dilution of the denaturant. Surprisingly, the refolded BioF was highly active. We also purified the B. sphaericus and E. coli BioF proteins by conventional means (Fig. 4). We used three different assays to monitor BioF activity: CoA release from pimeloyl-CoA measured by HPLC (Fig. 4), bioassay of AON using an E. coli ΔbioF strain, and gas chromatography-mass spectrometry of AON following derivation of its keto, carboxyl, and amino groups. The latter two assays can measure activity with either pimeloyl-CoA or pimeloyl-ACP. Utilization of pimeloyl-ACP by BioF could also be detected by a qualitative gel shift assay because pimeloyl-ACP migrates more slowly than nonacylated ACP (data not shown).
Purification and activity of the BioF proteins of E. coli, B. sphaericus (B. sph), and B. subtilis. Panels A and C show SDS-polyacrylamide gels of the BioF proteins whereas panels B and D show HPLC chromatograms of CoA release from pimeloyl-CoA, with the enzymes given on the panels. The apparent molecular masses of the E. coli, B. sphaericus, and B. subtilis BioF proteins were 43.1, 42.5, and 42.5 kDa, respectively.
We first tested the ability of B. subtilis BioF to use pimeloyl-ACP using bioassay of AON (Fig. 5). In this assay the plates contain biotin-free medium supplemented with a tetrazolium indicator and seeded with an E. coli ΔbioF assay strain. Samples for assay are pipetted onto small paper disks and the plates are incubated. If the samples contain AON it diffuses from the disks and allows growth of the assay strain. Growth results in reduction of the tetrazolium indicator to its insoluble red formazan. The diameter of the formazan deposit denotes the amount of AON applied to the disk. The AON bioassay is much less sensitive than bioassay of biotin and dethiobiotin (μmol versus pmol) (25). probably due to its poor uptake, which limits its use as a quantitative assay. However, neither the B. subtilis nor the B. sphaericus BioF proteins utilized pimeloyl-ACP, whereas E. coli BioF was active with this substrate. All three enzymes showed robust utilization of pimeloyl-CoA (Fig. 5).
Bioassay of AON synthesis. AON samples produced by in vitro BioF reactions with either pimeloyl-CoA or pimeloyl-ACP as the substrate were spotted onto sterile disks on minimal agar plates lacking biotin that contained the E. coli ΔbioF reporter strain STL115 and a redox indicator (16). The upper left panel shows a control lacking enzyme and the upper right panel shows a positive control with 1 μmol of the AON standard. The lower panel on the left shows results for B. subtilis BioF. The upper and lower panels on the right show results for E. coli and B. sphaericus BioF, respectively.
To obtain a more quantitative assay of ACP substrate utilization we attempted to make UV-absorbing or fluorescent derivatives of AON for HPLC assays. However, the diverse functional groups of AON led to complex products. We then turned to gas chromatography with detection by selected-ion monitoring of mass spectra. This required modification of the AON keto, carboxyl, and amino groups to obtain a volatile derivative. Using this assay, we measured AON synthesis over a concentration range of 1 to 30 μM using either pimeloyl-CoA or pimeloyl-ACP (the ACP moiety was that of E. coli). E. coli BioF had essentially the same activity with both substrates with a Michaelis constant of about 10 μM, whereas the refolded B. subtilis BioF gave good data for pimeloyl-CoA with a Michaelis constant of about 15 μM, although the refolded nature of the protein dictates a strong caveat to this value. For unknown reasons, parallel assays of B. subtilis BioF activity with pimeloyl-ACP gave erratic results showing only a very weak dependence on substrate concentration. However, consistent with the bioassays, pimeloyl-ACP gave only about 7% of the activity seen with pimeloyl-CoA at 5-μM substrate concentration. The ACP substrate gave higher activities at higher concentrations but these data were compromised by the lack of dependence on substrate concentration. One possible explanation for these results is that pimeloyl-ACP may have acted both as a substrate and as a stabilizer of the refolded B. subtilis BioF. The refolding process may not have restored a fully native structure and active site. Note that ACP seems to be a rather “sticky” protein in that it often copurifies with proteins expressed in E. coli. Two examples are a cysteine desulfurase (26) and an anion transporter (27). ACP is also reported to stimulate in vitro transposition reactions (28, 29) suggesting that it may have a chaperonin-like activity.
A BioW-BioF fusion protein is active in both reactions.Genomes of several highly diverse bacteria encode putative BioW-BioF fusion proteins. The 656-residue protein of Desulfosporosinus orientis DSM 765 (UniProtKB G7W906) aligns with BioW and BioF of B. subtilis with 35% identity (residues 15 to 252) and 40% identity (residues 271 to 656), respectively, with only a single gap of two residues (excepting the ∼20-residue linker region). If both domains have enzymatic activity, fusion of BioF to BioW would argue that the acyl thioester substrate of these BioF domains is almost certainly pimeloyl-CoA. We expressed the D. orientis DSM 765 fusion protein in an E. coli ΔbioF mutant strain and found that it allowed growth in biotin-free medium but only when exogenous pimelic acid was supplied (Fig. 6). The exogenous pimelate requirement demonstrated that the D. orientis fusion protein had both pimeloyl-CoA synthetase and AON synthase activities. Expression of the fusion protein also allowed growth of a B. subtilis ΔbioW strain in a medium lacking biotin and pimelate (data not shown).
Complementation abilities of the BioWF fusion protein from Desulfosporosinus orientis DSM 765. In the upper left quadrant, strain STL121 (the E. coli ΔbioF strain STL115 carrying a plasmid encoding E. coli BioF) was induced with 1 mM IPTG, whereas in the lower quadrants, the ΔbioF strain expressing the BioWF fusion protein was induced with 0.2% arabinose.
DISCUSSION
The in vivo results of Fig. 2 and 3 clearly demonstrate that B. subtilis BioF cannot utilize pimeloyl-ACP as a substrate either in its native setting or in E. coli. In contrast, pimeloyl-ACP is an excellent substrate for E. coli BioF in both bacterial species. E. coli BioF also shows robust activity with pimeloyl-CoA generated by BioW from exogenous pimelate.
Indeed, in vitro assays of E. coli BioF show essentially identical activities with the two substrates, thereby validating the use by other workers of pimeloyl-CoA as a BioF substrate (6, 8, 11). The problematic results obtained with pimeloyl-ACP as the substrate for the refolded B. subtilis BioF (but not for E. coli BioF) indicate that the in vivo results are the more valid indication of B. subtilis BioF substrate specificity.
Although the biotin pimelate moiety is made by the ACP-dependent fatty acid synthetic pathway (18), the B. subtilis biotin synthetic pathway requires BioW and hence pimeloyl-CoA synthesis. The parsimonious pathway would be to condense pimeloyl-ACP with l-alanine to form AON, as does E. coli. Instead B. subtilis switches thioester moieties from ACP to CoA and this switch is now shown to be essential due to the specificity of B. subtilis BioF. Two plausible pathways for the thioester switch were (i thioesterase cleavage of pimeloyl-ACP followed by BioW-catalyzed activation to pimeloyl-CoA or ii) a transthioesterification similar to that catalyzed by the FabD malonyl-CoA:ACP transacylase of fatty acid synthesis. However, this second pathway is precluded by the requirement of the BioW pimeloyl-CoA synthetase for biotin synthesis (18). Hence we are left with a pathway that is wasteful of ATP, which is required for conversion of free pimelate to pimeloyl-CoA. However, the demand for biotin is very low and thus the waste of ATP would be trivial.
A prior example of substrate divergence in the B. subtilis and E. coli biotin synthesis pathways is that of BioA (2,8-amino-7-oxononanoate synthase), the enzyme that directly follows BioF in the pathway (30) (Fig. 1). E. coli BioA (and the other characterized BioA proteins) use S-adenosyl-l-methionine as an amino donor, whereas B. subtilis BioA uses l-lysine, a much less expensive donor (S-adenosyl-l-methionine synthesis consumes three ATP equivalents, which are lost upon destruction of the deamination product of the canonical BioA reaction) (31). It seems noteworthy that both BioF and BioA are pyridoxal 5′-phosphate-dependent enzymes. Indeed, the overall structure of 2,8-amino-7-oxononanoate synthase is quite similar to that of AON synthase, suggesting that the two enzymes could be evolutionarily related and perhaps share a ancestor (32).
A plausible molecular explanation for the inability of B. subtilis BioF to utilize pimeloyl-ACP is a lack of the basic residues required to interact with the strongly acidic ACP moiety. In the available ACP-enzyme crystal structures, ACP helix II, the helix having the 4′-phosphopantetheine prosthetic group attached to Ser-36 at its amino terminus, is involved in most interactions with enzymes (33). The ACP-enzyme interactions are predominately hydrophilic in nature, with almost all being salt bridges between acidic residues of helix II and enzyme arginine or lysine residues. These data gave rise to the hypothesis that ACP-requiring enzymes will have an arginine/lysine-rich “positive patch” that will provide the primary sites of interaction with ACP helix II (33). E. coli BioF appears to contain such a patch. The E. coli BioF crystal structure shows that the protein is shaped like a half-open hand with the active site (identified by the pyridoxal 5′-phosphate bound to Lys-236) located at roughly the palm/fingers interface (Fig. 7A). Our attempts to obtain cocrystals of E. coli BioF with pimeloyl-ACP have not yet been successful. However, in silico docking studies suggest that a properly oriented ACP molecule could readily fit into the E. coli BioF lumen and five basic residues (four Arg and one Lys) are present on the lumenal surface and could readily form salt bridges with ACP (Fig. 7A). The B. subtilis BioF structure is readily aligned with and modeled on the E. coli BioF structure, with a Swiss Model global quality estimation score of 0.68 (34) (Fig. 7B). The modeled B. subtilis BioF has the open-hand structure found in all extant BioF crystal structures. However, the surface of the B. subtilis BioF lumen lacks the basic residues present in E. coli BioF. Indeed, the basic residues of the E. coli BioF lumen are generally replaced with hydrophobic residues or glutamine in B. subtilis BioF (Fig. 7C). The same picture is seen for the BioF proteins of other members of the B. subtilis group, e.g., Bacillus amyloliquefaciens and Bacillus licheniformis. However, the BioF of Bacillus cereus, the archetype of the other major group of Bacillus species, retains the basic residues (data not shown) consistent with its utilization of the ACP-dependent E. coli pimelate synthesis pathway (15, 35). The “hardwired” nature of the D. orientis BioWF protein would argue that the basic residues would not be conserved upon alignment of the BioF segment of the fusion with E. coli BioF and this was the case. The fusion protein residues found at these positions were often valine (data not shown). As noted above the Bacilli, which based on other criteria (36) separate into two well-studied groups termed the B. subtilis group and the B. cereus group, diverge in their pimelate synthesis pathways in that B. cereus uses the E. coli pathway. The B. cereus biotin operon encodes BioC and BioH proteins that functionally replace the E. coli proteins (15, 35).
Structure of E. coli BioF (6) (PDB accession no. 1dj9) used to model B. subtilis BioF plus alignment of the two proteins. Modeling was performed using the Swiss Model website (34) and the structural alignment resulted from the modeling. The basic residues are shown in blue on the E. coli structure and the analogous regions of the B. subtilis model are in magenta. The asterisks on the alignment mark the basic residues of the E. coli lumenal surface. The catalytically important residues are marked with a boldface V, whereas Lys236, which forms a Schiff base with the pyridoxal 5′-phosphate cofactor, is marked with a boldface Sb.
The B. subtilis bioI gene presents an enigma in that its cognate BioF, the next enzyme in the pathway, cannot utilize the product of the enzyme it encodes. Moreover, bioI is found only in B. subtilis and its close relatives, B. amyloliquefaciens and B. licheniformis. Other BioW-encoding bacteria (e.g., Staphylococcus aureus, Aquifex aeolicus) lack BioI. Although we previously demonstrated that bioI is not required for biotin synthesis (18), its presence allows E. coli BioF, but not B. subtilis BioF, to function in a B. subtilis strain lacking BioW (Fig. 3). Moreover, bioI expression in Corynebacterium glutamicum strains that also express the E. coli bioF and bioB genes resulted in biotin prototrophy (37, 38). Therefore, the bioI mRNA detected in prior transcriptional analyses of B. subtilis (39, 40) is translated into a functional enzyme in its native host and in foreign hosts. Previously, the only data indicating BioI function in B. subtilis were indirect, i.e., increased biotin production in strains in which the sequences between bioB and bioI had been manipulated (19). However, other explanations remain possible for those data, such as increased lifetime of the mRNA encoding the upstream genes and perhaps a role for the gene of unknown function (ytbQ) that is cotranscribed with bioI.
MATERIALS AND METHODS
Media and strains.E. coli and B. subtilis strains (Table 1) were grown in LB medium. Minimal medium for E. coli contained M9 salts, 0.4% glucose, 1 mM MgSO4, 1 μg/ml thiamine, and 0.1% Casamino acids. General defined media for B. subtilis contained Spizizen's salts (ammonium sulfate, 0.2%; dipotassium phosphate, 1.4%; monopotassium phosphate, 0.6%; sodium citrate [2H2O], 0.1%; and magnesium sulfate [7H2O], 0.02%) (41), trace elements (MgCl2, CaCl2, FeCl2, MnCl2, ZnCl2, CuCl2, CoCl2, and NaMoO4), 0.5% glucose, 0.04% potassium glutamate and 1 mM MgSO4. Supplements required for growth included 0.01% tryptophan for B. subtilis 168. Genetic complementation experiments requiring gene expression under Pspac and PBAD promoters were performed in minimal media with 0.5% glycerol as the carbon source. Induction of the respective promoters was done with 1-mM IPTG or 0.2% arabinose. Antibiotics were used in the following concentrations (μg/ml): chloramphenicol, 30; erythromycin sulfate, 1: lincomycin HCl, 25; spectinomycin sulfate, 100; sodium ampicillin, 100; and kanamycin sulfate, 50. AON was purchased from Cayman Chemical as KAPA-HCl. Chromosomal integration of pDR111 carrying E. coli bioF under lac control was done as previously described (18, 42, 43).
Bacterial strains and primers
Purification of E. coli BioF.The E. coli bioF gene was amplified from the strain MG1655 genome by Steven Lin of this laboratory with primers EcBioF For and EcBioF Rev, which introduced a BspHI site overlapping the initiation codon and an XhoI site downstream of the coding sequence. The PCR product was digested with BspHI and XhoI and ligated to expression vector pET28b cut with NcoI and XhoI to give a gene encoding a C-terminally hexahistidine-tagged BioF (44). The recombinant plasmid was transformed into strain BL21(DE3) (Novagen) with selection on kanamycin-containing plates. The recombinant expression strain was grown overnight in LB medium and subcultured into 1 liter of LB. The culture was induced with 0.1 mM IPTG at an optical density at 600 nm (OD600) of 0.7 and the protein was allowed to overexpress for 4 h at 37°C. The cells were harvested and resuspended in 50 mM Tris-HCl, 300 mM NaCl, 10 mM imidazole, 1 mM Tris (2-carboxyethyl) phosphine hydrochloride (TCEP), and 10% glycerol at pH 8. The cells were then lysed by passage through a French press and the clear supernatant was harvested after centrifuging the lysate at 39,000 × g for 30 min. One milliliter of 50% slurry of Ni-NTA agarose (Qiagen) was equilibrated in lysis buffer. The protein sample was passed through the agarose and washed with buffers containing 20-mM and 60-mM imidazole. The BioF protein was eluted with a buffer containing 250-mM imidazole. After SDS-PAGE analysis, fractions of eluate containing the purest protein were pooled and concentrated using a 30-kDa Amicon concentrator (EMD Millipore). The concentrated protein was dialyzed in 20 mM K2HPO4, 100 mM NaCl, 1 mM TCEP, 100 μM pyridoxal 5′-phosphate (PLP), and 20% glycerol at pH 7.5 and flash frozen at −80°C.
Purification of B. sphaericus BioF.The bioF gene of B. sphaericus IFO3525 was amplified from the genome with the primers BsphBioF For and BsphBioF Rev, which introduced an NdeI site overlapping the initiation codon and an XhoI site downstream of the termination codon. The PCR product was digested with NdeI and XhoI and ligated to vector pET30a cut with the same enzymes. The resulting plasmid was transformed into BL21(DE3) Tuner (Novagen) and colonies were selected using kanamycin. The expression strain containing the plasmid was grown overnight in LB medium and subcultured into 1 liter of LB and grown to an OD600 of ∼1 at room temperature. Protein expression was induced with 1 mM IPTG and the culture was grown at room temperature for 15 h. The cells were harvested by centrifugation and resuspended in 20 mM Tris-HCl (pH 7.8), 1 mM tris(2-carboxyethyl)phosphine (TCEP), 100 μM pyridoxal phosphate (PLP), and 10% glycerol. They were then lysed by passage through a French pressure cell. The supernatant obtained after centrifugation of cell lysate at 39,000 × g for 30 min at 4°C was applied to a 5-ml HiTrap Q ion-exchange chromatography column (GE Healthcare) equilibrated with the buffer used to suspend the harvested cells. The column was treated with a linear gradient from 0 to 1 M NaCl in 20 mM Tris-HCl (pH 7.8), 1 mM TCEP, 100 μM PLP, and 10% glycerol, and the protein eluted at about 150 mM NaCl. Fractions containing the highly purified protein were pooled, concentrated, and resuspended in 20 mM K2HPO4 (pH 7.5), 100 mM NaCl, 1 mM TCEP, 1 M ammonium sulfate, 100 μM PLP, and 10% glycerol and applied to a 5-ml HiTrap Phenyl HP hydrophobic interaction chromatography column (GE Healthcare) equilibrated with the same buffer. The column was washed with 20 mM K2HPO4 (pH 7.5), 100 mM NaCl, 1 mM TCEP, and 10% glycerol and the protein was eluted with a linear gradient from 1 to 0 M ammonium sulfate. SDS-PAGE analysis revealed that the protein eluted in the flowthrough, while most of the contaminants bound to the column. The flowthrough was concentrated using a 10-kDa Amicon concentrator and flash frozen at −80°C after dialysis against a storage buffer of 20 mM K2HPO4 (pH 7.2), 100 mM NaCl, 100 μM PLP, 1 mM TCEP, and 10% glycerol.
Purification of B. subtilis BioF under denaturing conditions.The bioF gene of B. subtilis 168 was amplified from the genome with primers BsBioF For and BsBioF Rev, which introduced an NdeI site overlapping the initiation codon and an XhoI site downstream of the termination codon. The PCR product was digested with NdeI and XhoI and ligated to vector pET30a cut with the same enzymes. The resulting plasmid was transformed into BL21(DE3) Tuner (Novagen) and colonies were selected using kanamycin. A culture started from a single colony was grown overnight in LB medium and subcultured into 1 liter of LB to allow growth to an OD600 of 1.0. The culture was induced with 0.7 mM IPTG for 4 h at 37°C. The cells were harvested by centrifugation and washed with 10 mM Tris-HCl, 100 mM NaCl, and 1 mM EDTA (pH 7.5). Cell pellets were suspended in 50 mM Tris-HCl, 200 mM NaCl, 1% Triton X-100, 8% sucrose, and 1 mM phenylmethylsulfonyl fluoride at pH 8. The cells were lysed by passage through a French press and the lysate was centrifuged at 39,000 × g for 30 min. The insoluble inclusion bodies were washed in the same buffer but lacking Triton X-100 and again centrifuged to remove the buffer. The inclusion bodies were then dissolved in 6 M guanidine hydrochloride and 50 mM Tris-HCl (pH 7.5) to a concentration of 10 mg/ml and stirred slowly at 4°C for 4 to 6 h to allow equilibration. The solubilized protein was added dropwise (1 drop/5 s) into 500 ml of a folding buffer of 20 mM K2HPO4, 150 mM NaCl, 1 mM TCEP, 5 mM EDTA, and 0.1 mM phenylmethylsulfonyl fluoride, 40 μM PLP, 0.5 mM l-arginine, and 20% glycerol at pH 7.5 at 4°C while stirring and the solution was allowed to equilibrate overnight. The solution was centrifuged to remove insoluble debris and dialyzed against 50 mM K2HPO4, 150 mM NaCl, 1 mM TCEP, 100 μM PLP, and 10% glycerol (pH 7.5). The protein was concentrated using a 10-kDa Amicon concentrator (EMD Millipore) and flash frozen at −80°C.
Expression of the D. orientis DSM 765 BioWF fusion protein.The gene was amplified from the D. orientis DSM 765 genomic DNA with primers DoBioWF For and DoBioWF Rev, which introduced a SmaI site immediately upstream of the initiation codon and an SphI site downstream of the termination codon. The PCR product was digested with SmaI and SphI and ligated to vector pBAD322 (45) cut with the same enzymes.
High-performance liquid chromatography enzyme assay.BioF activity was assayed by UV detection of CoA at 260 nm. The reaction mixtures contained 50 mM K2HPO4 (pH 7.1) buffer, 10 mM MgCl2, 10 mM l-alanine, 200 μM PLP, and 200 μM pimeloyl-CoA and the reactions were initiated by adding enzyme to 5 μM. After incubation at 37°C for 1 h, the reaction was quenched by addition of trichloroacetic acid to 2% to precipitate the protein, which was removed by centrifugation. The clear supernatant was loaded onto a μBondapak C18 cartridge (Waters) previously equilibrated with 50 mM ammonium acetate (pH 5). Solvent A was 50 mM ammonium acetate (pH 5) and solvent B was acetonitrile. The sample was eluted with 0% B for 10 min, and then 30 min of a gradient from 0% B to 50% B followed by a 5 min gradient from 50% B to 80% B, which was maintained for 10 min. The retention times of the peaks were confirmed by analysis of CoA and pimeloyl-CoA standards.
Bioassay of AON synthesis.E. coli ΔbioF strain STL115, which requires 8-amino-7-oxononanoate (AON) (or a later compound in the biotin synthetic pathway) for growth, was used as the indicator strain. The bioassay minimal medium was as previously described (18). AON synthesis was analyzed in a 50 μl reaction mixture that contained 50 mM K2HPO4 (pH 7.1) buffer, 10 mM MgCl2, 10 mM l-alanine, 200 μM PLP, 20 μM pimeloyl-CoA (or 20 μM pimeloyl-ACP), and 5 μM BioF (from E. coli, B. sphaericus, or B. subtilis). After incubation at 37°C for 90 min, the samples were heated at 99°C for 10 min to precipitate the proteins. The clear supernatant obtained after centrifugation was concentrated under nitrogen and suspended in 10 μl of sterile water. The samples were spotted onto sterile 6-mm disks (BD BBL blank test discs) placed on top of the solidified agar and the plates were incubated at 30°C overnight.
Gas chromatography-mass spectral detection of AON.Assay mixtures were dried and the AON ketone group was derivatized with 50 μl methoxyamine hydrochloride (40 mg/ml in pyridine; Sigma-Aldrich) for 90 min at 50°C, then with 50 μl of N-(tert-butyldimethylsilyl)-N-methyltrifluoroacetamide containing 1% tert-butyldimethylsilyl chloride (Thermo Scientific) at 50°C for another 120 min to derivatize the AON amino and carboxyl groups. Chromatograms were acquired using a gas chromatography-mass spectrometer system (Agilent Inc.) consisting of an Agilent 7890 gas chromatograph, an Agilent 5975 mass spectrum detector, and a HP 7683B autosampler. Gas chromatography was performed on a ZB-5MS (60 m × 0.32 mm inner diameter and 0.25 μm film thickness) capillary column (Phenomenex). The inlet and Mass Selective Detector (MSD) interface temperatures were 250°C, and the ion source temperature was adjusted to 230°C. An aliquot of 5 μl was injected with the split ratio of 7:1. The helium carrier gas was kept at a constant flow rate of 2.4 ml · min−1. The temperature program was 5 min isothermal heating at 70°C, followed by an oven temperature increase of 5°C min−1 to 270°C. The mass spectrometer was operated in positive electron impact mode at 69.9 eV ionization energy in the m/z 50 to 800 scan range. Mass spectra were recorded in the combined scan/selected-ion monitoring (SIM) mode. In SIM mode a fragment of m/z 387 was tracked (loss of a tert-butyldimethylsilyl group from the fully derivatized AON of m/z 444). The spectra and retention time of target peaks were evaluated using MSD Chemstation E.02.01.177 (Agilent) programs and compared with mass spectra of an authentic AON standard. This standard was used to obtain a calibration curve of 100, 50, 25, 12.5, 6.25, 3.125, 1.563, 0.781 μM. The reactions and analyses were performed by the University of Illinois at Urbana-Champaign, Roy J. Carver Biotechnology Center (Metabolomics).
ACKNOWLEDGMENTS
This work was supported by National Institutes of Health grant AI15650 from the National Institute of Allergy and Infectious Diseases.
We thank A. Ulanov of the University of Illinois at Urbana-Champaign, Roy J. Carver Biotechnology Center (Metabolomics) for the mass spectral analyses.
FOOTNOTES
- Received 21 September 2017.
- Accepted 15 October 2017.
- Accepted manuscript posted online 20 October 2017.
- Copyright © 2017 American Society for Microbiology.
