To Be or Not To Be a Poly(3-Hydroxybutyrate) (PHB) Depolymerase: PhaZd1 (PhaZ6) and PhaZd2 (PhaZ7) of Ralstonia eutropha, Highly Active PHB Depolymerases with No Detectable Role in Mobilization of Accumulated PHB

Abstract

The putative physiological functions of two related intracellular poly(3-hydroxybutyrate) (PHB) depolymerases, PhaZd1 and PhaZd2, of Ralstonia eutropha H16 were investigated. Purified PhaZd1 and PhaZd2 were active with native PHB granules in vitro. Partial removal of the proteinaceous surface layer of native PHB granules by trypsin treatment or the use of PHB granules isolated from ΔphaP1 or ΔphaP1-phaP5 mutant strains resulted in increased specific PHB depolymerase activity, especially for PhaZd2. Constitutive expression of PhaZd1 or PhaZd2 reduced or even prevented the accumulation of PHB under PHB-permissive conditions in vivo. Expression of translational fusions of enhanced yellow fluorescent protein (EYFP) with PhaZd1 and PhaZd2 in which the active-site serines (S190 and Ser193) were replaced with alanine resulted in the colocalization of only PhaZd1 fusions with PHB granules. C-terminal fusions of inactive PhaZd2(S193A) with EYFP revealed the presence of spindle-like structures, and no colocalization with PHB granules was observed. Chromosomal deletion of phaZd1, phaZd2, or both depolymerase genes had no significant effect on PHB accumulation and mobilization during growth in nutrient broth (NB) or NB-gluconate medium. Moreover, neither proteome analysis of purified native PHB granules nor lacZ fusion studies gave any indication that PhaZd1 or PhaZd2 was detectably present in the PHB granule fraction or expressed at all during growth on NB-gluconate medium. In conclusion, PhaZd1 and PhaZd2 are two PHB depolymerases with a high capacity to degrade PHB when artificially expressed but are apparently not involved in PHB mobilization in the wild type. The true in vivo functions of PhaZd1 and PhaZd2 remain obscure.

INTRODUCTION

Ralstonia eutropha H16 is a chemolithoautotrophic betaproteobacterium that has become famous because of its ability to accumulate large amounts of poly(3-hydroxybutyrate) (PHB). R. eutropha is used for the commercial production of bioplastics (polyhydroxyalkanoates [PHAs] such as PHB and copolymers of 3-hydroxybutyrate and 3-hydroxyvalerate) and is considered a model organism for PHA research (1–5). Investigation of the composition of the surface layer of PHB granules of R. eutropha H16 revealed an astonishingly high number of polypeptides that are predicted or have already been shown to be specifically attached to the PHB granule surface. These proteins include enzymes involved in biosynthesis (PHB synthase PhaC1) (6), in granule structure integrity (phasins PhaP1-PhaP7), in PHB mobilization (PhaZa1 to PhaZa5, PhaZb, PhaZc, PhaZd), and in other functions (PhaR, PhaM). For reviews, see references 1, 5, 7, and 8. Other PHA-accumulating bacteria and Archaea are also known to have proteins specifically attached to the PHA surface (9–15). For an overview and more references, see Table 2 in reference 5. Unfortunately, designation of PHB granule-associated proteins in R. eutropha is not uniformly used in literature: for alternative designations of PHB depolymerases and PHB oligomer hydrolases in R. eutropha, see Table 1. We wondered about the high number of genes (seven intracellular PHB [iPHB] depolymerases and two oligomer hydrolases) with predicted functions in the hydrolysis of PHB in R. eutropha. Convincing and independent evidence from several labs of a function as a physiological iPHB depolymerase is available only for PhaZa1 (16–18). PhaZa1 (and its isoenzymes) has a cysteine in the catalytic active site of the lipase boxes, and thiolysis (19) of PHB to 3-hydroxybutyryl-coenzyme A by PhaZa1 has been demonstrated (17, 20, 21). Moreover, PhaZa1 is the only PHB depolymerase on whose in vivo localization data are available. Fusions of PhaZa1 with enhanced yellow fluorescent protein (EYFP) were clearly attached to PHB granules (22). Remarkably, insertion in or deletion of phaZa1 reduced but did not completely inhibit the mobilization of accumulated PHB, and this indicated that other PHB-degrading enzymes must exist and should be active during PHB mobilization in R. eutropha. However, no direct evidence of the contribution of another iPHB depolymerase in PHB mobilization is available. Deletion of phaZa2, phaZa3, or phaZd had no detectable effect on PHB mobilization (18, 23). Solid data on the function of other iPHB depolymerase are rare. Transcriptomics of R. eutropha genes indicated that besides phaZa1, some of the postulated iPHB depolymerases (phaZa2, phaZa4, phaZa5, and phaZd2) were expressed (24–26), but expression on the level of protein activity has never been demonstrated. PhaZd is of special interest, as the purified protein in vitro has the highest PHB depolymerase activity of all of the iPHB depolymerases investigated (23). The R. eutropha genome predicts that the H16_B2401 gene product is an isoenzyme of PhaZd, but no information on the properties of the gene product is available. Because of the similarity of H16_B2401 (PhaZ7) to PhaZd (H16_B2073, PhaZ6), we speculated that H16_B2401 could also be a highly active PHB depolymerase that could suppress a phenotype of a ΔphaZd mutant. We therefore performed a comprehensive analysis of the potential physiological functions of PhaZd (PhaZd1, PhaZ6) and the H16_B2401 gene product (PhaZd2, PhaZ7) in the PHB metabolism of R. eutropha.

TABLE 2.

Strains and plasmids used in this study

Strain or plasmid	Relevant characteristic	Source or reference
Escherichia coli strains
JM109	Cloning strain
S17-1	Conjugation strain	51
BL21(DE3)/pLys	Expression strain	Novagen
Ralstonia eutropha strains
H16	Wild-type strain	DSMZ 428
H16 ΔphaP1	Chromosomal deletion of phaP1	13
H16 Δ(phaP1-phaP4)	Chromosomal deletions of phaP1-phaP4	13
H16 Δ(phaP1-phaP5)	Additional deletion of phaP5 in ΔphaP1-phaP4 background	52, this study
H16 ΔphaZd1	Chromosomal deletion of phaZd1	This study
H16 ΔphaZd2	Chromosomal deletion of phaZd2	this study
H16 ΔphaZd1ΔphaZd2	Chromosomal deletion of phaZd1, phaZd2	this study
Plasmids
pBBR1MCS2	Broad-host-range vector	53
pBBR1MCS2-PphaC-eyfp-c1	Universal vector for construction of fusions C terminal to EYFP under control of PphaC promoter	52
pBBR1MCS2-PphaC-eyfp-n1	Universal vector for construction of fusions N terminal to EYFP under control of PphaC promoter	27
pBBR1MCS-2-PphaC-eyfp-phaZd1	N-terminal fusion of PhaZd1 to EYFP	This study
pBBR1MCS-2-PphaC-eyfp-phaZd2	N-terminal fusion of PhaZd2 to EYFP	This study
pBBR1MCS-2-PphaC-eyfp-phaZd1(S190A)	N-terminal fusion of inactive PhaZd1 to EYFP	This study
pBBR1MCS-2-PphaC-eyfp-phaZd2(S193A)	N-terminal fusion of inactive PhaZd2 to EYFP	This study
pBBR1MCS-2-PphaC-phaZd1-eyfp	C-terminal fusion of PhaZd1 to EYFP	This study
pBBR1MCS-2-PphaC-phaZd2-eyfp	C-terminal fusion of PhaZd2 to EYFP	This study
pBBR1MCS-2-PphaC-phaZd1(S190A)-eyfp	C-terminal fusion of inactive PhaZd1 to EYFP	This study
pBBR1MCS-2-PphaC-phaZd2(S193A)-eyfp	C-terminal fusion of inactive PhaZd2 to EYFP	This study
pBBR1MCS2-PBAD-eyfp-c1	Universal vector for construction of fusions C terminal to EYFP under control of PBAD promoter	This study
pBBR1MCS2-PBAD-eyfp-n1	Universal vector for construction of fusions N terminal to EYFP under control of PBAD promoter	This study
pBBR1MCS-2-PBAD-eyfp-phaZd1	N-terminal fusion of PhaZd1 to EYFP	This study
pBBR1MCS-2-PBAD-eyfp-phaZd2	N-terminal fusion of PhaZd2 to EYFP	This study
pBBR1MCS-2-PBAD-eyfp-phaZd1(S190A)	N-terminal fusion of inactive PhaZd1 to EYFP	This study
pBBR1MCS-2-PBAD-eyfp-phaZd2(S193A)	N-terminal fusion of inactive PhaZd2 to EYFP	This study
pBBR1MCS-2-PBAD-phaZd1-eyfp	C-terminal fusion of PhaZd1 to EYFP	This study
pBBR1MCS-2-PBAD-phaZd2-eyfp	C-terminal fusion of PhaZd2 to EYFP	This study
pBBR1MCS-2-PBAD-phaZd1(S190A)-eyfp	C-terminal fusion of inactive PhaZd1 to EYFP	This study
pBBR1MCS-2-PBAD-phaZd2(S193A)-eyfp	C-terminal fusion of inactive PhaZd2 to EYFP	This study
pBBR1MCS-2-PphaC-phaZd1	Overexpression of PhaZd1	This study
pBBR1MCS-2-PphaC-phaZd2	Overexpression of PhaZd2	This study
pBBR1MCS-2-PphaC-phaZd1(S190A)	Overexpression of inactive PhaZd1	This study
pBBR1MCS-2-PphaC-PhaZd2(S193A)	Overexpression of inactive PhaZd2	This study
pLO3	Suicide vector; Tcr	28
pLO3-ΔphaZd1	Deletion vector for phaZd1, fragments up- and downstream of phaZd1 cloned between SacI and XbaI sites of pLO3	This study
pLO3-ΔphaZd2	Deletion vector for phaZd2, fragments up- and downstream of phaZd2 cloned between SacI and XbaI sites of pLO3	This study
pBBR1MCS-3-lacZ	lacZ reporter plasmid for determination of promoter activity on translational level	33
pBBR1MCS-3-lacZ-PphaC	lacZ reporter plasmid for determination of phaC promoter activity	This study
pBBR1MCS-3-lacZ-PphaZd1	lacZ reporter plasmid for determination of phaZd1 promoter activity	This study
pBBR1MCS-3-lacZ-PphaZd2	lacZ reporter plasmid for determination of phaZd2 promoter activity	This study
pET28a	His tag expression vector; Kmr	Novagen
pET28a-phaZd1	Plasmid for expression of His6-PhaZd1	This study
pET28a-phaZd2	Plasmid for expression of His6-PhaZd2	This study

TABLE 1.

Designation of PHB depolymerases and PHB oligomer hydrolases of R. eutropha H16^a

Protein (KEGG)	Locus tag (organization)	Length (amino acids)	Mol mass (kDa)	Active site	Characteristic	Reference(s)
PhaZa1 (PhaZ1)	A1150 (single gene)	419	47.3	SVC183QP	nPHB	16, 17, 20, 21
PhaZa2 (PhaZ2)	A2862 (A2863, galM?)	404	44.8	AIC173QP	?	18
PhaZa3 (PhaZ5)	B1014 (single gene)	407	45.2	AVC175QP	?	18
PhaZa4 (PhaZ4)	PHG178 (single gene)	245b	27.4b	?	?	48
PhaZa5 (PhaZ3)	B0339 (single gene)	412	45.2	?	?	32
PhaZb (PhaY1)	A2251 (A2250–A2259)	718	74.3	SVS320NG	Soluble/nPHB	49
PhaZc (PhaY2)	A1335 (A1335–A1338)	293	31.6	GTS108MG	Soluble	50
PhaZd1 (PhaZ6)	B2073 (single gene)	362	39.2	GMS190AG	Capacity to bind to nPHB	23, this study
PhaZd2 (PhaZ7)	B2401 (single gene)	365	38.4	GLS193AG	Capacity to hydrolyze nPHB	23, this study

^aThe designations of PHB depolymerase proteins in the literature and in the KEGG database, as well as the respective locus tags and important features are shown. Question marks indicate that experimentally verified data are not available.

^bCorrect gene start questionable.

MATERIALS AND METHODS

Bacterial strains, plasmids, and culture conditions.

The bacteria and plasmids used in this study are shown in Table 2. Escherichia coli JM109 and S17-1 were used for cloning experiments and as the donor strain in conjugation experiments, respectively. E. coli strains were grown on LB medium supplemented with the appropriate antibiotics at 37°C. R. eutropha H16 strains were routinely grown on nutrient broth (NB; 0.8%, [wt/vol]) medium at 30°C. Alternatively, PHB granule formation was monitored in mineral salts medium with 0.5% (wt/vol) fructose or 0.5 to 2% (wt/vol) gluconate. In experiments in which the formation and mobilization of PHB were studied, it was important to start with PHB-free cells. To this end, the respective R. eutropha strain was grown in two subsequent seed cultures (10 ml NB, 100-ml Erlenmeyer flask). The second seed culture was almost free of intermediately accumulated PHB after 24 to 30 h of growth at 30°C, although a few very long cells (frequency, ≤1%) with PHB granules remained. The majority of the cells were shortened rods without any detectable PHB granules. The total PHB content of the second seed culture was below 3%, and it was used to inoculate the main culture (5 to 10% [vol/vol]). Sodium gluconate (0.2% [wt/vol]) was added to promote PHB accumulation. In some experiments, 0.2% (wt/vol) l-arabinose was added to induce the P_BAD promoter. Recombinant R. eutropha strains harboring pBBR1-MCS-2 derivatives were grown in the presence of 150 μg/ml kanamycin. Formation of PHB granules was detectable by fluorescence microscopy (Nile red staining) within 10 min after the transfer of cells to fresh medium.

Construction of fusion proteins with EYFP.

Fusions of the gene for EYFP (eyfp) with genes of R. eutropha H16 (Table 2) were generally constructed as both N-terminal and C-terminal fusions with universal vectors (pBBR1MCS-2-P_phaC-eyfp-c1, pBBR1MCS-2-P_phaC-eyfp-n1) based on the broad-host-range plasmid pBBR1MCS-2. These vectors harbored the constitutively expressed (in R. eutropha) promoter of the phaCAB operon (P_phaC) or, alternatively, the arabinose-inducible P_BAD promoter and the coding region of eyfp. For details, see reference 27. The gene of interest was cloned in frame with the eyfp gene between the XhoI and BamHI restriction sites of pBBR1MCS-2-P_phaC-eyfp-c1 and between the NdeI and BamHI restriction sites of pBBR1MCS-2-P_phaC-eyfp-n1, respectively. All constructs were conjugatively transferred from recombinant E. coli S17-1 to R. eutropha H16, and selection was achieved by plating on mineral salts medium supplemented with 0.2% fructose and 350 μg ml⁻¹ kanamycin.

Construction of chromosomal knockouts.

Precise chromosomal deletions of phaZd1, phaZd2, or both genes were constructed in R. eutropha H16 by the sacB-sucrose selection method (15% sucrose used for selection) with pLO3 as the deletion vector as described elsewhere (28). For the sequences of the primer used, see Table 3. The genotype of the resulting deletion mutant was verified by PCR of the respective genomic region and determination of its DNA sequence. Only clones with correct DNA sequences were used.

TABLE 3.

Oligonucleotides used in this study

Oligonucleotide	Sequence (5′–3′)a
C-terminal fusions to eyfp
phaZd1_C_f_XhoI	CCGCTCGAGGCATGACCAAAAGCTTTGCCGC
phaZd1_C_r_BamHI	CGGGATCCTCAACGGCGGTGCTGG
phaZd2_C_f _XhoI	CCGCTCGAGGCATGCCCCGCTCATCCGG
phaZd2_C_r_BamHI	CGGGATCCTCACCCCCCAACCTCCAGA
N-terminal fusions to eyfp
phaZd1_N_f_NdeI	GGGAATTCCATATGACCAAAAGCTTTGCCGCTGACTGGC
phaZd2_N_r_BamHI	CGGGATCCCCACGGCGGTGCTGGCTGAAGAACTGC
phaZd1_N_f_NdeI	GGGAATTCCATATGCCCCGCTCATCCGG
phaZd2_N_r_BamHI	CGGGATCCCCCCCCCCAACCTCCAGA
Change from serine to alanine
phaZd1_S190A_F	GCCGGCATGGCCGCCGGCGCG
phaZd1_S190A_R	CGCGCCGGCGGCCATGCCGGC
phaZd2_S193A_F	CTGGGGCTGGCCGCGGGCGGC
phaZd2_S193A_R	GCCGCCCGCGGCCAGCCCCAG
Chromosomal deletions
phaZd1upstreamF	CGATCAGAGCTCCAGGGATTCGACCGAGACCGCACAAACC
phaZd1upstreamR	GCCGTTAATTAAGCCGGAGGACTCCTGATCGTGTGACGCGATCTCC
phaZd1downstreamF	CGGCTTAATTAACGGCAGTCGGGCAGCACCAATGCGCATCAAGC
phaZd1downstreamR	GCTCTAGATGATGACGCCCACCTTCCTGCAGAAGCTGTATGG
phaZd2upstreamF	CGATCAGAGCTCTGCCCGGGCCGGAAATCATGCAGC
phaZd2upstreamR	GCCGTTAATTAAGCCGGCGGCGTCCTGTCGGCTGGGGTAACC
phaZd2downstreamF	CGGCTTAATTAACGGCGGCAGGCTTGACGCCTCATGGATATCGC
phaZd2downstreamR	GCTCTAGACATGCCCGCAACCCAGCTCGACG
N-terminal His6 tag:
phaZd1_N-His_f_NdeI	GGGAATTCCATATGACCAAAAGCTTTGCCGCTGACTGGC
phaZd1_N-His_r_BamHI	CGGGATCCTCAACGGCGGTGCTGG
phaZd2_N-His_f_NdeI	GGGAATTCCATATGCCCCGCTCATCCGG
phaZd2_N-His_r_BamHI	CGGGATCCTCACCCCCCAACCTCCAGA
lacZ promoter fusions
phaZd1_SpeI-F	GGACTAGTATTGTGGCAAACGTGGCGAAAGAGG
phaZd1_NcoI-R	CATGCCATGGGGACTCCTGATCGTGTGACGCG
phaZd2_SpeI-F	GGACTAGTCAGTTGCGCGCGTACCGCGG
phaZd2_NcoI-R	CATGCCATGGGGCGTCCTGTCGGCTGG
phaC1_SpeI-F	GGACTAGTGCCGAGGCGGATTCCCGCATTG
phaC1_NcoI-R	CATGCCATGGGTTGATTGTCTCTCTGCCGTCACT

^aRestriction sites are underlined.

Site-directed mutagenesis of phaZd1 and phaZd2.

PCR was performed with TaKaRa PrimeStar polymerase. Insertion of mutations into the coding sequences of phaZd1 and phaZd2 resulting in exchange of the active-site serine for alanine (PhaZd1 [S190A] and PhaZd2 [S193A]) was done by overlap extension PCR with the primers shown in Table 3. E. coli cells were chemically transformed with plasmid DNA by standard procedures. Successful mutation introduction was confirmed by DNA sequencing.

Isolation of native PHA granules and activity assay for PhaZd1 and PhaZd2.

Native PHB (nPHB) granules were prepared from crude extracts (French press, three times) of PHB-rich cells of R. eutropha H16 by two sodium phosphate-buffered glycerol density gradient centrifugation steps as described previously (29, 30). The activities of PhaZd1 and PhaZd2 were assayed turbidimetrically at 650 nm. The assay mixture contained Tris-HCl (pH 8.5) (concentrations are indicated in Results) and nPHB granules (initial optical density at 650 nm [OD₆₅₀], ∼1). In some experiments, nPHB granules were “activated” with trypsin (1.25 mg/ml, 10 min) before the reaction was started by the addition of PHB depolymerase. One unit of nPHB depolymerase activity was defined as the amount required for the hydrolysis of 1 μg of PHB/min. The apparent extinction coefficient (ε) for the nPHB granule preparation was 1.2 μl μg⁻¹ cm⁻¹. Denatured PHB (dPHB) was prepared from sodium hypochlorite-digested, PHB-rich R. eutropha H16 cells. Alternatively, PHB depolymerase assays were performed by pH stat assay at a constant pH of 8.5 as described previously (30, 31).

Determination of nPHB hydrolysis products.

The soluble products of nPHB hydrolysis by PhaZd1 and PhaZd2 were derivatized with bromophenacyl bromide (BPB) and separated by high-performance liquid chromatography (HPLC) as described previously (30, 31). Different ratios of nPHB and purified PHB depolymerase and variation of incubation times were used to identify intermediates and hydrolysis products.

Determination of β-galactosidase activity.

PhaZd1 and PhaZd2 are encoded by singly expressed genes (32). The promoter regions of phaZd1 and phaZd2 (300 bp upstream of ATG) were amplified by PCR. The resulting DNA fragments were digested with SpeI and NcoI and cloned between the same sites in broad-host-range plasmid pBBR1MCS-3-lacZ (33). Constructs were conjugatively transferred from recombinant E. coli S17-1 to R. eutropha H16, and transconjugants were selected on mineral salts medium supplemented with 0.2% fructose and 15 μg ml⁻¹ tetracycline. All β-galactosidase activity assays were performed in triplicate, and the mean results were calculated. The strain of interest was grown in two subsequent seed cultures in NB medium supplemented with 15 μg ml⁻¹ tetracycline at 30°C and at 150 rpm. The main culture was grown in 96-well plates (1-ml volume/well) in NB medium. A 100-μl volume of cells diluted 1:5 was mixed with 900 μl of Z buffer (10.7 g liter⁻¹ Na₂HPO₄ · 2H₂O, 5.5 g liter⁻¹ NaH₂PO₄, 0.75 g liter⁻¹ KCl, 0.246 g liter⁻¹ MgSO₄ · 7H₂O, 2.7 ml liter⁻¹ β-mercaptoethanol, pH 7) before 20 μl of 0.1% (wt/vol) SDS and 20 μl of toluene were added for permeabilization. The toluene was evaporated by shaking at 150 rpm and 37°C for 40 min. A 30-μl volume of o-nitrophenyl-β-d-galactopyranoside solution (4 mg ml⁻¹ in Z buffer without β-mercaptoethanol) was added to 150 μl of permeabilized cells, and activity was measured by recording the A₄₂₀ for 40 min with a microplate reader (all steps with prewarmed [28°C] solutions). β-Galactosidase activities in Miller units were calculated according to reference 34. Additionally, 5 μl of the second seed culture was spotted onto NB agar containing 15 μg ml⁻¹ tetracycline and 40 μg 5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside (X-Gal) ml⁻¹. The colonies and the development of a blue color were observed for 2 to 8 days at 30°C.

Other methods.

Quantitative analysis of PHA content was done by gas chromatography after acid methanolysis of lyophilized cells according to reference 35. DNA manipulation and construction of plasmids were done by standard molecular biological methods and according to the supplier's instructions. Protein concentrations were routinely determined by the Bradford method (36) or by the bicinchoninic acid method in the cases of purified PhaZd1 and PhaZd2. Polyacrylamide gel electrophoresis was performed under denaturing (SDS) and reducing (β-mercaptoethanol) conditions. Gels were stained with Coomassie brilliant blue G250 or with silver. Western blot analysis for confirmation of EYFP-PHB depolymerase fusion products was performed with commercial polyclonal antibodies raised against green fluorescent protein. Fluorescence microscopy of Nile red-stained cells was done as described previously (37), except that ethanol was replaced with dimethyl sulfoxide (DMSO) as a cosolvent in some experiments (1 μg Nile red/ml DMSO).

RESULTS

Comparison of PhaZd (H16_B2073, PhaZ6) and H16_B2401 (PhaZ7).

PhaZd had been previously described as a highly active iPHB depolymerase by Abe and coworkers in T. Saito's lab (23). Although the authors described a slightly increased PHB content of a ΔphaZd1 mutant during the PHB accumulation phase, no detectable effect on the mobilization of PHB was observed and therefore the physiological function of PhaZd in the mobilization of PHB remained obscure. The R. eutropha H16 genome has another putative iPHB depolymerase (H16_B2401, PhaZ7) that is 33% identical and 37% similar to PhaZd (H16_B2073, PhaZ6). Both genes are singly expressed and are not part of an operon, as they are both framed by predicted coding regions on the other DNA strand. We suggest that H16_B2073 (PhaZd, PhaZ6) be designated PhaZd1 and that H16_B2401 (PhaZ7) be designated PhaZd2 (Table 1). No experimental data on the function, activity, or other properties of PhaZd2 are available. We therefore performed a thorough characterization of both PHB depolymerases.

Constitutive expression of PhaZd1 and PhaZd2 prevents accumulation of PHB.

We cloned the phaZd1 and phaZd2 genes each into broad-host-range vector pBBR1MCS2 under the control of the constitutive P_phaC promoter, transformed the constructs to E. coli S17-1, and then transferred the plasmids to R. eutropha H16 via conjugation. Accumulation of PHB by the resulting transconjugants was investigated by cultivation of the respective strains under PHB-permissive conditions, i.e., growth on NB medium that had been supplemented with 0.2% (wt/vol) gluconate to increase the C-to-N ratio. No or only a trace amount of PHB was accumulated by R. eutropha clones expressing PhaZd1 or PhaZd2 (Fig. 1), while the wild type harboring the empty pBBR1MCS2 plasmid alone accumulated considerable amounts of PHB, as revealed by staining of the cells with Nile red. We conclude that PhaZd1 and PhaZd2 have substantial PHB depolymerase activity in vivo that leads to instant intracellular hydrolysis of the accumulated PHB. PhaZd1 and PhaZd2 are both serine hydrolases with a lipase box sequence (GMS₁₉₀AG and GLS₁₉₃AG for PhaZd1 and PhaZd2, respectively) at the predicted active site. To verify that the enzyme activities of the cloned PhaZd1 and PhaZd2 depolymerases were indeed responsible for the PHB-negative phenotype of the strains, we exchanged the active-site serines of the lipase boxes of PhaZd1 (GMS₁₉₀AG) and PhaZd2 (GMS₁₉₃AG) for alanines and repeated the growth experiments. Crude extracts of cells expressing mutated PhaZd1(S190A) or PhaZd2(S193A) did not show any PHB depolymerase activity with nPHB granules, whereas cell extracts of clones with expression of the intact PhaZd1 or PhaZd depolymerase showed high PHB depolymerase activity (data not shown). These results confirmed that the predicted active-site serines (Ser190 and Ser193 in PhaZd1 and PhaZd2, respectively) are essential for PHB depolymerase activity.

FIG 1.

Accumulation of PHB in R. eutropha strains overexpressing PhaZd1, PhaZd2, catalytically inactive PhaZd1(S190A), or catalytically inactive PhaZd2(S193A) from phaC promoters. Samples were taken after 4 h of growth in NB medium with 0.2% sodium gluconate and examined by phase-contrast (left) and fluorescence (middle, right) microscopy after staining with Nile red. Bar, 2 μm. The empty plasmid is pBBR1MCS2-P_phaC. WT, wild type.

Remarkably, R. eutropha expressing the inactive variants of PhaZd1 or PhaZd2 accumulated PHB at levels comparable to those of the wild type and confirmed that the PHB-negative phenotype of R. eutropha was caused by expression of the intact PhaZd1 and PhaZd2 proteins (Fig. 1). Notably, cells expressing PhaZd2 tended to be longer than wild-type cells and grew more slowly than wild-type or PhaZd1-expressing cells.

Purification of PhaZd1 and PhaZd2 and determination of PHB depolymerase activity in vitro.

PhaZd1 and PhaZd2 were expressed as N-terminally hexahistidine-tagged proteins in recombinant E. coli and purified by standard affinity chromatography on Ni-agarose columns. Both proteins were subjected to PHB depolymerase assay with (amorphous) nPHB and denatured (crystalline) granules (dPHB) (29–31) that had been isolated from R. eutropha. The purified PHB depolymerases hydrolyzed nPHB at specific activities of 12,000 (PhaZd1) and 850 (PhaZd2) U/mg. No activity of either the two proteins was detected with dPHB. This result was in agreement with the postulated functions of PhaZd1 and PhaZd2 as iPHB depolymerases and confirmed previously published data for PhaZd1 (23).

Biochemical characterization of PhaZd1 and PhaZd2.

Purified PhaZd1 showed considerably higher nPHB depolymerase activity than PhaZd2 in 0.1 M Tris-HCl buffer, pH 8.5. PhaZd1 hydrolyzed a PHB suspension with an OD₆₅₀ of 1 almost completely in only 20 min (Fig. 2A), similar to extracellular PHB depolymerase PhaZ7 of Paucimonas lemoignei (29, 38). Notably, a lag phase of ∼5 min was observed until maximal hydrolysis occurred. In contrast, PhaZd2 had low specific PHB depolymerase activity when the same buffer conditions were used (850 U/mg). However, when the experiments were repeated at a higher ionic strength, such as in buffer with 0.5 M Tris-HCl at pH 8.5 (instead of 0.1 M Tris) or when nPHB granules from an R. eutropha ΔphaP1 or Δ(phaP1-phaP5) mutant strain were used, the PHB depolymerase activity of PhaZd2 was strongly increased up to 21-fold (16,800 U/mg) (Fig. 2A; Table 4). It is known that some PHB depolymerases have higher nPHB depolymerase activity if a part of the proteinaceous surface layer of nPHB granules is removed by treatment with trypsin (39–43). We therefore tested the influence of trypsin treatment of nPHB granules on the activities of PhaZd1 and PhaZd2. PhaZd1 hydrolyzed trypsin-treated nPHB granules with almost double activity (24,000 U/mg), and PhaZd2 revealed 15-fold higher activity than untreated nPHB granules (13,000 U/mg) in 0.1 M Tris-HCl buffer, pH 8.5 (Fig. 2B; Table 4). Remarkably, no lag phase was observed for the hydrolysis of trypsin-activated granules by either PhaZd1 or PhaZd2 (Fig. 2B). Ionic and nonionic detergents completely inhibited the depolymerase activity of PhaZd1 and PhaZd2 (Table 4). Somewhat surprising results were obtained when the dependence of PhaZd1 or PhaZd2 on salts was determined. Low concentrations (1 to 5 mM) of divalent cations (MgCl₂ or CaCl₂) increased the depolymerase activity of PhaZd1 2- to 4-fold and that of PhaZd2 2- to 11-fold (Fig. 2C; Table 4). The activity of PhaZd1 at 10 mM divalent salts was lower than that at 5 mM, but the activity of PhaZd2 was still higher (2- to 3-fold higher than that at 5 mM and 30- to 37-fold higher than that with buffer without salts). At 100 mM MgCl₂ or CaCl₂, PhaZd1 was almost inactive and the activity of PhaZd2 was reduced to the values determined in the absence of salts (Fig. 2D). Monovalent ions (KCl) had less drastic effects at 10 or 100 mM. In conclusion, PhaZd1 and PhaZd2 both required salts (1 to 10 mM) for maximal depolymerase activity but were inhibited by a high salt concentration (100 mM). The activation of PhaZd2 depolymerase activity by salts was significantly higher than that of PhaZd1. Remarkably, salts affected the activities of PhaZd1 and PhaZd2 with nPHB granules as the substrate but had no significant influence on the hydrolysis of p-nitrophenyl esters such as p-nitrophenylbutyrate (data not shown). The effects of other substances on the activities of PhaZd1 and PhaZd2 are summarized in Table 4.

FIG 2.

Hydrolysis of nPHB granules by purified PhaZd1 and PhaZd2. PHB depolymerase activities were determined in the presence of different concentrations of buffer or ions as indicated. If not stated otherwise, nPHB granules without trypsin treatment and 100 mM Tris-HCl buffer, pH 8.5, were used.

TABLE 4.

Specific PHB depolymerase activities of purified PhaZd1 and PhaZd2^a

Condition	Relative activity (%)
	PhaZd1	PhaZd2	PhaZd1	PhaZd2
Substrate
n-PHB in:
0.1 M Tris	100	100
0.5 M Tris			40 ± 1	2,100 ± 56
n-PHB + trypsine in:
0.1 M Tris	190 ± 1	1,540 ± 34
0.5 M Tris			NDb	ND
d-PHB in:
0.1 M Tris	<4	<4
0.5 M Tris			ND	ND
Detergents
Triton X-100 at:
0.01%	<4	<4
0.1%			<4	<4
SDS at:
0.01%	<4	<4
0.1%			<4	<4
Salts
MgCl2 at:
1 mM	210 ± 5	690 ± 23
5 mM			250 ± 15	870 ± 17
CaCl2 at:
1 mM	380 ± 24	260 ± 54
5 mM			380 ± 48	1,100 ± 620
MgCl2 at:
10 mM	140 ± 10	3,700 ± 320
100 mM			<4	140 ± 2
CaCl2 at:
10 mM	360 ± 17	3,000 ± 760
100 mM			<4	110 ± 8
KCl at:
10 mM	100 ± 1	110 ± 12
100 mM			190 ± 10	330 ± 4
Chelators
EDTA at:
1 mM	120 ± 3	130 ± 4
5 mM			260 ± 4	180 ± 11
EDTA + MgCl2 (2 mM) at:
1 mM	160 ± 5	100 ± 33
5 mM			150 ± 1	29 ± 1
EDTA + CaCl2 (2 mM) at:
1 mM	130 ± 2	260 ± 16
5 mM			110 ± 4	190 ± 1
Redox-active agents
DTEc,e at:
1 mM	17 ± 1	100 ± 6
4 mM			10 ± 1	41 ± 3
DTTd,e at:
1 mM	18 ± 2	84 ± 8
4 mM			11 ± 1	83 ± 9
Alkylating agent iodoacetamidee at:
1 mM	61 ± 2	83 ± 6
10 mM			63 ± 3	76 ± 7
Serine hydrolase inhibitor PMSFe,f at:
1 mM	31 ± 1	100 ± 9
10 mM			12 ± 1	64 ± 6

^aAll experiments were performed by turbidometric assay with native R. eutropha H16 PHB granules suspended in 0.1 M Tris-HCl, pH 8.5, if not stated otherwise. In some experiments, 100% activities correspond to 12,000 U/mg (PhaZd1) and 850 U/mg (PhaZd2) with nPHB granules and 0.1 M Tris-HCl, pH 8.5.

^bND, not determined.

^cDTE, dithioerythritol.

^dDTT, dithiothreitol.

^eTrypsinized nPHB granules were used.

^fPMSF, phenylmethylsulfonyl fluoride.

The PHB depolymerase activities of PhaZd1 and PhaZd2 were assayed with the sensitive pH stat assay. About 83 and 190 μmol acid/min/mg protein were determined for PhaZd1 and PhaZd2, respectively, when trypsinized nPHB granules were used as the substrate. The hydrolysis rates were linear for at least 20 min (data not shown). To determine the nature of the products of the PHB depolymerase reactions, nPHB granules were hydrolyzed by PhaZd1 and PhaZd2 until the opaque nPHB granule suspension became clear. Subsequently, the solubilized products were identified by HPLC assay after derivatization with BPB with the derivatized 3-hydroxybutyrate monomer (R1) and dimer (R2) as standards. The BPB method increases the sensitivity of the detection method without hydrolyzing any potentially produced 3-hydroxybutyrate oligomers (30, 31). Both PHB depolymerases produced a mixture of the 3-hydroxybutyrate dimer and the monomer in a 3:1 ratio in the case of PhaZd1 and in a 4:1 ratio in the case of PhaZd2 (Fig. 3), similar to previous data for PhaZd1 (23). Trimers and higher oligomers were not detected if the reaction was run to completion.

FIG 3.

HPLC chromatogram of products of PhaZd1- and PhaZd2-catalyzed hydrolysis of nPHB granules (in 1 ml 0.1 M Tris-HCl, pH 8.5; initial OD₆₅₀, ∼1). Soluble products after a 20-min reaction time were derivatized with bromophenacylbromide and separated by HPLC. The elution times of the 3-hydroxybutyrate monomer (R1) and dimer (R2) derivatives were 13.0 and 15.4 min, respectively. Note that PhaZd1 and PhaZ2 produce a mixture of 3-hydroxybutyrate monomers and dimers.

Subcellular localization of PhaZd1 and PhaZd2.

If the physiological function of PhaZd1 and PhaZd2 is the depolymerization of accumulated PHB, the enzymes should be able to specifically bind to PHB in vivo, as has been previously described for other iPHB depolymerases (22, 44). Attachment to and detachment from PHB granules of proteins with a PHB depolymerase function could be a way to regulate the accumulation and mobilization of accumulated PHB. We therefore constructed fusions of PhaZd1 and PhaZd2 with EYFP as described in Materials and Methods. EYFP was fused to both the C and N termini of PhaZd1 and PhaZd2, and all four fusions were expressed from the constitutive phaC promoter of broad-host-range plasmid pBBR1MCS2 after conjugative transfer of the plasmids to R. eutropha H16. The fusion proteins were catalytically active, as shown by the high PHB depolymerase activities of the respective cell extracts (data not shown). R. eutropha expressing PhaZd1-EYFP, EYFP-PhaZd1, PhaZd2-EYFP, or EYFP-PhaZd2 was grown on NB medium that had been supplemented with 0.2% sodium gluconate to promote PHB accumulation. Staining of the cells with Nile red revealed that most of the cells contained hardly any PHB in the first 10 h of growth. Only at 12 h and at later stages of growth was a low number of PHB granules detectable in the cells. Apparently, the expressed PHB depolymerase fusion proteins had an in vivo activity similar to that of the proteins without the fusion (Fig. 1) and accumulated PHB was rapidly degraded by the depolymerase. In principle, the same results were obtained when PhaZd1 and PhaZd2 were expressed from an arabinose-inducible promoter (P_BAD). Depending on the time point of arabinose addition, the PHB content of the cells rapidly decreased for both PhaZd1 and PhaZd2 (data not shown). Because of rapid PHB hydrolysis by expressed PhaZd1 and PhaZd2 depolymerases, a potential colocalization of PHB granules with the depolymerases PhaZd1 and PhaZd2 could not be tested. We therefore changed the depolymerase-active sites (GMS₁₉₀AG [PhaZd1] and GLS₁₉₃AG [PhaZd2]) to alanine residues in all four fusion proteins, respectively, and repeated expression with the catalytically inactive EYFP fusions. When the recombinant strains were grown in NB-gluconate medium, fluorescence microscopic investigation showed that the cells contained Nile red-stainable granules comparable in size and number to those of the wild type and indicated that substantial amounts of PHB had been accumulated. Apparently, the inactive PhaZd1 and PhaZd2 fusions could no longer hydrolyze the accumulated PHB. When the subcellular localization of the PhaZd1(S₁₉₀A)-EYFP and EYFP-PhaZd1(S₁₉₀A)-EYFP fusions was determined, fluorescent foci that colocalized with PHB granules were found (Fig. 4), indicating that PhaZd1 has the ability to bind to the PHB granule surface in vivo. However, weak fluorescence was also observed in the cytoplasm of the cells, indicating that not all of the fusion protein was PHB granule bound. Presumably, constitutive expression of the PhaZd1 fusion proteins from the constitutive promoter resulted in the expression of more PHB depolymerase molecules than could be bound to the surface of the PHB granules.

FIG 4.

Fluorescence microscopy of R. eutropha H16 expressing PhaZd1 or inactive PhaZd1(S190A) fused to EYFP (C- and N-terminal fusions) (top left), PhaZd2 or inactive PhaZd2(S193A) fused to EYFP (C- and N-terminal fusions) (top right). Bar, 2 μm. Colocalization of PhaZd1(S-A)-EYFP with PHB is visible as dark structures in the phase-contrast image (bottom left). Blue-colored images and the bottom right images show single cells expressing EYFP-PhaZd2(S193A) at a higher magnification. Note the spindle-like fluorescent structures of EYFP-PhaZd2(S193A). The second image from the right in the bottom row shows a single cell additionally stained with Nile red. Samples were taken after 4 h of growth in NB medium with 0.2% sodium gluconate.

When the same experiment was performed with the inactive C-terminal fusion of PhaZd2 (PhaZd2-EYFP), fluorescence was detected only in the cytoplasm and PHB granules were not labeled (Fig. 4). Occasionally, fluorescent foci were observed that did not, however, colocalize with PHB granules and might represent aggregates or inclusion bodies. An interesting observation was made when we investigated the subcellular localization of the inactive N-terminal fusion of PhaZd2 (EYFP-PhaZd2); in this case, spindle-like fluorescent structures were frequently and reproducibly observed (Fig. 4). The spindles were located in the cell center at the start of growth. At later stages of growth, the spindles were located mostly near one cell pole and often reached a neighboring cell that was still connected to the other cell shortly after cell division (Fig. 4). Colocalization with PHB granules was not observed. When the same experiment was performed with the N-terminal (active) fusion of PhaZd2 (EYFP-PhaZd2), a homogeneous distribution of the fluorescence in the cytoplasm was observed.

In conclusion, both PHB depolymerases PhaZd1 and PhaZd2 were able to hydrolyze PHB in vivo but only for the inactive variant of PhaZd1 could clear colocalization with PHB granules be confirmed in vivo. The meaning of the spindle-like structures of EYFP-PhaZd2 is unknown.

Deletion of phaZd1 and/or phaZd2 has no impact on PHB mobilization.

Abe et al. had previously shown that deletion of phaZd1 led to slightly increased PHB content during the PHB accumulation phase but mobilization of PHB under starvation conditions was not significantly affected (23). We hypothesized that a phenotype of the ΔphaZd1 mutant strain could be suppressed by the other PHB depolymerase (PhaZd2). Our data shown above clearly indicate that PhaZd1 and PhaZd2 have PHB depolymerase activity in vivo and can reduce the iPHB content. However, the experiments were performed with plasmids in which phaZd1 or phaZd2 was expressed from the phaC promoter that is constitutive in R. eutropha or from the arabinose-inducible P_BAD promoter. In order to study the impact of wild-type expression of phaZd1 and phaZd2 on PHB mobilization and to test for a possible suppressive effect of phaZd2 on a phaZd1 deletion, we constructed chromosomal ΔphaZd1 and ΔphaZd2 single-deletion mutants and a ΔphaZd1 ΔphaZd2 double-deletion mutant. The three mutant strains and the wild type were grown in NB-gluconate medium, and ODs and PHB contents were monitored for 48 h. As shown in Fig. 5, the growth curves and PHB contents of all of the mutant strains were practically identical to those of the wild type. Two repetitions of the growth experiment gave no indication of significant differences in the OD values of the four strains during growth. An increase in the PHB content from <10% to the maximum PHB content (∼37%) at ∼12 h and a subsequent decrease to levels below 5% at 48 h were determined for all of the strains. Notably, all of the mutant strains were able to mobilize PHB in the stationary growth phase in a manner similar to that of the wild type. A slightly higher level of PHB content was observed for the ΔphaZd1 mutant and the double mutant at 4 and 8 h of growth, respectively, compared to the wild type. However, these slight differences vanished within the PHB accumulation phase of 12 h. Our data show that neither phaZd1 nor phaZd2 has a significant effect on PHB accumulation and neither of them is required for PHB mobilization.

FIG 5.

Growth and PHB contents of WT and ΔphaZd1, ΔphaZd2, and ΔphaZd1 ΔphaZd2 mutant R. eutropha. Cells were cultured in NB medium with 0.2% sodium gluconate. The values in these graphs are averages of two biological and two technical replicate experiments. Error bars indicate standard deviations.

PhaZd1 and PhaZd2 are not expressed.

Our results clearly showed that PhaZd1 and PhaZd2 have high PHB depolymerase activity in vitro and in vivo when they are overexpressed from plasmids but apparently have little effect on PHB content in wild-type cells and do not detectably contribute to mobilization of PHB during the stationary growth phase. To assess the physiological function of PhaZd1 and PhaZd2, we determined the expression of the genes that encode them by constructing translational lacZ fusions. R. eutropha cells harboring the promoter regions of phaZd1 and phaZd2 in translational fusions with lacZ were spotted onto solid NB-gluconate agar that had been supplemented with X-Gal. No significant color development was detected within 3 days of incubation at 30°C for phaZd1 and phaZd2 fusions (Fig. 6). In contrast, a lacZ fusion under the control of the phaC promoter turned blue within 24 h. The same results were obtained when the cells were grown in liquid NB-gluconate medium and tested for β-galactosidase activity after 8 h (Fig. 6). The activities of PhaZd1 and PhaZd2 were in the range of the negative control or only a little higher (∼25 Miller units). These data are in line with the of results proteome analysis, in which PhaZd1 and PhaZd2 were not detected in the PHB granule fraction of an NB-gluconate culture (5). We conclude that neither PhaZd1 nor PhaZd2 is significantly expressed during growth on NB-gluconate medium.

FIG 6.

Analysis of phaZd1 and phaZ2 transcription in R. eutropha. Cells harboring a translational fusion of phaZd1 or phaZd2 with lacZ were grown in liquid culture in NB or NB-gluconate medium for 8 h or on solid NB medium for 3 days at 30°C. A promoterless fusion and the constitutive phaC promoter of the phaCAB operon were used as negative and positive controls, respectively. Experiments were performed in triplicate.

DISCUSSION

In this study, we thoroughly characterized two PHB depolymerase proteins, PhaZd1 and PhaZd2, both in vitro and in vivo. Both enzymes have unusually high PHB depolymerase activities for iPHB depolymerase. The only known iPHB depolymerases with comparably high specific activities were those of Bacillus thuringiensis and Bacillus megaterium (42, 43). However, the PHB depolymerase activities strongly depended on the ionic strength of the buffer and on the presence or absence of PHB granule-associated proteins. Especially PhaZd2 showed little activity in 0.1 M Tris or in the absence of salts (Table 4). The buffer composition had a high impact only on PHB depolymerase activity and not on the hydrolysis of p-nitrophenylbutyrate. This indicated that the integrity of the PHB granules was altered by the salts. We therefore assume that the variation in the PHB depolymerase activities is more likely caused by salt effects on the substrate (nPHB granules) and apparently does not reflect a true dependency of the two depolymerase proteins on salts. Since constitutive expression of PhaZd1 and/or PhaZd2 drastically reduced the PHB content of the cells, we assume that the in vivo salt concentrations allow high in vivo activity of PhaZd1 and PhaZd2.

For a potential PHB depolymerase to be classified as a physiological iPHB depolymerase, it must meet four criteria. (i) The candidate protein must have PHB depolymerase activity with nPHB granules, (ii) the depolymerase should be able to bind to PHB in vivo, (iii) deletion of the depolymerase gene should lead to a detectable reduction in PHB mobilization, and (iv) the respective PHB depolymerase must be expressed during a PHB accumulation and mobilization cycle. PhaZd1 and PhaZd2 surely meet the first criterion, and at least PhaZd1 is able to bind to PHB granules in vivo. Moreover, both PHB depolymerases drastically reduced the iPHB content when they were artificially expressed. However, deletion of phaZd1, phaZd2, or both genes had no effect on PHB mobilization and none of the depolymerases was detectably expressed during growth on NB-gluconate medium. Therefore, we have to conclude that despite high PHB depolymerase activity in vitro and despite a strong reduction of the PHB contents in vivo after artificial depolymerase expression, PhaZd1 and PhaZd2 apparently have no physiological function in PHB metabolism during growth of R. eutropha on NB-gluconate medium. We cannot exclude a possible function of PhaZd1 and PhaZd2 in PHB metabolism under conditions in which PhaZd1 and/or PhaZd2 are significantly expressed. However, these conditions have not yet been identified and even the faint PhaZd1 band that was previously detected in Western blot assays of R. eutropha cell extracts by Abe and coworkers apparently was not sufficient to significantly reduce the PHB content in comparison to that of a ΔphaZd1 mutant (23). One should also consider the possibility that an enzyme with depolymerase activity on amorphous PHB could be involved in the hydrolysis of complex PHB (cPHB) or PHBylated proteins (oligo-PHB). cPHB and oligo-PHB are present in small amounts in all prokaryotic and eukaryotic organisms that have been looked at and can have important functions, e.g., in the modulation of enzyme activities and/or channel activities (45–47).