Identification and Characterization of a Novel Class of Extracellular Poly(3-Hydroxybutyrate) Depolymerase from Bacillus sp. Strain NRRL B-14911

Abstract

The catalytic, linker, and denatured poly(3-hydroxybutyrate) (dPHB)-binding domains of bacterial extracellular PHB depolymerases (PhaZs) are classified into several different types. We now report a novel class of extracellular PHB depolymerase from Bacillus sp. strain NRRL B-14911. Its catalytic domain belongs to type 1, whereas its putative linker region neither possesses the sequence features of the three known types of linker domains nor exhibits significant amino acid sequence similarity to them. Instead, this putative linker region can be divided into two distinct linker domains of novel types: LD1 and LD2. LD1 shows significant amino acid sequence similarity to certain regions of a large group of PHB depolymerase-unrelated proteins. LD2 and its homologs are present in a small group of PhaZs. The remaining C-terminal portion of this PhaZ can be further divided into two distinct domains: SBD1 and SBD2. Each domain showed strong binding to dPHB, and there is no significant sequence similarity between them. Each domain neither possesses the sequence features of the two known types of dPHB-binding domains nor shows significant amino acid sequence similarity to them. These unique features indicate the presence of two novel and distinct types of dPHB-binding domains. Homologs of these novel domains also are present in the extracellular PhaZ of Bacillus megaterium and the putative extracellular PhaZs of Bacillus pseudofirmus and Bacillus sp. strain SG-1. The Bacillus sp. NRRL B-14911 PhaZ appears to be a representative of a novel class of extracellular PHB depolymerases.

INTRODUCTION

Poly(3-hydroxybutyrate) (PHB) is synthesized and accumulated intracellularly as a carbon and energy storage material by a wide variety of bacteria (14, 28, 29). When nutrient availability is restricted or unbalanced, the accumulated PHB is subject to degradation by the intracellular PHB depolymerase(s) (PhaZ) to produce 3-hydroxybutyrate (3HB) oligomers and/or monomers. These products subsequently are metabolized by the intracellular 3HB-oligomer hydrolase and/or the 3HB dehydrogenase as carbon and energy sources (21, 22, 30, 36, 38). The intracellular PHB exists in an amorphous state. When the amorphous PHB is subjected to solvent extraction or exposed to the extracellular environment due to cell death or lysis, the amorphous PHB is transformed into denatured semicrystalline PHB (dPHB) (13, 16).

A wide variety of bacteria can produce and secrete extracellular PHB depolymerases for the specific degradation of dPHB (12, 16, 40), with some exceptions. One exception is the extracellular PHB depolymerase PhaZ7 of Paucimonas lemoignei, which can only degrade amorphous PHB (7). Another known exception is the periplasm-located PHB depolymerase of Rhodospirillum rubrum, which shows substrate specificity for amorphous PHB (8). Like intracellular PHB depolymerases, extracellular PHB depolymerases can degrade PHB into 3HB oligomers and/or monomers. The produced 3HB oligomers can be further degraded by extracellular 3HB-oligomer hydrolases into 3HB monomers or degraded by intracellular 3HB-oligomer hydrolases after the uptake of 3HB oligomers into bacterial cells (36, 43).

The mature forms of extracellular PHB depolymerases usually are composed of three domains: a catalytic domain, a linker domain, and one or two dPHB-binding domains (16, 35). The catalytic domain is classified into type 1 and type 2 subgroups, depending on the relative order of the histidine residue for the oxyanion hole and the catalytic triad residues. The linker domain is classified into three subgroups according to their sequence features or their similarities to other proteins: the cadherin-like (19, 25), the fibronectin type III-like (18, 20, 24, 34), and the threonine-rich linker domains (3, 15). The dPHB-binding domain is classified into two subgroups: type 1 and type 2 (16). There also are single-domain extracellular PHB depolymerases: the PhaZ7 of P. lemoignei (26) and the PhaZ of Penicillium funiculosum (10). These PhaZs contain only a unique catalytic domain. In this study, we have identified a novel class of extracellular PHB depolymerase from Bacillus sp. strain NRRL B-14911. This PhaZ contains two novel and distinct types of linker domains (LD1 and LD2), as well as two novel and distinct types of dPHB-binding domains (SBD1 and SBD2). Homologs of these novel domains also are present in the extracellular PhaZ of Bacillus megaterium and the putative extracellular PhaZs of Bacillus pseudofirmus and Bacillus sp. SG-1. The Bacillus sp. NRRL B-14911 PhaZ appears to be a representative of a novel class of extracellular PHB depolymerases.

MATERIALS AND METHODS

Bacterial strains and growth conditions.

The bacterial strains and plasmids used in this study are listed in Table 1. The oligonucleotide primers are listed in Table 2. Escherichia coli cells were grown at 37°C in Luria-Bertani (LB) medium (31). E. coli strain DH5α was used as a host for cloning purposes; E. coli strain BL21(DE3) and strain JM109 were used as hosts for the heterologous expression of proteins. Bacillus sp. NRRL B-14911 cells were grown at 37°C in LB medium for the isolation of the genomic DNA. Overlay agar plates that were made with M9 mineral salts medium (31) and supplemented with 0.6% dPHB in the top agar or overlay agar plates that were made with LB medium and supplemented with 0.6% dPHB in the top agar were prepared as described previously (17) and were used for the observation of halo formation by dPHB-degrading bacteria. Antibiotics were used at the following concentrations (μg/ml): ampicillin, 100 (for E. coli); tetracycline, 10 (for B. subtilis).

Table 1.

Bacterial strains and plasmids

Strain or plasmid	Descriptiona	Reference or sourceb
E. coli
DH5α	F− φ80dlacZΔM15 Δ(lacZYA-argF) recA1 gyrA endA1 relA1 supE44 hsdR17	Laboratory stock
JM109	recA1 supE44 endA1 hsdR17 gyrA96 relA1 thi Δ(lac-proAB) [F′ traD36 proAB+lacIqZΔM15]	Takara
BL21(DE3)	F−ompT hsdS gal dcm λ(DE3)	Novagen
B. subtilis
168	trpC2	Laboratory stock
BM1269	trpC2 (pGS2067)	This study
BM548	trpC2 (pHY300PLK)	This study
Bacillus sp.
NRRL B-14911		ARSCC
B. megaterium
ATCC 11561		ATCC
Plasmids
pHY300PLK	Expression vector; Apr, Tcr	Takara
pQE30	Expression vector for producing His-tagged proteins in E. coli JM109; Apr	Qiagen
pET22b	Expression vector for producing His-tagged proteins in E. coli BL21(DE3); Apr	Novagen
pET32a	Expression vector for producing His-tagged thioredoxin fusion proteins in E. coli BL21(DE3); Apr	Novagen
pGS1243	pQE30 carrying the NAD+-dependent (R)-3HB dehydrogenase gene of B. thuringiensis	40
pGS2067	pHY300PLK carrying DNA encoding the full-length PhaZ(BspNRRL)	This study
pGS2164	pET22b carrying DNA encoding the mature form of PhaZ(BspNRRL)	This study
pGS2183	pET22b carrying DNA encoding the C domain of PhaZ(BspNRRL)	This study
pGS2188	pET22b carrying DNA encoding the N domain of PhaZ(BspNRRL)	This study
pGS2218	pET22b producing the mature form of PhaZ(BspNRRL) with deletion of the linker domain LD1	This study
pGS2230	pET32a producing the fusion protein of TrxA and the C domain of PhaZ(BspNRRL) with deletion of the linker domain LD1	This study
pGS2247	pET32a producing the fusion protein of TrxA and the linker domain LD2 of PhaZ(BspNRRL)	This study
pGS2248	pET32a producing the fusion protein of TrxA and the dPHB-binding domain SBD1 of PhaZ(BspNRRL)	This study
pGS2249	pET32a producing the fusion protein of TrxA and the dPHB-binding domain SBD2 of PhaZ(BspNRRL)	This study
pGS2254	pET32a producing the fusion protein of TrxA and the linker domain LD1 of PhaZ(BspNRRL)	This study
pGS2281	pET22b producing the mature form of PhaZ(BspNRRL) with deletion of the linker domains LD1 and LD2	This study

^aAp^r, ampicillin resistant; Tc^r, tetracycline resistant.

^bARSCC, Agricultural Research Service Culture Collection; ATCC, American Type Culture Collection.

Table 2.

Oligonucleotide primers used in this study

Name	Sequence (5′–3′)a	Location
A098	GCGGAATTCGAATGAAAAGGAGGAATGG	PhaZ 5′ end
A099	GCCAAAGCTTCAAAGGGAAAAAAGAGAG	PhaZ 3′ end
A174	GCCACTCGAGGTTGGCGCGATTTACTAT	SBD2 3′ end
A397	GCCAACACATATGGCAGGACAATTTATTACAG	N 5′ end
A398	GCCAACACATATGGATGCCCCTGCTACATC	LD1 5′ end
A399	GCCACTCGAGGCCGGCATCTCCTCCA	N 3′ end
A467	GCGGAATTCGCCGGCATCTCCTCCA	N 3′ end
A468	GCGGAATTCAGCGCCATCTCAGG	LD2 5′ end
A551	GCGGGATCCGATGCCCCTGCTACATC	LD1 5′ end
A555	GCCAAAGCTTTCACGCTGTCTGTCCGAGC	LD2 3′ end
A556	GCGGAATTCATTGGGTCAGCAAAATTGC	SBD1 5′ end
A557	GCCAAAGCTTTCAAACGGCGGCCGGCTG	SBD1 3′ end
A558	GCGGAATTCTCTAATTACATTGATATAGATA	SBD2 5′ end
A568	GCCAAAGCTTTCATGTCAGCGAGTCCCCG	LD1 3′ end

^aUnderlined sequences represent restriction sites inserted.

Construction of plasmids.

To construct plasmid pGS2067, the insert DNA was amplified by PCR using the primer pair A098 plus A099, restriction digested, and inserted between the EcoRI and HindIII sites of pHY300PLK (Takara).

To construct plasmids pGS2164, pGS2183, and pGS2188, primer pairs A397 plus A174, A398 plus A174, and A397 plus A399, respectively, were used in PCR. The DNA fragments were individually cloned between the NdeI and XhoI sites of pET22b (Novagen).

To construct plasmid pGS2218, primer pairs A397 plus A467 and A468 plus A174 were used in PCR. After restriction digestion with NdeI plus EcoRI and EcoRI plus XhoI, respectively, these two DNA fragments were ligated together into the NdeI and XhoI sites of pET22b. Plasmid pGS2281 was constructed in a similar manner by using the primer pairs A397 plus A467 and A556 plus A174.

To construct plasmids pGS2230, pGS2247, pGS2248, and pGS2249, primer pairs A468 plus A099, A468 plus A555, A556 plus A557, and A558 plus A099, respectively, were used in PCR. The DNA inserts were individually cloned between the EcoRI and HindIII sites of pET32a (Novagen). To construct plasmid pGS2254, the primer pair A551 plus A568 was used in PCR. The insert DNA was cloned between the BamHI and HindIII sites of pET32a.

All sequences of PCR-amplified DNA fragments were confirmed by DNA sequencing.

Overproduction and purification of His-tagged proteins.

For the purification of His-tagged proteins overproduced from E. coli strain BL21(DE3) bearing plasmid pGS2164, pGS2183, pGS2188, pGS2230, pGS2249, or pGS2254, cells were grown in LB medium at 37°C to an absorbance at 600 nm of 0.5 and then treated with 0.3 mM IPTG (isopropyl-β-d-thiogalactopyranoside) at 37°C for 3 h. Cells were collected, disrupted by sonication, and centrifuged at 15,000 × g for 10 min. The resulting supernatant was used for the purification of His-tagged proteins by affinity chromatography on an Ni-nitrilotriacetic acid (NTA) agarose column according to the instructions of the matrix manufacturer (Qiagen Inc.). A similar procedure was followed to overproduce and purify His-tagged thioredoxin itself from E. coli strain BL21(DE3) bearing the expression vector pET32a (Novagen).

For the purification of His-tagged proteins overproduced from E. coli strain BL21(DE3) bearing plasmid pGS2218, pGS2247, or pGS2248, cells were grown in LB medium at 37°C to an absorbance at 600 nm of 0.5 and then treated with 0.02 mM IPTG at 16°C for 16 h. The purification of His-tagged NAD⁺-dependent (R)-3HB dehydrogenase of B. thuringiensis from E. coli strain JM109 bearing the plasmid pGS1243 was carried out exactly as described previously (41).

Preparation of dPHB.

dPHB was isolated from B. megaterium cells that were grown at 37°C in LB medium supplemented with 2% sodium acetate for 8 h by a procedure involving sodium hypochlorite digestion and subsequent solvent extraction with acetone-ether (2:1, vol/vol) as described previously (6).

Determination of the enzyme activities of PHB depolymerase and its derivatives.

The enzyme activity of PHB depolymerase or its derivative was determined by monitoring the decrease in the turbidity of a suspension made from dPHB due to the hydrolysis of dPHB as described previously (9). The turbidity of the reaction mixture was assayed spectrophotometrically at 650 nm by using a SpectraMax 190 microplate reader (Molecular Devices Corp.). The reaction mixture (200 μl) contained 100 mM Tris-HCl (pH 8.0), 0.7 mg of dPHB, and 0.24 μM purified PHB depolymerase or its derivative, and it was incubated at 37°C for 30 min. One unit of activity is defined as a decrease of 1 U in the optical density at 650 nm (OD₆₅₀) in 1 min.

To investigate the effects of various chemical reagents on the dPHB-hydrolyzing activity of PhaZ, the chemical reagent was preincubated with PhaZ in the reaction mixture for 10 min prior to the addition of dPHB.

The spectrophotometric measurement of yellow p-nitrophenol generated from the hydrolysis of p-nitrophenylbutyrate by PhaZ or its derivative was performed exactly as described previously (41).

Quantitation of dPHB and 3HB monomers.

The determination of the amount of dPHB was carried out as described previously (5). The amount of 3HB monomers produced from the hydrolysis of dPHB by PHB depolymerase was quantified by the enzymatic method using NAD⁺-dependent (R)-3HB dehydrogenase according to the description in a previous report (42) and as reported previously by us (41).

Western blot analysis of the dPHB- or chitin-binding ability.

PHB depolymerase or its derivative was pretreated with 1 mM PMSF (phenylmethylsulfonyl fluoride) for 10 min to inhibit the dPHB-degrading activity. The dPHB- or chitin-binding assay mixture (13 μl) contained 10 mM Tris-HCl (pH 7.4), 200 μg of dPHB or chitin (Sigma), and 1 to 3 μg of PMSF-treated PHB depolymerase or its derivative. After incubation at 30°C for 15 min, the assay mixture was separated into the supernatant and the dPHB- or chitin-containing pellet by centrifugation. The pellet was washed three times with 40 μl of 10 mM Tris-HCl (pH 7.4) at room temperature. Proteins in the washed pellet and in the supernatant from the assay mixture were analyzed by sodium dodecyl sulfate-12% polyacrylamide gel electrophoresis (SDS-12% PAGE) and Western blotting with anti-His tag polyclonal antibody (Santa Cruz Biotechnology, Inc.) as the probe. Western blotting was carried out exactly as described previously (5).

Protein sequence analysis.

Amino acid identities and similarities between proteins were calculated with the CLUSTAL W software (39). Signal peptide sequences were analyzed by the SignalP software (23). The isoelectric point (pI) was calculated by using the ExPASy proteomics server of the Swiss Institute of Bioinformatics.

Other methods.

The genomic DNA of Bacillus sp. NRRL B-14911 was isolated by the method described previously (27). The transformation of B. subtilis cells by the protoplast method was performed as described previously (4). Protein concentrations were determined by the bicinchoninic acid protein assay method according to the instructions of the assay kit manufacturer (Pierce Biotechnology, Inc.) with bovine serum albumin as the standard.

Amino acid sequence accession numbers.

Amino acid sequences for the following proteins have been submitted to GenBank (accession numbers are in parentheses): PHB depolymerases of Bacillus sp. NRRL B-14911 (ZP_01169502), Bacillus sp. SG-1 (ZP_021858765), B. pseudofirmus OF4 (YP_003425754), and B. megaterium QM B1551 (YP_003560966); NHL repeat-containing protein of Geobacter sp. FRC-32 (YP_002537077); FG-GAP repeat-containing protein of Arthrobacter aurescens TC1 (YP_946834); and Ig domain protein group 2 domain protein of Thermincola sp. JR (YP_003641592).

RESULTS

Identification of a putative extracellular PHB depolymerase from Bacillus sp. NRRL B-14911.

The genomic sequence of Bacillus sp. NRRL B-14911 has been partially determined and is available from GenBank. It encodes a putative PHB depolymerase (GenBank accession number ZP_01169502) of 592 amino acids (see Discussion) that shows 53% identity and 68% similarity to the extracellular PHB depolymerase of B. megaterium N-18-25-9 (590 amino acids; BAF35850) (37). The analysis of the putative PhaZ of Bacillus sp. NRRL B-14911 by the SignalP 3.0 software revealed the presence of a typical signal peptide (100% probability) from amino acid position 1 to 28. The most likely cleavage site is predicted to be located between Ala28 and Ala29 (99.7% probability). These features suggest that this putative PhaZ is an extracellular enzyme. When growing on an opaque overlay agar plate that was prepared with M9 mineral salts medium and supplemented with 0.6% dPHB in the top agar, Bacillus sp. NRRL B-14911 cells produced a clearing zone (Fig. 1A). This observation indicates the presence of dPHB depolymerase activity in Bacillus sp. NRRL B-14911. When the pHY300PLK-based plasmid pGS2067, which constitutively expressed the full-length PhaZ of Bacillus sp. NRRL B-14911 from a promoter present in pHY300PLK, was introduced into B. subtilis cells, the transformants produced a clearing zone on an opaque LB agar plate supplemented with 0.6% dPHB in the top agar (Fig. 1B). This is in contrast to that observed in the B. subtilis transformants harboring the control vector pHY300PLK (Fig. 1B). These observations indicate that this putative enzyme of Bacillus sp. NRRL B-14911 has dPHB-degrading activity.

Fig. 1.

(A) Growth of Bacillus sp. NRRL B-14911 on a dPHB-containing M9 agar plate for 48 h. (B) Growth of B. subtilis harboring the PhaZ(BspNRRL)-producing plasmid pGS2067 or the control plasmid pHY300PLK on a dPHB-containing LB agar plate for 24 h.

Characterization of the mature form of PhaZ of Bacillus sp. NRRL B-14911.

To examine the dPHB-hydrolyzing activity of the mature form of Bacillus sp. NRRL B-14911 PhaZ, we constructed plasmid pGS2164 (Fig. 2A) that could overproduce the His-tagged mature form of PhaZ in E. coli and then purified this His-tagged protein by affinity chromatography on an Ni-NTA agarose column. A turbidimetric method that monitored the decrease in the turbidity of a suspension made from dPHB was used to examine the degradation of dPHB by the purified PhaZ. The result showed that this PhaZ could rapidly hydrolyze dPHB (Fig. 3). We also analyzed the products of hydrolysis of dPHB by this PhaZ with purified NAD⁺-dependent (R)-3HB dehydrogenase as described in Materials and Methods. The yield of 3HB monomers produced from dPHB degradation by this PhaZ corresponded to approximately 99% of total 3HB equivalents present in dPHB. This result suggests that the Bacillus sp. NRRL B-14911 PhaZ works, at least in part, as an exotype enzyme.

Fig. 2.

(A) Domain structures of the Bacillus sp. NRRL B-14911 PhaZ and plasmid constructs producing the mature form of PhaZ, its deletion derivatives, or various TrxA fusion proteins. The starting and ending positions of each domain within full-length PhaZ are numbered relative to the translational start site of PhaZ. The amino acid length of each domain is indicated in parentheses. SP, signal peptide. (B) SDS-12% PAGE of purified His-tagged proteins. (C) SDS-12% PAGE of purified His-tagged TrxA fusion proteins. M, molecular mass standards.

Fig. 3.

Turbidimetric determination of the dPHB-degrading activity of PHB depolymerase or its derivative. A decrease in the turbidity of a reaction mixture (200 μl) containing dPHB (0.7 mg) and purified His-tagged protein (0.24 μM) was detected spectrophotometrically at 650 nm. Symbols: solid circles, the mature form of PhaZ; solid squares, the mature form of PhaZ with the deletion of the linker domain LD1; solid triangles, the N-domain of PhaZ; hollow circles, no protein.

The dPHB-degrading activity of PhaZ was strongly inhibited by the freshly prepared serine esterase inhibitor PMSF (96 and 99% inhibition at 1 and 10 mM, respectively), suggesting the presence of a nucleophilic serine residue at the active center of this PhaZ. The reducing agent DTT (dithiothreitol) also significantly inhibited the dPHB-degrading activity of PhaZ (50 and 75% inhibition at 1 and 10 mM, respectively), suggesting the requirement of a disulfide bond for the enzyme activity. PhaZ showed maximum activity toward dPHB at pH 9.0 and 70°C. The specific dPHB-degrading activity of PhaZ is approximately 23 U/mg of protein at pH 8.0 and 37°C and 220 U/mg of protein at pH 9.0 and 70°C. Its enzyme activity was not increased by the addition of 1 mM CaCl₂ to the reaction mixture and was not inhibited by the presence of 10 mM EDTA, suggesting that the calcium ion is not required as a cofactor for the enzyme activity. This is in striking contrast to the B. megaterium N-18-25-9 PhaZ (37).

To analyze the dPHB-binding ability of PhaZ, the mature form of PhaZ was pretreated with 1 mM PMSF to inhibit the dPHB-degrading activity. After the incubation of PMSF-treated PhaZ with dPHB, the assay mixture was separated into the supernatant and the dPHB-containing pellet by centrifugation. The pellet was washed with the incubation buffer three times. The washed pellet and the supernatant from the assay mixture then were subjected to SDS-PAGE followed by Western blotting with anti-His tag antibody as the probe. The result showed that PhaZ was present in the pellet (Fig. 4A), suggesting that PhaZ could bind to dPHB. In a control experiment, PhaZ could not bind to chitin (Fig. 4B), indicating that the binding is specific.

Fig. 4.

Western blot analysis of results of dPHB- and chitin-binding assays. dPHB or chitin (200 μg) was incubated with PMSF-treated PHB depolymerase (the mature form) or its derivative (1 to 3 μg) at 30°C for 15 min. The assay mixture then was separated into the supernatant and the dPHB- or chitin-containing pellet by centrifugation. The pellet was washed three times. Proteins in the washed pellet and in the supernatant from the assay mixture were analyzed by SDS-12% PAGE and Western blotting with anti-His tag polyclonal antibody as the probe. The experiments were performed at least twice. A representative experiment is shown. In, input His-tagged protein; T, total proteins before separation into fractions by centrifugation; S, supernatant; P, pellet.

Catalytic domain of PhaZ.

The N-terminal catalytic domain (from position 29 to 335; 307 amino acids) of Bacillus sp. NRRL B-14911 PhaZ (Fig. 2A) showed homology with those of the PhaZs of B. megaterium N-18-25-9 (level of identity, 61%), Bacillus sp. SG-1 (ZP_021858765) (80%), B. pseudofirmus OF4 (YP_003425754) (61%), and B. megaterium QM B1551 (YP_003560966) (61%) (Fig. 5). Interestingly, it also displayed homology with the intracellular PhaZ1 (323 amino acids; GenBank accession no. FJ175152) of B. megaterium ATCC 11561 (49%) (Fig. 5). PhaZ1 has been demonstrated to be capable of degrading dPHB (5). The analysis of the amino acid sequence alignment of the catalytic domains of these PhaZs suggests that Ser146, Asp222, and His295 of the Bacillus sp. NRRL B-14911 PhaZ constitute the putative catalytic triad, and that His63 is the putative oxyanion hole (Fig. 5). The sequential order of these residues indicates that the catalytic domain of the Bacillus sp. NRRL B-14911 PhaZ belongs to type 1. A pentapeptide sequence (G-L-S₁₄₆-A-G) present in its catalytic domain is similar to the lipase box motif (G-X-S-X-G) (where X represents any amino acid residue) (11) identified in many serine hydrolases.

Fig. 5.

Alignment of amino acid sequences of the putative catalytic domains of extracellular PhaZs from several Bacillus species and the full-length intracellular PhaZ1 from B. megaterium ATCC 11561. The sequences of the putative catalytic domains of extracellular PhaZs from Bacillus sp. NRRL B-14911 [PhaZ(BspNRRL)], Bacillus sp. SG-1 [PhaZ(BspSG)], B. pseudofirmus OF4 [PhaZ(Bps)], B. megaterium QM B1551 [PhaZ(BmeQM)], and B. megaterium N-18-25-9 [PhaZ(BmeN)], and the sequence of the full-length intracellular PhaZ1 from B. megaterium ATCC 11561 [PhaZ1(BmeATCC)] are shown. All sequences are numbered relative to the translational start sites of the proteins. The conserved histidine residues for the predicted oxyanion holes, the putative catalytic triads, and two conserved cysteine residues are shown in white on black. The lipase box-like sequences are shaded in gray. Identical residues are marked below with asterisks. GenBank accession numbers are the following: PhaZ(BspNRRL), ZP_01169502; PhaZ(BspSG), ZP_01858765; PhaZ(Bps), YP_003425754; PhaZ(BmeQM), YP_003560966; PhaZ(BmeN), BAF35850; and PhaZ1(BmeATCC), FJ175152.

The purified His-tagged catalytic domain (Fig. 2B) of the Bacillus sp. NRRL B-14911 PhaZ had very weak dPHB-degrading activity (Fig. 3), whereas its ability to hydrolyze the synthetic substrate p-nitrophenylbutyrate (ca. 0.085 μmol/min/mg of protein) was comparable to that of the mature form of PhaZ (ca. 0.088 μmol/min/mg of protein). Western blotting showed that this catalytic domain had no dPHB-binding ability (Fig. 4C), whereas the remaining C domain could bind to dPHB (Fig. 4D). These observations are in agreement with the idea that the catalytic domain of this PhaZ possesses a full structural requirement for the esterase activity against p-nitrophenylbutyrate, but it still requires the presence of the remaining C domain for dPHB degradation.

Linker domains LD1 and LD2.

We next attempted to define the linker domain of PhaZ. A BLAST search with the amino acid sequence of the Bacillus sp. NRRL B-14911 PhaZ from position 301 to 592 as a probe against the nonredundant protein databases of NCBI revealed an interesting result: a region of the Bacillus sp. NRRL B-14911 PhaZ (roughly from position 336 to 425) is significantly similar in amino acid sequence to certain regions (ca. 90 amino acids long) of a small group of PhaZs and a large group of PHB depolymerase-unrelated proteins. These PHB depolymerase-unrelated proteins include the FG-GAP repeat-containing protein of Arthrobacter aurescens TC1 (YP_946834), the NHL repeat-containing protein of Geobacter sp. FRC-32 (YP_002537077), and the Ig domain protein group 2 domain protein of Thermincola sp. JR (YP_003641592) (Fig. 6A). This putative linker domain of the Bacillus sp. NRRL B-14911 PhaZ, designated LD1 (Fig. 2A and 6A), neither possesses the sequence features of the three known types of linker domains of PhaZs (cadherin-like, fibronectin-like, and threonine-rich linker domains) nor shows significant amino acid sequence similarity with them. These features indicate that the putative linker domain LD1 of the Bacillus sp. NRRL B-14911 PhaZ is a representative of a novel type of linker domain.

Fig. 6.

(A) Alignment of amino acid sequences of the putative linker domains LD1 of extracellular PhaZs from several Bacillus species plus the homologous regions of several PHB-unrelated proteins. The sequences of the putative linker domains LD1 of the extracellular PhaZs from Bacillus sp. NRRL B-14911 [PhaZ(BspNRRL)], Bacillus sp. SG-1 [PhaZ(BspSG)], B. pseudofirmus OF4 [PhaZ(Bps)], B. megaterium QM B1551 [PhaZ(BmeQM)], and B. megaterium N-18-25-9 [PhaZ(BmeN)] and the sequences of the homologous regions of Ig domain protein group 2 domain protein from Thermincola sp. JR [Ig(Tsp)], NHL repeat-containing protein from Geobacter sp. FRC-32 [NHL(Gsp)], and FG-GAP repeat-containing protein from Arthrobacter aurescens TC1 [FG-GAP(Aau)] are shown. (B) Alignment of amino acid sequences of the putative linker domains LD2 of extracellular PhaZs from several Bacillus species. (C) Alignment of amino acid sequences of the putative dPHB-binding domain SBD1 of extracellular PhaZs from several Bacillus species. (D) Alignment of amino acid sequences of the putative dPHB-binding domain SBD2 of extracellular PhaZs from several Bacillus species. All sequences are numbered relative to the translational start sites of the proteins. Amino acid residues that are fully conserved (B, C, and D) or conserved in at least six out of the eight sequences (A) are shown in white on black. GenBank accession numbers are the following: NHL(Gsp), YP_002537077; FG-GAP(Aau), YP_946834; and Ig(Tsp), YP_003641592. Other accession numbers are indicated in the legend to Fig. 5.

To test the dPHB-binding ability of the putative linker domain LD1, we constructed a fusion protein of thioredoxin (TrxA) with LD1 as shown in the plasmid pGS2254 (Fig. 2A). It turned out that the purified His-tagged fusion protein TrxA-LD1 showed weak binding to dPHB in a reproducible manner (Fig. 4E). In a control experiment, the purified His-tagged TrxA protein itself was found to be unable to bind to dPHB (Fig. 4J). We also constructed a fusion protein of TrxA with the remaining C-terminal portion of PhaZ as shown in the plasmid pGS2230 (Fig. 2A). This purified His-tagged fusion protein exhibited a strong dPHB-binding ability (Fig. 4F).

Extracellular PHB depolymerases usually contain a dPHB-binding domain of 40 to 60 amino acids in length at their C termini (1, 16, 25, 35). Exceptionally, the extracellular PHB depolymerases of Pseudomonas stutzeri (25) and Marinobacter sp. strain NK-1 (19) contain two similar dPHB-binding domains at their C termini. To identify the possible dPHB-binding domain of PhaZ, we divided the remaining C domain of PhaZ into three parts (LD2, SBD1, and SBD2, of 52, 56, and 59 amino acids in length, respectively) (Fig. 2A) roughly on the basis of the amino acid alignment (Fig. 6B, C, and D). We then constructed fusion proteins of TrxA with each of them (Fig. 2A) to test their dPHB-binding abilities. It turned out that, like the fusion protein TrxA-LD1, the purified His-tagged fusion protein TrxA-LD2 also showed reproducibly weak binding to dPHB (Fig. 4G). There is no significant amino acid sequence similarity between LD1 and LD2. LD2 neither possesses the sequence features of the three known types of linker domains of PhaZs nor shows significant amino acid sequence similarity with them (Fig. 6B). However, homologs of LD2 are present in the extracellular PhaZs of B. megaterium, B. pseudofirmus, and Bacillus sp. strain SG-1 (Fig. 6B). These features indicate that the putative linker domain LD2 of the Bacillus sp. NRRL B-14911 PhaZ is also a representative of a novel type of linker domain.

It is generally believed that the linker domain plays a structural role by providing suitable spacing between the catalytic domain and the substrate-binding domain for dPHB degradation. It has been demonstrated previously that the deletion of the fibronectin-like linker domain of the Ralstonia pickettii T1 PhaZ abolished its dPHB-degrading ability, whereas its dPHB-binding ability still was maintained (24). This result indicates the importance of the presence of the linker domain for dPHB degradation by PhaZ. In this study, we found that the deletion of the putative linker domain LD1 from the mature form of PhaZ as shown in the plasmid pGS2218 (Fig. 2A) only partially reduced its dPHB-degrading ability (Fig. 3). Its dPHB-binding ability still was maintained (Fig. 4K). Therefore, we next attempted to delete LD1 and LD2 simultaneously from the mature form of PhaZ, as shown in the plasmid pGS2281 (Fig. 2A). Unexpectedly, the resulting PhaZ was insoluble and found only in inclusion bodies (data not shown). This prompted us to construct a fusion protein of the soluble protein TrxA with this deletion derivative. It was found that this fusion protein still was insoluble, even when the induction of protein production was carried out with 0.04 mM IPTG and at 16°C (data not shown). This problem hindered our analysis of the effect of the deletion of both LD1 and LD2 on the dPHB-degrading ability of this PhaZ.

dPHB-binding domains SBD1 and SBD2.

We next tested the dPHB-binding abilities of purified His-tagged TrxA-SBD1 and TrxA-SBD2. Like the mature form of PhaZ, both TrxA-SBD1 and TrxA-SBD2 showed strong binding to dPHB (Fig. 4H and I). There is no significant amino acid sequence similarity between SBD1 and SBD2 (Table 3). Neither SBD1 nor SBD2 shows significant amino acid sequence similarity to the two known types of dPHB-binding domains of PhaZs (Table 3). There is no cysteine residue in SBD1 and SBD2 (Fig. 6C and D). SBD1 and SBD2 do not contain the sequence motif HXXAGR (where X represents any amino acid residue) (Fig. 6C and D), which is conserved in the two known types of dPHB-binding domains (16). Nevertheless, homologs of SBD1 and SBD2 are present in the extracellular PhaZs of B. megaterium, B. pseudofirmus, and Bacillus sp. SG-1 (Fig. 6C and D). These features indicate that SBD1 and SBD2 of the Bacillus sp. NRRL B-14911 PhaZ are representatives of two novel and distinct types of dPHB-binding domains. The presence of two dPHB-binding domains in the Bacillus sp. NRRL B-14911 PhaZ may provide an advantage for PhaZ to more easily gain access to dPHB in the environment.

Table 3.

Amino acid sequence identities and similarities among the dPHB-binding domains of various PhaZs

dPHB-binding domain	Length of dPHB-binding domain (amino acids)	Position of dPHB-binding domain (residue)	% Identity/% similarity to PhaZ(BspNRRL) domain:
			SBD1	SBD2
PhaZ(BspNRRL) SBD1	56	478−533	100/100	14/20
PhaZ(BspSG) SBD1	56	478−533	58/77	2.9/5.8
PhaZ(BmeQM) SBD1	57	475−531	36/59	1.9/4.8
PhaZ(Bps) SBD1	57	482−538	40/62	2.9/5.7
PhaZ(BmeN) SBD1	57	475−531	36/59	1.9/4.8
PhaZ(BspNRRL) SBD2	59	534−592	14/20	100/100
PhaZ(BspSG) SBD2	60	534−593	12/19	62/73
PhaZ(BmeQM) SBD2	59	532−590	19/31	38/57
PhaZ(Bps) SBD2	59	539−597	13/24	37/60
PhaZ(BmeN) SBD2	59	532−590	15/31	40/58
PhaZ(Pst) SBDII	60	455−514	16/30	12/15
PhaZ(Pst) SBDI	50	527−576	11/19	13/18
PhaZ(Msp) SBDII	59	450−508	14/27	14/19
PhaZ(Msp) SBDI	50	529−578	12/15	10/19
PhaZ(Rpi)	53	440−492	1.9/2.9	16/28
PhaZ(Sex)	50	439−488	2.1/4.3	0.9/1.9
PhaZ2(Ple)	53	381−433	14/24	4.9/5.8
PhaZ3(Ple)	53	367−419	11/18	16/24

DISCUSSION

The annotated size of the extracellular PhaZ of Bacillus sp. NRRL B-14911 (ZP_01169502) was 595 amino acids. However, there is no typical Shine-Dalgarno (SD) sequence (33) present with an appropriate spacing upstream from the predicted translational start site ATG (5-CCGAAGAATGAAAAGGAGGAATG-3′). Therefore, the translation of this PhaZ is more likely to be initiated from a second ATG located 6 nucleotides downstream of the first one, because the second ATG is preceded by a typical SD sequence (5′-AGGAGG-3′) with a spacing of 10 nucleotides. Thus, this second ATG is predicted to work as the translational start site of PhaZ. If so, this PhaZ should comprise 592 amino acids, which is what was used throughout this paper.

Bacillus sp. NRRL B-14911 is a marine bacterium that was isolated from seawater at a depth of 10 m. Tools for the genetic manipulation of Bacillus sp. NRRL B-14911 have not been successfully developed so far. Therefore, it is not feasible now to disrupt the phaZ gene in the chromosome to examine its influence on the ability of Bacillus sp. NRRL B-14911 to form a clearing zone on an opaque plate. It has been noted that the PhaZs of Alcaligenes faecalis AE122, Ralstonia pickettii T1, and Pseudomonas stutzeri, which contain type 1 catalytic domains, could degrade dPHB to produce a mixture of 3HB oligomers and monomers, whereas those PhaZs from Acidovorax sp., Comamonas sp., C. acidovorans, C. testosteroni, and Streptomyces exfoliates, which contain type 2 catalytic domains, could yield 3HB monomer as a main product (32). However, the PhaZs from Bacillus sp. NRRL B-14911 and P. stutzeri (25) contain type 1 catalytic domains, and almost all of the end products from the degradation of dPHB by these PhaZs are 3HB monomers. Therefore, there seems to be no absolute correlation between the types of the catalytic domains of PhaZs and the composition of end products of dPHB degradation by PhaZs.

It has been proposed that the fibronectin type III-like linker domains present in several bacterial water-insoluble polymer hydrolases (including extracellular PHB depolymerases) were acquired initially from an animal source and spread later by horizontal transfers between distantly related bacteria (2). It is of interest that the putative linker domain LD1 of the Bacillus sp. NRRL B-14911 PhaZ is significantly similar in amino acid sequence to certain regions of a large group of PHB depolymerase-unrelated proteins in the databases. It is unclear whether the regions that are present in many bacterial proteins and homologous to the linker domain LD1 of the Bacillus sp. NRRL B-14911 PhaZ also were acquired initially from an animal source and spread by horizontal transfers among distantly related bacteria.

In this study, dPHB-binding assays revealed that each of the two linker domains, LD1 and LD2, showed weak binding to dPHB. This is somewhat different from the cadherin-like linker domains of the extracellular PhaZs of P. stutzeri and Marinobacter sp. NK-1, which showed no binding to dPHB (19, 25). The significance of the presence of two distinct linker domains in the Bacillus sp. NRRL B-14911 PhaZ and the biological implication of the very weak binding of the linker domains LD1 and LD2 to dPHB are unclear.

It is worth mentioning that the calculated pI is 4.02 and 4.86 for the linker domains LD1 and LD2 of the Bacillus sp. NRRL B-14911 PhaZ, respectively, and is 6.20 and 9.52 for its dPHB-binding domains SBD1 and SBD2, respectively. Its dPHB-binding domains appear to be more basic than its linker domains. This phenomenon also is manifested in the extracellular PHB depolymerases of other types: the calculated pI is 3.73 for the linker domain of P. stutzeri PhaZ (ACG63776), 5.55 for SDB2, and 8.95 for SDB1; 3.54 for the linker domain of Marinobacter sp. NK-1 PhaZ (BAC15574), 4.14 for SDB2, and 4.75 for SDB1; 4.78 for the linker domain of R. pickettii PhaZ (BAA04986) and 5.98 for its SDB; 4.58 for the linker domain of S. exfoliates PhaZ (AAB02914) and 8.19 for its SDB; 8.76 for the linker domain of P. lemoignei PhaZ2 (AAB17150) and 9.25 for its SDB; 5.08 for the linker domain of P. lemoignei PhaZ3 (AAB48166) and 9.36 for its SDB; 4.36 for the linker domain of Comamonas sp. PhaZ (AAA87070) and 8.89 for its SDB; and 6.08 for the linker domain of C. acidovorans PhaZ (BAA19791) and 8.93 for its SDB. The biological significance of this phenomenon is not yet clear.