Simultaneous Improvements of Pseudomonas Cell Growth and Polyhydroxyalkanoate Production from a Lignin Derivative for Lignin-Consolidated Bioprocessing

With the recent advances in processing carbohydrates in lignocellulosics for bioproducts, almost all biological conversion platforms result in the formation of a significant amount of lignin by-products, representing the second most abundant feedstock on earth. However, this resource is greatly underutilized due to its heterogeneity and recalcitrant chemical structure. Thus, exploiting lignin valorization routes would achieve the complete utilization of lignocellulosic biomass and improve cost-effectiveness. The culture conditions that encourage cell growth and polyhydroxyalkanoate (PHA) accumulation are different. Such an inconsistency represents a major hurdle in lignin-to-PHA bioconversion. In this study, we traced and compared transcription levels of key genes involved in PHA biosynthesis pathways in Pseudomonas putida A514 under different nitrogen concentrations to unveil the unusual features of PHA synthesis. Furthermore, an inducible strong promoter was identified. Thus, the molecular features and new genetic tools reveal a strategy to coenhance PHA production and cell growth from a lignin derivative.

ABSTRACT

Cell growth and polyhydroxyalkanoate (PHA) biosynthesis are two key traits in PHA production from lignin or its derivatives. However, the links between them remain poorly understood. Here, the transcription levels of key genes involved in PHA biosynthesis were tracked in Pseudomonas putida strain A514 grown on vanillic acid as the sole carbon source under different levels of nutrient availability. First, enoyl-coenzyme A (CoA) hydratase (encoded by phaJ4) is stress induced and likely to contribute to PHA synthesis under nitrogen starvation conditions. Second, much higher expression levels of 3-hydroxyacyl-acyl carrier protein (ACP) thioesterase (encoded by phaG) and long-chain fatty acid-CoA ligase (encoded by alkK) under both high and low nitrogen (N) led to the hypothesis that they likely not only have a role in PHA biosynthesis but are also essential to cell growth. Third, 40 mg/liter PHA was synthesized by strain A_phaJ4C1 (overexpression of phaJ4 and phaC1 in strain A514) under low-N conditions, in contrast to 23 mg/liter PHA synthesized under high-N conditions. Under high-N conditions, strain A_alkKphaGC1 (overexpression of phaG, alkK, and phaC1 in A514) produced 90 mg/liter PHA with a cell dry weight of 667 mg/liter, experimentally validating our hypothesis. Finally, further enhancement in cell growth (714 mg/liter) and PHA titer (246 mg/liter) was achieved in strain A_{xyl_alkKphaGC1} via transcription level optimization, which was regulated by an inducible strong promoter with its regulator, XylR-P_xylA, from the xylose catabolic gene cluster of the A514 genome. This study reveals genetic features of genes involved in PHA synthesis from a lignin derivative and provides a novel strategy for rational engineering of these two traits, laying the foundation for lignin-consolidated bioprocessing.

IMPORTANCE With the recent advances in processing carbohydrates in lignocellulosics for bioproducts, almost all biological conversion platforms result in the formation of a significant amount of lignin by-products, representing the second most abundant feedstock on earth. However, this resource is greatly underutilized due to its heterogeneity and recalcitrant chemical structure. Thus, exploiting lignin valorization routes would achieve the complete utilization of lignocellulosic biomass and improve cost-effectiveness. The culture conditions that encourage cell growth and polyhydroxyalkanoate (PHA) accumulation are different. Such an inconsistency represents a major hurdle in lignin-to-PHA bioconversion. In this study, we traced and compared transcription levels of key genes involved in PHA biosynthesis pathways in Pseudomonas putida A514 under different nitrogen concentrations to unveil the unusual features of PHA synthesis. Furthermore, an inducible strong promoter was identified. Thus, the molecular features and new genetic tools reveal a strategy to coenhance PHA production and cell growth from a lignin derivative.

INTRODUCTION

Polyhydroxyalkanoates (PHA) are a class of natural biopolymers that are accumulated by microbes as carbon and energy storage materials (1). They can be used as alternatives to petroleum-based plastics due to their environmentally friendly aspects of sustainability and low CO₂ emissions. In addition, their biocompatibility and biodegradability allow them to be bioimplant materials for medical and therapeutic applications (2). Moreover, their controllable thermal and mechanical properties, as well as diverse molecular weights, allow PHA to function as chemical precursors and methyl ester-based fuels (1, 3, 4). The high PHA content (21% of the cellular dry weight) observed in Escherichia coli was biosynthesized from fatty acids as early as 1997 (5). However, due to the high cost of fatty acids and their toxicity to bacteria, researchers have attempted to bioproduce PHA from inexpensive and renewable unrelated carbon sources, e.g., glucose and glycerol (6–8). Lignocellulose is the most abundant resource for sustainable fuels and chemical production. With the development of cellulosic biofuel production, substantial quantities of lignin by-products are expected to be generated, representing the second most abundant feedstock on earth (9). However, the recalcitrant nature of lignins and certain technological barriers limit lignin mainly to production of process heat and power via combustion during cellulosic biofuel production (10–12). Thus, finding an efficient conversion route for lignin to PHA would offer a significant opportunity for the viability and sustainability of modern lignocellulosic biorefineries and offset global energy shortages and climate change (13). A few studies have investigated the feasibility of lignin-to-PHA bioconversion routes, including bacterial screening for lignin-consolidated bioprocessing (14, 15) and integrated biological funneling and chemical catalysis for lignin-to-PHA production (3). However, few studies have attempted to study the PHA biosynthetic pathways when strains are grown using either lignin or its derivatives as a sole carbon source (16), most likely because the complex process of lignin depolymerization in lignin-to-PHA bioconversion has hampered investigations and comparisons between PHA biosynthesis pathways. Thus, the molecular mechanisms of PHA biosynthesis, in which various metabolic pathways of fatty acids interact to mediate carbon flux to PHA biosynthesis pathways under conditions where strains are grown on lignin (or its derivatives), remain elusive.

Pseudomonas species are well-known medium-chain-length PHA (mcl-PHA)-producing organisms (17), and recent studies have shown that they are capable of lignin degradation (18). They secrete oxidative enzymes (e.g., laccases and dye peroxidases) to trigger lignin decomposition and generation of lignin oligomers (e.g., β-aryl-ether and coniferyl aldehyde), which are subsequently catabolized via versatile aromatic compound pathways (19–21). In addition, they adapt easily to different environments and have readily available genetic manipulation systems (16, 22). Taken together, these traits indicate they could play a profound role in lignin-to-PHA conversion. Pseudomonas species readily accumulate mcl-PHA under nutrient imbalances, e.g., nitrogen starvation conditions (1). Under nitrogen limitation, strains usually exhibit poor growth (e.g., lower cell density, lighter cell dry weight, or lower growth rate) (14, 23). It should be noted that cell growth (cell biomass in particular) is a critical factor for total PHA production (7). Such an inconsistency between cell growth and PHA production represents one of three major hurdles in lignin-to-PHA bioconversion, the others being lignin depolymerization and aromatic compound degradation. In fact, numerous studies have investigated the PHA biosynthesis pathways in Pseudomonas, where PHA is synthesized via either the β-oxidation pathway or by the de novo fatty acid biosynthesis pathway when the organisms are grown on related (e.g., fatty acids) or unrelated (e.g., glucose and glycerol) carbon sources, respectively (1, 24, 25). Moreover, a large number of strategies have been developed to enhance PHA production, such as reconstructing PHA synthetic pathways in model hosts without the ability for PHA accumulation (e.g., E. coli) (6, 7, 25–27). However, simultaneous improvement of both cell growth and PHA production in Pseudomonas has not been demonstrated, especially when lignin or its derivatives are used as sole carbon sources. Devising a rational strategy to counter this hurdle may require a mechanistic understanding of the links between the two traits.

Here, we employed the previously identified lignin-utilizing Pseudomonas putida strain A514 as a research model to investigate the PHA biosynthesis pathways. To simplify the lignin depolymerization process, we utilized a lignin substitute, vanillic acid, a key intermediate metabolite that connects several peripheral pathways to the central pathway (β-ketoadipate pathway) during lignin depolymerization, as the sole carbon source (Fig. 1A) (16, 18). The transcription levels of key genes involved in PHA biosynthesis were tracked and compared when A514 cells were grown on M9 medium with low (65 mg/liter) and high (1 g/liter) nitrogen (N) concentrations, unveiling unusual features of PHA synthesis. Exploiting these molecular features and developing genetic tools enabled us to devise a rational strategy to coenhance PHA production and cell growth.

FIG 1.

Proposed metabolic pathways for lignin-to-PHA bioconversion. Key genes and metabolic intermediates are highlighted in red. Identifications (IDs) of the corresponding genes in A514 are shown. (A) Lignin degradation pathways. Dashed arrows indicate multiple steps. (B) De novo fatty acid biosynthesis. (C) β-Oxidation pathway. After each turn of the cycle, an acyl-CoA (indicated as C_n) that is two carbons shorter than the initial one (indicated as C_{n + 2}) is generated. The dashed line connects acyl-CoA intermediates of different chain lengths. (D) PHA biosynthesis pathways. (E, F, and G) Expression ratios for alkK homologs and phaG and phaJ genes, as well as for PHA synthetic genes, in A514 that were grown under the low-N conditions, compared to those grown under the high-N conditions. The ratios that were >1 were defined as induced. (H) Expression levels of the phaJ4, phaG, and alkK (PputA514_1839) genes in A514 under both high-N and low-N conditions. The inset in panel H shows the expression level of phaJ4 at an appropriate scale. The expression levels and expression ratios are defined in Materials and Methods.

RESULTS

Transcriptional comparison revealed specific features of PHA synthesis in A514 when a lignin derivative was used as the sole carbon source.

P. putida A514 was grown on either high-N or low-N M9 medium, supplemented with 15 mM vanillic acid as the sole carbon source. First, it was observed that the doubling time of A514 cells grown on high-N medium (4.8 h) was higher than that of cells grown on low-N medium (3.8 h) during the exponential phase (Table 1 and Fig. S1 in the supplemental material). Second, cell dry weight (CDW) under high-N conditions was approximately 6-fold higher than under low-N conditions (Table 1). Third, 15 mM vanillic acid was completely consumed under high-N growth conditions, whereas only 6.5 mM vanillic acid was utilized under low-N conditions (Table 1 and Fig. S1). Taken together, these results indicate a likely change in gene expression in A514 between the high-N and low-N growth conditions. Thus, the transcription levels of genes involved in PHA synthesis under these two conditions were subsequently examined by reverse transcription-quantitative PCR (qRT-PCR) (Table 2).

TABLE 1.

Cell growth and mcl-PHA accumulation in the recombinant P. putida A514 strains

Culturea	Strain	Relevant plasmid	Doubling timeb (h)	VAc utilization (mM)	CDWd (mg/liter)	PHA titer (mg/liter)	PHA content (wt %)
High N (1 g/liter)	A514 WT	None	4.8 ± 0.2	14.9 ± 0.01	664.1 ± 3.6	11 ± 2.0	1.7 ± 0.3
	APvan	pPvan	4.7 ± 0.3	14.8 ± 0.01	664.6 ± 2.2	12 ± 1.2	1.8 ± 0.2
	AphaJ4C1	pTJ4C1	4.6 ± 0.3	15.0 ± 0.02	668.0 ± 3.5	23 ± 1.0	3.4 ± 0.2
	AalkKphaGC1	pTphaGC1	4.5 ± 0.2	14.9 ± 0.01	667.5 ± 4.8	90 ± 2.1	13.5 ± 0.3
	AR-PxylA	pTRPxylA	4.9 ± 0.4	15.0 ± 0.01	694.7 ± 11.1	12 ± 0.6	1.7 ± 0.2
	Axyl_phaJ4C1	pTPxylAphaGC1	4.8 ± 0.1	15.0 ± 0.02	702.7 ± 19.4	26 ± 1.7	3.7 ± 0.3
	Axyl_alkKphaGC1	pTPxylAphaJ4C1	4.8 ± 0.2	15.0 ± 0.01	714.8 ± 12.9	246 ± 17.3	34.4 ± 0.3
Low N (65 mg/liter)	A514 WT	None	3.8 ± 0.3	6.5 ± 0.03	133.8 ± 10.9	23 ± 1.0	17.3 ± 1.8
	APvan	pPvan	3.7 ± 0.2	6.6 ± 0.06	135.5 ± 5.8	23 ± 3.8	17.1 ± 2.1
	AphaJ4C1	pTJ4C1	3.7 ± 0.2	7.4 ± 0.05	133.0 ± 7.8	43 ± 1.5	32.6 ± 1.2
	AalkKphaGC1	pTphaGC1	3.9 ± 0.1	7.6 ± 0.05	145.2 ± 7.3	62 ± 6.1	42.4 ± 2.2
	AR_PxylA	pTRPxylA	3.8 ± 0.2	7.5 ± 0.04	148.7 ± 6.1	24 ± 0.6	16.6 ± 1.1
	Axyl_phaJ4C1	pTPxylAphaGC1	3.7 ± 0.3	7.9 ± 0.04	141.3 ± 12.3	63 ± 2.3	44.5 ± 4.9
	Axyl_alkKphaGC1	pTPxylAphaJ4C1	3.5 ± 0.1	8.7 ± 0.04	175.3 ± 13.3	116 ± 4.5	66.3 ± 4.7

^aThe strains were grown in 15 mM vanillic acid-M9 medium under either high-N (1 g/liter) or low-N (65 mg/liter) conditions.

^bThe doubling time (DT) was calculated during the exponential phase, based on the formula DT = t · [log 2/(log N_t − log N₀)].

^cVA, vanillic acid.

^dCDW, cell dry weight.

TABLE 2.

Primers used in qRT-PCR

Primer	Sequences
PputA514_5834	F: 5′-TCACCGCCGACATCTACTC-3′; R: 5′-CTGCTGGACAACACGAACTC-3′
PputA514_5836	F: 5′-CACTTCGCATTCGCCTTGT-3′; R: 5′-TGTTGCCGCCTGAGTTGA-3′
PputA514_5834	F: 5′-GCGACCGTATCCTTGAATGT-3′; R: 5′-GAAGTGGTAGTAGAGGTTGCC-3′
PputA514_5838	F: 5′-GCGTCTCAACAGTGCTAT-3′; R: 5′-TTTGCTTGGTCAGGGTATC-3′
PputA514_5839	F: 5′-AGTAATGTCTCAAGCGTCAA-3′; R: 5′-TGCCGATTCGATTCAAGG-3′
PputA514_4767	F: 5′-ATTATGTCGGCAAGGAACT-3′; R: 5′-GACGTGGATGAACTGATGA-3′
PputA514_4070	F: 5′-TGTTCAAGGAGCGTATCG-3′; R: 5′-GGCTTCTGGAAGGTCATC-3′
PputA514_5233	F: 5′-CAAAGGTCTTTGGCTTTCGC-3′; R: 5′-CAGGAAGATTGGCAGCTTGA-3′
PputA514_1318	F: 5′-GCCTGTATCCGCAATTCAAC-3′; R: 5′-CCTTCGTCAGCATCTTCTCAT-3′
PputA514_1839	F: 5′-CCAGATCGCCGTTGTTACT-3′; R: 5′-GCCTGCCATTCAATGATGTC-3′
PputA514_2970	F: 5′-TCAGGATCAGGACCTTGC-3′; R: 5′-GCGTAACACAGCGACATA-3′
PputA514_4067	F: 5′-CCACATCTATGCCTTCACCTT-3′; R: 5′-CTTCCACTTCGACAGTTCCTT-3′
PputA514_4068	F: 5′-GGTTCGTGAGCATTGTTG-3′; R: 5′-CGTCTTCGATCTCGTTGG-3′
PputA514_5230	F: 5′-CAAAGACCCGAACCTGACTC-3′; R: 5′-TGCTGCGAAACTCCACATAC-3′
PputA514_3207	F: 5′-ACCACAACATCCTCAACAA-3′; R: 5′-ATACCGAAGCAGTGATACAA-3′
PputA514_0580	F: 5′-TGTGGAACTGGCTGGTGTC-3′; R: 5′-GCTGATTGTCTCGGCATCG-3′
PputA514_3979	F: 5′-TGAAGATGGCTACTGGTGGAT-3′; R: 5′-TCGGCGACTTTAGGGTGAG-3′
PputA514_2337	F: 5′-CCATCCGATCACCTTCTACGA-3′; R: 5′-GGCAATCAGCAGGCGATAC-3′
PputA514_4978	F: 5′-GGCAACGAAGTGGATGAA-3′; R: 5′-CTGGCAACGGATGATGTC-3′

Similarly to that of P. putida strain KT2440, the A514 pha gene cluster consists of phaC1 and phaC2, encoding two polymerases, phaZ, encoding a depolymerase, phaD, encoding a transcriptional regulator, and two genes encoding polyhydroxyalkanoate granule-associated proteins (phaF and phaI) (see Fig. S2 in the supplemental material). The organization of these genes in the A514 genome is also similar to that of P. putida KT2440, consistent with previous studies showing that the PHA biosynthetic gene cluster is well conserved among the Pseudomonas PHA producer strains (17) (Fig. S2). Comparison of gene transcription under high-N and low-N conditions revealed several characteristics of PHA synthesis in A514 when employing a lignin derivative (vanillic acid) as the sole carbon source. First, genes located in the PHA synthesis cluster were upregulated, including phaC1 (PputA514_5834), phaC2 (PputA514_5836), phaD (PputA514_5837), phaF (PputA514_5838), and phaI (PputA514_5839) (Fig. 1D and G), confirming the phenotype that PHA synthesis was induced under nitrogen deprivation (Fig. S1). Second, phaG (PputA514_1318) and alkK (PputA514_1839) were induced (Fig. 1B and E). phaG (encoding 3-hydroxyacyl-ACP thioesterase), combined with alkK (encoding fatty acid-CoA ligase), provides 3-hydroxyacyl-CoA, the precursor for PHA synthesis (6), thus connecting the de novo fatty acid biosynthesis pathway and the PHA biosynthesis pathway. The induced phaG gene demonstrated its role in PHA synthesis in the presence of vanillic acid as the carbon source. In addition, nine paralogs that are homologous to the alkK gene from P. putida KT2440 were identified in A514, and their transcription levels were examined (6, 28). Among them, PputA514_1839 (encoding a long-chain fatty acid-CoA ligase) was prominent in that it not only encoded the highest protein identity (89%; E value, 0) to AlkK of P. putida KT2440, but also showed the highest level of transcription (Fig. 1E), suggesting its probable contribution in the production of 3-hydroxyacyl-CoA for PHA production. Third, phaJ4, which connects the β-oxidation pathway and the PHA synthesis pathway, was highly upregulated (∼54-fold) (Fig. 1C and F). phaJ [encoding (R)-specific enoyl-CoA hydratase] produces 3-hydroxyacyl-CoA from fatty acid oxidation as the precursor for PHA biosynthesis (7). A514 harbors three phaJ paralogs, namely, phaJ1 (PputA514_4070), phaJ3 (PputA514_5233), and phaJ4 (PputA514_4767). Among them, phaJ4 exhibited the highest induction (Fig. 1F), suggesting an important role for this gene in directing carbon flux from the β-oxidation pathway to PHA biosynthesis. This is consistent with our previous study, where overexpressing phaJ4 and phaC1 increased PHA production when lignin was used as the sole carbon source (16). Fourth, compared to phaJ4, phaG and alkK sustained much higher expression levels under both high- and low-N conditions, indicating that, in addition to PHA synthesis, they were likely to be important in cell growth (Fig. 1H). Furthermore, comparison of low-N and high-N conditions showed that phaJ4 was highly induced (∼54-fold), in contrast to phaG and alkK, which showed only ∼6-fold activation (Fig. 1E and F). This was distinct from the PHA-producing model strain, P. putida KT2440, where it was phaG that presented the greatest induction (∼220-fold versus 3.8-fold for phaJ4) when grown on low-N medium compared to high-N medium (6), suggesting specific features of PHA synthesis in A514 when utilizing vanillic acid as the carbon source.

Increasing carbon flux from de novo fatty acid biosynthesis to PHA synthesis can simultaneously improve PHA production and cell growth.

Our previous study demonstrated that phaJ4 and phaC1 are important for PHA biosynthesis from lignin under nitrogen starvation conditions in A514 cultures (16). However, the unique transcriptional feature of phaG and alkK, in this study, suggested their role in both PHA biosynthesis and cell survival, thus raising the feasibility of enhancing PHA production without impairing cell growth. To validate this hypothesis, the plasmid pPvan, plasmid pPROBE-TT carrying the P_van promoter, was employed as the expression vector (Table 3) (29). phaG (PputA514_1318), alkK (PputA514_1839), and phaC1 (PputA514_5834) were cloned into pPvan to construct the plasmid pTphaGC1, while phaJ4 (PputA514_4767) and phaC1 (PputA514_5834) were cloned into pPvan, generating the pTJ4C1 vector. The three plasmids were subsequently transformed into strain A514, generating the recombinant strains A_Pvan, A_alkKphaGC1, and A_phaJ4C1, respectively (Table 3). A_alkKphaGC1 was subsequently characterized for cell growth and PHA production, while A_phaJ4C1 and A_Pvan were used as reference strains.

TABLE 3.

Plasmids and bacterial strains used in this study

Strain or plasmid	Relevant characteristic(s)	Source and/or reference
Plasmids
pPROBE-TT		30
pPvan	pPROBE-TT derivative, P. putida A514 promoter of vanAB (Pvan)	This study
pTJ4C1	pPROBE-TT derivative, Pvan, phaJ4, and phaC1	This study
pTphaGC1	pPROBE-TT derivative, Pvan, phaG, alkK, and phaC1	This study
pTPxylA	pPROBE-TT derivative, P. putida A514 promoter of xylA (PxylA)	This study
pTRPxylA	pPROBE-TT derivative, PxylA, and xylR	This study
pTPxylAphaGC1	pPROBE-TT derivative, PxylA, phaG, alkK, and phaC1	This study
pTPxylAphaJ4C1	pPROBE-TT derivative, PxylA, phaJ4, and phaC1	This study
Strains
P. putida A514	Wild type	Dennis Gross lab (TAMU) (17)
AT	A514 carrying pPROBE-TT	This study
APvan	A514 carrying pPvan	This study
AphaJ4C1	A514 carrying pTJ4C1	This study
AalkKphaGC1	A514 carrying pTphaGC1	This study
APxylA	A514 carrying pTPxylA	This study
AR-PxylA	A514 carrying pTRPxylA	This study
Axyl_phaJ4C1	A514 carrying pTPxylAphaJ4C1	This study
Axyl_alkKphaGC1	A514 carrying pTPxylAphaGC1	This study

Under nitrogen excess conditions, A_alkKphaGC1 showed significant improvement in PHA titer, which increased to 90 mg/liter, although levels of cell growth and vanillic acid utilization for all three strains were similar (Table 1 and Fig. S3A). Under nitrogen limitation conditions, A_alkKphaGC1 showed improvement in cell biomass compared to that of A_Pvan (P = 0.016; Table 1 and Fig. S3B). In addition, A_alkKphaGC1 produced 54% and 166% more PHA than did A_phaJ4C1 and A_Pvan, respectively, reaching ∼62 mg/liter (Table 1 and Fig. S3B). Corresponding to the higher CDW and PHA titer, more vanillic acid was consumed by A_alkKphaGC1 than by A_Pvan (P = 0.006, Table 1 and Fig. S3B).

These data demonstrate two aspects of the mechanism for PHA biosynthesis from vanillic acid. First, although increasing 3-hydroxyacyl-CoA from either the de novo fatty acid biosynthesis pathway or β-oxidation pathway improved PHA production, overexpression of phaG and alkK contributed more to PHA production in strain A514, especially under conditions of excess nitrogen (Table 1 and Fig. S3). Second, when A514 cells were grown under conditions favoring PHA accumulation (nutrition imbalance), enhancing the expression levels of phaG and alkK coimproved cell growth and PHA biosynthesis. On the basis of these results, we experimentally validated the role of phaG and alkK in PHA production and cell growth, and we suggest that this property could be further exploited to coenhance cell growth and PHA production.

Screening an inducible strong promoter to regulate the transcription levels of phaG and alkK from strain A514.

In A_alkKphaGC1, the PHA synthetic gene cluster from pTphaGC1 was constitutively expressed in the presence of vanillic acid, as the promoter P_van was induced by vanillic acid (Table 3). PHA synthesis, thereby, was turned on during the initial cell growth period, likely causing a burden on A_alkKphaGC1. Consequently, to further increase PHA production, an alternative inducible strong promoter was required to regulate the relevant gene expression.

For this purpose, a putative xylose catabolism gene cluster (xylAFGH), which appeared to be specifically induced by xylose, was targeted based on our previous genomics data (Fig. 2A) (16). The promoter sequence (P_xylA) is located in the 160-bp intergenic region shared by the xylR (xylose operon regulatory protein) and xylA (xylose isomerase) genes. It was predicted by the PePPER webserver and cloned into the promoter probe vector pPROBE-TT, with a gfp (green fluorescent protein gene) reporter gene, generating the pTP_xylA construction (29, 30) (Table 3). When A_PxylA (carrying pTP_xylA in A514) was grown in M9 medium supplemented with 15 mM vanillic acid (rather than with xylose), the measured green fluorescence was close to background levels observed in A514 cells harboring the plasmid pPROBE-TT (designated A_T) (Fig. 2B, bars 1 and 2, and Table 3). In contrast, an obvious green fluorescent signal was detected with the addition of xylose, confirming that the activity of P_xylA was specifically induced by xylose (Fig. 2B, bar 3).

FIG 2.

Effect of the regulator, regulator-induced expression time, and inducer concentrations on the activity of the promoter P_xylA. (A) Organization of the putative A514 xylose catabolism locus, as well as reporter constructs used to detect activity of P_xylA. P, promoter; gfp, green fluorescent protein gene. (B) Activities of P_xylA and P_van under various inducing conditions. For bars 1 to 5, either 0 mM or 0.2 mM (final concentration) xylose was added when strains were at the mid-exponential phase. After 6 h of induction, green fluorescence intensity was measured to determine the promoter activity. Bars 6 and 7 represent the P_van activity in A_Pvan at the mid-exponential phase and the stationary phase, respectively. Bars 8 and 9 represent the XylR-P_xylA activity in A_R-PxylA at the stationary phase, which was exposed to 0 mM or 2 mM xylose for 12 h, respectively. pTT, pPROBE-TT; Con, concentration. (C) Time point-dependent induction of XylR-P_xylA activity. Response of XylR-P_xylA to different growth phases at which 0.2 mM xylose (final concentration) was added. Fluorescence intensity was measured every 3 h until A_R-PxylA entered the stationary phase. (D) Xylose concentration-dependent induction of XylR-P_xylA activity. A_R-PxylA at the mid-exponential phase was exposed to different concentrations of xylose for 6 h. For panels B, C, and D, all of the strains were cultivated in M9 medium supplemented with 15 mM vanillic acid. Fluorescence intensity, which was detected to determine the promoter activity, is expressed in arbitrary units normalized for 10⁶ CFU.

To maximize the activity of P_xylA, three crucial factors were investigated, as follows: the regulator, induced expression time, and inducer concentrations. First, xylR (PputA514_3011) enhanced the activity of P_xylA. xylR is located immediately upstream of the xylAFGH cluster, with its transcription oriented in the opposite direction, suggesting a role in modulating xylAFGH expression in response to xylose (Fig. 2A). Therefore, we created a recombinant strain, A_R-PxylA, bearing pTRP_xylA (Table 3). When assayed in A_R-PxylA, the promoter activity was distinctly enhanced, 2.4-fold higher than that with P_xylA (Fig. 2B, bars 3 and 5), confirming that XylR is a positive regulator for P_xylA. Second, a higher promoter activity was found when xylose was added during the mid-exponential phase. To examine the effect of the induced expression time on the P_xylA activity, 0.2 mM xylose (final concentration) was added at three time points (during the early-to-mid exponential phase), and the level of green fluorescence was measured from time of addition to stationary phase. When xylose was added during the early exponential phase (optical density at 600 nm [OD₆₀₀], ∼0.36), promoter activity in A_PxylA was quickly switched on, reaching highest activity at 3 h and sustained at this level until stationary phase (15 h after induction began). In contrast, when xylose was added during the mid-exponential phase (OD₆₀₀, ∼0.56 and ∼0.78), promoter activity increased along with the induced expression time and reached the highest activity at 12 h (stationary phase). Since higher green fluorescence intensity was detected in A_PxylA when xylose was supplemented at the mid-exponential phase than at the early exponential phase, xylose was added at the mid-exponential phase (e.g., at OD₆₀₀ of ∼0.78) for the following study (Fig. 2C). Third, there was an inducer concentration-dependent effect on promoter activity, with a final concentration of 2 mM xylose being determined as the appropriate concentration. To assess the effect of different xylose concentrations, A_R-PxylA was grown in M9 medium with 15 mM vanillic acid, reaching an OD₆₀₀ of ∼0.78, and then exposed to various inducer concentrations (from 0.02 mM to 20 mM) for 6 h (Fig. 2D). It was determined that 2 mM xylose was sufficient to activate P_PxylA.

Therefore, the optimized promoter activity of XylR-P_xylA in the expression vector pTRP_xylA was achieved when adding 2 mM xylose at an OD₆₀₀ of ∼0.78 and was approximately 4-fold higher than that of A_Pvan under 15 mM vanillic acid conditions (Fig. 2B, bars 7 and 9).

PHA production was further enhanced via transcription level optimization.

Next, the expression vector pTRP_xylA was employed to coexpress the two PHA synthetic gene clusters in A514. Thus, two recombinant strains, A_{xyl_phaJ4C1} and A_{xyl_alkKphaGC1}, were generated. Substrate utilization, cell growth, and PHA titers of these two strains were assessed (Table 1 and Fig. S4). (i) More vanillic acid was utilized by A_{xyl_alkKphaGC1} than by A_alkKphaGC1 under nitrogen deprivation conditions (P = 6 × 10⁻⁶, Table 1 and Fig. S3 and S4). A similar trend was observed in A_{xyl_phaJ4C1}, in contrast to in A_phaJ4C1 (P = 1 × 10⁻⁴). (ii) In comparison with A_alkKphaGC1, cell growth was further improved in A_{xyl_alkKphaGC1}, especially under nitrogen starvation conditions. The growth rate and cell biomass of A_{xyl_alkKphaGC1} were both enhanced under the low-N conditions (P = 0.008 for doubling time and P = 0.019 for CDW; Table 1). In addition, a higher cell biomass of A_{xyl_alkKphaGC1} was also obtained under the high-N conditions (P = 0.003; Table 1). (iii) A_{xyl_alkKphaGC1} and A_{xyl_phaJ4C1} produced more PHA titer than A_alkKphaGC1 and A_phaJ4C1, respectively (P values < 0.05), whereas A_Pvan and A_R-PxylA produced PHA titers similar to that of A514 (P values > 0.05; Table 1 and Fig. S3 and S4). This indicated that the PHA titer was improved by regulating the transcription levels of key genes. A514 and A_Pvan cultures (carrying promoter P_van in the plasmid pPROBE-TT) were grown in vanillic acid-M9 medium, while A_R-PxylA cells (carrying promoter R-P_xylA in the plasmid pPROBE-TT) were grown in vanillic acid-M9 medium with 2 mM xylose inducer added at the mid-exponential phase (OD₆₀₀, ∼0.78). The similar PHA titers among the three strains (P values > 0.05) suggested that 2 mM xylose has very little contribution to PHA synthesis. (iv) Although the highest PHA content (wt %, 66) was reached in A_{xyl_alkKphaGC1} under low-N conditions, the greatest PHA titer was produced by A_{xyl_alkKphaGC1} under high-N conditions, reaching ∼250 mg/liter in a single fermentation batch (Table 1 and Fig. S4B). This would be the result of more cell biomass being obtained in A_{xyl_alkKphaGC1} under high-N conditions, demonstrating the role of the cell biomass in PHA production. (v) Gas chromatography-mass spectrometry (GC-MS) composition analysis revealed that the PHA produced were medium-chain-length PHA (mcl-PHA) (C₆ to C₁₄). Compared to that in A_R-PxylA, the major units in A_{xyl_alkKphaGC1} PHA polymers were shifted from shorter-carbon-chain units to longer-carbon-chain units. Under nitrogen starvation conditions, 3-hydroxydecanoate (3HD, C₁₀) (∼33%) and 3-hydroxytetradecanoate (3HTD, C₁₄) (∼23%) were the major units in A_{xyl_alkKphaGC1}, while the 3HD and 3HTD content in mcl-PHA synthesized in A_{xyl_phaJ4C1} also increased. In contrast, the major PHA component of A_R-PxylA was 3-hydroxyhexanoate (3HHx, C₆), ∼45% (Table 4). Under nitrogen excess conditions, 3HD (∼41%) and 3-hydroxyoctanoate (3HO, C₈) (∼31%) were the major monomers in A_{xyl_alkKphaGC1}, while 3HHx accounted for approximately ∼57% of the PHA polymers of A_R-PxylA. In particular, 3HTD (∼3%) was detected in mcl-PHA produced by A_{xyl_alkKphaGC1}, whereas it was not found in mcl-PHA produced by A_R-PxylA or A_{xyl_phaJ4C1} (Table 4).

TABLE 4.

GC compositional analysis of mcl-PHA produced by recombinant P. putida A514 strains^a

Strain	Culture	PHA composition (mol%)b
		3HHx (C6)	3HO (C8)	3HD (C10)	3HDD (C12)	3HTD (C14)
APvan	High N (1 g/liter)	56.0 ± 4.8	ND	21.7 ± 1.2	22.3 ± 1.0	ND
AphaJ4C1		53.6 ± 2.1	ND	25.7 ± 1.5	20.7 ± 1.7	ND
AalkKphaGC1		15.0 ± 1.0	20.0 ± 1.9	47.8 ± 3.1	14.9 ± 2.2	2.3 ± 0.6
AR-PxylA		57.4 ± 5.0	ND	19.9 ± 0.6	22.7 ± 1.0	ND
Axyl_phaJ4C1		56.8 ± 1.0	ND	20.6 ± 0.8	22.6 ± 1.1	ND
Axyl_ alkKphaGC1		10.2 ± 0.8	30.9 ± 1.8	41.2 ± 1.5	15.2 ± 0.7	2.5 ± 0.1
APvan	Low N (65 mg/liter)	44.2 ± 0.3	8.3 ± 1.0	20.2 ± 1.2	13.4 ± 2.0	13.9 ± 0.6
AphaJ4C1		33.2 ± 0.5	11.9 ± 0.8	25.8 ± 2.3	12.1 ± 1.5	17.0 ± 1.0
AalkKphaGC1		17.0 ± 1.1	18.3 ± 1.0	34.1 ± 2.1	12.1 ± 2.9	18.5 ± 0.3
AR-PxylA		45.4 ± 2.8	7.3 ± 0.2	16.6 ± 2.0	14.9 ± 0.7	15.8 ± 0.4
Axyl_phaJ4C1		32.5 ± 0.9	9.8 ± 0.9	24.7 ± 2.3	13.8 ± 1.2	19.2 ± 0.1
Axyl_ alkKphaGC1		16.0 ± 0.8	15.2 ± 2.6	32.8 ± 1.7	13.1 ± 0.4	22.9 ± 0.1

^aVanillic acid (15 mM) was used as the carbon source.

^b3HHx, 3-hydroxyhexanoate; 3HO, 3-hydroxyoctanoate; 3HD, 3-hydroxydecanoate; 3HDD, 3-hydroxydodecanoate; 3HTD, 3-hydroxytetradecanoate; ND, not detected.

DISCUSSION

PHA are polyesters that are synthesized by microbes as storage materials and are substantially accumulated under conditions where fatty acids are employed as the feedstock, without the requirement of inorganic ion deprivation. However, the high cost of fatty acids hinders commercialization of PHA production. Therefore, lignin, a by-product generated in large volumes by lignocellulosic biorefineries, is the ideal substrate for PHA production. When lignin is utilized as a carbon source, PHA accumulation usually occurs under conditions of fluctuating nutrient availability due to its function as a carbon and energy storage molecule. However, this nutrition imbalance generally limits cell growth, which is a critical factor for the total PHA titer and thus represents one of the major hurdles in a consolidated lignin-to-PHA bioconversion scheme. Previous studies have reported strategies for heterologous PHA production without compromising cell biomass in E. coli, with either glucose or glycerol as the sole carbon source (6, 7, 31–33), but not, as yet, with lignin or its derivatives. This is probably due to E. coli's inability to degrade lignin, restricting it to glycerol or other carbohydrates. Therefore, testing and comparing the PHA synthetic pathways when lignin or its derivatives are provided as the main source of carbon may serve as an essential foundation for engineering these two traits in Pseudomonas species, organisms known to be capable of both lignin degradation and PHA accumulation.

First, our work extended molecular tools for Pseudomonas strains by identifying a xylose-inducible promoter, P_xylA, which is specifically induced by xylose at a concentration (2 mM) that is sufficient to stimulate its activity without interfering with cell growth when a lignin derivative is utilized as the sole carbon source. Moreover, the adjacent xylose operon regulatory protein, XylR, was validated to stimulate its activity. Thus, XylR-P_xylA might be superior for the control of strongly transcribed genes, complementary to an inducible weak promoter, P_van (16, 34).

Second, our study unveiled unique features of the P. putida A514 mcl-PHA biosynthetic genes when this strain is grown on a lignin derivative. Comparison of low-N versus high-N conditions revealed that phaJ4 exhibited the highest induction (Fig. 1F). This contradicts the current notion that phaG, but not phaJ4, is significantly upregulated when P. putida KT2440 is grown with glycerol (6). The stress-induced phaJ4 gene should play an important role in PHA biosynthesis under low-N conditions, which is consistent with the result where A_{xyl_phaJ4C1} generated more mcl-PHA under low-N conditions than did A_{R_PxylA} (Table 1). In addition, comparison of the expression levels of phaG and alkK to phaJ4 leads to the hypothesis that phaG and alkK not only have a role in PHA biosynthesis but are also likely to be essential in cell growth (Fig. 2). This hypothesis was validated by the experiment where A_{xyl_alkKphaGC1} coenhanced the PHA titer and cell growth under both low-N and high-N conditions (Table 1). Furthermore, total PHA titer reached the highest level (∼250 mg/liter) under high-N conditions, although the PHA content in A_{xyl_alkKphaGC1} was not the highest (∼35 wt% CDW), possibly a consequence of A_{xyl_alkKphaGC1} having the highest cell dry weight.

Third, the mcl-PHA, synthesized in A_{xyl_phaJ4C1} and A_{xyl_alkphaGC1}, have a greater number of longer-chain monomers (Table 4). This phenotype may be ascribed to the functions of phaJ4, phaG, and alkK. phaJ4_Pa in P. aeruginosa encodes an (R)-specific enoyl-CoA hydratase that shows specificity for the hydration of longer-chain-length enoyl-CoAs (C₆ to C₁₂) (35). The higher contents of 3HO, 3HDD, and 3HTD in A_{xyl_phaJ4C1} cells under nitrogen deprivation indicated the homologous phaJ4 gene in A514 likely has greater preference for C₈, C₁₀, and C₁₄ substrates. In addition, the function of PhaG (in P. putida KT2440) has been shown as a 3-hydroxyacyl-ACP thioesterase to produce 3-hydroxy fatty acids, while AlkK (in Pseudomonas oleovorans) is involved in catalyzing the conversion of 3-hydroxy fatty acids to the PHA precursor 3-hydroxyacyl-CoA (6, 28). The closest homolog in A514 to the AlkK gene is the gene encoding a long-chain fatty acid-CoA ligase (PputA514_1839). Thus, the longer chain PHA monomers (C₁₀ and C₁₄) produced in A_{xyl_alkphaGC1} appear to be the result of alkK substrate specificity in A514. This is consistent with the previous report that the mcl-PHA synthesized in recombinant E. coli, carrying the vector to coexpress PhaG, PhaC1, and the medium-chain fatty acid-CoA ligase (AlkK-like protein in KT2440), mainly contain C₈ and C₁₀ (6).

Finally, compared to other studies on PHA biosynthesis from lignin or its derivatives, higher PHA production (246 mg/liter) and cell dry weights (715 mg/liter) were generated in batch cultivation (Table 5). Although a slightly higher PHA titer (252 mg/liter) was produced by P. putida KT2440 from alkaline pretreated liquor, it was obtained through fed-batch fermentation and high substrate concentrations (including those of glucose, acetate, coumaric acid, and ferulic acid) (3). Additionally, the monomeric composition and content of PHA produced by bacterial strains from lignin or its derivatives are very important, as they can greatly affect the properties of PHA copolymers (36). Both Cupriavidus basilensis B-8 and Pandoraea sp. strain ISTKB produce short-chain-length PHA (scl-PHA, C₃ to C₅), whereas the monomeric content of mcl-PHA generated by P. putida KT2440 from alkaline pretreated liquor was not reported (Table 5) (10, 37, 38). In contrast, longer-chain monomers produced by A514 in general and the large amount of longer monomers (e.g., 3HTD) synthesized by A_{xyl_alkphaGC1} in particular would have higher crystallinity and better tensile strength (36), which may endow mcl-PHA with novel mechanical properties and wider array of potential applications. This demonstrated that increasing the expression levels of phaG and alkK is a more efficient strategy to coenhance PHA titer and cell growth, and that P. putida A_{xyl_alkKphaGC1} is a very competitive strain for the production of PHA from lignin derivatives.

TABLE 5.

Comparison of the mcl-PHA or scl-PHA (PHB) production from lignin or its derivatives

Strain	Carbon substrate	Cultivation mode	Cultivation time (h)	CDW (mg/liter)	mcl-PHAa titer (mg/liter)	scl-PHAb titer (mg/liter)	PHA content (wt%)	Source or reference
Cupriavidus basilensis B-8	5 g/liter soluble lignin	Shake flask batch	120	736	NDc	128	17	38
Pandoraea sp. ISTKB	2 g/liter vanillic acid	Shake flask batch	96	215	ND	72	33	37
P. putida JCM13063	10 g/liter vanillic acid	Shake flask batch	72	210	ND	trd	<1	10
P. putida Gpo1	10 g/liter vanillic acid	Shake flask batch	72	140	ND	tr	<1	10
P. putida KT2440	2 g/liter coumaric acid	Shake flask batch	48	470	160	ND	34	3
P. putida KT2440	2 g/liter ferulic acid	Shake flask batch	48	436	170	ND	39	3
P. putida Axyl_alkKphaGC1	2.5 g/liter vanillic acid	Shake flask batch	50	715	246	ND	34	This study
P. putida KT2440	90% alkaline-pretreated liquor	Fed-batch fermentation	48	787	252	ND	32	3

^amcl-PHA, medium-chain-length PHA.

^bscl-PHA, short-chain-length PHA.

^cND, not determined.

^dtr, trace.

Conclusions.

In summary, this study revealed specific features of the PHA synthesis pathways for P. putida A514 when grown using a lignin derivative and demonstrated a rational strategy to coengineer cell growth and PHA production. As improvements in cell growth and bioproduct production are crucial and shared goals in the sustainable development of lignocellulosic biorefineries (11), our findings should be valuable not just to the production of PHA, but also to that of a wide variety of biofuels and biochemicals. Further genetic strategies, including optimization of the carbon flux either from de novo fatty acid biosynthesis or β-oxidation pathways to PHA synthesis (e.g., deletion, insertion, or substitution of the genes involved in fatty acid metabolism, global nitrogen regulation, and so on), and fermentation optimization at flask level or bioreactor level in the near future would further improve the PHA content in cells, which then could be integrated into lignin-consolidated bioprocessing to promote the production of bioproducts.

MATERIALS AND METHODS

Bacterial strains and growth conditions.

The strains and plasmids used in this study are summarized in Table 3. E. coli DH5α was used for all molecular manipulations to construct plasmids. This strain was grown on Luria-Bertani (LB) medium at 37°C. During recombinant plasmid construction, 15 μg/ml tetracycline was added to the medium. For P. putida A514 and cultivation of its recombinant strains, all strains were grown on LB agar for 24 h at 30°C. A single colony was inoculated into 5 ml of 15 mM vanillic acid-M9 mineral medium (with a high concentration NH₄Cl [1 g/liter]) and grown at 30°C with shaking at 200 rpm (39). When the stationary phase was reached, 0.2 ml of culture solution was transferred into 20 ml of fresh vanillic acid-M9 mineral medium with either high-concentration NH₄Cl (1 g/liter) or low-concentration NH₄Cl (65 mg/liter) at 30°C, 200 rpm, and passaged two times (with 1% [vol/vol] inoculation) under the same culture condition. Subsequently, 1% (vol/vol) of seed inoculum for shake flask cultivation was incubated in 100 ml of vanillic acid-M9 mineral medium with high-N or low-N conditions at 30°C and 200 rpm. Tetracycline (15 μg/ml) was added to the medium when culturing P. putida A514 recombinant strains. Cell growth was spectrophotometrically monitored by measuring the optical densities at 600 nm (OD₆₀₀). For the cultivation of strains A_R-PxylA, A_{xyl_phaJ4C1}, and A_{xyl_alkKphaGC1}, xylose was introduced to a final concentration of 2 mM when the OD₆₀₀ was either ∼0.78 (in the high-nitrogen culture medium) or ∼0.22 (in the low-nitrogen culture medium). The doubling times (DT) of all the strains were calculated during the exponential phase, based on the formula DT = t · [log 2/(log N_t − log N₀)], where t = duration, N_t = final cell concentration, and N₀ = initial cell concentration. To determine the number of living cells of N₀ and N_t, 100 μl of fermentation culture was serially diluted and plated on LB agar plates. CFU/ml was assessed as cell concentration. All of the experiments were performed in biological triplicate and technical duplicate.

Determination of gene transcription levels by qRT-PCR.

qRT-PCR was performed as previously described (40). Briefly, a total volume of 10 ml of P. putida A514 cell culture was harvested from either high-nitrogen (OD₆₀₀, ∼0.78) or low-nitrogen (OD₆₀₀, ∼0.22) culture medium during the mid-exponential phase. The total RNA was isolated using a TransZol Up Plus RNA kit (Transgen Biotech). cDNA libraries were synthesized with a TransScript one-step genomic DNA (gDNA) removal and cDNA synthesis supermix kit (Transgen Biotech), employing 0.5 μg of total RNA as the template for each sample. The qRT-PCRs were performed using the SYBR Premix Ex Taq II kit (TaKaRa) with the real-time PCR detection system (Applied Biosystems VII A7). The primers used in qRT-PCR are listed in Table 2. Gene expression levels were assessed by comparing the ratios of the housekeeping gene rpoD (PputA514_4978, encoding the housekeeping σ⁷⁰ factor) threshold cycle (C_T) values to target gene C_T values using the following equation: ratio of rpoD/target = 2^{CT(rpoD) − CT(target)} (6). The n-fold induction was determined by the expression levels of the genes of cells grown on low-N M9 medium versus those of cells grown on high-N M9 medium. All of the samples were prepared and analyzed in biological triplicate and technical duplicate.

Plasmid and strain construction.

A 315-bp PCR fragment containing the promoter fragment of the vanAB genes (encoding vanillate demethylase A and vanillate O-demethylase oxidoreductase in the vanillic acid degradation pathway), P_van, was amplified from P. putida A514 genomic DNA and subsequently ligated into the plasmid pPROBE-TT (29) through HindIII and EcoRI, to generate the plasmid pPvan. pPvan was used as the expression vector for coexpression of PHA synthesis genes. The putative 3-hydroxyacyl-ACP thioesterase gene (phaG, PputA514_1318), long-chain fatty acid-CoA ligase gene (alkK, PputA514_1839), enoyl-CoA hydratase gene (phaJ4, PputA514_4767), and poly(3-hydroxyalkanoate) polymerase 1 gene (phaC1, PputA514_5834) were amplified by PCR from the genome of A514. These three DNA fragments (phaG, alkK, and P_van) were fused by gene splicing by overlap extension (SOE) PCR to produce a 2.8-kb PCR fragment, whereas the DNA fragments of phaJ4 and P_van were fused to produce a 0.78-kb PCR fragment. The two overlap PCR fragments were digested with HindIII and KpnI, respectively, while the phaC1 fragment was digested with KpnI and EcoRI. DNA fragments containing P_van, phaG, alkK, and phaC1 were cloned into the pPROBE-TT to construct plasmid pTphaGC1. Similarly, DNA fragments containing P_van, phaJ4, and phaC1 were cloned into the pPROBE-TT to construct the pTJ4C1 vector (Table 3).

For the inducible promoter investigation, a chromosomal fragment comprising the 140-bp xylR-xylA intergenic region was PCR amplified and cloned into the plasmid pPROBE-TT through HindIII and XbaI, generating the pTP_xylA vector (29). To construct pTRP_xylA, a genomic fragment containing the ORF region of xylR and the xylR-xylA intergenic region was PCR amplified, digested by HindIII and XbaI, and then ligated into the pPROBE-TT (Table 3).

The plasmid pTRP_xylA was used as the expression vector to coexpress the PHA synthesis genes. A DNA fragment containing phaC1 and either phaG and alkK or phaJ4 was amplified and cloned into pTRP_xylA to construct pTP_xylAphaGC1 and pTP_xylAphaJ4C1, respectively (Table 3).

All plasmids were verified by DNA sequencing and were subsequently transformed into A514 cells through chemical transformation (41), followed by selection on LB agar supplemented with 15 μg/ml tetracycline.

Green fluorescence signal detection.

Cells from 15 mM vanillic acid-M9 medium cultures were harvested, centrifuged, washed once with 10 mM phosphate buffer, and then resuspended in phosphate buffer. Fluorescence was measured on an LS50B luminescence spectrometer (PerkinElmer Instruments, Norwalk, CT) at an excitation wavelength of 490 nm, an emission wavelength of 510 nm, and emission/excitation slit widths of 8 nm. Intensity readings are represented by arbitrary units and were normalized to a cell density of 10⁶ CFU/ml. All of the experiments were performed in biological triplicate and technical duplicate.

Analytical procedures.

To measure the concentration of vanillic acid in the medium, 1-ml cell samples were centrifuged for 5 min at 16,000 × g, and the concentration of vanillic acid in the supernatant was measured via absorbance at 289 nm, as previously described (16). For PHA production analysis, PHA biosynthesis and monomer compositions were analyzed by gas chromatography-tandem mass spectrometry (GC-MS/MS; Agilent, Santa Clara, CA). PHA titer was defined as the PHA concentration (mg) per one liter of shake flask batch culture, while PHA content was defined as the percentage of PHA concentration (mg) to cell dry weight (mg). Liquid cultures (100 ml) were harvested at the stationary phase and centrifuged at 9,000 × g for 15 min at 4°C, washed twice with 15 ml Nanopure water, lyophilized (Lyophilizer Alpha 1-4 LSC; Martin Christ Gefriertrocknungsanlagen GmbH, Osterode am Harz, Germany) at −59°C and 0.140 × 10⁵ mPa for a minimum of 24 h, and weighed. The lyophilized cells were dissolved in 2 ml of methanol-sulfuric acid (85:15) solution and 2 ml chloroform (containing a final concentration of 0.01 mg/ml 3-methylbenzoic acid as the internal standard) and then incubated at 100°C for 4 h to methanolysis. After cooling, 1 ml of demineralized water was added to the organic phase, which contained the resulting methyl esters of monomers. The organic phase was filtered and analyzed by GC-MS/MS (Agilent series 7890B GC system; Santa Clara, CA), coupled with a 7000D MS detector (extractor ion [EI]; 70 eV). An aliquot (1 μl) of the organic phase, which was properly diluted, was injected into the gas chromatography column. Separation of compounds was achieved using an HP-5 capillary column (5% phenyl-95% methyl siloxane, 30 m by 0.32-mm inside diameter [i.d.] by 0.25-μm film thickness). Helium was used as the carrier gas at a flow rate of 1.1 ml/min. The injector and transfer line temperatures were set at 280°C and 300°C, respectively. The oven temperature program was as follows: initial temperature of 60°C for 3 min, then from 60°C to 200°C at a rate of 5°C per min, then 200°C for 1 min, and finally from 200°C to 280°C a rate of 15°C per min. The EI mass spectrum was recorded in full scan mode (m/z 40 to 550). PHA standard samples (3HHx, 3HO, 3HD, 3HDD, and 3HTD), which were purchased from Sigma-Aldrich, Larodan, and TRC, were also analyzed by GC-MS/MS according to the method above. The retention time (RT), patterns of fragment ion peaks for known PHA standards, and resulting mass spectra available compared with the GC-MS library database (NIST08s) were all employed to determine the PHA monomers in the study samples (7, 23). All of the experiments were performed in biological triplicate and technical duplicate. Differences between groups (including cell dry weight, vanillic acid concentration, and PHA titer) were evaluated using two-tailed unpaired t tests. Those with P values of <0.05 are considered significant.