Effects of Granule-Associated Protein PhaP on Glycerol-Dependent Growth and Polymer Production in Poly(3-Hydroxybutyrate)-Producing Escherichia coli

Bacterial strains and plasmids.The bacterial strains, plasmids, and primers used in this work are summarized in Table 1.

Culture conditions and growth media.For DNA manipulations and strain construction, cultures were grown at 37°C in LB. Ampicillin (100 μg/ml), kanamycin (50 μg/ml), and chloramphenicol (20 μg/ml) were added when indicated. For growth and PHB accumulation analysis bacteria were grown in one of the following media: M9-lactose medium, containing (per liter of deionized water) 1 g NH₄Cl, 6 g Na₂HPO₄, 3 g KH₂PO₄, 0.5 g NaCl, 0.2 g MgSO₄·7H₂O, 10 mg CaCl₂, and 30 g lactose, or MYA medium, containing (per liter of deionized water) 6 g Na₂HPO₄, 3 g KH₂PO₄, 1.4 g (NH4)₂SO₄, 0.5 g NaCl, 0.2 g MgSO₄·7H₂O, 10 g yeast extract, and 5 g casein amino acids. MYA was supplemented with either glucose (MYA-Glu) or glycerol (MYA-Gly) at 30 g/liter.

DNA manipulations.Plasmid and genomic DNA isolation, restriction enzyme digestion, agarose gel electrophoresis, and DNA ligations were performed by standard procedures (15) and following specific instructions from the manufacturers. Transformations were carried out as previously described (7). E. coli strain DH5α was routinely used as a host for most of the recombinant plasmids.

Cloning of phaP and phaF from Azotobacter sp. strain FA8.A 5.1-kb XhoI genomic fragment from Azotobacter sp. strain FA8 containing genes phaF, phaP, and phaR, together with parts of two flanking insertion sequences (18), was cloned into pBlueScript, giving rise to plasmid pRX23. This plasmid was cut with SmaI and HindIII, and the 1.8-kb fragment containing phaF and phaP was cloned into pBlueScript cut with the same enzymes, resulting in plasmid pBSK-PF. To obtain plasmid pAD-PF, the insert from pBSK-PF was subcloned into the Km^r vector pBBR1MCS-2 using the BamHI and HindIII restriction sites. A 1.1-kb fragment containing phaP was cut from pBSK-PF with SalI and cloned into the Cm^r vector pBBR1MCS, resulting in plasmid pAD-P.

Construction of phaP expression vector.Primers phaPup and phaPlow (Table 1) were used to obtain a 591-bp amplification fragment using plasmid pRX23 as a template. The resulting amplification fragment was cut with HindIII and BamHI and ligated into vector pQE31. The resulting plasmid, pQP, confers resistance to ampicillin and expresses phaP from a promoter operator element consisting of the phage T5 promoter and two lac operator sequences. The insert was cloned in such a way that the amino-terminal end of the phaP gene product is fused to the first six amino acids of the lacZ gene, followed by six histidine residues, resulting in a fusion protein. The phasin protein was expressed in strain M15/pQP and purified by affinity chromatography as recommended by the supplier, using the QIAexpress system kit (Qiagen). The purified protein was used to obtain an anti-PhaP antiserum in mice.

Western analysis.For immunoblotting experiments, cells were pelleted and sonicated in 1 ml of 100 mM phosphate buffer, pH 7.5, plus 1 mM phenylmethylsulfonyl fluoride. Protein concentration was measured according to the Bradford method (3). Equal amounts (30 μg) of total protein were used for a Western immunoblotting assay. Proteins were separated by sodium dodecyl sulfate-12% polyacrylamide gel electrophoresis and transferred to a nitrocellulose membrane (BA85; Schleicher & Schuell). For Western blotting, the membrane was probed with mouse anti-PhaP polyclonal serum diluted 1:2,500, followed by alkaline phosphatase-conjugated coat anti-mouse immunoglobulin G (NEN) diluted 1:5,000. The blots were developed with the Phototope-Star chemiluminescent detection kit (New England Biolabs).

Bioreactor cultivation.Batch cultures were carried out at 37.0 ± 0.2°C in a 5.6-liter stirred tank reactor equipped with six flat-bladed disk turbines (BioFlo110; New Brunswick Scientific Co., Edison, NJ) with a working volume of 3 liters MYA-Gly, containing 50 μg/ml kanamycin and 20 μg/ml chloramphenicol for plasmid maintenance. Culture pH was controlled at 7.20 by automatic addition of either 3 N KOH or 5 N H₂SO₄. Dissolved oxygen concentration was maintained above 30% of air saturation throughout the fermentation by the automatic control of the agitation speed (up to 1,000 rpm) while sparging the fermentor with 1 vessel volume of air per minute. Dissolved oxygen was measured using an Ag/AgCl polarometric oxygen probe (Mettler Toledo, Greifensee, Switzerland). Foam was suppressed by adding 30 μl/liter Antifoam 289 (Sigma-Aldrich). Residual glycerol concentration was determined using an enzymatic kit (Roche Diagnostics, Germany) at the end of the fermentation.

Determination of plasmid stability.The amount of plasmid-containing cells was determined by plating samples from the corresponding cultures on LB with or without the appropriate antibiotics. Plasmid stability was recorded as the percentage of antibiotic-resistant cells.

Biomass determination.Samples taken from the bioreactor were immediately chilled at 0°C by placement in an ice bath. Cells from 10-ml samples were washed twice with deionized water, recovered by centrifugation, dried at 85°C for 36 h, and weighed. Biomass content was defined as grams (dry weight) of cells (CDW) per liter. All results are indicative of replicate experiments.

Analysis of PHB production.For qualitative detection of PHB inclusion bodies, cells were observed by fluorescence microscopy after staining with the basic oxazine Nile Blue A as previously described (17). PHB content in flask cultures was determined gravimetrically after alkaline treatment with 0.2 N NaOH as described previously (5). PHB content in bioreactor cultures was quantitatively determined by gas chromatography using a slight modification of the method described by Braunegg et al. (4, 16). Pure PHB from C. necator was used as a standard. PHB concentration was defined as g polymer per liter of culture broth. PHB content was defined as a percentage of CDW. All results are indicative of duplicate or triplicate experiments.

Purification of PHB.PHB produced at the end of the fermentation experiments was extracted from lyophilized cells with hot CHCl₃ using a Soxhlet apparatus, ethanol precipitated, and recovered by filtration. The precipitate was dried, dissolved in CHCl₃, filtered to remove contaminating particles, and dried on a glass petri dish to obtain a thin film. The resulting polymer was characterized by gas chromatography as described above, and used for differential scanning calorimetry (DSC) measurements.

DSC measurements.Glass transition temperature (T_g), melting temperature (T_m), and crystallinity of purified PHB samples were determined by DSC using Mettler 822 and STARe thermal analysis system version 6.1 software (Mettler Toledo AG, Switzerland) as previously described (16). Crystallinity of PHB was estimated from the enthalpy of fusion obtained by DSC. The fusion enthalpy of a theoretical 100% crystalline sample was assumed to be 146 J/g (2). All results are indicative of duplicate or triplicate experiments.

Transmission electron microscopy.Cells were fixed by adding 2.5% (vol/vol) glutaraldehyde and embedded in agar. One-millimeter pieces of the agar were fixed for 30 min in phosphate-buffered 2.5% (vol/vol) glutaraldehyde, rinsed three times with phosphate buffer, and postfixed in phosphate-buffered 1% (wt/vol) osmium tetroxide for 1 h. The agar pieces were rinsed with water again and fixed for 1 h in 1% (wt/vol) aqueous uranyl acetate. All fixations were carried out at room temperature. After dehydration in a graded series of ethanol and two changes in propylene oxide, the agar pieces containing bacterial cells were embedded in Epon 812 resin (Spi Supplies, New Chester, PA). Thin sections stained with uranyl acetate and lead citrate were examined in a Philips EM 301 transmission electron microscope.

Effect of phaP and phaF on PHB synthesis in E. coli carrying phaBAC from Azotobacter sp. strain FA8.K24 is an E. coli strain carrying the PHB biosynthetic genes from Azotobacter sp. strain FA8 in Ap^r plasmid pJP24 (16). As a first approach to analyze the possible effects of putative PHA regulatory genes found in Azotobacter sp. strain FA8 on recombinant E. coli, pAD-PF, a Km^r plasmid containing phaP and phaF, was introduced into K24. Both recombinant plasmids have compatible replication origins and different selection markers, so they can coexist in the recombinants, allowing the combined expression of the heterologous genes. Plasmid pAD-PF contains phaP and phaF, situated in opposite (converging) orientations and approximately 300 bases upstream from each gene. phaP is situated colinearly with the lac promoter. The growth behavior and ability to accumulate the polymer in M9-lactose medium were analyzed in K24 containing pAD-PF (K24PF) or the Km^r vector without an insert (K24vk). Results obtained in three independent experiments (in which each parameter was measured in duplicate) showed that although CDW concentration at the end of the cultivation was similar (1.1 ± 0.1 g/liter and 1.1 ± 0.2 g/liter for K24vk and K24PF, respectively), the PHB content was higher for the strain bearing pAD-PF (24.6% ± 1.1% and 31.8% ± 3.8% for K24vk and K24PF, respectively). To verify if the effects observed were due to the expression of phaP, a plasmid containing only this gene was constructed.

Construction of a stable PHA-accumulating recombinant strain that expresses PhaP.Strain K24 was not able to accumulate high concentrations of PHB due to the instability of its Ap^r recombinant plasmid, but pJP24K, a Km^r derivative of pJP24 analyzed in a previous study (16), was efficiently maintained in bioreactor cultures of the corresponding strain, K24K. In view of this, we decided to study the effects of PhaP in strain K24K. Plasmid pAD-P, a Cm^r plasmid that expresses phaP under lac control, was constructed and introduced into strain K24K. This plasmid contains the coding region of phaP including 392 bases from its upstream region.

phaP expression in the recombinants was verified by Western blot experiments (Fig. 1). The phaP-containing insert was cloned in both orientations, but only the orientation colinear with the lac promoter allowed the expression of PhaP, as detected in Western blot assays (data not shown).

• Open in new tab
• Download powerpoint

FIG. 1.

Immunoblot assays of PhaP in E. coli strains containing different plasmids. Lane 1, M15/pQP; lane 2, K24Kvc cultured in MYA-Gly in flask cultures, 24 h; lane 3, K24KP cultured in MYA-Gly in flask cultures, 24 h. Equal amounts (30 μg) of total protein were used for each lane.

The effect of PhaP in K24K was analyzed in agitated flask cultures (50 ml in 500-ml Erlenmeyer flasks) in MYA medium supplemented with glucose or glycerol. Cell growth and PHB content were monitored in the cultures of K24K bearing pAD-P (K24KP) or the Cm^r plasmid without insert (K24Kvc) (Table 2). In both carbon sources the PhaP-bearing strains grew more and contained more polymer. Cultures of K24KP grown in glucose achieved higher biomass and PHB accumulation levels than those grown in glycerol, but growth of K24Kvc was similar on the two carbon sources. Plasmid loss was less than 1% for both cultures. Cells were observed by transmission electron microscopy, revealing PHA granules of similar size, present in similar numbers in the two strains, but more cells containing polymer granules were observed for K24KP (data not shown).

As a result of enhanced growth and polymer accumulation, K24KP produced several times more polymer than the strain without PhaP, and this difference was more evident when using glycerol as the carbon source.

Effect of PhaP on cell growth and PHB production from glycerol in bioreactor cultures.Strain K24K bearing plasmid pBBR1MCS (K24Kvc) or pAD-P (K24KP) was grown in MYA-Gly medium. Cell growth, PHB production, and cell morphology were analyzed. Table 3 shows cell growth and PHB content of a representative bioreactor culture. The strain expressing PhaP consumed more substrate, grew 1.9 times more (15.3 versus 8.0 g/liter), and produced 2.6-times-more PHB (7.9 versus 3.1 g/liter) than the strain with the vector alone. Plasmid loss was less than 1% for both cultures.

The cells were microscopically observed to detect changes in cell morphology, as other authors have reported a tendency for some recombinant strains to form filaments when grown in bioreactors (28). Cells expressing PhaP were slightly bigger, probably because of their increased polymer content, but no filamentation was observed in any of the cultures (data not shown). K24KP cells from the 24-h bioreactor cultures were observed in a transmission electron microscope. Large granules occupying a great portion of the cells could be observed (see Fig. S1 in the supplemental material).

Physical properties of PHB obtained from glycerol bioreactor cultures were investigated by DSC. The polymer from K24Kvc had a lower T_g and T_m than the phasin-bearing recombinant (46.3 ± 0.3 and 169.8 ± 0.4 for K24Kvc and 56.9 ± 1.0 and 177.3 ± 0.5 for K24KP, respectively). The PHB obtained from K24Kvc also showed a lower percent crystallinity (52.6 ± 0.4) than the polymer from K24KP (66.4 ± 1.5). These results show that the recombinant strains produce PHB with different physical characteristics.