Article pubs.acs.org/Biomac
Swapping of Phasin Modules To Optimize the In Vivo Immobilization of Proteins to Medium-Chain-Length Polyhydroxyalkanoate Granules in Pseudomonas putida Nina Dinjaski and M. Auxiliadora Prieto* Department of Environmental Biology, Centro de Investigaciones Biológicas, CSIC, C/Ramiro de Maeztu, 9, 28040 Madrid, Spain S Supporting Information *
ABSTRACT: PhaF is a bimodular protein of Pseudomonas putida KT2442 exhibiting multiple functions within the polyhydroxyalkanoate (PHA) apparatus. It behaves as phasin or PHA granule binding protein (by BioF domain) and also as nucleoid-associated protein involved in granule localization and segregation during cell division (by C-terminal domain). This work addresses the function of the PhaI phasin in the PHA granule formation machinery. Epifluorescence microscopy and flow cytometry studies of P. putida phasin mutant cells producing recombinant phasin domains fused to GFP protein demonstrated a balanced granule distribution after cell division only when low dosage of PhaF, and BioF domain or PhaI, are expressed together, revealing the exchangeability of phasins modules. These findings show the precise combination of phasin module production that leads to the optimal PHA production and granule localization and distribution, concomitantly to in vivo immobilization of recombinant proteins to PHA (22 mg of protein/g PHA).
■
immobilization of the GAP fusions.14−16 In vivo approach consists of GAP fusion immobilization onto the PHA granule surface while the granules are being formed inside the PHA producer cell (Figure 1).17 Different strategies using GAPs, such as synthases, depolymerases, and phasins as fusion tags, have been described.12−15,18 Among them, phasins are very attractive due to their diversity in terms of primary structure compared to the other GAPs, especially as an affinity tag for protein purification.8 Phasins are the most abundant protein found at the granule surface, thereby exerting a severe influence on the size and number of granules in the cell.5 Different approaches for proteins anchoring to the granule surface using phasins are reported, most of them using Escherichia coli as host, a non-natural PHB producer, carrying heterologous genes for PHB metabolism.12,19 A particular system for protein immobilization/purification into mcl-PHA, based on the use of the N-terminal domain of the phasin PhaF of Pseudomonas putida as polypeptide tag (BioF tag) was described (Figure 1).17 BioF tag was applied to deliver proteins of environmental interest as for instance Bacillus thuringiensis Cry1Ab toxin1 providing a useful tool for in vivo immobilization of active insecticidal proteins to a biodegradable support.11 One characteristic of this method is the simplicity of the downstream process; once the fermentation is accomplished in P. putida, the mcl-PHA granules carrying the BioF-proteins
INTRODUCTION Polyhydroxyalkanoate (PHA) granules are considered as prokaryotic subcellular organelles, taking over the control of the carbon and energy storage of the cells.1 They consist of phospholipid-coated polyesters and granule-associated proteins (GAPs) on their surfaces such as PHA synthases, involved in the polymerization of the biopolyester, PHA depolymerases, responsible for mobilization,2 phasins, the main structural components of GAPs,3−5 and other proteins such as enzymes related to the synthesis of PHA monomers,6 as well as transcriptional regulators (Figure 1).3,7 Independent of the type of PHA (short-chain-length PHA (scl-PHA) with C4−C5 monomers as polyhydroxybutyrate or medium-chain-length PHA (mcl-PHA) with C6−C12 monomers as polyhydroxyoctanoate), granules autonomously act upon several functions, including PHA synthesis, storage, and mobilization, and therefore are proposed to be denominated carbonosomes.1 Due to the PHA properties such as biocompatibility, biodegradability and production from renewable carbon sources, these granules are recognized as potential functionalized beads for biotechnological and medical applications.8 Recombinant bacterial cells were used to produce tailor-made functionalized micro- or nanobeads in which GAPs attached to the PHA granule have been engineered to display fusion proteins of interest. Examples of high-affinity bioseparation,9 enzyme immobilization,10 protein delivery in natural environments,11 diagnostics,12 and as an antigen delivery system13 have been described. The technology can be applied in vitro or in vivo; in the first case, chemical extraction of the PHA is followed by in vitro bead production and in vitro purification/ © 2013 American Chemical Society
Received: June 18, 2013 Revised: July 18, 2013 Published: July 25, 2013 3285
dx.doi.org/10.1021/bm4008937 | Biomacromolecules 2013, 14, 3285−3293
Biomacromolecules
Article
Figure 1. PHA producing cell, scheme of PHA granule structure in P. putida KT2442 synthesing BioF-GFP fusion protein and a model of BioF-GFP fusion protein. (A) Transmission electron microscopy (TEM) image of mcl-PHA producing P. putida KT2442 cells. PHA reserve granules are located running lengthwise to the cell, forming a characteristic needle array, as previously described;4 (B) Model of PHA granule structure and BioF system. The PHA granule is composed of PHA core coated with phospholipid monolayer (gray) where granule-associated proteins GAPs (phasins, synthases, depolymerase, and acyl-CoA synthetase) are embedded or attached. PhaF phasin is depicted as tetramer interacting with PHA granule via N-terminal domain composed of α helix (in blue) and leucine zipper (in red) and followed by C-terminal (in yellow) that provides interaction with DNA. BioF tag, composed of PhaF N-terminal domain plus leucine zipper, is shown fused to GFP protein (in green). PhaI phasin is composed of α helix (in purple) and leucine zipper (in pink), whereas synthase is depicted as white square, depolymerase as black square and acyl-CoA synthetase as pink circle; C, Model of BioF fusion protein. BioF system with granule binding capacity consists of a long amphipatic α helix (depicted in blue), followed by short leucine zipper (in red) involved in protein oligomerization. BioF system was fused to the reporter GFP protein (in green).
Moreover, we determine the minimal necessary amount of natural phasin proteins to achieve the optimal PHA production.
fusions can be isolated from the crude cell lysate by soft centrifugation and directly used for further applications. Two phasins, PhaF and PhaI, were characterized in P. putida strain (Figure 1B).3,4,17,20 The PhaF functions as nucleoidassociated protein involved in several roles linked to the mclPHA metabolism.4 According to the three-dimensional model validated by thermodynamic, hydrodynamic, and spectroscopic techniques,20 PhaF is a tetramer made up from elongated monomers, which are composed of a long, amphipathic Nterminal helix with PHA binding capacity, followed by a short leucine zipper, a proposed signature motif for protein−protein interaction, and a superhelical C-terminal domain wrapped around the chromosomal DNA.20 BioF tag consists of 142 amino acids comprising the N-terminal helix and the putative oligomerization zipper domain (Figure 1C). Much evidence supports the structural and functional independence of the Nand C-terminal domains. First, PhaI phasin shares considerable sequence similarity with the N-terminal region of PhaF, including the putative oligomerization linker,20 and is capable of acquiring a folded, stable, and functional structure by itself.3,17 Also, the N-terminal domain is conserved in other PHA-binding phasins with a different C-terminal part.3,4 Moreover, very diverse fusion proteins containing the Nterminal moiety of PhaF (the BioF affinity tag) can also be adsorbed onto PHA granules without compromising their function.11,17 In this work we address the impact of the lack of PhaI phasin on PHA granule accumulation/formation, heterogeneity of the cell population in terms of granule content and segregation, and the possibility of PhaI phasin replacement by BioF fusion proteins. Our findings result in an optimization of BioF system by overcoating the nanobeads with the fusion protein.
■
MATERIALS AND METHODS
Bacterial Strains, Media, and Growth Conditions. The bacterial strains used throughout this study are listed in Table 1. Further details of the constructions of the Pseudomonas putida strains are shown in Figure 2A. Unless otherwise stated, Escherichia coli and P. putida strains were grown in lysogeny broth (LB) medium26 at 37 and 30 °C, respectively. The appropriate selection antibiotics, kanamycin (50 μg/mL) or ampicillin (100 μg/mL) were added when needed. For PHA production P. putida strain was grown in 0.1 N M63, a nitrogenlimited minimal medium (13.6 g of KH2PO4/L, 0.2 g of (NH4)2SO4/ L, 0.5 mg of FeSO4·7H2O/L, adjusted to pH 7.0 with KOH) at 30 °C and 250 rpm as previously described.17 This medium was supplemented with 1 mM MgSO4 and a solution of trace elements (composition 1000 × 2.78 g of FeSO4·7H2O/L, 1.98 g of MnCl2· 4H2O/L, 2.81 g of CoSO4·7H2O/L, 1.47 g of CaCl2·2H2O/L, 0.17 g of CuCl2·2H2O/L, 0.29 g of ZnSO4·7H2O/L). Sodium octanoate 15 mM was used as carbon source. Growth was monitored with a Shimadzu UV-260 spectrophotometer at 600 nm. DNA Manipulations, Plasmid, and Strain Constructions. P. putida KT42I, phaI gene deleted strain, was constructed by disruption of phaI gene using plasmid pK18mobsacB.27 DNA manipulations and other molecular biology techniques were performed according to Stambrook and Russel (2001).26 All oligonucleotides used for PCR amplification are listed in Table 1. The 481bp and 339bp fragments upstream and downstream phaI gene were PCR-amplified with F5mutI and F3mutI or X5mutI and X3mutI primers, respectively. Total DNA of P. putida KT2442 strain was used as template. DNA fragments were purified by standard procedures using Gene Clean (BIO 101, Inc.). The resulting fragments were digested with the appropriate restriction enzymes and ligated using T4 ligase resulting in a single 820bp fragment carrying a deletion of the phaI gene which was cloned into the unique BamHI and SmaI sites of pK18mobsacB to yield pKI18mobsacB. Plasmid pKI18mobsacB was used to deliver the phaI 3286
dx.doi.org/10.1021/bm4008937 | Biomacromolecules 2013, 14, 3285−3293
Biomacromolecules
Article
To construct P. putida KT42C1ZC2F strain, the phaF gene was inactivated by marker exchange using the mobilizable suicide plasmid pKNGFdel, as previously described.4 Briefly, biparental filter-mating technique was performed28 using E. coli SM10λpir (pKNGFdel) as donor strain and P. putida KT42C1ZC2 as recipient strain. Transconjugants (SucR, SmS) were isolated. The second crossover event was confirmed by PCR using primers D5mutF and I3mutF. Insertion of GFP and BioF (N-Terminal of PhaF Protein) Fusion as a Monocopy into the Chromosome of P. putida Strains. To study localization of BioF in vivo, the mobile cassette carrying the biof-gf p fusions were inserted as a monocopy into the chromosome of different P. putida strains via mini-Tn5 transposons. A 1730bp DNA NotI/NotI fragment containing Ptac::BioF-GFP-LYTAG was obtained after digestion from pMAB20-GFP-LYTAG plasmid, kindly provided by Biomedal S.L. Plasmid pCNB5 was used as cloning vector where biof-gf p was inserted as NotI fragment to yield pCNB5BioF-GFP allowing to drive the expression of the fusion under the control of lacIq-Ptrc regulatory system. E. coli CC118λpir cells were transformed and positive clones were selected in LB plates supplemented with kanamycin. Resulting construction (pCNB5BioF-GFP) was transferred into P. putida KT2442, P. putida KT42F, P. putida KT42I, P. putida KT42C1ZC2, and P. putida KT42C1ZC2F by triparental mating technique to give rise to P. putida KT42-BG, P. putida KT42F-BG, P. putida KT42I-BG, P. putida KT42C1ZC2-BG, and P. putida KT42C1ZC2F-BG, respectively (Figure 2A). Transconjugants were selected on 0.1 N M63 plates supplemented with 0.2% citrate, kanamycin, 0.1 mM IPTG (isopropyl-1-thio-β-Dgalactopyranoside) and confirmed by SDS-PAGE and Western Blot analysis as previously described.2,4 Western blot analysis was performed with the ECL Western Blotting Detection Kit (Amersham Biosciences) according to the protocol described by the manufacturer. Rabbit polyclonal antiserum against PhaI and BioF was generated as previously described.2 Colonies were picked to LB plate with and without IPTG. Selection of fluorescent green colonies was done from LB plate supplemented with IPTG by fluorescent magnifying lamp. Complementation of P. putida KT42I and P. putida KT42I-BG Strains. Plasmid pPF61, harboring the phaF gene from P. putida GPo1 under the control of the Ptrc promoter3 was introduced into P. putida KT42I and P. putida KT42I-BG chromosome by triparental mating. The resulting strains KT42I−F and KT42I-BGF, respectively were cultivated in PHA production medium in the presence of 5 mM IPTG as previously described. PHA Quantification. Polyhydroxyalkanoate monomer composition and cellular PHA content were determined by gas chromatography−mass spectrometry (GC-MS) following previously described protocol.22 Briefly, samples were subjected to methanolysis in the presence of 15% (w/v) sulfuric acid, and resulting methyl esters of monomers were analyzed by injecting 1 μL of sample into PerkinElmer AutoSystem gas chromatograph equipped with SPB1 Supelco capillary column (25 m × 0.25 mm inner diameter × 0.22 μm) and ionization detector. Biomass was calculated as previously described.22 An additional, quantification of the PHA content was performed by flow cytometry (flow cytometer Coulter EPICS XL). Analysis was done as described before.4 Briefly, PHA was stained adding 3 μL of Nile Red stock solution (1 mg/mL in dimethyl sulfoxide) to 1 mL of harvested and washed cells, previously diluted to OD600 of 0.2. The mixture was incubated in the dark for 15 min and analyzed. P. putida KT42C1 minus strain (phaC1 mutant which produces
