Characterization of a Novel cis-3-Hydroxy-l-Proline Dehydratase and a trans-3-Hydroxy-l-Proline Dehydratase from Bacteria

ABSTRACT

Hydroxyprolines, such as trans-4-hydroxy-l-proline (T4LHyp), trans-3-hydroxy-l-proline (T3LHyp), and cis-3-hydroxy-l-proline (C3LHyp), are present in some proteins including collagen, plant cell wall, and several peptide antibiotics. In bacteria, genes involved in the degradation of hydroxyproline are often clustered on the genome (l-Hyp gene cluster). We recently reported that an aconitase X (AcnX)-like hypI gene from an l-Hyp gene cluster functions as a monomeric C3LHyp dehydratase (AcnX_{Type I}). However, the physiological role of C3LHyp dehydratase remained unclear. We here demonstrate that Azospirillum brasilense NBRC 102289, an aerobic nitrogen-fixing bacterium, robustly grows using not only T4LHyp and T3LHyp but also C3LHyp as the sole carbon source. The small and large subunits of the hypI gene (hypI_S and hypI_L, respectively) from A. brasilense NBRC 102289 are located separately from the l-Hyp gene cluster and encode a C3LHyp dehydratase with a novel heterodimeric structure (AcnX_{Type IIa}). A strain disrupted in the hypI_S gene did not grow on C3LHyp, suggesting its involvement in C3LHyp metabolism. Furthermore, C3LHyp induced transcription of not only the hypI genes but also the hypK gene encoding Δ¹-pyrroline-2-carboxylate reductase, which is involved in T3LHyp, d-proline, and d-lysine metabolism. On the other hand, the l-Hyp gene cluster of some other bacteria contained not only the AcnX_{Type IIa} gene but also two putative proline racemase-like genes (hypA1 and hypA2). Despite having the same active sites (a pair of Cys/Cys) as hydroxyproline 2-epimerase, which is involved in the metabolism of T4LHyp, the dominant reaction by HypA2 was clearly the dehydration of T3LHyp, a novel type of T3LHyp dehydratase that differed from the known enzyme (Cys/Thr).

IMPORTANCE More than 50 years after the discovery of trans-4-hydroxy-l-proline (generally called l-hydroxyproline) degradation in aerobic bacteria, its genetic and molecular information has only recently been elucidated. l-Hydroxyproline metabolic genes are often clustered on bacterial genomes. These loci frequently contain a hypothetical gene(s), whose novel enzyme functions are related to the metabolism of trans-3-hydroxyl-proline and/or cis-3-hydroxyl-proline, a relatively rare l-hydroxyproline in nature. Several l-hydroxyproline metabolic enzymes show no sequential similarities, suggesting their emergence by convergent evolution. Furthermore, transcriptional regulation by trans-4-hydroxy-l-proline, trans-3-hydroxy-l-proline, and/or cis-3-hydroxy-l-proline significantly differs between bacteria. The results of the present study show that several l-hydroxyprolines are available for bacteria as carbon and energy sources and may contribute to the discovery of potential metabolic pathways of another hydroxyproline(s).

INTRODUCTION

Hydroxyproline, one of the major proline-analogous compounds, is a nonstandard amino acid and has two chiral carbon atoms, each with four isomeric forms. Among these eight isomers, several l-hydroxyproline (l-Hyp) compounds have been detected in certain proteins, including collagen, the cell wall of plants, and some peptide antibiotics. In the first two of these proteins, the l-proline residue is posttranslationally hydroxylated to trans-4-hydroxy-l-proline [(2S,4R)-4-hydroxyproline; T4LHyp] and/or trans-3-hydroxy-l-proline [(2S,3S)-3-hydroxyproline; T3LHyp]. These are synthesized by α-ketoglutarate and Fe(II)-dependent prolyl 4-hydroxylase and propyl 3-hydroxylase, respectively (Fig. 1) (1). In the case of the so-called hydroxyproline-rich glycoprotein (HRGP) of plant cell walls, l-Hyp (T4LHyp) residues are subsequently glycosylated by hydroxyproline O-arabinosyltransferase (2). Major sources of l-Hyp in nature are the degradation of collagen and HRGP by several proteases, including collagenase, prolidase, and/or peptidase produced by mammals and microorganisms. On the other hand, direct biosynthesis from free l-proline to T4LHyp (3), T3LHyp (4), cis-4-hydroxy-l-proline [(2S,4S)-4-hydroxyproline; C4LHyp] (5), and cis-3-hydroxy-l-proline [(2S,3R)-3-hydroxyproline; C3LHyp] (6) has been identified in a few bacteria and fungi (Fig. 1). One of the physiological roles of these hydroxylases may be the biosynthesis of peptide antibiotics containing l-Hyp such as telomycin (C3LHyp and T3LHyp) (7), microcolin (C4LHyp) (8), pneumocandins (T4LHyp and T3LHyp) (9), etamycin (T4LHyp) (10), and empedopeptin (T3LHyp) (11). Therefore, free l-Hyp may be alternatively produced by the degradation of these antibiotics through protease (peptidase), particularly in soil and water environments. The catabolism of l-Hyp, including T4LHyp, T3LHyp, and C3LHyp, by (micro)organisms has recently been investigated at the molecular level (Fig. 2A to C and Table 1; see also Table S1 in the supplemental material) (12–21).

FIG 1.

A hypothetical schematic of the relationship between l-Hyp and microorganisms in nature. Dashed arrows indicate (potential) degradation of a microorganism by proteases.

FIG 2.

l-Hyp metabolism in bacteria. T4LHyp (A), T3LHyp (B), and C3LHyp (C) pathways are shown. Characterizations of HypA2 (this study) and HypJ (17) proteins indicate the hypothetical pathways of C3DHyp and T3DHyp [(2S,3S)-3-hydroxyproline], respectively (dashed arrow). (D) Schematic gene clusters related to the metabolism of l-Hyp. Enzyme names and Enzyme Commission (EC) numbers are listed in Table 1. Homologous genes are indicated in the same color and correspond to those shown in panels A to C. Gray and TR genes are putative l-Hyp transporters and transcriptional regulators, respectively. Putative genes in the boxes were purified and characterized in this study. (E) Growth curves of A. brasilense NBRC 102289 on l-proline, T4LHyp, C4DHyp, T3LHyp, and C3LHyp as the sole carbon source (30 mM) (left panel) and concentrations of T4LHyp, T3LHyp, and C3LHyp in medium estimated using an amino acid analyzer (right panel). Similar results were obtained in two independent experiments. (F) Enzyme activities of cell extracts prepared from A. brasilense NBRC 102289 cells grown on several carbon sources. Values are the averages ± standard deviations (n = 3). Black and gray bars in Pyr2C reductase indicate NADH- and NADPH-dependent activities, respectively.

TABLE 1.

List of enzymes involved in the metabolism of l-Hyp by organism group

Pathway	Presence of the pathway in:a			Enzyme	EC no.	Gene(s)e
	Bacteria	Archaea	Mammals
T4LHyp	+ (15, 17)	Δ (12)b	×c	Hydroxyproline 2-epimerase	5.1.1.8	hypA
				C4DHyp dehydrogenase	1.5.99.−	hypB, hypC, hypD
				HPC deaminase	3.5.4.22	hypE
				KGSA dehydrogenase	1.2.1.26	hypF
T3LHyp	+ (20)	Δ (21)d	+ (19)	T3LHyp dehydratase	4.2.1.77	hypG
				Pyr2C reductase	1.5.1.1	hypH, hypK
C3LHyp	+ (24, 28)	×	×	C3LHyp dehydratase	4.2.1.−	hypI, hypIS, hypIL, hypJ
				Pyr2C reductase	1.5.1.1	hypH, hypK

^aThe presence (+) or absence (×) of the pathway is indicated. The triangle indicates uncertainty about whether the pathway is operative. Reference numbers are given in parentheses.

^bThe OCC_00372 protein from Thermococcus litoralis is a bifunctional proline racemase/hydroxyproline 2-epimerase, whereas no growth on T3LHyp is found.

^cAnother pathway, in which T4LHyp is finally converted to pyruvate and glyoxylate, is present (1).

^dT. litoralis possesses two T3LHyp metabolic enzymes, whereas no growth on T3LHyp is found.

^ehypA, hypG, and hypJ are homologous, and the hypI, hypI_S, and hypI_L genes are homologous (Fig. 2A to C).

In contrast to mammals (1), bacteria metabolize free T4LHyp to α-ketoglutarate via three intermediates in four enzymatic steps (Fig. 2A). Hydroxyproline 2-epimerase, encoded by the hypA gene, initially catalyzes the isomerization of T4LHyp to cis-4-hydroxy-d-proline [(2R,4S)-4-hydroxyproline; C4DHyp] (13–15); therefore, bacteria using this pathway have the ability to grow with not only T4LHyp but also C4DHyp as the sole carbon source (Fig. 2E; see also Fig. 5A), and C4DHyp contained in peptide antibiotics, such as etamycin (10), may be provided by hydroxyproline 2-epimerase. C4DHyp is then oxidized to Δ¹-pyrroline-4R-hydroxy-2-carboxylate (HPC) by a flavin-containing C4DHyp dehydrogenase. Interestingly, there are two types of C4DHyp dehydrogenases: heteromeric-type enzymes encoded by the hypB, hypC, and hypD genes and homomeric-type enzymes encoded by the hypB gene (15, 16). Both of the hypB genes are sequentially homologous with each other, and the HypB protein encoding the heteromeric-type enzyme exhibits full dehydrogenase activity by itself (the so-called catalytic subunit). This finding strongly suggests that the T4LHyp pathway evolved convergently in bacteria. Finally, HPC is converted to α-ketoglutarate via α-ketoglutaric semialdehyde (KGSA) by HPC deaminase and KGSA dehydrogenase, encoded by the hypE and hypF genes, respectively (15, 17, 18). In the metabolism of T3LHyp (Fig. 2B and Tables 1 and S1), T3LHyp dehydratase, encoded by the hypG gene, initially catalyzes the dehydration of T3LHyp to Δ¹-pyrroline-2-carboxylate (Pyr2C) via a putative Δ²-pyrroline-2-carboxylate intermediate (19–21). Pyr2C is then converted by NAD(P)H-dependent Pyr2C reductase, encoded by the hypH or hypK gene, to yield l-proline, which is metabolized by the general degradation of l-proline. T3LHyp dehydratase belongs to the same proline racemase superfamily as the archetype proline racemase and hydroxyproline 2-epimerase, in spite of the different enzyme functions (19). There is no sequential similarity between the HypH and HypK proteins, suggesting that the T3LHyp pathway also evolved convergently in bacteria (20).

FIG 5.

Physiological role of AbAcnX. (A) Growth of the wild type and the mutants of A. brasilense NBRC 102289 on minimal medium agar plates containing the indicated carbon source (30 mM). (B) Transcriptional analysis. Numbers are threshold cycle values that were measured by qRT-PCR, and red boxes indicate that each of the indicated genes was more strongly induced than with l-proline. Results were taken from three independent experiments. Standard deviations are shown in parentheses.

The genes hypA to hypK often cluster together with the putative l-Hyp (ABC-type) transporter genes (22), the transcriptional regulator gene (13, 23), the enolase-like gene (hypJ) (24), the aconitase X-like gene (hypI), and the proline racemase-like gene (hypL) on bacterial genomes (Fig. 2D). Furthermore, l-Hyp gene clusters of several bacteria contain a putative prolidase, collagenase, and/or (metallo)peptidase gene(s), supporting the hypothesis that l-Hyp is derived from collagen, HRGP, and/or peptide antibiotics, as described above. Among these hypothetical genes, the HypJ protein was recently shown to catalyze the dehydration of C3LHyp to Pyr2C, which is subsequently converted into l-proline by Pyr2C reductase (Fig. 2C and Tables 1 and S1), and bacteria that possess this protein may metabolize C3LHyp as the sole carbon source; however, the growth rate on C3LHyp (and T3LHyp) is significantly lower than that on l-proline (14, 24).

Aconitase X (AcnX) was initially discovered in a comparative analysis of archaeal genomes in 2003 (25) and was registered in the aconitase superfamily consisting of aconitase, 2-methylcitrate dehydratase, homoaconitase, and isopropylmalate isomerase (Fig. S1) (25–28). These four enzymes catalyze the homologous stereospecific isomerization of α- to β-hydroxyl acids by sequential dehydration and hydration (anti-elimination/addition). On the other hand, we recently revealed that AcnX proteins, encoded by the hypI gene from bacteria (Pseudomonas aeruginosa PAO1 and Agrobacterium tumefaciens C58) and fungi (Trichoderma reesei QM6a), catalyze the same dehydration of C3LHyp to Pyr2C as the HypJ protein by a reaction that is not homologous to reactions of other aconitase enzymes; this is the first functional annotation of AcnX (Fig. 2C and D and Table 1) (28). AcnX (subfamily) has been further classified into AcnX_{Type I}, consisting of a single polypeptide from bacteria and fungi, and AcnX_{Type II} from bacteria (AcnX_{Type IIa}) and archaea (AcnX_{Type IIb}), which probably consists of (fragmented) small and large polypeptide chains: the former corresponds to the first ∼120 amino acid residues of AcnX_{Type I} (Fig. S1). Among the three AcnX groups, the HypI protein corresponds to AcnX_{Type I}, whereas the function of AcnX_{Type II} currently remains unclear due to its absence within the l-Hyp gene cluster.

Azospirillum brasilense is one of the most well-studied aerobic nitrogen-fixing bacteria and is found in rhizospheres of several grasses. We previously revealed that its type strain NBRC 102289 (ATCC 29145) possesses alternative pathways of l-arabinose (29–31) and acid sugars (d-galactarate and d-glucarate) (17) which differed from those in other bacteria, in which KGSA is provided by the enzymatic dehydration of l-2-keto-3-deoxyarabinonate and decarboxylating dehydration of d-5-keto-4-deoxyglucarate, respectively (Fig. S2). Furthermore, this bacterium was the first (micro)organism found to metabolize not only T4LHyp but also T3LHyp significantly (20), indicating the possibility of another interesting pathway(s).

The gene context in the bacterial genome was of considerable help in estimating the potential substrate of AcnX_{Type I}, whereas the physiological role is currently unclear due to a lack of bacterial growth on C3LHyp (28). In this study, we first found that C3LHyp is rapidly degraded by A. brasilense NBRC 102289 as well as T4LHyp and T3LHyp and that C3LHyp dehydratase activity induced by C3LHyp is derived from heterodimeric AcnX_{Type IIa} protein. On the other hand, l-Hyp gene clusters of a few bacteria including Micromonospora viridifaciens contained not only the (AcnX_{Type IIa}) C3LHyp dehydratase gene but also two putative proline racemase-like genes (hypA1 and hypA2). Among them, the HypA2 protein was identified as a bifunctional T3LHyp dehydratase/2-epimerase, in which the dehydratase activity was significantly dominant. The physiological roles of l-Hyp in bacteria and convergent evolution of the metabolic pathways between bacteria are also discussed.

RESULTS

Discovery of C3LHyp-metabolizing bacteria.

Although the potential metabolism of C3LHyp and T3LHyp by several bacteria, including Labrenzia aggregata IAM 12614, Paracoccus denitrificans PD1222, and Bacillus cereus ATCC 14579, has already been reported (14, 24, 28), their growth rates were significantly lower than those with T4LHyp and l-proline (14). We here demonstrated that A. brasilense NBRC 102289 robustly grew on not only T4LHyp and T3LHyp (20) but also C3LHyp as the sole carbon source; to the best of our knowledge, this is the first clear evidence of the metabolism of C3LHyp by a (micro)organism (Fig. 2E). We then investigated whether the C3LHyp pathway plays a role in the metabolism of T4LHyp and/or T3LHyp because of structural similarities (Fig. 2F). Hydroxyproline 2-epimerase, which is involved in the metabolism of T4LHyp (Fig. 2A and Table 1), was induced by T4LHyp (and C4DHyp) but not by C3LHyp (and T3LHyp). On the other hand, both C3LHyp and T3LHyp induced activities of C3LHyp dehydratase (HypI and/or HypJ protein), T3LHyp dehydratase (HypG protein), and Pyr2C reductase (HypH and/or HypK protein). Among these, the C3LHyp dehydratase activity acquired no cofactors, similar to the HypI protein but unlike the HypJ protein (absolutely dependent on Mg²⁺) (24). On the other hand, the Pyr2C reductase activity was dually specific between NADPH and NADH (specific for the HypK protein); the NADPH-dependent activity induced by T4LHyp (and C4DHyp) is derived from the HypH protein (20). These results strongly suggest that C3LHyp dehydratase (HypI-type protein) and Pyr2C reductase (HypK protein) are physiologically involved in the C3LHyp pathway.

Candidates for the C3LHyp dehydratase gene.

Although the genome sequence of A. brasilense NBRC 102289 is currently unavailable, the nucleotide sequences of several genes from this bacterium show very high similarities (∼98%) to those of Burkholderia lata (17, 20, 29–31). Therefore, a homology search using the protein BLAST program was performed against the genome sequence of B. lata using the hypI and hypJ genes as the probe protein sequences. In the case of the latter, ∼20 homologous genes were found as putative dehydratases for several acid sugars and dipeptide epimerase in the enolase superfamily, and their sequence identities were only 20 to 30%. On the other hand, Bcep18194_B2340 and Bcep18194_B2339 were associated with aconitase subunits 1 and 2, and each showed ∼40% sequence identity with the last ∼400 and first ∼120 amino acid residues of HypI, respectively (referred to as A. brasilense hypI_L [AbhypI_L] and AbhypI_S genes). Additionally, the 3′ part of AbhypI_L was slightly overlapped by the 5′ part of AbhypI_S. Therefore, we considered the AbhypI_L-AbhypI_S genes to be the first candidates for a C3LHyp dehydratase, and included each subunit in the heteromeric structure of AcnX_{Type IIa}.

Expression of the recombinant protein.

The AbhypI_S and AbhypI_L genes, in which a His₆ tag sequence was attached at the N terminus of the AbhypI_S gene, were coexpressed in Escherichia coli and purified with an Ni²⁺-chelating affinity column. Two major distinct bands with molecular masses at 24 (17,034.98) and 42 kDa (44,416.16), respectively, were observed in the SDS-PAGE analysis (values in parentheses indicate the calculated molecular mass of the enzyme) (see Fig. S3A in the supplemental material), and Western blotting using the anti-His₆ tag antibody revealed that the latter corresponded to the HypI_S protein (Fig. S3B). Since the molar ratio of HypI_S to HypI_L in the SDS-PAGE gel was ∼1:1 and since the native molecular mass, estimated by gel filtration, was approximately 80 kDa (Fig. S3C), this protein molecule may be composed of a heterodimeric structure of subunits, confirming that AcnX_{Type I} fused with both subunits is a monomer (referred to as AbAcnX).

AcnX_{Type IIa} functions as a C3LHyp dehydratase.

In order to estimate dehydration activity with C3LHyp, we previously developed a conventional spectrophotometric assay method using NADPH-dependent Pyr2C reductase from P. aeruginosa PAO1 (P. aeruginosa HypH [PaHypH]) as a coupling enzyme, as described in Materials and Methods (28). When the eluent of the recombinant AbAcnX protein from the Ni²⁺-chelating affinity (spine) column was directly used in the assay, a significant decrease in absorbance at 340 nm was observed, with specific activity at 2.29 μmol · min⁻¹ · mg⁻¹. On the other hand, this activity was gradually lost on ice, and only 5.5% of activity remained after 8 h, which differed from the activity of the AcnX_{Type I} enzyme (Fig. 3H); difficulties were associated with directly identifying the reaction product by ¹H nuclear magnetic resonance (NMR). In addition, concentration measurements by an amino acid analyzer revealed that C3LHyp was consumed in a time-dependent manner (Fig. 3A), and when the reaction was performed in the copresence of PaHypH, l-proline was produced in a time-dependent manner (Fig. 3B); specific activity values were similar to those obtained in the spectrophotometric assay. Kinetic parameters with C3LHyp are shown in Fig. 3G. The k_cat/K_m values of AbAcnX were 27- and 7.0-fold lower than those of PaAcnX and A. tumefaciens AcnX (AtAcnX), and the k_cat values were 20- and 8.3-fold lower, respectively; the enzymes may have been gradually inactivated during the purification steps. Although 2,3-cis-3,4-cis-3,4-dihydroxy-l-proline was an additional substrate for AbAcnX, its specific activity was more than 30-fold lower than that of C3LHyp (Fig. 3C). Since C4LHyp was produced by coupling with PaHypH (Fig. 3D), the reaction product must be a Δ¹-pyrroline-4S-hydroxy-2-carboxylate. Pyrrole-2-carboxylate, a structural analog of the (putative) transition state of C3LHyp (Δ²-pyrroline-2-carboxylate) (Fig. 2C), inhibited activity (50% inhibitory concentration [IC₅₀], 1.91 mM) (Fig. 3I). These properties were similar to those of bacterial AcnX_{Type I} (28). Collectively, these results suggest that the AbAcnX protein, belonging to heteromeric AcnX_{Type IIa} (subfamily), functions as a C3LHyp dehydratase.

FIG 3.

Characterization of the AbAcnX protein as a C3LHyp dehydratase. The reaction products of C3LHyp (A, B, E, and F) and 2,3-cis-3,4-cis-dihydroxy-l-proline (C and D) with AbAcnX (A and C), AbAcnX plus PaHypH and NADPH (B and D), MvAcnX (E), or MvAcnX plus PaHypH and NADPH (F) were determined using an amino acid analyzer. PaHypH is an NADPH-dependent Pyr2C reductase (20). Concentrations were expressed as the percentage of the amount initially present. Similar results were obtained in two independent experiments. (G) Lineweaver-Burk plots and kinetic parameters of AbAcnX (this study), MvAcnX (this study), PaAcnX (28), and AtAcnX (28) toward C3LHyp. Results were taken from three independent experiments. Standard deviations are shown in parentheses. (H) Comparison of stability for AcnX_{Type IIa} (AbAcnX and MvAcnX) and AcnX_{Type I} (PaAcnX and AtAcnX). His₆-tagged PaAcnX and AtAcnX proteins were prepared using previously described methods (Fig. S3A and B) (28). Enzyme activities were measured after incubation for the indicated time on ice and are shown as relative values expressed as a percentage of the controls without the treatment. Values are the averages ± standard deviations (n = 3). (I) Inhibition study on the C3LHyp dehydratase activities of AbAcnX (this study), MvAcnX (this study), PaAcnX (28), and AtAcnX (28) by pyrrole-2-carboxylate (inset). Relative specific activity values (averages ± standard deviations; n = 3) are expressed as percentages of the values obtained in the absence of an inhibitor. IC₅₀s (in parentheses) were calculated by curve fitting using ImageJ software (http://rsb.info.nih.gov/ij/).

Characterization of other AcnX_{Type IIa} enzymes.

A homology search using the protein BLAST program revealed that a large number of bacteria possess the AcnX_{Type IIa} homologous protein. The AcnX_{Type IIa} genes from a few bacteria such as Intrasporangium calvum, Micromonospora viridifaciens, and Knoellia sp. strain Soil729 are located within the l-Hyp gene cluster (Fig. 2D). Therefore, in order to elucidate the catalytic functions of other AcnX_{Type IIa} proteins, we enzymatically investigated INTCA_RS16085–RS16090 (hypI_L-hypI_S) from I. calvum (IcAcnX) and GA0074695_2670–2671 (hypI_S-hypI_L) from M. viridifaciens (MvAcnX); IcHypI_L and MvHypI_L showed 52.7% and 51.5% sequence identities with AbHypI_L, respectively. Only the His₆-tagged MvAcnX protein was successfully expressed in E. coli, purified, and characterized (Fig. S3A and B). Regarding the results obtained, C3LHyp was also an active substrate for this protein (Fig. 3E and F), and kinetic parameters (Fig. 3G) and inhibition by pyrrole 2-carboxylate (Fig. 3H) were similar to those of AbAcnX. Since rapid inactivation of the purified enzyme was also observed (Fig. 3I), it is likely that this lability is specific for AcnX_{Type IIa}. These results suggest that the universal function of AcnX_{Type IIa} is as a C3LHyp dehydratase, the same as for AcnX_{Type I}, confirming that most of the amino acid residues of AcnX_{Type I} possibly related to Fe(III) binding, substrate recognition, and/or structural folding are conserved in AcnX_{Type IIa} (Fig. 4A) (28) and also that AcnX_{Type IIa} is more closely related to monomeric AcnX_{Type I} than the same heteromeric AcnX_{Type IIb} (and AcnX_{Type IIc}; see below) (Fig. 4B).

FIG 4.

Phylogenetic analysis of the AcnX protein family. (A) Partial multiple-sequence alignment of deduced amino acid sequences. Numbers and colors with the sequences correspond to those in panel B. Three putative ligands for Fe(III) bound in the active center are shown as white letters in black boxes. Gray-shaded letters indicate highly conserved amino acid residues. Substrate binding sites are in bold letters. (B) Phylogenetic trees based on the small (left) and large (right) subunits, respectively. A complete list of organisms included in the analysis and sequence accession numbers are shown in Table S3 in the supplemental material. The number on each branch indicates the bootstrap value. The sources of the enzymes are indicated in accordance with the legend on the figure.

Gene regulation and disruption analysis.

In order to estimate the physiological roles of the AcnX_{Type IIa} protein, we performed gene disruption experiments on A. brasilense NBRC 102289 by introducing a kanamycin resistance (Km^r) cassette into the hypI (hypI_L) gene (Fig. S4). The hypI mutant strain obtained was distinct from the wild-type strain in that C3LHyp did not support growth as a sole carbon source, whereas no significant difference was observed in growth on T3LHyp (and T4LHyp and C4DHyp) between the two strains (Fig. 5A). On the other hand, in comparisons with l-proline in a quantitative real-time PCR (qRT-PCR) analysis, the hypI and hypG genes were induced by C3LHyp and T3LHyp, whereas the hypA gene was induced by T4LHyp (and C4DHyp) only (Fig. 5B). A. brasilense NBRC 102289 possesses two different Pyr2C reductase enzymes encoded by the hypK and hypH genes (Fig. 2D), and of these the NAD(P)H-dependent HypK protein is involved in the metabolism of T3LHyp, d-proline, and d-lysine (Fig. S5) (20). We herein demonstrated that the hypK mutant strain, which was constructed previously, did not grow on d-proline, d-lysine, or on C3LHyp (Fig. 5A) and that C3LHyp induced the expression of the hypK gene (Fig. 5B). The phenotypes of mutants and the results of the qRT-PCR analysis were consistent with changes in enzyme activities in the cell extract in the experiment shown in Fig. 2F. Therefore, we concluded that the hypI and hypK genes function as C3LHyp dehydratase and Pyr2C reductase genes, respectively, involved in C3LHyp metabolism. Since T3LHyp, d-proline, and d-lysine are not substrates for AbAcnX and since the hypI mutant strain grew on these amino acids, the transcription of hypI, hypK, and hypG genes may be commonly upregulated by a ketamine compound such as Pyr2C and Δ¹-piperidine-2-carboxylate (Fig. S5).

Discovery of a unique T3LHyp dehydratase.

As described above, in the case of M. viridifaciens (and I. calvum and Knoellia sp. Soil729), MvhypI_s-MvhypI_L genes and MvhypK genes clustered together with the MvhypB-F genes (Fig. 2D), indicating that this gene cluster is related to the metabolism of not only T4LHyp but also C3LHyp and/or T3LHyp. Hydroxyproline 2-epimerase (Cys/Cys) and T3LHyp dehydratase (Cys/Thr, Ser/Cys, or Ser/Thr) are considered to belong to the same proline racemase superfamily, together with the archetype proline racemase (Cys/Cys), and have been classified into four types based on two specific residues at the active sites (shown in parentheses) (12, 19–21). Both the MvHypA1 (GA0074695_2673) and MvHypA2 (GA0074695_2680) proteins containing the l-Hyp gene cluster possessed two cysteine residues at their active sites (Cys⁹¹/Cys²⁵⁴ and Cys⁹⁰/Cys²⁵⁶, respectively) specific for hydroxyproline 2-epimerase (and proline racemase) but not T3LHyp dehydratase (Fig. 6A). Therefore, MvHypA1 and MvHypA2 were characterized enzymatically using purified (His₆-tagged) recombinant proteins (Fig. S3A and B). A spectrometric assay analysis using a coupling enzyme revealed that MvHypA1 and MvHypA2 proteins exhibited hydroxyproline 2-epimerase and T3LHyp dehydratase activities, respectively; this is the first report of the Cys/Cys type of T3LHyp dehydratase. Kinetic parameters of MvHypA1 and MvHypA2 with T4LHyp and T3LHyp, respectively, are shown in Fig. 7A. The K_m value of MvHypA2 was more than 10-fold lower than values of the HypG proteins (Cys/Thr) (16, 20).

FIG 6.

Phylogenetic analysis of the proline racemase superfamily including MvHypA1 and MvHypA2. (A) Partial multiple-sequence alignment of deduced amino acid sequences. Numbers and colors with the sequences correspond to those in panel B. Catalytic cysteine, threonine, and/or serine residues are shown as white letters in black boxes. Gray-shaded letters indicate highly conserved amino acid residues, with seven substrate binding sites in bold. (B) Phylogenetic tree. A complete list of organisms included in the analysis and sequence accession numbers are shown in Table S4 in the supplemental material. The number on each branch indicates the bootstrap value. The sources of the enzymes are indicated in accordance with the legend on the figure.

FIG 7.

Characterization of MvHypA1 and MvHypA2 proteins. (A) Kinetic parameters of MvHypA1 (this study), MvHypA2 (this study), AbHypG (20), and HypG from Colwellia psychrerythraea 34H (CpHypG) (20) (Fig. 2D) toward T3LHyp or T4LHyp. Results were taken from three independent experiments. Standard deviations are shown in parentheses. AbHypG and CpHypG are the Cys/Thr-type of T3LHyp dehydratase. (B and C) ¹H NMR spectra of the reaction product of T3LHyp with the wild-type (B) or C261T mutant enzyme of MvHypA2 (C) and Pyr2C and C3DHyp are shown. Asterisks are peaks derived from an internal standard. Left panels show the assignments of protons in D₂O. (D to G) Reaction products of T3LHyp (D and E) and C3DHyp (F and G) with MvHypA2 and with MvHypA2 plus PaHypH and NADPH, as indicated, were determined using an amino acid analyzer.

Bifunctional T3LHyp dehydratase/2-epimerase.

In order to examine the activity of MvHypA2 in more detail, the enzyme reaction was analyzed using NMR and an amino acid analyzer. T3LHyp was incubated with MvHypA2 in D₂O for various times. ¹H NMR spectra showed a progressive loss of H₁ peaks and additionally contained resonances associated with only Pyr2C (Fig. 7B). On the other hand, concentration measurements using an amino acid analyzer revealed that the reaction, performed in the copresence of PaHypH, produced not only l-proline but also cis-3-hydroxy–d-proline [(2R,3S)-3-hydroxyproline; C3DHyp], a potential product of 2-epimerization (Fig. 7E). The specific activity values for the disappearance of T3LHyp and production of C3DHyp were estimated to be 23.2 and 2.55 μmol · min⁻¹ · mg⁻¹, respectively; the former needs to be derived from dehydration and 2-epimerization (Fig. 7D).

In order to correctly analyze 2-epimerase activity, we enzymatically synthesized C3DHyp (unavailable commercially) from T3LHyp (12). Specific activity values for the disappearance of C3DHyp (2.08) and production of T3LHyp (2.27) were similar to the value obtained for the production of C3DHyp from T3LHyp (Fig. 7F), indicating that 2-epimerization (but not dehydration) is reversible and that its rate is approximately 9-fold lower than that of dehydration. When the reaction was performed in the copresence of PaHypH, not only T3LHyp but also l-proline appeared; C3DHyp was converted to Pyr2C via T3LHyp (Fig. 7G). These results indicate that MvHypA2 is a novel bifunctional T3LHyp dehydratase/2-epimerase; however, it currently remains unclear whether the latter activity is involved in the metabolism of C3DHyp by bacteria (Fig. 2B). Zhao et al. (14) characterized 51 proline racemase-like proteins from bacteria, among which no enzyme with dual activities was identified.

DISCUSSION

Functional annotation of AcnX_{Type IIa} as a C3LHyp dehydratase.

Since a few AcnX_{Type IIa} genes are located within the l-Hyp gene cluster in the bacterial genome as well as most of the AcnX_{Type I} genes (Fig. 2D), their common function as a C3LHyp dehydratase is reasonable. Furthermore, since AcnX_{Type IIa} shows higher sequential similarity to monomeric AcnX_{Type I} than the same heteromeric AcnX_{Type IIb} (and AcnX_{Type IIc} [see below]) (Fig. 4B), AcnX_{Type I} has recently appeared through the fusion of the hypI_S and hypI_L (hypI_SL) genes encoding AcnX_{Type IIa} after acquirement of the enzyme function in the evolutionary stage. Starkeya novella DSM 506 possesses hypJ and hypI_SL genes. Only the former is located within the l-Hyp gene cluster (Fig. 2D), whereas the latter is sequentially closer to AcnX_{Type IIb} than to AcnX_{Type IIa} (referred to as AcnX_{Type IIc}) (Fig. 4B). In the case of Pandoraea thiooxidans DSM 25325, two hypI_SL genes encoding AcnX_{Type IIa} and AcnX_{Type IIc} were separately located on the genome, and only the gene encoding the former was located within the l-Hyp gene cluster (Fig. 2D). No AcnX_{Type IIc} (or AcnX_{Type IIb}) gene clustered together with a gene(s) related to the l-Hyp and/or proline metabolism, indicating another physiological function(s) of the AcnX protein.

The greatest difference observed in the properties of AcnX_{Type IIa} and AcnX_{Type I} as C3LHyp dehydratases was that the former was significantly labile (Fig. 3H). We hypothesized that the main reason for this is that bound Fe(III) is easily lost from the active center, similar to the [4Fe-4S] cluster in known aconitase enzymes (26, 27, 32). All amino acid residues of AcnX_{Type I} (possibly) related to Fe(III) binding, substrate recognition, and/or structural folding are completely conserved in AcnX_{Type IIa}, except for a serine residue at the position corresponding to Cys68 of PaAcnX (Fig. 4A, pink-shaded letters). This position may be located near to bound Fe(III), and the alanine mutant of PaAcnX (C68A) shows only 0.4% specific activity of that of the wild-type enzyme (28). We attempted to characterize the S69C and S70C mutants of AbAcnX and MvAcnX, respectively. Although we successfully prepared a recombinant protein of the latter, the mutation had no significant effect on stability (Fig. 3H). Therefore, we are now in the process of characterizing the C68S mutant of PaAcnX.

Novel evolutionary insights into the proline racemase superfamily.

The phylogenetic tree of the proline racemase superfamily consists of four subfamilies (proline racemase, hydroxyproline 2-epimerase, and T3LHyp dehydratases 1 and 2), and the proline racemase-like proteins recently identified from archaea are clearly distinct from these subfamilies (Fig. 6B) (12). Among them, T3LHyp dehydratase subfamily 1 contains enzymes from eukaryotes (19), bacteria (20), and archaea (21), and two specific residues commonly found at the active sites are Cys/Thr. Furthermore, MvHypA1 (Cys/Cys) is closely related to T3LHyp dehydratase subfamily 2 (Ser/Thr and Ser/Cys) and is deeply branched and forms the novel hydroxyproline 2-epimerase group with HypA1 proteins from I. calvum and Knoellia sp. Soil729. Therefore, the characterization of MvHypA1 and MvHypA2 in this study may strengthen the hypothesis that T3LHyp dehydratase subfamilies 1 and 2 convergently evolved from an ancestral enzyme of the same Cys/Cys type as proline racemase and hydroxyproline 2-epimerase.

Another question is why dehydration is significantly more dominant in MvHypA2 than 2-epimerization. In order to elucidate the reaction mechanism in more detail, we characterized the C261T mutant of MvHypA2 by NMR and kinetic constant analyses (using large amounts of the protein). Although Pyr2C was also identified as a reaction product of T3LHyp (Fig. 7C), the k_cat/K_m value was 826-fold lower than that of the wild-type enzyme, and this was mainly attributed to a 162-fold increase in K_m (Fig. 7A), strongly indicating the involvement of Cys²⁶¹ in catalysis. This may indicate an evolutionary potential to acquire new enzyme functions without mutations at the active site(s) (24, 33, 34).

Involvement of l-Hyp metabolism.

As described above, A. brasilense NBRC 102289 grew robustly on T4LHyp, T3LHyp, and C3LHyp (Fig. 2E). The transcriptional regulation of genes related to 4-hydroxy-l-proline metabolism and that of genes related to 3-hydroxy-l-proline metabolism are independent of each other, and there are significant metabolic networks between 3-hydroxy-l-proline, d-proline, and d-lysine pathways (Fig. 2F and 5; see also Fig. S5 in the supplemental material). On the other hand, the l-Hyp gene clusters of L. aggregata IAM 12614 (possibly also S. novella DSM 506) and Sinorhizobium meliloti 2011 consist of genes for T4LHyp, T3LHyp, and C3LHyp metabolism (Fig. 2D), and the l-Hyp gene cluster of L. aggregata IAM 12614 is induced by all of these l-Hyp proteins; however, they metabolize 3-hydroxy-l-prolines very poorly (13, 14, 18, 24). Furthermore, the functional characterization of the AcnX, HypA1, and HypA2 proteins from M. viridifaciens indicated that the l-Hyp gene cluster is additionally involved in the metabolism of C3DHyp (Fig. 2B). Therefore, it is possible that each l-Hyp gene cluster is responsible for the metabolism of l-Hyp(s) produced from different natural sources (Fig. 1): collagen (containing only T4LHyp and T3LHyp) (1), HRGP (only T4LHyp) (2), and peptide antibiotics (T4LHyp, C4DHyp, C4LHyp, T3LHyp, and/or C3LHyp) (7–11). In fact, collagenase, peptidase, and/or the peptidase gene are located (closely) within the l-Hyp gene cluster of several bacteria (Fig. 2D).

To our knowledge, C3LHyp is found only in telomycin together with T3LHyp produced by Streptomyces canus ATCC 12646 (7). The telomycin biosynthesis is dependent on the tem1-tem34 gene cluster, in which both l-proline cis-3-hydroxylase (tem32) and l-proline trans-3-hydroxylase genes (tem29) are contained (35) (Fig. 8A). This bacterium also possesses the l-Hyp gene cluster related to T4LHyp metabolism (telomycin contains no T4LHyp) but not the 3-hydroxy-l-proline pathway and l-proline trans-4-hydroxylase gene (36). Since A. brasilense NBRC 102289 may not possess the ability to produce any peptide antibiotic or to hydroxylate free l-proline (no l-proline hydroxylase gene on the genome), T4LHyp, T3LHyp, and C3LHyp transported into cells are used only as energy sources (Fig. 8B). On the other hand, S. meliloti 2011 possesses the l-proline cis-4-hydroxylase gene (p4h; SMc03253) (5) even though it does not use C4LHyp as a carbon source (18); the physiological role of this is currently unclear (Fig. 8C). To our knowledge, there is no report that bacteria utilize l-Hyp(s) that is synthesized from l-proline as an energy source.

FIG 8.

Schematic of l-Hyp utilization from S. canus ATCC 12646 (A), A. brasilense NBRC 102289 (B), and S. meliloti 2011 (C). Dashed arrows indicate (potential) intercellular uptake of l-Hyp(s).

As described in the Introduction, among metabolic enzymes involved in l-Hyp pathways, there are several examples of convergent evolution: for C4DHyp dehydrogenase, HypB and HypBCD (15, 16); for C3LHyp dehydratase, HypI (AcnX_{Type I}) and HypG (24, 28); for Pyr2C reductase, HypK and HypH (21). Furthermore, T3LHyp dehydratase appeared twice in the same proline racemase-superfamily: HypG and HypL (Fig. 6B) (14). In different bacterial species, these enzymes are involved in l-Hyp pathways with several combinations, suggesting that even the same l-Hyp pathway(s) evolved independently, at least partially (Fig. 2D) (15, 28). Therefore, the discovery of the l-Hyp gene cluster containing hypI_SL (AcnX_{Type IIa}) and hypA1 genes, instead of hypI and hypA, as C3LHyp dehydratase and hydroxyproline 2-epimerase, respectively, is consistent with this hypothesis.

MATERIALS AND METHODS

Materials.

C3LHyp and T3LHyp were purchased from Sigma-Aldrich (USA) and Kanto Chemical (Tokyo, Japan), respectively. A. brasilense NBRC 102289, M. viridifaciens NBRC 101887, and I. calvum NBRC 12989 were purchased from the National Institute of Technology and Evaluation (Chiba, Japan).

General procedures.

Basic recombinant DNA techniques were performed as described by Sambrook et al. (37). Bacterial genomic DNA was prepared using a DNeasy tissue kit (Qiagen). PCR was performed using a GeneAmp PCR system 2700 (Applied Biosystems) for 30 cycles in 50 μl of a reaction mixture containing 1 U of KOD FX DNA polymerase (Toyobo), appropriate primers (15 pmol), and template DNA under the following conditions: denaturation at 98°C for 10 s, annealing at 50°C for 30 s, and extension at 68°C for time periods calculated at an extension rate of 1 kbp · min⁻¹. DNA sequencing was conducted using a BigDye cycle sequencing kit, version 3.1 (Applied Biosystems), and appropriate primers with a Genetic Analyzer 3130 (Applied Biosystems). High-pressure liquid chromatography (HPLC) was performed using an Agilent 1120 compact LC system (Tosoh). Protein concentrations were measured by the method of Lowry et al. (38) with bovine serum albumin as the standard. SDS-PAGE was performed as described by Laemmli (39), and protein concentrations in the gel bands were estimated using NIH ImageJ, version 1.45s, software.

Bacterial strain, culture conditions, and preparation of cell extracts.

A. brasilense NBRC 102289 was cultured aerobically with vigorous shaking at 30°C in minimal medium (20, 29) supplemented with a 30 mM concentration of a carbon source. The cells that grew were harvested by centrifugation at 30,000 × g for 20 min, suspended in 50 mM Tris-HCl (pH 8.0), disrupted by sonication for 20 min at appropriate intervals on ice using an Ultra Sonic disruptor UD-211 (Tomy Seiko Co., Ltd., Tokyo, Japan), and then centrifuged at 108,000 × g at 4°C for 20 min to obtain cell extracts.

Plasmid construction for the expression of recombinant proteins.

The primer sequences used in this study are shown in Table S2 in the supplemental material. In this study, the prefixes Ab (A. brasilense NBRC 102289), Mv (M. viridifaciens), Ic (I. calvum), Pa (P. aeruginosa PAO1), and At (A. tumefaciens C58) have been added to gene symbols or protein designations where required for clarity.

MvhypI_S (GA0074695_2670), MvhypI_L (GA0074695_2671), IchypI_S (INTCA_RS16090), IchypI_L (INTCA_RS16085), MvhypA1 (GA0074695_2673), and MvhypA2 genes (GA0074695_2680) were amplified by PCR using primers containing appropriate restriction enzyme sites at the 5′ and 3′ ends and the genome DNA of M. viridifaciens or I. calvum as a template. AbhypI_S and AbhypI_L genes were amplified by PCR using primers designed from putative aconitase subunit 2 (Bcep18194_B2339) and aconitase subunit 1 (Bcep18194_B2340) genes from B. lata and the genome DNA of A. brasilense NBRC 102289 as a template, and amplified products were sequenced.

The pETDuet-1 vector (Novagen), with two multiple cloning sites (MCS), was used for the cloning and expression of the putative heteromeric AcnX (HypI_SL) protein. In this expression vector, the gene inserted in MCS-I yielded a recombinant protein with a His₆ tag at the N terminus. The other recombinant protein corresponding to the gene located in MCS-II did not have a fused affinity tag. The hypI_L gene was introduced into the NdeI-KpnI site in MCS-II, whereas the hypI_S gene was introduced into the BamHI-HindIII (for AbhypI_S and IchypI_S) or EcoRI-HindIII (for MvhypI_S) site in MCS-I in order to obtain pET/AbhypI_SL, pET/IchypI_SL, and pET/MvhypI_SL, respectively (Fig. S4A). The MvhypA1 or MvhypA2 gene was introduced into the BamHI-HindIII site in MCS-I of the pETDuet-1 vector in order to obtain pET/MvhypA1 and pET/MvhypA2, respectively.

A site-directed mutation was introduced into the AbhypI_L, MvhypI_L, or MvhypA2 gene by sequential steps of PCR using sense and antisense primers and pET/AbhypI_SL or pET/MvhypA2 as a template. Mutant proteins were expressed and purified by the same procedures as each wild-type enzyme.

Expression and purification of the recombinant protein.

Escherichia coli strain BL21(DE3) harboring pET/AbhypI_SL, pET/MvhypI_SL, pET/IchypI_SL, pET/MvhypA1, or pET/MvhypA2 was grown at 37°C to a turbidity of 0.6 at 600 nm in super broth medium containing ampicillin (50 mg/liter). After the addition of 1 mM isopropyl-β-d-thiogalactopyranoside (IPTG), cells in the culture were grown at 37°C for a further 6 h (for MvHypI_SL and MvHypA1) or at 18°C for 18 h (for AbHypI_SL, IcHypI_SL, and MvhypA2) in order to induce the expression of the His₆-tagged protein. The cells that grew were harvested, suspended in buffer A (50 mM sodium phosphate buffer [pH 8.0] containing 300 mM NaCl and 10 mM imidazole), disrupted by sonication, and then centrifuged. The supernatant was loaded onto a Ni-nitrilotriacetic acid (NTA) spin column (Qiagen) equilibrated with buffer A. The column was washed three times with buffer B (50 mM sodium phosphate buffer [pH 8.0] containing 300 mM NaCl, 10% [vol/vol] glycerol, and 50 mM imidazole). The enzymes were then eluted with buffer C (pH 8.0, buffer B containing 250 mM imidazole instead of 50 mM imidazole). For AbHypI_SL and MvIcHypI_SL, the eluent was directly used in further experiments, whereas for MvHypA1 and MvHypA2, enzymes were dialyzed against 50 mM Tris-HCl buffer (pH 8.0) containing 50% (vol/vol) glycerol and stored at −35°C until used. The native molecular masses of recombinant proteins were estimated by gel filtration using a HiLoad 16/60 Superdex 200 prep grade column (GE Healthcare) equilibrated with 50 mM Tris-HCl buffer (pH 8.0). A Gel Filtration Markers kit for protein molecular masses of 29,000 to 700,000 Da (Sigma) was used for molecular markers.

Western blot analysis.

The purified recombinant protein was separated by SDS-PAGE, and proteins on the gels were transferred onto a polyvinylidene difluoride (PVDF) membrane (Hybond-P; GE Healthcare). Western blot analysis was performed using an enhanced chemiluminescence (ECL) Western blotting system (GE Healthcare), a penta-His antibody (Qiagen), and goat anti-mouse IgG(H+L) secondary antibody conjugated with horseradish peroxidase (HRP) (Invitrogen).

Spectrophotometric enzyme assay.

C3LHyp dehydratase, T3LHyp dehydratase, hydroxyproline 2-epimerase, and Pyr2C reductase were spectrophotometrically assayed using a previously described method (16, 20, 28). Briefly, the first and second enzymes were assayed by coupling their reactions to Pyr2C reductase from P. aeruginosa PAO1 (PaHypH) (20) and monitoring the oxidization rate of NADPH at 340 nm. On the other hand, the third enzyme was assayed by coupling its reaction to flavin-containing C4DHyp dehydrogenase (AbHypBCD) (16) and monitoring the reduction rate of p-iodonitrotetrazolium violet (artificial electron acceptor) in the copresence of phenazine methosulfate (electron transfer intermediate) at 490 nm.

Enzyme assay by HPLC.

The purified protein was added to 50 mM Tris-HCl buffer (pH 8.0) (1 ml) containing 10 mM substrate and incubated at 30°C. After various incubation times, the enzyme reaction was stopped by brief incubation at −80°C. Substrate consumption was estimated by HPLC using a Hitachi L-8900 PH amino acid analyzer (Hitachi, Tokyo, Japan). C3DHyp (unavailable commercially) was enzymatically synthesized from T3LHyp by proline racemase from Thermococcus litoralis (TlProR) (Fig. 7); this enzyme utilizes not only l-proline but also T4LHyp, T3LHyp, and C3LHyp as substrates (12). The reaction mixture (50 ml) consisted of 50 mM Tris-HCl buffer (pH 8.0) containing 30 mM T3LHyp. After the addition of TlProR (50 mg), the mixture was left at 50°C overnight and treated by ultrafiltration with an Amicon Ultra-15 filter (Millipore) to remove TlProR.

Identification of reaction products.

In order to remove glycerol from the storage buffer, purified MvHypA2 (wild-type, ∼0.6 mg; C261T mutant, ∼6.8 mg) was dialyzed at 4°C overnight in 50 mM K₂HPO₄/KH₂PO₄ buffer (pH 7.0), lyophilized, and dissolved in D₂O (1.2 ml) containing 20 mM T3LHyp. NMR spectra were recorded at 25°C on a JEOL JNM-EC400 NMR spectrometer (JEOL, Ltd., Tokyo, Japan) operating at 400 MHz. 2,2-Dimethyl-2-silapentane-5-sulfonate was used as an internal standard.

Mutant construction.

A schematic diagram of the plasmid construction for the disruption of the AbAcnX gene is shown in Fig. S4A in the supplemental material. Briefly, the Tn5-derived EcoRI 1.3-kbp kanamycin resistance (Km^r) cassette was inserted into the two EcoRI sites in the coding sequence of the AbhypI_L gene of pET/AbhypI_SL to yield pET/AbhypI_SL::Km^r. In order to introduce the restriction site for SphI at the 5′ end of the DNA fragment containing the Km^r gene in the AbhypI_L gene, PCR was performed using pET/AbhypI_SL::Km^r as a template and the two primers P2 and P17 (Table S2). The ∼2-kbp DNA fragment was digested by SphI and subcloned into the SphI site in a tetracycline resistance (Tc^r) cassette of the suicide vector pSUP202 (40) in order to yield pSUP/AbhypI_L::Km^r. E. coli S17-1 (40) was transformed with pSUP/AbhypI_L::Km^r, and the transformant was then mobilized to A. brasilense NBRC 102289 by biparental mating. Transconjugants were selected on a minimal medium agar plate supplemented with 30 mM T4LHyp and 50 μg of kanamycin per liter using the Km^r (presence of the Km^r cassette) and Tc^s (loss of pSUP202) phenotypes. The construction was confirmed by genomic PCR using the P3 and P4 primers outside the target gene (Fig. S4B and Table S2).

Gene expression analysis.

The preparation of RNA samples from A. brasilense NBRC 102289 cells and one-step RT-PCR were performed as previously described (41). The primers used are listed in Table S2 in the supplemental material.

Phylogenetic analysis.

Protein sequences were analyzed using the protein BLAST and Clustal W programs distributed by the DNA Data Bank of Japan (DDBJ).

Accession number(s).

The nucleotide sequences of the AbhypI_S and AbhypI_L genes were submitted to GenBank under accession number LC221834.