Characterization of a Novel Subgroup of Extracellular Medium-Chain-Length Polyhydroxyalkanoate Depolymerases from Actinobacteria

Abstract

Nineteen medium-chain-length (mcl) poly(3-hydroxyalkanoate) (PHA)-degrading microorganisms were isolated from natural sources. From them, seven Gram-positive and three Gram-negative bacteria were identified. The ability of these microorganisms to hydrolyze other biodegradable plastics, such as short-chain-length (scl) PHA, poly(ε-caprolactone) (PCL), poly(ethylene succinate) (PES), and poly(l-lactide) (PLA), has been studied. On the basis of the great ability to degrade different polyesters, Streptomyces roseolus SL3 was selected, and its extracellular depolymerase was biochemically characterized. The enzyme consisted of one polypeptide chain of 28 kDa with a pI value of 5.2. Its maximum activity was observed at pH 9.5 with chromogenic substrates. The purified enzyme hydrolyzed mcl PHA and PCL but not scl PHA, PES, and PLA. Moreover, the mcl PHA depolymerase can hydrolyze various substrates for esterases, such as tributyrin and p-nitrophenyl (pNP)-alkanoates, with its maximum activity being measured with pNP-octanoate. Interestingly, when poly(3-hydroxyoctanoate-co-3-hydroxyhexanoate [11%]) was used as the substrate, the main hydrolysis product was the monomer (R)-3-hydroxyoctanoate. In addition, the genes of several Actinobacteria strains, including S. roseolus SL3, were identified on the basis of the peptide de novo sequencing of the Streptomyces venezuelae SO1 mcl PHA depolymerase by tandem mass spectrometry. These enzymes did not show significant similarity to mcl PHA depolymerases characterized previously. Our results suggest that these distinct enzymes might represent a new subgroup of mcl PHA depolymerases.

INTRODUCTION

Biodegradability of polymers has drawn much attention as a solution to problems concerning the global environment and biomedical technologies. Several aliphatic polyesters showing properties comparable to those of conventional plastics have been developed and used as biodegradable plastics, such as poly(3-hydroxyalkanoate) (PHA), poly(ε-caprolactone) (PCL), poly(l-lactide) (PLA), and poly(ethylene succinate) (PES). They can be synthesized from petrochemicals (PES and PCL) or from renewable resources (PLA and PHA) (58). Among these biodegradable plastics, PHA is the only one that is completely synthesized by microorganisms and accumulates intracellularly during unbalanced growth conditions (30). Additionally, PHA is suitable for a broad range of applications in medicine, the pharmaceutical industry, and industry due to its biocompatibility and biodegradability (2). Moreover, all of the PHA monomers are enantiomerically pure and in the R configuration (3, 40, 44). More than 150 hydroxyalkanoic acids (HAs) have been identified as constituents of these microbial polyesters (6, 57). Interestingly, these monomers are valuable intermediates that can be used as starting materials for the synthesis of antibiotics, vitamins, flavors, and pheromones (1). Since chiral (R)-HAs are normally difficult to synthesize by chemical means (2), the study of enzymatic PHA hydrolysis has attracted much attention.

The ability to degrade extracellular PHA in the environment depends on the release of extracellular PHA depolymerases (17) that could be specific for either short-chain-length (scl) PHA (3 to 5 carbon atoms) (EC 3.1.1.75) or medium-chain-length (mcl) PHA (6 to 14 carbon atoms) (EC 3.1.1.76) (17). Depending on the depolymerase activity, the end products are only monomers, both monomers and dimers, or a mixture of oligomers as a result of the enzymatic PHA degradation (17).

Extracellular PHA depolymerase-producing microorganisms are widely distributed and have been isolated from various environments (32, 51). Currently, very few mcl PHA depolymerases have been characterized in comparison to the number of scl PHA depolymerases studied (26). To date, most of the mcl PHA depolymerases reported belong to Gram-negative bacteria, predominantly Pseudomonas species (22, 31). The poly(3-hydroxyoctanoate) depolymerase from Pseudomonas fluorescens GK13 (PhaZ_PflGK13) was the first mcl PHA depolymerase studied in detail at the molecular level (49–51). Additionally, several biotechnological applications of this enzyme have been reported, including the construction of fusion proteins with affinity to mcl PHAs (13), the production of (R)-3HAs (8), and the synthesis of polyesters (48). Thus, this enzyme is considered the prototype enzyme of extracellular mcl PHA depolymerases. In general, these enzymes consist of a signal peptide, an N-terminal substrate binding domain, and a C-terminal catalytic domain (15, 22). In a recent study, the identification of a significantly different mcl PHA depolymerase gene from the thermophilic bacterium Thermus thermophilus HB8 has been reported (36). Recently, the isolation and identification of Streptomyces venezuelae SO1 as a novel mcl PHA depolymerase (PhaZ_SveSO1) producer have been reported by our group (47). However, the molecular characteristics of the genes encoding mcl PHA depolymerases from Streptomyces origins have not yet been cleared. In this paper, we report the isolation of several novel extracellular mcl PHA-degrading microorganisms, predominantly Streptomyces species. Two of the isolates, SL3 and SO2, have been identified to be Streptomyces roseolus and Streptomyces omiyaensis, respectively. Furthermore, the mcl PHA depolymerase from S. roseolus SL3 (PhaZ_SroSL3) has been biochemically characterized. In addition, we provide for the first time information about the primary structure of the mcl PHA depolymerases from Streptomyces bacteria.

MATERIALS AND METHODS

Chemicals.

Poly(3-hydroxyoctanoate-co-3-hydroxyhexanoate [11%]) [P(3HO) or mcl PHA] was supplied by Biopolis, S.A. (Valencia, Spain), and CPI (Newcastle, United Kingdom). Accurel MP-1000 was purchased from Membrana GmbH (Obenburg, Germany). Poly(3-hydroxypropionic acid) [P(3HP)] was donated by CIBA (Manchester, United Kingdom). Chromatography media were obtained from GE Healthcare (Uppsala, Sweden). Molecular weight standards, p-nitrophenyl (pNP)-alkanoates, poly(3-hydroxybutyric acid) [P(3HB)], poly(3-hydroxybutyric acid-co-3-hydroxyvaleric acid) {P(3HB-HV[12%])}, PCL, PES, and PLA were obtained from Sigma-Aldrich (St. Louis, MO). All other chemicals were supplied by Merck (Darmstadt, Germany).

Preparation of biopolymer suspensions.

Latex suspensions of PCL and P(3HO) were prepared as described by Schirmer and Jendrossek (50). In the case of PES and PLA, 4 volumes of water were poured into 1 volume of polymer suspension in methylene chloride with stirring. The suspensions were emulsified by an ultrasonic homogenizer, and the solvent was then evaporated. P(3HB), P(3HP), and P(3HB-HV[12%]) suspensions of similar concentration (10 mg/ml) were prepared by dispersing each polymer powder in water by ultrasonic treatment.

Isolation and identification of mcl PHA-degrading microorganisms.

Several mcl PHA-degrading bacterial strains were isolated in our laboratory from natural environmental samples (soil, sludge, and water) taken at different places on the Leioa campus of the University of the Basque Country, Vizcaya, Spain. Serial dilutions of the homogenized samples were spread on P(3HO)-mineral agar plates consisting of P(3HO) latex-covered petri plates with mineral medium such as M9 medium (45) and E medium (24). The plates were incubated for 2 to 3 days at 30°C. Those strains which showed clearing of the P(3HO) latex were selected and isolated.

The bacteria were identified by sequence analysis of the 16S rRNA gene.

For further identification, the cultural, morphological, and physiological characteristics of the SL3 and SO2 strains were obtained by following the methods given in the International Streptomyces Project (ISP) (54). Aerial spore mass color and substrate mycelium color were recorded using Inter-Society Color Council, National Bureau of Standards (NBS), color name charts (18), after incubation for 20 days at 30°C in oatmeal agar medium (ISP medium 3). Morphological observations of spores and mycelia were made by light microscopy (Nikon Eclipse 50i A light microscope) and scanning electron microscopy (model JEOL 6100 scanning electron microscope). The carbon utilization test was performed in ISP9 medium with the addition of d-glucose (positive control), l-arabinose, sucrose, d-xylose, myo-inositol, d-mannitol, d-fructose, rhamnose, or raffinose and in the absence of a carbon source (negative control), as described by Shirling and Gottlieb (54).

Strains SL3 and SO2 were identified from ISP (55, 56). The identified mcl PHA-degrading strains have been deposited in the Spanish Type Culture Collection (CECT, Valencia, Spain; www.cect.org) as Streptomyces roseolus SL3 CECT 7919 and Streptomyces omiyaensis SO2 CECT 7923.

Microorganisms and growth conditions.

The following microorganisms were used in this study: S. venezuelae SO1 CECT 7920, S. omiyaensis SO2 CECT 7923, and S. roseolus SL3 CECT 7919. All other strains are listed in Table 1. Polymer-degrading bacteria were routinely grown in solid M9 mineral medium (45) containing 1.5% (wt/vol) agar with the carbon sources indicated in the text. For enzyme production, S. roseolus SL3 and S. venezuelae SO1 cells were grown at 30°C in 250-ml Erlenmeyer flasks containing 100 ml of mineral medium supplemented with a film (0.15 g) of P(3HO) as the sole carbon and energy source, as described by Santos et al. (47). The strains were maintained as frozen spore suspensions in 15% (vol/vol) glycerol at −20°C, as described by Kieser et al. (19).

Table 1.

Microbial strains isolated, their closest relative bacteria based on 16S rRNA analysis, and their polyester-degrading abilities

Strain	Closest relative (GenBank accession no.)	Category	% similarity	Ability to degrade polyestera
				P(3HO)	P(3HP)	P(3HB)	P(HB-HV[12%])	PCL	PLA	PES
SL1	Pseudomonas alcaligenes (Z76653)	Gammaproteobacteria	99.6	+++	ND	−	−	+	−	−
SL2	Streptomyces atratus (DQ026638)	Actinobacteria	99.6	+	+	+	+	−	−	−
SL3	Streptomyces roseolus (AB184168)	Actinobacteria	99.8	+	++	+++	+++	+	−	−
SL6	Stenotrophomonas maltophilia (HQ406762.1)	Gammaproteobacteria	99.2	+	ND	−	ND	−	−	−
SL11	Streptomyces anulatus (AB184875)	Actinobacteria	99.5	+	+	+	+	++	−	−
SL15	Streptomyces beijiangensis (AB249973)	Actinobacteria	99.4	+++	ND	++	++	+	−	−
SO2	Streptomyces omiyaensis (AB184411)	Actinobacteria	99.5	++	+	++	++	+++	−	−
W1	Pseudomonas beteli (DQ299947.1)	Gammaproteobacteria	99.0	+	ND	−	ND	−	−	−
W2	Rhodococcus equi (X80614)	Actinobacteria	99.6	+	ND	−	−	−	−	−
W3	Streptomyces pulveraceus (AB184808)	Actinobacteria	99.8	+	ND	+	+	+	−	−
GK13	Pseudomonas fluorescens	Gammaproteobacteria		+++	−	−	−	ND	−	−

^aThe ability to degrade different polyesters was determined by clear-zone formation around the colony on the opaque plates after 2 to 3 days of growth at 30°C. Symbols and abbreviations: −, no clearing zone; +, small clearing zone; ++, medium clearing zone; +++, large clearing zone; P(3HP), poly(3-hydroxypropionate); P(3HB), poly(3-hydroxybutyrate); P(HB-HV[12%]), poly(3-hydroxybutyrate-co-3-hydroxyvalerate); PCL, poly-ε-caprolactone; PLA, poly(l-lactide); PES, poly(ethylene succinate); ND, not determined.

For the isolation of genomic DNA, the bacteria were grown for 3 days at 30°C in 250-ml Erlenmeyer flasks containing 100 ml of S-YEME medium (19) in an orbital incubator shaker at 250 rpm. Cultures were harvested at 4°C by centrifugation (10,000 × g for 20 min). The resulting pellet was used for DNA extraction. Genomic DNAs of Streptomyces strains were isolated as described by Kieser et al. (19).

Purification of mcl PHA depolymerases.

S. roseolus SL3 and S. venezuelae SO1 cells were grown in 2-liter Erlenmeyer flasks containing 800 ml of mineral medium (23) supplemented with a film (1.2 g) of P(3HO). Flasks were inoculated with 100 ml of a culture of mineral medium supplemented with glucose (0.4%, wt/vol) that had been grown for 72 h, and the cultures were grown for 3 days in an orbital incubator shaker at 250 rpm and 30°C. Cells were harvested by filtration, and the enzyme present in the supernatant was purified by adsorption onto porous polypropylene (Accurel MP-1000) as reported by Gangoiti et al. (8).

Enzyme assays.

Esterase activity was assayed using several pNP-alkanoates as the substrate (8). Blanks without enzyme were performed to determine spontaneous hydrolysis not due to enzymatic activity. One unit of esterase activity was the amount of enzyme that released 1 μmol of p-nitrophenol per min under standard conditions. The extinction coefficient (ε) for pNP at pH 9.5 was determined to be 16.635 mM⁻¹ · cm⁻¹.

Qualitative estimation of the hydrolytic activity of mcl PHA depolymerase from S. roseolus SL3 toward different polymers was performed by a drop test on indicator plates (14). Briefly, 5 ml of a 1% (wt/vol) polymer emulsion was mixed with 5 ml of 1% (wt/vol) agarose in 200 mM Tris-HCl buffer, pH 8.5, and poured on a glass plate. Samples (20 μl) were loaded in 5-mm-diameter holes made in the gel and incubated at 30°C for 24 h. Similarly, qualitative determination of esterase activity was performed on agarose plates using tributyrin as the substrate, as described by Gandolfi et al. (7). The diameters of the resulting clearing zones were semiquantitatively correlated with the enzyme activity.

Identification of hydrolysis products of P(3HO).

The hydrolysis products from the P(3HO) substrate catalyzed by PhaZ_SroSL3 were identified. For this purpose, reaction mixtures containing 250 μg of P(3HO) latex in 20 mM Tris-HCl buffer, pH 8.0, and 50 μg of the purified enzyme were incubated (in 2-ml tubes) at 30°C and in an orbital shaker at 160 rpm for various time intervals (3 h, 24 h, 48 h, and 72 h). The enzymatic reaction was stopped by incubating the tubes for 5 min at 100°C, and then the tubes were centrifuged at 4°C for 60 min at 14,000 × g. The degradation products were isolated from supernatants and derivatized using bromophenacyl bromide (BPB), as described by Gebauer and Jendrossek (11). The detection and quantification of the hydrolysis products were performed by high-pressure liquid chromatography (HPLC)-photo diode array (PDA), and the identity of the 3-hydroxyoctanoic acid (3-HO) oligomer peaks detected at 254 nm were determined by HPLC-mass spectrometry (MS) (8). The peak of 26.1 min corresponds to unreacted BPB.

Determination of the N-terminal protein sequences.

The pure PhaZ_SroSL3 and PhaZ_SveSO1 were electroblotted from an SDS-polyacrylamide gel to a polyvinylidene difluoride (PVDF) membrane (Biotrace; Pall Corporation). Edman degradation analysis was carried out in the Proteomics and Bioinformatics facility from UAB, a member of the ProteoRed network.

Identification of mcl PHA depolymerases genes.

In order to determine the mcl PHA depolymerase sequences from Streptomyces, PhaZ_SveSO1 was subjected to de novo peptide sequence analysis by mass spectrometry. For this purpose, a Coomassie blue-stained gel spot corresponding to the enzyme was excised, washed, reduced with dithiothreitol (DTT), and alkylated with iodoacetamide. The in-gel digest with trypsin was carried out at 37°C. The resultant peptides were analyzed by matrix-assisted laser desorption ionization (MALDI)–tandem time of flight (TOF/TOF) mass spectrometry (4700 proteomics analyzer; Applied Biosystems) in MS and MS/MS modes. To enhance the quality of the tandem MS/MS spectra for the de novo-sequencing, N-terminal chemical modification using 4-sulfophenyl isothiocyanate (SPITC) was carried out at 55°C for 30 min (10). The N-terminal derivatized peptides were desalted and concentrated using μZip-Tips C₁₈ (Millipore), as described by the manufacturer. The sample was spotted onto the MALDI target plate prespotted with alpha-cyano-4-hydroxycinnamic acid matrix.

Peptide de novo sequencing was carried out manually using the program mMass (http://www.mmass.org/). De novo-derived peptide sequences were combined in one search query and analyzed by MS-BLAST (53). Searches were performed against nonredundant proteins with PAM30MS as the search matrix. The sequences obtained were subjected to multiple alignments employing CLUSTALW (28).

The phaZ_SroSL3, phaZ_SveSO1, and S. omiyaensis SO2 phaZ (phaZ_SomSO2) genes were partially amplified by PCR using chromosomal DNA as the template. The degenerated PCR primers were designed in accordance with the N-terminal protein sequences, considering the codon usage in Streptomyces (19), as well as on the basis of the sequence of the hypothetical protein SVEN_7345 from S. venezuelae ATCC 10712 (GenBank accession no. CCA60631.1). PCR amplifications were performed in a Px2 thermal cycler (Thermo Hybaid, United Kingdom) using the TDPfu program, adjusted for the high G+C content of Streptomyces genomes, and employing Pfu DNA polymerase (Promega) (9). The phaZ_SveSO1 and phaZ_SomSO2 genes were partially amplified using primers VN1 (5′-CGAGGTGGACGTCGACATCGAGG-3′) and A4R (5′-GCGCAGCCACGCCGTGGTCGG-3′), whereas in the case of the phaZ_SroSL3 gene, primers NSL3 (5′-GTSGGSACSGACTGGGACCG-3′) and A4R were used.

DNA fragments (∼600 bp) amplified in each PCR were purified from the agarose bands. DNA sequences were determined by the dideoxy chain termination method (46) with an automated sequencer, DNA analyzer 3730 (Applied Biosystems).

Enzyme analysis.

SDS-PAGE was performed as described by Laemmli (27). Two-dimensional electrophoresis was performed by isoelectric focusing using immobilized pH gradient strips (pH 3 to 10) (first dimension) and SDS-polyacrylamide gel electrophoresis (second dimension). The protein concentration was determined by the method of Peterson (39), using bovine serum albumin as the standard.

Nucleotide sequence accession numbers.

The 16S rRNA gene sequences from isolates obtained in this study were deposited in GenBank under the accession numbers JX305978 to JX305987. The partial sequences of phaZ_SroSL3, phaZ_SveSO1, and phaZ_SomSO2 have been deposited in GenBank under the accession numbers JX305988, JX305989, and JX305990, respectively.

RESULTS

Screening of mcl PHA-degrading bacteria and their ability to hydrolyze other aliphatic polyesters.

Nineteen bacteria able to grow on P(3HO) as the sole source of carbon and energy were isolated from samples of soil, sludge, and water [the first letter(s) in the isolate designations, SO, SL, and W, respectively]. All the isolates produced a clearing zone surrounding the colony within 2 to 3 days of incubation on opaque P(3HO) agar at 30°C. Ten different bacteria were identified from their 16S rRNA gene sequences (Table 1), including three Gram-negative and seven Gram-positive bacteria. The closest relative strain to each isolate has also been included in Table 1. Interestingly, six of these bacteria belonged to the genus Streptomyces.

The isolated bacteria were screened for polymer-degrading capacity using the clear-zone method. None of the Gram-negative bacteria were able to hydrolyze scl PHA. In contrast, all Streptomyces strains showed rapid growth and degradation of scl PHA, as well as PCL (Table 1). However, none of the isolated bacteria were able to hydrolyze PES and PLA.

Characterization of strains SO2 and SL3.

On the basis of their great ability to degrade different polyesters, the mcl PHA-degrading SL3 and SO2 strains, isolated from sludge and soil, respectively, were selected to study the degradation of P(3HO) in detail. SL3 and SO2 are Gram-positive, aerobic, and nonmotile filamentous bacteria with branching vegetative hyphae embedded in the substrate and aerial hyphae bearing spores. The spores of both bacteria show a smooth surface and occur in rectiflexible chains containing more than 10 spores per chain (see Fig. S1 in the supplemental material). Strain SL3 developed an aerial mycelium in the red color series and a yellow-brownish substrate mycelium. In contrast, the color of the aerial mycelium of strain SO2 on ISP3 was gray, while that of the substrate mycelium was yellow-brownish. These bacteria did not produce diffusible pigments in any of the media tested. SL3 and SO2 strains utilized d-glucose, d-xylose, and rhamnose but were unable to use myo-inositol, d-mannitol, sucrose, and raffinose. SL3 utilized l-arabinose and d-fructose, whereas only a trace of growth was observed in the case of SO2 in the presence of these sugars. On the basis of phylogenetic analyses of the sequence of the 16S rRNA gene and morphological and physiological characteristics, strains SL3 and SO2 were identified as Streptomyces roseolus and Streptomyces omiyaensis, respectively (see Materials and Methods for details). The results shown in Table 1 suggest that S. roseolus SL3 and S. omiyaensis SO2 may synthesize at least two different PHA depolymerases specific for scl or mcl PHAs, as it has been suggested for S. exfoliatus (25).

Biochemical properties of mcl PHA depolymerase from S. roseolus SL3.

The molecular mass of the purified enzyme from S. roseolus SL3, determined by SDS-PAGE analysis, was approximately 28 kDa (Fig. 1). Nondenaturing (ND)-PAGE analyses showed only one enzyme form with an estimated native molecular mass of 28 kDa, indicating that this native enzyme consists of a single polypeptide chain. Besides, the isoelectric point of PhaZ_SroSL3 was about 5.2. The effect of pH on PhaZ_SroSL3 activity was examined at pH values ranging from 6.0 to 12.0, using pNP-octanoate (pNPO) as the substrate. This enzyme exhibited its maximum activity at pH 9.5 and retained more than 60% of this activity over a pH range from 8.0 to 10.5. The N-terminal amino acid sequence of the mature PhaZ_SroSL3 was determined by Edman degradation to be AIPPVGTDWDRP (Fig. 1). This sequence showed at least 50% identity only to that of PhaZ_SspKJ-72 (23). However, it showed low identity to the sequences corresponding to Pseudomonas species (22), indicating that the mcl PHA depolymerases produced by Streptomyces strains may be encoded by a different type of gene.

Fig 1.

SDS-PAGE analysis of the purified mcl PHA depolymerase from S. roseolus SL3. Proteins were separated in a homogeneous 12% (wt/vol) acrylamide gel and revealed by Coomassie brilliant blue R-250. Lane M, molecular mass markers; lane 1, purified enzyme. The N-terminal amino acid sequence of the enzyme, in one-letter code, was determined by Edman degradation.

Substrate specificity of PhaZ_SroSL3 depolymerase.

PhaZ_SroSL3 hydrolyzes mcl PHA and PCL (see Fig. S2 in the supplemental material), forming large clearing zones after 24 h of incubation at 30°C. These results suggest that the depolymerase is able to hydrolyze ester bonds of β- and ω-polyhydroxyalkanoates with a relatively long side chain. However, as expected, no hydrolytic activity was detected with scl PHA, such as P(3HB), P(3HP), and P(3HB-HV[12%]). Moreover, the enzyme was unable to hydrolyze PES and PLA, a poly(alkenedicarboxylate) and a polyester consisting of α-hydroxyalkanoate repeating units, respectively. Besides, PhaZ_SroSL3 showed slight activity toward tributyrin, which is a typical substrate for esterases. However, after 3 days of reaction at 30°C, the enzyme was unable to hydrolyze olive oil, which is a suitable substrate for lipases, indicating that this depolymerase does not show lipase activity (data not shown). Similar substrate specificity was observed with PhaZ_SveSO1 (47) and PhaZ_SomSO2 (unpublished data). As described before (14), the prototype PhaZ_PflGK13 did not hydrolyze scl PHA or PLA. In addition, similar to Streptomyces enzymes, in this study no hydrolytic activity was observed in the presence of PES or olive oil using PhaZ_PflGK13 as the catalyst. However, PhaZ_PflGK13 was not able to hydrolyze tributyrin, and only a small clearing zone was observed in PCL-agarose plates after 24 h at 30°C (data not shown).

Moreover, the esterase activity of PhaZ_SroSL3 was assayed using several pNP-alkanoates as the substrates (Table 2). The enzyme showed the highest esterase activity with pNPO (4.1 U/mg protein), whereas it was unable to hydrolyze pNP-hexadecanoate. On the other hand, its activity with scl pNP-alkanoates was significantly lower. Similar substrate specificities for pNP-alkanoates were described for PhaZ_SspKJ-72 (23) and PhaZ_SveSO1 (47). In contrast, PhaZ_PflGK13 showed maximum esterase activity when pNP-tetradecanoate was used as the substrate (8).

Table 2.

Relative activity of mcl PHA depolymerase of S. roseolus SL3^a

Substrate	Relative activity (%)
pNP-acetate	0.5
pNP-butyrate	3
pNP-valerate	30
pNP-octanoate	100
pNP-decanoate	93
pNP-dodecanoate	87
pNP-hexadecanoate	13
pNP-octadecanoate	4

^aThe pure enzyme was assayed with the indicated chromogenic substrates at a final concentration of 0.3 mM, in all cases. One hundred percent activity corresponded to 4.1 U/mg protein.

Products of extracellular mcl PHA depolymerase from S. roseolus SL3 reaction.

Enzymatic degradation of P(3HO) latex catalyzed by PhaZ_SroSL3 was followed by HPLC-PDA, and the identity of the resulting peaks was determined by HPLC-MS. The composition and relative amounts of the hydrolysis products identified were significantly dependent on the time of hydrolysis used (Fig. 2). Thus, during the early enzymatic period (3 h), trimer 3-HO-HO-HO (∼41%) was the main hydrolysis product detected. However, longer periods of incubation yielded higher concentrations of 3-HO monomers, whereas those of trimers markedly decreased. In fact, after 72 h of enzymatic hydrolysis, 3-HO monomers were the main degradation products (∼57%) obtained, while the trimers were almost absent (Fig. 2). The trimer 3-HO-HO-HX and the dimer 3-HO-HX could also be detected. However, it was difficult to determine the relative amount of the monomer 3-hydroxyhexanoic acid (3-HX) since it showed the same retention time as the unreacted BPB (26.1 min). When P(3HO) was incubated at 30°C for 72 h in the absence of the enzyme, no degradation occurred (data not shown).

Fig 2.

Evolution with the hydrolysis time of the abundance of P(3HO) products catalyzed by the S. roseolus SL3 mcl PHA depolymerase. The products are indicated as follows: HO, HO-HO, and HO-HO-HO, monomer, dimer, and trimer of 3-hydroxyoctanoic acid, respectively; HO-HX and HO-HO-HX, dimer and trimer of 3-hydroxyoctanoic acid and 3-hydroxyhexanoic acid, respectively. The P(3HO) used as a substrate was a copolymer composed of 89% 3-HO and 11% 3-HX.

Identification of mcl PHA depolymerases from Actinobacteria.

To identify the depolymerase-encoding genes from Actinobacteria, the amino acid sequences of four peptides of the purified PhaZ_SveSO1 (47) were determined by de novo sequencing analysis as VDLEHIGSAGHSQGGAAAVNAAIDAR, DSSHLPAVYGEVR, APTTAWIR, and RNWHNGDENAR. MS-BLAST analysis of these peptide sequences revealed a best match with a hypothetical protein from Streptosporangium roseum DSM43021 (34) (GenBank accession no. YP_003340976). The mcl PHA-degrading ability of this bacterium was confirmed by the clear-zone formation method (data not shown). Furthermore, this protein exhibited high amino acid similarity (more than 69%) with the hypothetical proteins of other Actinobacteria species, including two sequences from Rhodococcus erythropolis strains (Table 3).

Table 3.

Similarity between amino acid sequences of Sros_5476 and hypothetical proteins identified from sequenced microbial genomes

Protein	Source/microorganism	GenBank accession no.	% identity/% similarity	Reference
Hypothetical protein Sros_5476	Streptosporangium roseum DSM 43021	YP_003340976		Nolan et al. (34)
Hypothetical acetyl xylan esterase SPW_6174	Streptomyces sp. W007	ZP_09405870	78/85	Unpublished
Hypothetical acetyl xylan esterase SACT1_2252	Streptomyces griseus Xyleb KG-1	ZP_08235685	81/85	Grubbs et al. (12)
Hypothetical protein SGR_2003	Streptomyces griseus subsp. griseus NCBR 13350	YP_001823515	80/84	Ohnishi et al. (35)
Hypothetical protein SVEN_7345	Streptomyces venezuelae ATCC 10712	CCA60631	59/70	Pullan et al. (41)
Hypothetical protein RHOER0001_1689	Rhodococcus erythropolis SK121	ZP_04385744	57/69	Unpublished
Putative hydrolase RER_58150	Rhodococcus erythropolis PR4	YP_002769262	56/69	Sekine et al. (52)

In parallel with this work, the complete genome of S. venezuelae ATCC 10712 was elucidated (41). Although a putative P(3HB) depolymerase (GenBank accession no. CCA60573.1) was annotated, no open reading frame (ORF) encoding a mcl PHA depolymerase was identified. Interestingly, among the amino acid sequences obtained by BLAST analysis, the hypothetical protein SVEN_7345 (GenBank accession no. CCA60631.1) from this bacterium was found (Table 3). On the basis of the DNA sequence of this protein, as well as on the N-terminal sequences determined by Edman degradation (see Materials and Methods for details), DNA fragments of ∼600 bp of the phaZ_SroSL3, phaZ_SveSO1, and phaZ_SomSO2 genes were amplified using their corresponding isolated chromosomal DNAs as the template (Fig. 3). The deduced amino acid sequences shared significantly high similarity (71 to 94%) with all the hypothetical mcl PHA depolymerase proteins identified by de novo sequencing and homology search.

Fig 3.

Alignment of amino acid sequences of mcl PHA depolymerases. Identical amino acids are shaded in gray. The lipase consensus sequence is marked in bold. Amino acids that might constitute a catalytic triad and the possible oxyanion are indicated in bold and by asterisks. The signal peptides predicted by the SignalP (version 4.0) program are boxed. The N-terminal amino acid sequences determined by Edman degradation are indicated in italics. 1, S. venezuelae SO1; 2, S. omiyaensis SO2; 3, S. roseolus SL3; 4, S. venezuelae ATCC 10712 (GenBank accession no. CCA60631); 5, R. erythropolis SK121 (GenBank accession no. ZP_04385744); 6, R. erythropolis PR4 (GenBank accession no. YP_002769262); 7, S. griseus Xyleb KG-1 (GenBank accession no. ZP_08235685); 8, S. griseus subsp. griseus NCBR 13350 (GenBank accession no. YP_001823515); 9, Streptomyces sp. strain W007 (GenBank accession no. ZP_09405870); 10, S. roseum DSM 43021 (GenBank accession no. YP_003340976).

The identified mcl PHA depolymerase gene sequences (overall G+C content range, 65 to 74% mol) encoded proteins consisting of ∼279 to 293 amino acids (29.4 to 30.5 kDa). All these sequences included a classical N-terminal signal peptide of 25 to 35 amino acids, as predicted by the SignalP (version 4.0) program (38) (Fig. 3). The calculated molecular masses of the mature proteins ranged from 26.1 to 27.8 kDa. In addition, the high content of aromatic (7.2 to 9.4%) and uncharged aliphatic (44.3 to 50.8%) side chain residues in these amino acid sequences suggested that these enzymes were strongly hydrophobic. In general, these proteins showed a larger number of charged amino acids (17.8 to 21.8% for E, D, R, K, and H) than the mature enzyme of P. fluorescens GK13 (15%). On the other hand, these sequences did not show significant similarity to any of the already known extracellular mcl PHA depolymerases. In fact, no more than 32.5% and 22.1% similarities were observed between these proteins and PhaZ_PflGK13 and PhaZ_TthHB8, respectively (see Table S3 in the supplemental material). However, similar to all extracellular PHA depolymerases, the primary structure corresponding to Actinobacteria strains contained strictly conserved amino acids (Ser-Asp-His) that comprise a catalytic triad in the active center (Table 4). Moreover, the catalytic domain of these proteins contained the consensus lipase box pentapeptide of serine hydrolases G-X₁-S-X₂-G, in which X₁ is a His residue and X₂ is a Gln residue. Additionally, Table 4 shows those residues identified to be possible oxyanion hole amino acids, on the basis of the homology modeling of the mcl PHA depolymerase from S. venezuelae ATCC 10712 (see Fig. S4 in the supplemental material).

Table 4.

Alignment of amino acid sequences of mcl PHA depolymerases in neighborhood of putative active sites^a

Strain	Sequence at the following active-site positionb				Reference or source
	Ser (S)	Asp (D)	His (H)	Oxyanion hole
S. venezuelae ATCC 10712	154 VDLDHIASAGHSQGGAAA	198 YLAGQRDLTVW	228 RGAGHLSSIGDG	176 DTAVPIQPGPLTDPD	Pullan et al. (41)
R. erythropolis PR4	152 VDLEHIGASGHSQGGAAA	196 YLAGQADAIVW	226 RGATHFGTAING	174 DTAVAIQPGPLNDVD	Sekine et al. (52)
R. erythropolis SK112	152 VDLEHIGASGHSQGGAAA	196 YLAGQADAIVW	226 RGATHFGTAING	174 DTAVAIQPGPLNDVD	Unpublished
Streptosporangium roseum DSM 43021	144 VDLDRIGASGHSQGGAAA	188 ILAGQRDSIVW	218 RGADHFTVVGAP	166 DTVVPIQPGPLADAD	Nolan et al. (34)
S. griseus subsp. griseus NCBR 13350	99 VDLEHIGAVGHSQGGSAA	149 LLAGQRDSIVL	173 RGADHFTVVGDP	121 DTVLPIQPGPLADID	Ohnishi et al. (35)
S. griseus Xyleb KG-1	156 VDLEHIGAVGHSQGGSAA	200 LLAGQRDSIVL	230 RGADHFTVVGDP	105 DTVLPIQPGPLADID	Grubbs et al. (12)
Streptomyces sp. W007	147 VDLEHIGASGHSQGGAAA	191 LLAGQRDSIVF	221 RGADHFTVVGDP	169 DTILPIQPGPLANID	Unpublished
S. roseolus SL3	VDLARIGSAGHSQGGAAA	YLAGERDLTVW	RGAGHLSSIGDG	DTAVPIQPGPLTDPD	This study
S. venezuelae SO1	VDLEHIGSAGHSQGGAAA	YLAGQRDLTVW	RGAGHLSSIGDG	DTAVPIQPGPLTDPD	This study
S. omiyaensis SO2	VDLEHIGSAGHSQGGAAA	YLAGQRDLTVW	RGAGHLSSIGDG	DTAVPIQPGPLTDPD	This study
Consensus	VDL-:I+--GHSQGG-AA	-LAG--D--V*	RGA-H*--*--	DT***IQPGPL--*D	This study
P. fluorescens GK13	172 LNAQRQYATGISSGGYNT	228 FLHGFVDAVVP	260 PLGGHEWFAASP	111 VQNLLDHGYAVIAP	Schirmer and Jendrossek (50)
Consensus of lipases	-V-**GhS-G+---	-----D-*v	---H*------	--***HG*-----	Jendrossek (16)

^aThe positions of Ser, Asp, His, and the putative oxyanion hole in the premature depolymerase proteins are indicated. Consensus sequences are presented, in which amino acids conserved in all the analyzed sequences are indicated by capital letters and in which symbols are as follows: *, amino acids with hydrophobic side chains; +, amino acids with a small side chain; :, amino acids with charged side chains. The corresponding sequences of the P(3HO) depolymerase of P. fluorescens GK13 are given. The consensus sequences of lipases are also shown, in which amino acids conserved in all the analyzed sequences are indicated by capital letters, those which have been conserved in 10 or more proteins are marked by lowercase letters, and the symbols are as defined for the consensus sequences.

^bAmino acids that constitute the catalytic triad, the possible oxyanion, and the conserved amino acids of the lipase consensus sequence are indicated in bold.

DISCUSSION

In this work, 10 mcl PHA-degrading depolymerase producer bacteria were isolated from natural samples and their abilities to degrade different aliphatic biodegradable polyesters were evaluated. Among our identified bacteria, only three of them, Pseudomonas alcaligenes, Stenotrophomonas maltophilia, and Rhodococcus equi, have already been described to be extracellular mcl PHA-degrading bacteria (20, 29, 42). Previous work (22, 33) demonstrated that Gram-negative bacteria belonging to Pseudomonas and Stenotrophomonas species are the predominant mcl PHA degraders in soil and marine environments. However, six Streptomyces strains were identified among our isolates. Additionally, only those strains belonging to the Streptomyces genus showed the ability to degrade not only mcl PHA but also scl PHA and PCL. These results indicated that streptomycetes may play an important role in the degradation of polyesters. However, none of the isolates can degrade PLA and PES.

The majority of the PHA-degrading microorganisms are known to express only one type of PHA depolymerase that acts upon either scl PHA or mcl PHA (17). However, the ability to degrade scl PHA and mcl PHA by producing two types of depolymerases is rare and has been reported in only a few bacteria (5, 23, 25, 36, 47). In this work, the mcl PHA degraders S. roseolus SL3 and S. omiyaensis SO2 were also found to express scl PHA depolymerase in the presence of P(3HB). Additionally, when S. roseolus SL3 was grown in the presence of P(3HO) as the sole carbon source, it produced one polypeptide chain of mcl PHA depolymerase with a mass of ∼28 kDa and a pI of ∼5.2. These results are similar to those for several mcl PHA depolymerases characterized from other sources (47) but significantly different from those for the P(3HO) depolymerase from Pseudomonas fluorescens GK13 (dimer; 48 kDa; pI ∼7).

As previously reported by Santos et al. (47), it is likely that the mcl PHA depolymerases produced from Streptomyces strains have a wider range of substrate specificity. In this work, the substrate specificity of PhaZ_SroSL3 confirms this hypothesis. In fact, in contrast to the mcl PHA depolymerases from Pseudomonas, the enzyme degrades PCL and tributyrin but not olive oil. However, none of the mcl PHA depolymerases reported so far exhibited detectable activities against PLA (14, 20, 23, 47) and PES (47).

The pure PhaZ_SroSL3 mainly hydrolyzed P(3HO) to the monomeric unit of 3-hydroxyoctanoate (3-HO) after 72 h of reaction. In this sense, PhaZ_SroSL3 behaves like the extracellular mcl PHA depolymerases of P. alcaligenes LB19 (20) and S. venezuelae SO1 (47). On the other hand, PhaZ_PflGK13 (49) and PhaZ_SspKJ-72 (23) mainly hydrolyzed P(3HO) to the dimeric form of 3-hydroxyoctanoate. Thus, PhaZ_SroSL3 appears to have a promising potential for biotechnological application in the production of enantiomerically pure (R)-3-HO monomers. Several mcl PHA depolymerases have been biochemically characterized. However, only the PhaZ_PflGK13-coding gene and a few other homologous genes have been cloned and sequenced (21, 29, 37, 50), including an mcl PHA depolymerase from the predator Bdellovibrio bacteriovorus (31). Additionally, in a recent work, a significant different gene from a thermophilic bacterium, T. thermophilus HB8, has been identified (36). However, no gene sequence of the genus Streptomyces has been reported so far.

In this work, de novo sequencing of PhaZ_SveSO1 allowed the identification of a novel subgroup of mcl PHA depolymerases from Actinobacteria. These new types of mcl PHA depolymerases showed high sequence similarity (more than 60%) to each other (see Table S3 in the supplemental material), as well as to the deduced amino acid sequences of PhaZ_SroSL3, PhaZ_SveSO1, and PhaZ_SomSO2. Inspection of the amino acid sequences revealed no significant similarity to previously characterized mcl PHA depolymerases (less than 33%). The primary structure of these enzymes showed the signal peptide domain typical of mcl PHA depolymerases. Besides, like most serine hydrolases, these enzymes showed the catalytic triad amino acids (Ser, Asp, His) and the lipase consensus pentapeptide Gly-X₁-Ser-X₂-Gly. In all the enzymes identified in this work, X₁ was a His residue and X₂ was a Gln residue. Similarly, in true lipases, the X₁ residue is generally occupied by His or Tyr, whereas X₂ is variable (50). However, in all mcl PHA depolymerases of Pseudomonas strains analyzed so far, X₁ was an Ile residue and X₂ was a Ser residue. Interestingly, contrary to Pseudomonas enzymes, PhaZ_SroSL3, PhaZ_SomSO2, and PhaZ_SveSO1 can degrade PCL and tributyrin as bacterial lipases. The presence of a His residue instead of a hydrophobic one in the X₁ position would be a possible explanation for the differences in substrate specificities observed between mcl PHA depolymerases from Pseudomonas and Streptomyces origins. Moreover, the mcl PHA depolymerases described in this study contained a larger number of charged amino acids (∼18 to 22%) than the mature enzyme from P. fluorescens GK13 (15%). An increased number of intramolecular ion bonds by charged amino acids is known to contribute to the thermal stability of enzymes by conferring rigidity on them (36, 43). This fact is in accordance with previous results that demonstrated that PhaZ_SveSO1 showed higher thermostability than its PhaZ_PflGK13 counterpart (8, 47).

The three-dimensional (3D) model structure of the mcl PHA depolymerase of S. venezuelae ATCC 10712 was deduced by homology modeling using Pseudomonas mendocina lipase as the template (see Fig. S4 in the supplemental material). This model revealed that the enzyme consisted of an α/β hydrolase core with the catalytic triad (Ser¹²⁵-Asp¹⁶⁹-His¹⁹⁹) at its surface, being very exposed to the solvent, and Gln¹⁴⁷ as the oxyanion hole amino acid that stabilized the tetrahedral transition estate. Therefore, it is assumed that this enzyme does not undergo the typical phenomenon known as interfacial activation described for several lipases and for the intracellular mcl PHA depolymerase from Pseudomonas putida KT2442. Similar conclusions were deduced by de Eugenio et al. (4) on the basis of the 3D model of PhaZ_PflGK13. Similar to PhaZ_PflGK13, S. venezuelae ATCC 10712 PhaZ does not have a lid domain, and the two enzymes show a similar architecture and catalytic mechanism of ester hydrolysis. Moreover, a disulfide bridge was predicted by the model, explaining the previously observed inhibition of PhaZ_SveSO1 in the presence of DTT (47).

mcl PHA depolymerases are excellent candidate biocatalysts for environmental, industrial, and medical applications. This study provides novel information on mcl PHA depolymerases from Actinobacteria, in terms of molecular structure, revealing significant differences from Pseudomonas enzymes. Additionally, these results offer the possibility of cloning and expression of these distinct enzymes for their possible exploitation in biotechnological processes.