Overexpression of RuBisCO form I and II genes in Rhodopseudomonas palustris TIE-1 augments polyhydroxyalkanoate production heterotrophically and autotrophically
ABSTRACT
With the rising demand for sustainable renewable resources, microorganisms capable of producing bioproducts such as bioplastics are attractive. While many bioproduction systems are well-studied in model organisms, investigating non-model organisms is essential to expand the field and utilize metabolically versatile strains. This investigation centers on Rhodopseudomonas palustris TIE-1, a purple non-sulfur bacterium capable of producing bioplastics. To increase bioplastic production, genes encoding the putative regulatory protein PhaR and the depolymerase PhaZ of the polyhydroxyalkanoate (PHA) biosynthesis pathway were deleted. Genes associated with pathways that might compete with PHA production, specifically those linked to glycogen production and nitrogen fixation, were deleted. Additionally, RuBisCO form I and II genes were integrated into TIE-1’s genome by a phage integration system, developed in this study. Our results show that deletion of phaR increases PHA production when TIE-1 is grown photoheterotrophically with butyrate and ammonium chloride (NH4Cl). Mutants unable to produce glycogen or fix nitrogen show increased PHA production under photoautotrophic growth with hydrogen and NH4Cl. The most significant increase in PHA production was observed when RuBisCO form I and form I & II genes were overexpressed, five times under photoheterotrophy with butyrate, two times with hydrogen and NH4Cl, and two times under photoelectrotrophic growth with N2 . In summary, inserting copies of RuBisCO genes into the TIE-1 genome is a more effective strategy than deleting competing pathways to increase PHA production in TIE-1. The successful use of the phage integration system opens numerous opportunities for synthetic biology in TIE-1.
IMPORTANCE
Our planet has been burdened by pollution resulting from the extensive use of petroleum-derived plastics for the last few decades. Since the discovery of biodegradable plastic alternatives, concerted efforts have been made to enhance their bioproduction. The versatile microorganism Rhodopseudomonas palustris TIE-1 (TIE-1) stands out as a promising candidate for bioplastic synthesis, owing to its ability to use multiple electron sources, fix the greenhouse gas CO2, and use light as an energy source. Two categories of strains were meticulously designed from the TIE-1 wild-type to augment the production of polyhydroxyalkanoate (PHA), one such bioplastic produced. The first group includes mutants carrying a deletion of the phaR or phaZ genes in the PHA pathway, and those lacking potential competitive carbon and energy sinks to the PHA pathway (namely, glycogen biosynthesis and nitrogen fixation). The second group comprises TIE-1 strains that overexpress RuBisCO form I or form I & II genes inserted via a phage integration system. By studying numerous metabolic mutants and overexpression strains, we conclude that genetic modifications in the environmental microbe TIE-1 can improve PHA production. When combined with other approaches (such as reactor design, use of microbial consortia, and different feedstocks), genetic and metabolic manipulations of purple nonsulfur bacteria like TIE-1 are essential for replacing petroleum-derived plastics with biodegradable plastics like PHA.
INTRODUCTION
Recent improvements in genetic engineering tools have enabled scientists to systematically engineer organisms that produce various value-added chemicals, including biofuels, therapeutic products, food, and bioplastics (1–3). At first, most of these engineering efforts were focused on widely used model organisms, such as Escherichia coli, Saccharomyces cerevisiae, and Synechococcus sp. MIT9509 (4–8). This emphasis resulted in an array of genetic tools that have been effectively developed for valuable biomolecule biosynthesis and for conducting physiological studies (8–22). However, the reliance on organic carbon as the primary carbon source poses a limitation for heterotrophic model organisms, contributing to elevated bioproduction costs (15–17). Recent studies have highlighted numerous advantages in utilizing non-model organisms for bioproduction (23–25). Over the past decade, one such group of microbes that has gained attention is the purple nonsulfur bacteria exemplified by Rhodopseudomonas palustris TIE-1 (TIE-1) (26–28), Rhodospirillum rubrum (29, 30) and Rhodomicrobium (21, 31). TIE-1, a Gram-negative purple nonsulfur photosynthetic bacterium is renowned for its versatile metabolism, rendering it an excellent host for diverse bioproduction and pathway studies (26, 27). TIE-1 exhibits four primary metabolisms: chemoautotrophy, photoautotrophy, chemoheterotrophy, and photoheterotrophy (26, 27). These different metabolisms enable TIE-1 to use a wide variety of carbon sources such as carbon dioxide (CO2) and many organic acids. TIE-1 can use ammonium salts such as ammonium chloride (NH4Cl) or fix nitrogen from dinitrogen gas (N2) as a nitrogen source (26). Moreover, it can use multiple electron sources including hydrogen (H2) or ferrous iron (Fe (II)). One of the most appealing features of TIE-1 is its ability to acquire electrons directly from a poised electrode, which enables us to use it in microbial electrosynthesis (MES) (27, 32–35). MES is a system in which microorganisms are used to produce valuable compounds using their ability to obtain electrons from the bioelectrical reactor (33, 35, 36). Using electrons from a poised electrode or a solar panel-powered MES system, TIE-1 produces biodegradable plastic and biofuel using CO2 as a carbon source, N2 as a nitrogen source, and light as an energy source (27, 37). These represented the first steps toward a sustainable and carbon-neutral process for bioproduction using TIE-1 in MES. Besides its ability to utilize various substrates, TIE-1’s metabolic diversity also makes it an extraordinary model organism for pathway investigation (38, 39). For example, the use of RuBisCO mutants allowed us to study the association between the Calvin–Benson–Bassham (CBB) cycle in carbon fixation and extracellular electron transport (40). Similarly, a TIE-1 ∆pioABC mutant was used to investigate the electron uptake mechanism during photoferrotrophic and electrotrophic growth conditions (38). During these studies, mutants were first generated and grown under heterotrophic growth conditions in which the electron uptake machineries were not involved. These strains were then switched to autotrophic and electron uptake growth conditions to further understand their role in metabolism (38, 40). Not only do these studies provide insights into TIE-1’s metabolism, but they also open doors for a deeper understanding of other closely related purple nonsulfur bacteria, such as Rhodopseudomonas palustris CGA009, which has been studied extensively for biohydrogen production (41, 42).
The available genetic tools for TIE-1 are limited compared to widely used model organisms, with most tools being based on homologous recombination (34, 40). Fig. S1 illustrates the two-step homologous recombination process for achieving gene integration in TIE-1. Although this process results in markerless strains, it is time-consuming, with an efficiency lower than 50% (34). To reduce processing time and enhance efficiency, we explored phage recombination techniques for gene integration. This technique has gained attention in various bacteria, including Methanosarcina (43) and Mycobacterium smegmatis (44), due to its simple design and high efficiency (45, 46). The φC31 phage recombinase is commonly utilized due to its ability to function independently of a helper protein, and the recombination is unidirectional (47). For example, in Clostridium ljungdahlii, a whole butyric acid synthesis pathway was integrated into its genome by φC31 recombinase (45). In Methanosarcina spp., the φC31 recombinase achieved genome editing efficiency that is 30 times higher than that of homologous recombination (43) .
Among many value-added chemical products obtained from advanced genome engineering tools, bioplastics from microorganisms have become an attractive product (reviewed in (48)). This is particularly pertinent due to the detrimental environmental impact of excessive plastic usage in recent years. Bioplastics preserve the advantageous properties of petroleum-based plastics, such as high durability, moldability, water, and heat resistance, while also offering biocompatibility, emerge as a promising alternative (49–51). Moreover, numerous microorganisms possess the natural ability to degrade bioplastics, including the polyhydroxyalkanoate (PHA) family, either aerobically or anerobically, typically within a span of 5 to 6 weeks (50–52). However, the high feedstock cost remains an obstacle to the bioplastic’s competitiveness in the market (reviewed in (53, 54)). This issue can be addressed by using photoautotrophic microbes that can use cheap alternatives and waste feedstock (such as CO2) (27).
PHA stands out as the most extensively studied bioplastic (48, 54). The PHA pathway consists mainly of phaA, phaB, phaC, and phaZ genes. phaA encodes a ß-keto thiolase, while phaB encodes an acetoacetyl-CoA reductase (27). PhaC catalyzes the polymerization of PHA, whereas PHA depolymerase, PhaZ, catalyzes its mobilization during carbon starvation (27). Additional proteins, such as PhaP and PhaR, have also been reported to contribute to the maturation and regulation of PHA granules (55, 56). In Paracoccus denitrificans, PhaR is characterized as a repressor of PHA synthesis and acts by binding to the intergenic region of the phaC-phaP and phaP-phaR genes (55). However, PhaR is proposed to be an activator for PHA synthesis in Cupriavidus necator in a PhaP-dependent and -independent manner. Deletion of the phaR gene decreased PHA production in this organism (56). We have previously reported that TIE-1 possesses one phaR gene (Rpal_0531) using bioinformatics (27).
To improve PHA production, several gene manipulations have been undertaken directly on PHA pathway genes or genes from other pathways that may impact or compete with the PHA biosynthesis pathway (57–59). For instance, a study revealed that the deletion of the phaZ gene in Sinorhizobium meliloti increased PHA production compared to wild-type when utilizing formate generated through electrochemical CO2 reduction (57). This suggests that prevention of PHA degradation results in intracellular accumulation of PHA. Metabolic pathways such as glycogen production and nitrogen fixation are also potential competitors for bacterial PHA production (58). In Synechocystis sp. PCC 6803, PHA accumulation is reported to be linked to glycogen production under prolonged nitrogen starvation conditions. Mutants lacking the glycogen phosphorylase genes showed impaired PHA accumulation, supporting the link between glycogen and PHA synthesis under nitrogen starvation (NaNO3) conditions (58). Nitrogen fixation, which is a pathway highly demanding of electrons, could also present another competition to PHA production (60). Furthermore, PHA accumulation has also been reported to be induced by nitrogen deprivation ((N2) or NH4Cl) in many bacteria (41, 61, 62).
In another study, increase in PHA production was achieved by enhancing carbon fixation through the Calvin–Benson–Bassham (CBB) cycle in Ralstonia eutropha (now C. necator) (60). This enhancement involved the heterologous overexpression of the RuBisCO gene from Synechococcus sp. PCC 7002 in Ralstonia eutropha. Consequently, the overexpression resulted in a substantial increase in cell density, measured by optical density, by up to 89.2%. Additionally, there was a significant augmentation of the mass percent of PHA production, reaching up to 99.7% (60).
In this study, we used the φC31 integration system to integrate additional copies of the RuBisCO form I and II genes driven by the PaphII constitutive promoter into the TIE-1 genome. Additionally, to increase PHA accumulation in TIE-1, we created mutants that lack key genes in the PHA pathway, particularly focusing on the phaZ and the phaR genes. The behavior of the mutants lacking the phaR helped us elucidate whether phaR is an activator or repressor of the PHA pathway in TIE-1. In addition to exploring genes within the PHA pathway, we also examined the impact of deleting genes in pathways that might potentially compete with the PHA synthesis pathway. These include the mutants that have been previously studied in biobutanol production in the TIE-1: Δgly mutant lacking glycogen synthase and the mutant lacking the two NifA regulators nifA, Rpal_1624 & Rpal_5113, of the nitrogen fixation pathway of TIE-1 (37). We also hypothesize that like R. eutropha, overexpressing RuBisCO genes (form I and II) in TIE-1 could increase intracellular carbon abundance and hence increase PHA accumulation. We tested PHA production in all six strains and wild-type under a variety of growth conditions, including non-nitrogen-fixing conditions with NH4Cl (referred to as non-nitrogen-fixing conditions throughout) and nitrogen-fixing conditions with N2 gas (referred to as nitrogen-fixing conditions throughout).
Our results show that the deletion of the phaR gene increased PHA production (dry cell weight % wt/wt) when TIE-1 was grown photoheterotrophically with butyrate under non-nitrogen-fixing conditions. We also observed an increase in PHA production (dry cell weight % wt/wt) from the Δgly and ΔnifA under photoautotrophic growth conditions with H2 and under non-nitrogen-fixing conditions. PHA production increased in TIE-1 strains overexpressing RuBisCO form I and form I & II genes under photoheterotrophy with butyrate irrespective of the nitrogen source used and photoautotrophy with H2 under non-nitrogen-fixing conditions as well as photoelectrotrophically under nitrogen fixing conditions. We show that both gene deletion and overexpression can enhance PHA production by TIE-1. This work advances TIE-1 as a model organism for potential commercialization for PHA production and provides valuable insights for future genetic engineering endeavors aimed at enhancing bioplastic production in other purple nonsulfur bacteria.
RESULTS
Generating TIE-1 mutants and suicide plasmids with the different antibiotic markers
To improve polyhydroxyalkanoate (PHA) production, we constructed a mutant lacking the regulator of PHA pathway (ΔphaR) and a mutant lacking PHA depolymerase (ΔphaZ) (Fig. S1 and S2). The mutant lacking the glycogen synthesis gene (Δgly) as well as the double mutant lacking the nitrogen-fixing regulator nifA genes (ΔnifA1 ΔnifA2; referred as DnifA throughout) were constructed previously (37) . These strains are listed in Table 1, and the plasmids used to construct them are listed in Table 2.
TABLE 1
| Strain | Relative characteristics | Function | Source |
|---|---|---|---|
| AB415 | Wild-type Rhodopseudomonas palustris TIE-1 | Wild-type TIE-1 | (26) |
| WB065 | Wild-type TIE-1 with attB | For designing RuBisCO strains | This study |
| WB068 | Wild-type TIE-1 with φC31 integrase; Plac; lacIq; attB | For designing RuBisCO strains | This study |
| WB090 | Wild-type TIE-1 with mCherry; PaphII promoter; fd terminator; attL; attR | For designing RuBisCO strains | This study |
| WB091 | Wild-type TIE-1 with loxp-mCherry-loxp; PaphII promoter; fd terminator; attL; attR | For designing RuBisCO strains | This study |
| AB188 | TIE-1 Δgly mutant | TIE-1 mutant lacking the glycogen synthase gene | (37), this study |
| AB187 | TIE-1 ΔnifA double mutant | TIE-1 mutant lacking the nifA1&2 genes | (37), this study |
| AB186 | ΔphaR | TIE-1 mutant lacking the phaR gene | This study |
| AB189 | ΔphaZ | TIE-1 mutant lacking the phaZ gene | This study |
| AB199 | Wild-type TIE-1 with RuBisCO form I gene overexpressed | Engineered TIE-1 for PHA overproduction | This study |
| AB200 | Wild-type TIE-1 with RuBisCO form I and II genes overexpressed | Engineered TIE-1 for PHA overproduction | This study |
TABLE 2
| Plasmid | Relative characteristics | Source |
|---|---|---|
| pJQ200KS | ori P15A; Genr | (63) |
| pAB314 | pJQ200KS with GlmUSX_UP and GlmUSX_DN | (34) |
| pAB356 | Cmr | (64) |
| pAB357 | Kanr | (64) |
| pAB358 | Tcr | (64) |
| pAB359 | Ampr | (64) |
| pSRKGm | Plac; lacIq; ori pBBR1; Genr | (64) |
| pWB083 | pAB314 with attB | This study |
| pWB081 | pJQ200KS with mCherry; PaphII promoter; fd terminator and attP | This study |
| pWB084 | pSRKGm with φC31 integrase | This study |
| pWB086 | pAB314 with φC31 integrase; Plac; lacIq; attB | This study |
| pWB088 | pJQ200KS with φC31 integrase; PaphII promoter; fd terminator and attP | This study |
| pWB091 | pJQ200KS with Kanr | This study |
| pWB092 | pJQ200KS with Cmr | This study |
| pWB107 | RuBisCO form I and form II gene overexpression | This study |
| pWB108 | RuBisCO form I gene overexpression | This study |
| pAB863 | pJQ200KS with phaR 1 kb up and 1 kb down for deletion of phaR | This study |
| pAB868 | pJQ200KS with phaZ 1 kb up and 1 kb down for deletion of phaZ | This study |
All the mutants were constructed by using a suicide plasmid (pJQ200KS) carrying a gentamicin resistance cassette. Using gentamycin has two disadvantages: 1) wild-type TIE-1 has a remarkably high resistance to gentamicin with a minimum inhibitory concentration (MIC) of 200 µg/mL and 2) gentamicin is expensive (65). For enhanced competitiveness in large-scale production of PHA, opting for other antibiotics with lower MIC values than gentamicin could offer a more cost-effective alternative. Accordingly, we constructed suicide plasmids containing two widely used antibiotic resistance markers: chloramphenicol (MIC of 100 µg/mL) or kanamycin (MIC of 50 µg/mL), using pJQ200KS as the backbone, as listed in Table 2 and shown in Fig. S3.
Engineering TIE-1 using a phage integration system to increase PHA production in TIE-1
As previously discussed, homologous recombination presents a time-intensive process and remains so far the only successful genetic tool for TIE-1. To address these challenges, we devised a phage integration system for incorporation of genes into its genome. Fig. 1 delineates the essential components of the φC31 recombinase system: the attB site, attP site, and the φC31 integrase.
Fig 1
To compensate for the lack of an attB site in TIE-1, we inserted it into the genome, as described in Materials and Methods. The attP site was introduced into a suicide plasmid with a constitutively expressed mCherry gene under the PaphII promoter (pWB081). For the expression of φC31 integrase, the optimal way would be to use a temperature-sensitive plasmid (66). After the targeted genome editing, the removal of the φC31 integrase could be conveniently executed using this approach. Unfortunately, there is no known temperature-sensitive plasmid that replicates in TIE-1. Hence, we decided to build two different systems: (1) a plasmid-based system, where the integrase is introduced into TIE-1 by a plasmid (Fig. 1B), and (2) a genome-based system, where the integrase is integrated into the TIE-1 genome (Fig. 1C). The advantage of the plasmid-based system is its mobility, while the genome-based system is more stable and does not rely on antibiotics. Both an inducible promoter (Plac) and a strong constitutive promoter (PaphII) were tested. To summarize, as shown in Fig. 1, we have four different designs for the expression of the φC31 integrase: a) PaphII-driven φC31 integrase on a suicide plasmid (pWB088); b) Plac-driven φC31 integrase on a self-replicating plasmid (pWB084); c) PaphII-driven φC31 integrase on the TIE-1 genome (Fig. 1C(a)); and d) Plac-driven φC31 integrase on the TIE-1 genome (Fig. 1C(b)). After three separate trials, we were not able to obtain a genome-based system using the constitutive promoter (Fig. 1C(a)). Thus, we only present the results of the other three systems. The successful integration of mCherry was indicated by visualizing red fluorescence at an emission wavelength of 610 nm.
Upon confirming integration, we assessed the efficiency of the various systems by measuring the transformation efficiency (detailed calculation described in Materials and Methods) and the integration efficiency, which is defined as the percentage of colonies that have red fluorescence signals among all obtained colonies. As shown in Fig. 2A and B, the transformation efficiency normalized to the plasmid concentration is higher for the genome-based system (P < 0.001) (Fig. 1C). This higher efficiency could be due to the sufficiency of only one plasmid for the system to be functional. Between the two plasmid-based systems, the constitutively expressed φC31 reached higher efficiency (Fig. 1B, pWB088). This higher efficiency could be due to the independence from the use of the inducer, IPTG. For the integration frequency, the genome-based system (Fig. 1C) and the plasmid-based system both with the constitutive promoter (Fig. 1B, pWB088) and inducible promoter resulted in similar editing efficiencies between 80% and 100% (P = 0.13) (Fig. 2B). To summarize, the genome-based system with an inducible promoter results in highest transformation efficiency, while all three systems have similar editing efficiencies.
Fig 2
After successfully integrating mCherry, we wanted to use the phage integration tool to assist in improving PHA production in TIE-1. Previous research conducted on Ralstonia eutropha (now Cupriavidus necator) has shown that overexpression of RuBisCO resulted in improved PHA production (60). Thus, to improve PHA production, we integrated a PaphII -driven RuBisCO form I alone or RuBisCO form I and form II together into the TIE-1 genome to obtain two new TIE-1 strains: Ωrub(I) and Ωrub(I&II). We were unable to obtain a plasmid with PaphII -driven RuBisCO form II only. The successful integration of these genes and PaphII was analyzed by PCR amplification of the PaphII region and the RuBisCO gene form I and II, as indicated in Fig. S4. In addition, the whole genome of each constructed strain has been sequenced.
Deletion of phaR, phaZ, or overexpression of the RuBisCO form I impaired the growth of TIE-1 during photoheterotrophic growth with butyrate
Because TIE-1 has previously exhibited high PHA production under anoxic photoheterotrophic conditions with butyrate (27), we evaluated this growth condition using all the constructed strains. Under these anoxic growth conditions, light is used as an energy source and butyrate is used both as an electron and a carbon source. For all growth conditions explored, we determined the effect of nitrogen fixation by supplying N2 gas (nitrogen-fixing) or ammonium chloride (NH4Cl) salt (non-nitrogen-fixing). The growth parameters were obtained both from 1 mM (shown in Table 3; Fig. 3A and B) and 10 mM (shown in Table S1) butyrate. A 1 mM concentration was chosen to allow a clear growth trend without reaching saturation of the OD660. To collect enough biomass, PHA production was only obtained from growth with 10 mM butyrate. Our results show that the growth with 1 mM butyrate did not reveal a significant difference between the various constructed strains (Fig. 3A and B). Deleting phaR or phaZ increased the generation time of TIE-1 by around 3 hours compared to wild-type when grown photoheterotrophically with 10 mM butyrate under non-nitrogen-fixing conditions (Table S1) (P = 0.01). However, a slightly shorter generation time (around 3 hours shorter than that of wild-type) was obtained from the Δgly and the ΔnifA under these conditions (Table S1) (P = 0.007). Under nitrogen-fixing conditions, the ΔphaZ mutant continued to show a growth defect observed as a longer generation time (3 hours longer than wild type), whereas the ΔphaR and Δgly mutants showed the same generation time as that of wild-type (Table S1). As expected, ΔnifA was not able to grow under nitrogen-fixing conditions (Table 3; Table S1; Fig. 3B). We compared the growth of the engineered RuBisCO strains Ωrub(I) and Ωrub(I&II) under the same photoheterotrophic growth condition with butyrate. The engineered strain carrying only the overexpressed RuBisCO gene form I (Ωrub(I)) has a growth defect under both nitrogen-fixing and non-nitrogen-fixing conditions with 10 mM butyrate (Table S1), as well as with 1 mM butyrate (Table 3). This defect seems to be more noticeable under nitrogen-fixing growth conditions. The Ωrub(I) strains have an ~2X (P = 0.01) longer lag time than wild-type strains when grown photoheterotrophically with 10 mM butyrate under non-nitrogen-fixing conditions, and ~5X (P = 0.001) longer generation time under nitrogen-fixing conditions (Table S1). Lag time was also about 4 hours longer (P = 0.001) when the Ωrub(I) was grown with only 1 mM butyrate both under nitrogen and non-nitrogen-fixing conditions (Table 3).
TABLE 3
| ΔphaR | P | ΔphaZ | P | Δgly | P | ΔnifA | P | Ωrub(I) | P | Ωrub(I&II) | P | WT | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Generation time G (h) | |||||||||||||
| But (NH4Cl) | 6.58 (0.34) | 0.39 | 5.45 (0.42) | 0.00 | 6.81 (0.18) | 0.96 | 6.58 (0.37) | 0.40 | 6.79 (0.13) | 0.92 | 6.49 (0.36) | 0.25 | 6.80 (0.20) |
| But (N2) | 10.18 (0.52) | 0.002 | 9.57 (0.34) | 0.002 | 8.03 (1.42) | 0.99 | NG | NG | 8.48 (0.84) | 0.42 | 7.0 (0.79) | 0.094 | 8.04 (0.22) |
| H2 (NH4Cl) | 14.98 (1.32) | 0.01 | 13.20 (0.54) | 0.00 | 9.74 (0.4) | 0.442 | 11.16 (0.71) | 0.154 | 9.7 (0.3) | 0.37 | 10.51 (4.25) | 0.89 | 10.14 (0.71) |
| H2 (N2) | 88.80 (27.78) | 0.52 | 64.90 (9.2) | 0.36 | 75.29 (12.74) | 0.971 | NG | NG | 149.9 (67.03) | 0.13 | 99.77 (36.01) | 0.35 | 75.74 (15.84) |
| Fe(II) (NH4Cl) | 595 (30.31) | 0.65 | 4620 (2000.5) | 0.03 | 280.71 (57.66) | 0.021 | 369.35 (115) | 0.076 | 740.67 (265) | 0.61 | 644.36 (119) | 0.98 | 641.67 (160) |
| Fe(II) (N2) | ND | ND | NG | NG | 235.14 (55.63) | 0.498 | NG | NG | 371.49 (11.7) | 0.004 | 205.33 (155) | 0.57 | 262.1 (29.13) |
| PE (NH4Cl) | 517.3 (48.1) | 0.21 | 1212.75 (173.2) | 0.0048 | 1008 (0) | 0.016 | 822.9 (86.6) | 0.0001 | 605.02 (3.49) | 0.18 | 337.41 (10.7) | 0.40 | 262.32 (9.11) |
| PE (N2) | 646.8 (0) | 0.71 | 3482.73 (3211) | 0.28 | 42446.2 (84892) | 0.37 | NG | NG | 1060.8 (105) | 0.54 | 4207 (3316) | 0.27 | 800.25 (115) |
| Maximum OD660 | |||||||||||||
| But (NH4Cl) | 0.78 (0.07) | 0.35 | 0.86 (0.04) | 0.22 | 0.81 (0.06) | 0.79 | 0.89 (0.01) | 0.00 | 0.71 (0.06) | 0.03 | 0.69 (0.03) | 0.00 | 0.82 (0.01) |
| But (N2) | 0.653 (0.04) | 0.036 | 0.67 (0.04) | 0.113 | 0.70 (0.02) | 0.107 | NG | NG | 0.78 (0.05) | 0.32 | 0.80 (0.07) | 0.28 | 0.74 (0.03) |
| H2 (NH4Cl) | 2.48 (0) | 0.001 | 2.495 (0.01) | 0.001 | 1.36 (0.28) | 0.454 | 1.64 (0.24) | 0.12 | 1.98 (0.742) | 0.26 | 2.22 (0.403) | 0.07 | 1.17 (0.071) |
| H2 (N2) | 1.73 (0.19) | 0.642 | 1.963 (0.19) | 0.11 | 1.89 (0.42) | 0.409 | NG | NG | 1.73 (0.239) | 0.68 | 1.3 (0.482) | 0.31 | 1.65 (0.19) |
| Fe(II) (NH4Cl) | 0.04 (0) | 0.002 | 0.021 (0) | 0.016 | 0.14 (0.008) | 0.00 | 0.082 (0) | 0.00 | 0.11 (0.012) | 0.00 | 0.03 (0.001) | 0.01 | 0.03 (0.0017) |
| Fe(II) (N2) | 0.02 (0.01) | 0.013 | NG | NG | 0.10 (0.02) | 0.977 | NG | NG | 0.06 (0.002) | 0.09 | 0.11 (0.029) | 0.85 | 0.09 (0.012) |
| PE (NH4Cl) | 0.27 (0.00) | 0.332 | 0.160 (0.00) | 0.029 | 0.17 (0) | 0.036 | 0.17 (0.00) | 0.035 | 0.26 (0.006) | 0.29 | 0.29 (0.001) | 0.34 | 0.38 (0.003) |
| PE (N2) | 0.39 (0.005) | 0.218 | 0.19 (0.00) | 0.944 | 0.098 (0.002) | 0.01 | NG | NG | 0.16 (0.002) | 0.46 | 0.122 (0) | 0.04 | 0.19 (0.001) |
| Time to reach maximum OD (h) | |||||||||||||
| But (NH4Cl) | 98 (0.00) | 0 | 114 (0.00) | 0 | 74 (0.00) | 0 | 74 (0.00) | 0 | 82 (0.00) | 0 | 74 (0.00) | 0 | 74 (0.00) |
| But (N2) | 122 (0) | 0.0065 | 122.33 (0) | 0.000 | 114 (0) | 0 | NG | NG | 130 (0) | 0 | 114 (0) | 0 | 98 (0) |
| H2 (NH4Cl) | 90.5 (0.71) | 0.00 | 290.5 (0.70) | 0.00 | 338.5 (0.71) | 1 | 338.5 (0.71) | 1 | 243 (135.76) | 0.42 | 338.5 (0.71) | 1 | 338.5 (0.71) |
| H2 (N2) | 472.67 (0.58) | 1 | 472.7 (0.57) | 1 | 472.67 (0.58) | 1 | NG | NG | 472.67 (0.58) | 1 | 472.67 (0.58) | 1 | 472.67 (0.58) |
| Fe(II) (NH4Cl) | 1680.33 (0.5) | 0.00 | 1680.2 (0.29) | 0.00 | 1320.33 (0.58) | 0.00 | 1320.67 (1.1) | 0 | 1319.67 (0.5) | 0.00 | 936.33 (0.58) | 0.23 | 935.67 (0.58) |
| Fe(II) (N2) | 1704.33 (0.5) | 0.00 | NG | NG | 1200.33 (0.58) | 1 | NG | NG | 671.67 (0.58) | 0.00 | 1200.33 (0.5) | 0.37 | 1200.33 (0.58) |
| PE (NH4Cl) | 336 (0) | 0.00 | 336 (0) | 0.00 | 336 (0) | 0.00 | 336 (0) | 0.00 | 336 (0) | 0.00 | 336 (0) | 0.00 | 336 (0) |
| PE (N2) | 336 (0) | 0.00 | 336 (0) | 0.00 | 336 (0) | 0.00 | NG | NG | 336 (0) | 0.00 | 336 (0) | 0.00 | 336 (0) |
| Lag time (h) | |||||||||||||
| But (NH4Cl) | 10.80 (0.23) | 0.00 | 10.92 (0.71) | 0.00 | 10.11 (1.10) | 0.11 | 10.77 (0.28) | 0.00 | 14.32 (2.72) | 0.00 | 11.52 (2.00) | 0.022 | 9.02 (0.36) |
| But (N2) | 17.06 (0.33) | 0.0051 | 17.56 (0.10) | 0 | 17.26 (1.75) | 0.14 | NG | NG | 19.22 (0.82) | 0.0019 | 16.56 (0.96) | 0.12 | 15.4 (0.39) |
| H2 (NH4Cl) | 94.03 (22.7) | 0.1454 | 84.97 (1.66) | 0.009 | 81.14 (1.5) | 0.011 | 73.59 (5.47) | 0.063 | 73.57 (10.03) | 0.14 | 89.87 (3.64) | 0.01 | 56.25 (3.45) |
| H2 (N2) | 267.23 (123) | 0.21 | 192 (13.97) | 0.103 | 191.1 (20) | 0.143 | NG | NG | 185.85 (69.6) | 0.55 | 115.83 (27.5) | 0.11 | 158.82 (23.3) |
| Fe(II) (NH4Cl) | ND | ND | ND | ND | 416.54 (27.27) | 0.366 | 805.62 (165) | 0.0218 | 706.5 (13.28) | 0.01 | 296.7 (11.71) | 0.8 | 321.34 (159) |
| Fe(II) (N2) | ND | ND | NG | NG | 919.89 (146.9) | 0.874 | NG | NG | 714.13 (60.7) | 0.008 | 792.07 (194) | 0.376 | 905.17 (33.2) |
a
Growth parameters obtained from butyrate are from 1 mM butyrate. Values are averages from biological triplicates. Standard error values are in parentheses. P, P values against wild-type values. P values in bold indicate statistical significance. But, butyrate; Fe, iron; NH4Cl, ammonium chloride; N2, nitrogen gas; NG, no growth; ND, not determined—almost close to the whole growth time. Final OD obtained from 14 days of growth under PE was reported as max OD.
Fig 3
The engineered strain Ωrub(I&II) carrying overexpressed RuBisCO form I and II genes did not show any significant differences in growth compared to wild-type strains when grown under non-nitrogen- fixing or under nitrogen-fixing conditions (Table S1; Table 3; Fig. 3C and D).
Deletion of the phaR gene or overexpression of the RuBisCO form I and II genes increases PHA production in TIE-1 under photoheterotrophic growth with butyrate
Deletion of the phaR (regulator gene) increased PHA production (mg/L/cell and dry cell weight % wt/wt) ~2X (P = 0.00) compared to wild-type under non-nitrogen-fixing or nitrogen-fixing conditions (Fig. 3E; Table 4; Table S2). Deletion of the phaZ, gly, or nifA genes resulted in an overall decrease in PHA production (mg/L/cell and dry cell weight % wt/wt) of ~10X (P = 0.05) compared to wild-type under growth under non-nitrogen-fixing conditions. These deletions did not seem to affect PHA production when compared to wild-type under nitrogen-fixing conditions. We did not detect any PHA from the ΔnifA grown under nitrogen-fixing conditions as it did not show any growth. Except for ΔphaR, all strains exhibited a consistent increase in PHA production during nitrogen-fixing growth conditions in comparison to growth under non-nitrogen- fixing conditions.
TABLE 4
| ΔphaR | P | ΔphaZ | P | Δgly | P | ΔnifA | P | Ωrub(I) | P | Ωrub(I&II) | P | WT | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| PHA (mg/L) | |||||||||||||
| Butyrate (NH4Cl) | 30.37 (0.33) | 0.03 | 4.86 (0.69) | 0.116 | 5.25 (0.67) | 0.098 | 8.21 (0.3) | 0.29 | 36.03 (1.61) | 0.003 | 41.1 (0.52) | 0.016 | 12.061 (0.27) |
| Butyrate (N2) | 14.3 (0.46) | 0.003 | 13.73 (1.68) | 0.001 | 15.87 (1.36) | 0.07 | NG | NG | 49.71 (4.88) | 0.002 | 53.23 (2.91) | 0 | 25.138 (0.05) |
| H2 (NH4Cl) | 24.45 (2.08) | 0.349 | 24.51 (1.33) | 0.387 | 29.18 (0.14) | 0.754 | 34.65 (3.1) | 0.58 | 40.15 (4.25) | 0.234 | 32.8 (3.57) | 0.819 | 31.089 (1.37) |
| H2 (N2) | 32.49 (1.25) | 0.505 | 30.82 (1.21) | 0.556 | 16.09 (0.76) | 0.151 | NG | NG | 32.72 (1.16) | 0.655 | 13.06 (1.16) | 0.115 | 25.757 (0.26) |
| Fe(II) (NH4Cl) | 1.25 (0.04) | 0.013 | ND | ND | 1.46 (0.04) | 0.092 | 1.7 (0.01) | 0.38 | 1.85 (0.034) | 0.374 | 1.79 (0.075) | 0.394 | 2.19 (0.03) |
| Fe(II) (N2) | 0.24 (0.012) | 0 | 0.29 (0.02) | 0.002 | 0.42 (0.008) | 0.02 | NG | NG | 0.28 (0.018) | 0.005 | 0.80 (0.004) | 0.663 | 0.96 (0.03) |
| PE (NH4Cl) | 0.40 (0.46) | 0.0002 | 1.23 (0.05) | 0.000 | 0.735 (0.18) | 0.000 | 0.59 (0.008) | 0.00 | 0.62 (0.17) | 0.000 | 0.64 (0.037) | 0.000 | 2.57 (0.058) |
| PE (N2) | 0.422 (0.096) | 0.604 | 0.82 (0.009) | 0.01 | 0.59 (0.03) | 0.14 | NG | NG | 0.67 (0.17) | 0.14 | 0.33 (0.012) | 0.94 | 0.34 (0.02) |
| PHA (mg/L/cell) (x10−11) | |||||||||||||
| Butyrate (NH4Cl) | 2.32 (0) | 0 | 0.079 (0.03) | 0.051 | 0.11 (0.03) | 0.056 | 0.1 (0.03) | 0.05 | 2.91 (0.22) | 0.007 | 5.29 (0.1) | 0.001 | 1 (0.01) |
| Butyrate (N2) | 1.89 (0.07) | 0.02 | 1.89 (0.23) | 0.606 | 1.4 (0.12) | 0.374 | NG | NG | 5.16 (0.45) | 0.00 | 8.41 (0.47) | 0.00 | 2.65 (0.02) |
| H2 (NH4Cl) | 2.17 (0.14) | 0.098 | 2.63 (0.18) | 0.127 | 3.22 (0.01) | 0.056 | 3.47 (0.27) | 0 | 3.58 (0.21) | 0.055 | 3.01 (0.27) | 0.012 | 1.52 (0.04) |
| H2 (N2) | 4.89 (0.22) | 0.517 | 3.84 (0.15) | 0.952 | 1.66 (0.07) | 0.098 | NG | NG | 4.07 (0.18) | 0.862 | 1.74 (0.31) | 0.111 | 3.76 (0.05) |
| Fe(II) (NH4Cl) | 2.32 (0.06) | 0 | ND | ND | 1.9 (0.05) | 0 | 4.37 (0.03) | 0.37 | 3.36 (0.71) | 0.033 | 5.64 (0.26) | 0.98 | 5.62 (0.09) |
| Fe(II) (N2) | 0.85 (0.03) | 0.082 | 0.45 (0.04) | 0.004 | 0.89 (0.03) | 0.075 | NG | NG | 0.35 (0.038) | 0.002 | 0.68 (0.009) | 0.111 | 1.66 (0.08) |
| PE (NH4Cl) | 0.10 (0.02) | 0.006 | 0.5 (0.02) | 0.803 | 0.35 (0.09) | 0.085 | 0.25 (0.00) | 0.25 | 0.18 (0.04) | 0.01 | 0.17 (0.006) | 0.005 | 0.55 (0.006) |
| PE (N2) | 0.124 (0.009) | 0.89 | 0.34 (0.023) | 0.023 | 0.49 (0.036) | 0.011 | NG | NG | 0.307 (0.11) | 0.05 | 0.207 (0.007) | 0.05 | 0.11 (0.016) |
| PHA (dry cell weight % wt/wt) | |||||||||||||
| Butyrate (NH4Cl) | 5.03 (0.00) | 0 | 0.15 (0.05) | 0.05 | 0.24 (0.23) | 0.05 | 0.22 (0.22) | 0.05 | 6.30 (0.48) | 0.007 | 11.45 (0.20) | 0.001 | 2.15 (0.02) |
| Butyrate (N2) | 4.08 (0.15) | 0.02 | 4.075 (0.25) | 0.606 | 3.03 (0.26) | 0.374 | NG | NG | 11.18 (0.98) | 0 | 18.23 (1.00) | 0 | 5.73 (0.03) |
| H2 (NH4Cl) | 4.70 (0.30) | 0.098 | 5.7 (0.375) | 0.127 | 6.98 (0.02) | 0.056 | 7.53 (0.58) | 0 | 7.75 (0.45) | 0.055 | 6.52 (0.58) | 0.012 | 3.28 (0.08) |
| H2 (N2) | 10.60 (0.48) | 0.517 | 8.325 (0.32) | 0.952 | 3.57 (0.15) | 0.098 | NG | NG | 8.83 (0.38) | 0.862 | 3.75 (0.50) | 0.111 | 8.15 (0.10) |
| Fe(II) (NH4Cl) | 5.03 (0.14) | 0 | ND | ND | 4.10 (0.13) | 0 | 9.4 8(0.07) | 0.37 | 7.28(1.50) | 0.033 | 12.23 (0.58) | 0.98 | 12.20 (0.20) |
| Fe(II) (N2) | 1.86 (0.08) | 0.082 | 0.9825 (0.1) | 0.004 | 1.93 (0.07) | 0.075 | NG | NG | 0.75 (0.80) | 0.002 | 1.48 (0.02) | 0.111 | 3.60 (0.19) |
| PE (NH4Cl) | 0.22 (0.03) | 0.006 | 1.075 (0.02) | 0.803 | 0.76 (0.18) | 0.085 | 0.53 (0.00) | 0.25 | 0.38 (0.08) | 0.01 | 0.35 (0.01) | 0.005 | 1.18 (0.018) |
| PE (N2) | 0.25 (0.00) | 0.89 | 0.725 (0.02) | 0.023 | 1.05 (0.08) | 0.011 | NG | NG | 0.65 (0.24) | 0.05 | 0.43 (0.02) | 0.05 | 0.24 (0.03) |
a
Values are averages from biological triplicates, except for those from photoelectrotrophic (PE) growth conditions, which are duplicates. PHA from growth on butyrate are from 10 mM concentrations. Standard error values are in parentheses. P, P values against wild-type values. P values in bold indicate statistical significance. Fe, iron; NH4Cl, ammonium chloride; N2, nitrogen gas; PE. photoelectrotrophic growth; NG, no growth; ND, not detectable.
When grown photoheterotrophically with butyrate, the two strains harboring the Ωrub(I) and Ωrub(I&II) demonstrated increased PHA production, regardless of the nitrogen source. Under non-nitrogen-fixing conditions, the Ωrub(I) strain had 2X (P = 0.00) higher PHA production (mg/L/cell and dry cell weight % wt/wt) than wild-type, while the Ωrub(I&II) exhibited a 5X increase (P = 0.001). This increase in PHA was also observed under nitrogen-fixing conditions where an increase of 1.9X (P = 0.007) was observed from the Ωrub(I) and 3X (P = 0.001) from the Ωrub(I&II) (Fig. 3F; Table 4; Table S2). Like the deletion mutants, an overall increase in PHA production was observed from the engineered and wild-type strains under nitrogen-fixing vs non-nitrogen-fixing conditions (Fig. 3F).
Deletion of the phaR, phaZ, or overexpression of RuBisCO form I and II genes increased the final optical density of TIE-1 under photoautotrophic growth with H2 under non-nitrogen-fixing conditions
Achieving production from a low-cost and abundant carbon source is one of the most impactful pathways for bioplastic production. Accordingly, we tested the growth of all the strains under photoautotrophic conditions using H2 as an electron source and CO2 as a carbon source in a freshwater basal liquid medium. Light is used as the energy source as TIE-1 is a photosynthetic bacterium. Fig. 4A; Table 3 show that although ΔphaR and ΔphaZ mutants showed a slightly longer generation time (~3 hours slower) (P = 0.001), they reached the maximum OD faster than wild-type strains when grown with H2 under non-nitrogen-fixing conditions.
Fig 4
Although the strains Ωrub(I&II) showed an extended lag time (1.6 X slower than wild-type P = 0.01), they reached a higher final OD (0.52 higher) than wild-type (Table 3) when grown with H2 and under non-nitrogen-fixing conditions. The Ωrub(I) did not show any significant growth difference compared to wild-type (Table 3; Fig. 4C). No difference was observed between the two engineered strains when grown with H2 under nitrogen-fixing conditions (Table 3; Fig. 4D).
Deletion of the glycogen synthase (gly), nifA genes, or overexpression of RuBisCO form I or I & II increased PHA production under photoautotrophic growth with H2 under non-nitrogen-fixing conditions
We tested PHA production (mg/L/cell and dry cell weight % wt/wt) from all the constructed strains under photoautotrophic growth conditions with H2. The mutants ΔphaR and ΔphaZ did not show any significant difference in PHA production compared to wild-type under growth with H2 regardless of the nitrogen source. The Δgly and the ΔnifA strains showed a 2X increase (P = 0.05 and P = 0, respectively) in PHA production when grown with H2 under non-nitrogen-fixing conditions (Table 4; Fig. 4E). The deletion of phaR or phaZ did not affect PHA production compared to wild-type strains. No PHA was produced by the ΔnifA during growth under nitrogen-fixing conditions due to its inability to grow. Unlike growth with butyrate, switching from non-nitrogen-fixing to nitrogen-fixing conditions increased PHA production only in wild-type TIE-1 but not in the mutants. The mutants showed lower PHA production compared to wild-type under nitrogen-fixing conditions with H2 as an electron source.
Like growth with butyrate, both engineered Ωrub(I) and Ωrub(I&II) strains showed an increase in PHA production (mg/L/cell and dry cell weight % wt/wt) when grown with H2 under non-nitrogen-fixing conditions. PHA production was nearly double in both engineered strains compared to wild-type (P = 0.05 and 0.01, respectively), (Table 4; Fig. 4F). Switching to nitrogen-fixing conditions with H2 did not affect PHA production (mg/L/cell and dry cell weight % wt/wt) of the two engineered strains Ωrub(I) and when compared to wild-type (Table 4; Fig. 4F). In contrast, a decrease in PHA production was observed from the Ωrub(I&II) under nitrogen-fixing conditions (Table 4; Fig. 4F).
Deletion of the phaR and phaZ genes impaired the ability of TIE-1 to oxidize Fe(II)
TIE-1 can use electrons produced by the oxidation of Fe(II) for photoautotrophy (photoferrotrophy). Under these difficult growth conditions, the main carbon source is CO2 and the energy source is light. We tested the growth of all the strains we constructed under photoferrotrophic growth conditions. All the strains were first pre-grown in H2 to allow the expression of the genes involved in Fe(II) oxidation in TIE-1, as performed previously (27). We observed a defect in the ability of TIE-1 to oxidize Fe(II) and grow when the phaZ or phaR genes were deleted (Fig. 5A and B). Under non-nitrogen-fixing conditions, the phaZ mutant was not able to oxidize Fe(II) or grow even after 50 days (P = 0.003 at day 50 compared to the wild-type). A significant delay in Fe(II) oxidation was observed in the phaR mutant under non-nitrogen-fixing conditions. ΔphaR was able to fully oxidize Fe(II) only after 65 days versus 40 days (P < 0.001 at day 40) for wild-type and grow to wild-type levels. In contrast, Δgly showed faster Fe(II) oxidation ability, which occurred after 15 days of growth (<0.001 at day 15) as opposed to ~40 days for wild-type when grown under non-nitrogen-fixing conditions. ΔnifA showed a similar iron oxidation pattern as wild-type under non-nitrogen-fixing conditions (Fig. 5A). Under non-nitrogen-fixing conditions, the oxidation ability of Δgly was similar to that of the wild-type. As expected, no Fe(II) oxidation occurred from ΔnifA as the strain was not able to grow under nitrogen-fixing conditions. Overexpression of RuBisCO form I appears to delay the ability of TIE-1 to oxidize Fe(II) under non-nitrogen-fixing conditions. However, when RuBisCO form I and II genes are both overexpressed, the strain oxidizes Fe(II) at the same time as wild-type (~40 days) (Fig. 5C). During nitrogen fixation, the Ωrub(I) showed faster Fe(II) oxidation compared to wild-type (P < 0.001 at ~day 30), whereas the Ωrub(I&II) initiated Fe(II) oxidation at the same time as wild-type (P = 0.42 at ~day 30), (Fig. 5D).
Fig 5
Deletion of the various genes or overexpression of RuBisCO form I and II genes did not improve PHA production in TIE-1 under photoferrotrophy
To test PHA production obtained from growth during photoferrotrophy, samples were collected right after complete Fe(II) oxidation occurred. Lower PHA production (mg/L/cell and dry cell weight % wt/wt) was observed across all the mutants under growth with Fe(II) and non-nitrogen-fixing conditions compared to wild-type (Table 4; Fig. 5E). Among all the mutants, the ΔnifA had the highest PHA production when grown under non-nitrogen-fixing conditions, with comparable levels to wild-type (P = 0.37).
ΔphaR and Δgly showed the lowest production, about half that observed in wild-type when grown under non-nitrogen-fixing conditions (P = 0). No PHA was detected from the ΔphaZ mutant from the non-nitrogen-fixing conditions, which is linked to its inability to oxidize Fe(II). Like growth under non-nitrogen-fixing conditions, PHA productions from the ΔphaR, ΔphaZ, and Δgly (P = 0) strains during nitrogen-fixing conditions were overall lower than that obtained from wild-type (Fig. 5E and F). PHA production (mg/L/cell and dry cell weight % wt/wt) obtained from Δphaz under nitrogen-fixing conditions was ~3X (P = 0.004) smaller than the wild-type (Table 4; Fig. 5E). No PHA was obtained from ΔnifA when grown under nitrogen-fixing conditions because this strain is incapable of growth under such conditions.
A decrease in PHA production of about half was observed from the single engineered Ωrub(I) strain under Fe(II) under non-nitrogen-fixing growth conditions (P = 0.033) (Table 4; Fig. 5F). However, PHA production obtained from the Ωrub(I&II) was similar to the wild-type under growth with Fe(II) and non-nitrogen-fixing conditions (P = 0.98). Like the trend obtained from the mutants, a decrease in PHA production was observed in the RuBisCO overexpressing strains under nitrogen-fixing conditions compared to the growth under non-nitrogen-fixing conditions. PHA production obtained from the Ωrub(I) was ~4X (P = 0.002) less than what was obtained from wild-type under nitrogen-fixing conditions.
Growth under photoelectrotrophy under non-nitrogen-fixing conditions showed increased electron uptake from the engineered RuBisCO strains
In addition to its ability to grow autotrophically using hydrogen and Fe(II) as electron sources, TIE-1 can uptake electrons directly from poised electrodes (32). We grew the cells in bioelectrochemical reactors as described in Materials and Methods with freshwater media under nitrogen-fixing or non-nitrogen-fixing conditions. All mutant strains except ΔphaR showed longer generation time compared to wild-type strains when grown under non-nitrogen-fixing conditions (Table 3) (P < 0.05). ΔphaZ and Δgly exhibited significantly slower growth, as evidenced by a generation time ~4X longer than that of wild-type. ΔnifA similarly exhibited an extended generation time, ~3X longer than that of wild-type (Table 3) (P = 0.0001). ΔphaR did not exhibit a significant difference in its generation time compared to wild-type when grown under non-nitrogen-fixing conditions. Ωrub(I) and Ωrub(I&II) did not show a defect in generation time compared to wild-type under photoelectrotrophic non-nitrogen-fixing conditions.
We also measured the final OD of the strains at the end of the 14-day incubation. These values are reported as max OD in Table 3. The final OD values were ~1/2 lower than those of wild-type from ΔphaZ, Δgly, and ΔnifA under non-nitrogen-fixing growth conditions (P < 0.05). ΔphaR as well as the engineered strains Ωrub(I) and Ωrub(I&II) had the same final OD as wild-type (Table 3) when grown under non-nitrogen-fixing conditions. Under photoelectrotrophic and nitrogen-fixing conditions, the generation time of ΔphaR, ΔphaZ, as well as Ωrub(I) and Ωrub(I&II) did not show any significant differences compared to wild-type (Table 1). The ΔnifA did not show growth as expected. Due to the intricate nature of the experimental setup in photoelectrotrophic growth, we could only measure the cell density at initial and final times, preventing us from reporting lag time. We then assessed the ability of each constructed strain to capture electrons from a poised electrode. Under non-nitrogen-fixing conditions, ΔphaR, ΔphaZ, ΔnifA, and Ωrub(I) did not show a significant difference in current uptake when compared to wild-type. However, Δgly showed ~6X (P = 0.001) lower current uptake, while Ωrub(I&II) showed a decrease of ~2X (P = 0.009) in electron uptake when compared to wild-type (Fig. 6; Table 5).
Fig 6
TABLE 5
| Strain | Current uptake (C) | |
|---|---|---|
| Growth under non-nitrogen-fixing conditions | Growth under nitrogen-fixing conditions | |
| WT TIE-1 | −0.7891 ± 0.02 | −1.2359 ± 0.1027 |
| ΔphaR | −0.5257 ± 0.2167 (P = 0.00018) | −0.0666 ± 0.2182 (P = 0.0017) |
| ΔphaZ | −0.6311 ± 0.2118 (P = 0.0029) | −0.3019 ± 0.1356 (P = 0.00092) |
| Δgly | −0.1241 ± 0.0015 (P = 0.0001) | −0.4695 ± 0.014 (P = 0.00036) |
| ΔnifA | −0.8461 ± 0.0595 (P = 0.00026) | −0.0938 ± 0.0356 (P = 0.00045) |
| Ωrub(I) | −0.9922 ± 0.014 (P = 0.00177) | −5.6128 ± 0.1182 (P = 0.00332) |
| Ωrub(I &II) | −0.3513 ± 0.0283 (P = 0.0009) | −20.2604 ± 0.2492 (P = 0.0008) |
| Abiotic control | −0.1159 ± 0.067 (P = 0.00021) | −0.1474 ± 0.0741 (P = 0.00067) |
a
Values are from two biological replicates, ± standard deviation values.
Under nitrogen fixation conditions, electron uptake varied among strains. Notably, ΔphaR exhibited an 18X reduction compared to wild-type (P = 0.001). Additionally, both ΔphaZ and Δgly displayed 4- and 2.6X lower uptake, respectively (P < 0.001). The ΔnifA, which expectedly was unable to grow under nitrogen fixation conditions, demonstrated an electron uptake like that of abiotic control. In contrast, Ωrub(I) and Ωrub(I&II) displayed 4.5 and 16X higher electron uptake than wild-type, respectively (Table 5) (P < 0.005).
Overexpressing the RuBisCO form I and form II of TIE-1 increased its PHA production under photoelectrotrophic nitrogen-fixing growth conditions but not under non-nitrogen-fixing conditions
When grown under non-nitrogen-fixing conditions, PHA values obtained from the wild-type are sufficiently low so as to hinder us from comparing between strains (Table 4; Fig. 6E and F).
When grown under nitrogen-fixing conditions, we observed that the PHA production of phaR is similar to that of wild type. However, PHA production increased by 3X in ΔphaZ and Δgly compared to wild-type (Table 4; Fig. 6E) (P < 0.05). PHA production by Ωrub(I) and Ωrub(I&II) went 2X when compared to wild-type (P = 0.05). However, under nitrogen-fixing conditions, PHA produced from Ωrub(I)= 0.24% and Ωrub(I&II) =0.16% (dry cell weight % wt/wt) did not show statistically significant differences (P = 0.21) when evaluated against each other.
DISCUSSION
In this study, we demonstrated that TIE-1 can be modified using genetic engineering and synthetic biology tools to enhance polyhydroxyalkanoate (PHA) production. We found that deleting the phaR regulator gene of PHA biosynthesis increased PHA production under non-nitrogen-fixing conditions with butyrate. Deletion of the glycogen synthase and the regulators of the nitrogen fixation pathways nifA1 and nifA2 increased PHA accumulation when grown under non-nitrogen-fixing conditions with hydrogen. More importantly, overexpressing RuBisCO form I and II using the φC31 integrase system in TIE-1 enhanced PHA production under photoheterotrophic growth with butyrate, regardless of the nitrogen source. This was also the case for autotrophic growth of these TIE-1 strains with hydrogen in non-nitrogen-fixing conditions. These results indicate that genetic engineering techniques, including our newly developed φC31 integration system, are effective for genetic manipulation in TIE-1 and increasing PHA production.
Effect of deletions of genes in the PHA pathway on growth and PHA production in TIE-1
The specific regulatory role of PhaR (Rpal_0531) has not been explored in TIE-1 before. Here, we discovered that deleting phaR did not significantly affect PHA production in TIE-1 across all growth conditions tested. However, the ΔphaR mutant showed growth defects under non-nitrogen-fixing photoheterotrophic growth with butyrate and photoautotrophic growth with H2 (Table S1; Table 3). Additionally, it exhibited decreased Fe(II) oxidation and electron uptake under nitrogen-fixing conditions compared to the wild-type (Fig. 5). In Bradyrhizobium diazoefficiens USDA110, inactivation of phaR decreased PHA accumulation and affected processes such as exopolysaccharide biosynthesis and heat stress tolerance (67). Our results suggest that PhaR’s role in TIE-1 may extend to other pathways. Further transcriptomic and proteomic analyses of this mutant are needed to understand PhaR’s broader regulatory roles.
We hypothesized that deleting the PHA depolymerase (PhaZ) in TIE-1 would inhibit PHA degradation, potentially increasing total PHA storage (Fig. 7). This expectation was based on findings from other purple nonsulfur bacteria, such as Rhodobacter sphaeroides, where the phaZ mutant showed improved PHA accumulation under nitrogen-rich conditions (68). However, deleting phaZ did not significantly elevate PHA production in TIE-1. Unexpectedly, the absence of phaZ affected TIE-1′s ability to oxidize Fe(II) during photoferrotrophic growth and decreased electron uptake under nitrogen-fixing photoelectrotrophic growth conditions. These findings suggest that disrupting PHA depolymerase alters the electron balance within TIE-1, meriting further bioinformatics and proteomics studies.
Fig 7
Effects of deletion of a potential competitive pathway on PHA production in TIE-1
We hypothesized that deleting the glycogen synthase (gly) gene would prevent glucose accumulation as glycogen, allowing glucose to be converted into pyruvate via glycolysis and subsequently into acetyl-CoA by pyruvate dehydrogenase, as illustrated in Fig. 7. Increased availability of acetyl-CoA could enhance PHA production. Contrary to our expectations, deletion of the glycogen synthase gene increased PHA production only under non-nitrogen-fixing conditions with hydrogen. A decrease in PHA production was observed under photoheterotrophic growth with butyrate and photoautotrophic growth with Fe(II) under nitrogen-fixing growth conditions. Additionally, a shorter generation time under non-nitrogen-fixing conditions with butyrate and Fe(II) suggests that blocking glycogen accumulation might channel carbon into biomass production rather than PHA production. Similar findings were reported in Synechocystis sp. PCC 6803, where glycogen was hypothesized to serve as a carbon source for PHA synthesis under nitrogen starvation conditions (58). Although autotrophic growth with hydrogen could be considered stressful compared to photoheterotrophic growth, the Δgly strain seems to prioritize biomass increase over PHA production under these conditions.
Effect of nitrogen fixation and nifA gene deletion on PHA production in TIE-1
Nitrogen fixation has been reported to increase PHA production (27, 41, 61). We observed this trend under photoheterotrophic growth with butyrate and photoautotrophic growth with hydrogen across most strains. However, under photoautotrophic growth with Fe(II), PHA production decreased under nitrogen-fixing compared to non-nitrogen-fixing conditions, with similar values obtained from photoelectrotrophic conditions under the two nitrogen sources. These observations align with those of our previous studies (27). The distinct responses under different growth conditions suggest potential variation in the regulatory mechanisms employed by TIE-1 under nitrogen fixation conditions.
Nitrogen fixation requires substantial reducing power (37). We investigated the impact of deleting two key activators of the nitrogenase genes, nifA1 and nifA2, to redirect NADPH toward PHA biosynthesis (Fig. 7). Contrary to our expectation, deleting nifA genes mainly decreased PHA production. However, under non-nitrogen-fixing photoheterotrophic growth with 10 mM butyrate, this strain showed a decreased generation time by approximately 25% (P = 0.007, Table 1; Table S2) suggesting that it favored growth over PHA production. Due to the highly reduced nature of butyrate and the presence of sodium bicarbonate in our media, it is possible that the electrons intended for nitrogen fixation were instead diverted to CO2 fixation for redox balance. This diversion could have contributed to biomass production rather than PHA synthesis. It is worth mentioning though that like the increase in n-butanol production observed in ΔnifA (37), we noticed an increase in PHA production from this strain grown under non-nitrogen-fixing photoautotrophic conditions with hydrogen (37).
Effect of overexpressing RuBisCO genes on PHA production in TIE-1
By increasing CO2 fixation through the Calvin–Benson–Bassham, we aim to increase acetyl-CoA abundance, leading to higher PHA production as shown in Fig. 7. Similar observations were reported in Ralstonia eutropha, where expressing a heterologous cyanobacterial RuBisCO gene resulted in a significant 99.7% increase in PHA accumulation under autotrophic growth. In TIE-1, overexpression of RuBisCO form I led to a ~ 2X increase in PHA production under both non-nitrogen- and nitrogen-fixing conditions. Overexpression of both RuBisCO forms I and II resulted in a ~ 5X increase under photoheterotrophic growth with butyrate and a ~ 3X increase under nitrogen-fixing conditions. The combined overexpression of both forms had a cumulative effect on PHA production. Surprisingly, overexpression of RuBisCO form I alone showed a growth defect under photoheterotrophy with butyrate and impaired Fe(II) oxidation under non-nitrogen-fixing conditions. This defect disappeared when both RuBisCO forms were overexpressed, suggesting a compensatory effect.
Future directions
We conclude that overexpressing RuBisCO form I and II increased PHA production in TIE-1 more efficiently than deleting competitive pathways. These (ΩrubI) and Ωrub(I&II) engineered strains could be envisioned as potential candidates for PHA production in larger scale, both under heterotrophic and autotrophic growth conditions. Larger reactors like those developed for biodiesel production by Awogbemi et al. (69) could be envisioned for these purposes. Using the φC31 system, it would be highly valuable to investigate the effect of overexpressing other genes that could potentially enhance PHA production in TIE-1. Given that the deletion of phaR, glycogen synthase (gly), and nifA genes did not result in a consistent increase in PHA production, it would be intriguing to initiate overexpression experiments with each of these genes. Additionally, gene overexpression could be extended to the PHA polymerase (phaC) gene using the integration system developed in this study to further boost PHA production in TIE-1.
Furthermore, it would be beneficial to assess the effects of creating combinatorial strains by overexpressing phaR, phaZ, glycogen synthase, and the nifA regulator genes, alongside the simultaneous overexpression of both native and non-native RuBisCO form I and II genes in TIE-1. This combinatorial approach may provide deeper insights into the regulatory networks and metabolic pathways influencing PHA synthesis in this organism. We also suggest that metabolic flux balance analysis and other machine learning methods might help us determine optimal mutant and overexpression strain construction for maximizing PHA production.
MATERIALS AND METHODS
Bacterial strains, media, and growth conditions
Table 1 lists all the strains used in the study. Lysogeny broth (LB) was used for growth of all E. coli strains at 37°C. Rhodopseudomonas palustris TIE-1 was grown in the medium containing 3 g/L yeast extract, 3 g/L peptone, 10 mM MOPS [3 N (morpholino) propanesulfonic acid] (pH 7.0), and 10 mM succinate (YPSMOPS) at 30°C. LB and YPSMOPS agar plates were prepared with the addition of 15 g/L agar. When needed, an antibiotic or sucrose was added as indicated in Table S3. All E. coli strains were grown on lysogeny broth (LB) at 37°C. Table 1 contains a list of the strains used in the study. Table S3 shows the concentration of antibiotics used as positive and negative selection components.
Anerobic growth of TIE-1 with hydrogen, Fe(II), or using poised electrodes was performed as previously described (27). Because transitioning between heterotrophic and photoautotrophic growth conditions in H2 and Fe(II) requires a metabolic shift in TIE-1 (27), to obtain stable growth, cells were first pre-grown in the yeast extract and then transferred into freshwater basal medium with H2 for an additional pre-growth before the growth study (27). For growth under photoelectrotrophy, the cells were grown with YP, washed three times with freshwater media, and used directly to innoculate the reactors to reduce contamination. The bioelectrochemical systems were set up as previously described (27) with the electrode modified to carbon felt (dimension 1 × 1×.5 cm).
Plasmid construction
All the plasmids used in this study are listed in Table 2. The kanamycin and chloramphenicol gene sequences were PCR-amplified from pSRKKm and pSRKCm, respectively. All these antibiotic resistance marker genes were then cloned into the pJQ200KS plasmid separately to replace the gentamicin resistance gene, resulting in pWB091 and pWB092. All the primers used are listed in Table S4.
Construction of engineered TIE-1 overexpressing RuBisCO form I and form II
Strains were constructed as described previously using markerless integration (34). pWB107 and pWB108 were individually conjugated into TIE-1 through E. coli S17-1/λ. After two successive homologous recombination processes, successful integrants were screened by the PCR, as shown in Fig. S4 and S5.
TIE-1 electroporation
To prepare electrocompetent cells, the TIE-1 strain was inoculated in 500 mL YPSMOPS and then incubated at 30°C. After reaching an OD660 of 0.5–0.6, the culture was centrifuged and washed at 4°C at 4,000 X g for 10 minutes. After washing five times with 10% glycerol, the cell pellet was resuspended in 2 mL of ice-cold 10% glycerol. This resuspension was aliquoted by 50 µL per sample and then saved in −80°C. For every electroporation, 0.2 µg of plasmid was added to 50 µL thawed electrocompetent cells and mixed well. This mixture was added to 1-mm gap electroporation cuvettes, and then cells were electroporated at 1.8 kV using a Biorad gene Micropulser. After electroporation, the mixture was added to 2 mL warmed Super Optimal Broth (SOB) or LB and grown for 45 minutes. Ten microliters, 100 µL, 500 µL, and the remainder of the culture were plated on the selective medium. Because these plates were further imaged by Nikon A1, glass Petri dishes were used instead of plastic Petri dishes.
Imaging of mCherry
After 4 to 5 days of incubation, the plated electroporated cultures expressing mCherry were imaged using the Nikon A1 confocal Eclipse Ti2 Microscope. For each plate, an image of the whole plate was captured using a camera through both the brightfield and the Texas-red channel (excitation 586 nm; emission 603 nm) with a 10X objective. The colony number of each plate was quantified using NIS-Elements AR Analysis 5.11.01 64-bit software.
Calculations of transformation efficiency and colony-forming unit per optical density
We evaluated our φC31 phage integration systems by calculating the transformation efficiency as calculated by the following equation:
The editing efficiency is calculated by:
Construction of TIE-1 mutants
The following strains were constructed using the markerless deletion method, as illustrated in Fig. S1: phaZ mutant ΔRpal_0578, and phaR mutant (Δrpal_0531). The two mutants Δgly and the ΔnifA double mutant have been described previously (37). The mutant constructions were as follows: 1 kb upstream and 1 kb downstream of each of the genes of interest were PCR-amplified and cloned into the cloning vector pJQ200KS. The constructed plasmids were then electroporated into the donor strain E. coli S17-1/λ. The E. coli donor strains carrying the constructed plasmids were conjugated with the TIE-1 wild-type. After two consecutive homologous recombination events, successful mutant candidates were screened using the PCR (Fig. S2) and the primers listed in Table S5.
PHA extraction and analysis
PHA samples were extracted from 5 mL of the liquid bacterial culture, and the organic phases were analyzed using LC-MS using a method we developed as previously reported (27). About 325 µL of methanol, 400 µL chloroform, and 75 µL of sulfuric acid were simultaneously added to the sample pellet. The mixtures were incubated at 95°C for only 1 hour to minimize the potential loss of methylated products in the aqueous phase. Subsequently, phase separation and acid elimination were performed by the addition of 500 µL of LC-MS-grade water. The organic phases were dried using a speed vacuum. Dried crotonic acid was then resuspended in 50% acetonitrile/50% water and analyzed using an Agilent Technologies 6420 Triple Quad LC/MS. Crotonic acid was detected with a mass-to-charge ratio (m/z) of 87. Standard curves of 1, 10, 50, and 100 ppm concentrations were created from a polyhydroxybutyrate (PHB) standard purchased from Sigma Aldrich. Standards were prepared the same way as the samples and were used for the determination of PHA from different samples.
PHA (mg/L) concentrations were determined using the standard curve generated from the counts obtained from the following dry powder PHB concentrations: 1, 10, 50, and 100 ppm.
PHA (mg/L/Cell) was obtained from the following equation:
PHA productivity in (mg/L/cell/hour) was obtained from the following equation:
OD measurements, cell counts, and cell volume determinations
OD measurements were performed using a Spectronic 200 (Thermo Fisher Scientifc, USA) at 660 nm. Cell enumerations were performed and have been used previously to report PHA production (27, 70). The cell volume was determined by staining TIE-1 with FM 1–43FX membrane stain and visualized under a Nikon, A1 confocal microscope Model Eclipse Ti2. Stained cells were 3-D-imaged using the FM 1–43 laser channel at 471 nm. Cells were stained according to the manufacturer’s recommendations with a slight modification. Briefly, 1 mL of the culture cell of OD660 ~1 was spun down at 15 k rpm for 1 minute, washed once with 1X PBS, and resuspended in 100 µL of ice-cold methanol for 2 minutes. An additional washing with 100 µL ice-cold acetone was performed. Cells were finally resuspended in 1.5 mL 1X PBS buffer and stained with the FM-1–43FX dye for 10 minutes in the dark. Cells were mounted on a coverslip coated with 10% poly-L-lysine and air-dried before observation under the microscope. The TIE-1 volume was determined as 1.048 pm3 using ~1,000 individual cells. PHA cell % vol/vol was obtained by dividing the volume of PHA per cell with the cell volume assuming that PHA density is 1.22 g/cm3 (71), and the cell % wt/wt fresh weight was obtained by dividing PHA mass by the fresh cell mass assuming that the cell density is 1.1 g/cm3 (72). The dry cell weight was estimated to be 40% of the fresh cell weight (73).
Genomic DNA sequencing of the engineered RuBisCo strains Ωrub(I) and Ωrub(I&II)
DNA extracted from each strain was submitted to Plasmidsaurus. Bacterial genome sequencing was performed by Plasmidsaurus using Oxford Nanopore Technology. Raw reads were downloaded from Plasmidsaurus and were analyzed and annotated on KBase (74) using the following pipeline.
Read quality filtering was performed using Filtlong (75), keeping the best 90% of reads by quality score, with a minimum length of 1,000 bp. Quality control of filtered reads was performed using FastQC (76). Assembly of reads was performed using the long-read assemblers MiniASM (77) and Flye (78). Quality of assemblies was checked using Quast (79). Finished assemblies were annotated for features (e.g., genes and RNA-coding regions) using RAST (80). Genomes were uploaded to the RAST server (81) viewed using the SEED viewer (82), and construct integration sites were confirmed using BLAST (83).
ACKNOWLEDGMENTS
We would like to thank Dr. Joshua Blodgett for facilitating PHA analysis using LC-MS. We would like to thank Jungwoo Lee for proofreading the manuscript.
This work was supported by the following grants to A.B.: The David and Lucile Packard Foundation Fellowship (201563111), the U.S. Department of Energy (grant number DESC0014613), and the U.S. Department of Defense, Army Research Office (grant number W911NF-18-1-0037), Gordon and Betty Moore Foundation, National Science Foundation (Grant Number 2021822, Grant Number 2124088, Grant Number 2117198, and Grant Number 2300081), the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract DEAC5207NA27344 (LLNL-JRNL-812309), NIGMS grant (NIHR01GM141344), and a DEPSCoR grant (FA9550-21-1-0211). A.B. was also funded by a Collaboration Initiation Grant, an Office of the Vice-Chancellor of Research Grant, an International Center for Energy, Environment, and Sustainability Grant, and a SPEED grant from Washington University in St. Louis.
T.O.R., W.B., M.S., J.O., H.S., and E.C. performed all the experiments. R.K. assisted with the bioelectrochemical experiments. BG assisted with the genome assembly. T.O.R., W.B., and A.B. wrote the manuscript with input from all authors.
Footnote
This article is a direct contribution from Arpita Bose, a member of the Applied and Environmental Microbiology Editorial Board, who arranged for and secured reviews by Baptiste Leroy, Universite de Mons, and Daniel Puyol, Universidad Rey Juan Carlos.
SUPPLEMENTAL MATERIAL
Supplemental material - aem.01438-24-s0001.docx
Figures S1 to S5; Tables S1 to S5.
- Download
- 927.97 KB
ASM does not own the copyrights to Supplemental Material that may be linked to, or accessed through, an article. The authors have granted ASM a non-exclusive, world-wide license to publish the Supplemental Material files. Please contact the corresponding author directly for reuse.
REFERENCES
1.
Keasling JD. 2010. Manufacturing molecules through metabolic engineering. Science 330:1355–1358.
2.
Müller T, Schick S, Beck J, Sprenger G, Takors R. 2023. Synthetic mutualism in engineered E. coli mutant strains as functional basis for microbial production consortia. Eng Life Sci 23.
3.
Panda S, Zhou K. 2023. Engineering microbes to overproduce natural products as agrochemicals. Synth Syst Biotechnol 8:79–85.
4.
Kyriakou M, Patsalou M, Xiaris N, Tsevis A, Koutsokeras L, Constantinides G, Koutinas M. 2020. Enhancing bioproduction and thermotolerance in Saccharomyces cerevisiae via cell immobilization on biochar: application in a citrus peel waste biorefinery. Renewable Energy.
5.
Saini M, Hong Chen M, Chiang C-J, Chao Y-P. 2015. Potential production platform of n-butanol in Escherichia coli. Metab Eng 27:76–82.
6.
Wang J, Lu X, Ying H, Ma W, Xu S, Wang X, Chen K, Ouyang P. 2018. A novel process for cadaverine bio-production using A consortium of two engineered Escherichia coli. Front Microbiol 9:1312.
7.
Li H, Gao S, Zhang S, Zeng W, Zhou J. 2021. Effects of metabolic pathway gene copy numbers on the biosynthesis of (2S)-naringenin in Saccharomyces cerevisiae. J Biotechnol 325:119–127.
8.
Orozco PA. 2023. Lanthipeptide biosynthesis in marine Synechococcus: expression, characterization, and bioengineering of a wide variety of synechococsins. https://research.rug.nl/en/publications/lanthipeptide-biosynthesis-in-marine-synechococcus-expression-cha.
9.
Zhang F, Rodriguez S, Keasling JD. 2011. Metabolic engineering of microbial pathways for advanced biofuels production. Curr Opin Biotechnol 22:775–783.
10.
Miller EM, Nickoloff JA. 1995. Escherichia coli electrotransformation, p 105–113. In Electroporation protocols for microorganisms. Springer.
11.
Mäkelä J, Sherratt DJ. 2020. Organization of the Escherichia coli chromosome by a MukBEF axial core. Mol Cell 78:250–260.
12.
van Rossum HM, Kozak BU, Pronk JT, van Maris AJA. 2016. Engineering cytosolic acetyl-coenzyme A supply in Saccharomyces cerevisiae: pathway stoichiometry, free-energy conservation and redox-cofactor balancing. Metab Eng 36:99–115.
13.
DiCarlo JE, Norville JE, Mali P, Rios X, Aach J, Church GM. 2013. Genome engineering in Saccharomyces cerevisiae using CRISPR-Cas systems. Nucleic Acids Res 41:4336–4343.
14.
Cakar ZP, Turanli-Yildiz B, Alkim C, Yilmaz U. 2012. Evolutionary engineering of Saccharomyces cerevisiae for improved industrially important properties. FEMS Yeast Res 12:171–182.
15.
den Haan R, Kroukamp H, Mert M, Bloom M, Görgens JF, van Zyl WH. 2013. Engineering Saccharomyces cerevisiae for next generation ethanol production. J Chem Techn Biotech 88:983–991.
16.
Abbott DA, Zelle RM, Pronk JT, van Maris AJA. 2009. Metabolic engineering of Saccharomyces cerevisiae for production of carboxylic acids: current status and challenges. FEMS Yeast Res 9:1123–1136.
17.
Cao X, Yu W, Chen Y, Yang S, Zhao ZK, Nielsen J, Luan H, Zhou YJ. 2023. Engineering yeast for high-level production of diterpenoid sclareol. Metab Eng 75:19–28.
18.
Parapouli M, Vasileiadis A, Afendra A-S, Hatziloukas E. 2020. Saccharomyces cerevisiae and its industrial applications. AIMS Microbiol 6:1–31.
19.
Luo H, Zheng P, Bilal M, Xie F, Zeng Q, Zhu C, Yang R, Wang Z. 2020. Efficient bio-butanol production from lignocellulosic waste by elucidating the mechanisms of Clostridium acetobutylicum response to phenolic inhibitors. Sci Total Environ 710:136399.
20.
Wang Q, Yu H, Xia Y, Kang Z, Qi Q. 2009. Complete PHB mobilization in Escherichia coli enhances the stress tolerance: a potential biotechnological application. Microb Cell Fact 8:1–9.
21.
Ferreira S, Pereira R, Wahl SA, Rocha I. 2020. Metabolic engineering strategies for butanol production in Escherichia coli. Biotechnol Bioeng 117:2571–2587.
22.
Gong C, Cao L, Fang D, Zhang J, Kumar Awasthi M, Xue D. 2022. Genetic manipulation strategies for ethanol production from bioconversion of lignocellulose waste. Bioresour Technol 352:127105.
23.
Calero P, Nikel PI. 2019. Chasing bacterial chassis for metabolic engineering: a perspective review from classical to non-traditional microorganisms. Microb Biotechnol 12:98–124.
24.
Brown B, Wilkins M, Saha R. 2022. Rhodopseudomonas palustris: a biotechnology chassis. Biotechnol Adv 60:108001.
25.
Azambuja SPH, Goldbeck R. 2020. Butanol production by Saccharomyces cerevisiae: perspectives, strategies and challenges. World J Microbiol Biotechnol 36:48.
26.
Jiao Y, Kappler A, Croal LR, Newman DK. 2005. Isolation and characterization of a genetically tractable photoautotrophic Fe(II)-oxidizing bacterium, Rhodopseudomonas palustris strain TIE-1. Appl Environ Microbiol 71:4487–4496.
27.
Ranaivoarisoa TO, Singh R, Rengasamy K, Guzman MS, Bose A. 2019. Towards sustainable bioplastic production using the photoautotrophic bacterium Rhodopseudomonas palustris TIE-1. J Ind Microbiol Biotechnol 46:1401–1417.
28.
Karthikeyan R, Singh R, Bose A. 2019. Microbial electron uptake in microbial electrosynthesis: a mini-review. J Ind Microbiol Biotechnol 46:1419–1426.
29.
Cabecas Segura P, Onderwater R, Deutschbauer A, Dewasme L, Wattiez R, Leroy B. 2022. Study of the production of poly(Hydroxybutyrate-co-hydroxyhexanoate) and Poly(Hydroxybutyrate-co-Hydroxyvalerate-co-Hydroxyhexanoate) in Rhodospirillum rubrum. Appl Environ Microbiol 88:e0158621.
30.
Bayon-Vicente G, Zarbo S, Deutschbauer A, Wattiez R, Leroy B. 2020. Photoheterotrophic assimilation of valerate and associated polyhydroxyalkanoate production by Rhodospirillum rubrum Appl Environ Microbiol 86:e00901-20.
31.
Conners E, Rengasamy K, Ranaivoarisoa T, Bose A. 2024. The phototrophic purple non-sulfur bacteria Rhodomicrobium spp. are novel chassis for bioplastic production. Microb Biotechnol. https://doi.org/10.1111/1751-7915.14552.
32.
Bose A, Gardel EJ, Vidoudez C, Parra EA, Girguis PR. 2014. Electron uptake by iron-oxidizing phototrophic bacteria. Nat Commun 5:3391.
33.
Rengasamy K, Ranaivoarisoa T, Singh R, Bose A. 2018. An insoluble iron complex coated cathode enhances direct electron uptake by Rhodopseudomonas palustris TIE-1. Bioelectrochemistry 122:164–173.
34.
Bose A, Newman DK. 2011. Regulation of the phototrophic iron oxidation (pio) genes in Rhodopseudomonas palustris TIE-1 is mediated by the global regulator, FixK. Mol Microbiol 79:63–75.
35.
Rengasamy K, Ranaivoarisoa T, Bai W, Bose A. 2021. Magnetite nanoparticle anchored graphene cathode enhances microbial electrosynthesis of polyhydroxybutyrate by Rhodopseudomonas palustris TIE-1. Nanotechnology 32:035103.
36.
Rabaey K, Girguis P, Nielsen LK. 2011. Metabolic and practical considerations on microbial electrosynthesis. Curr Opin Biotechnol 22:371–377.
37.
Bai W, Ranaivoarisoa TO, Singh R, Rengasamy K, Bose A. 2021. N-butanol production by Rhodopseudomonas palustris TIE-1. Commun Biol 4:1257.
38.
Gupta D, Sutherland MC, Rengasamy K, Meacham JM, Kranz RG, Bose A. 2019. Photoferrotrophs produce a PioAB electron conduit for extracellular electron uptake. MBio 10.
39.
Singh R, Ranaivoarisoa TO, Gupta D, Bai W, Bose A. 2020. Genetic redundancy in iron and manganese transport in the metabolically versatile bacterium Rhodopseudomonas palustris TIE-1. Appl Environ Microbiol 86:e01057-20.
40.
Guzman MS, Rengasamy K, Binkley MM, Jones C, Ranaivoarisoa TO, Singh R, Fike DA, Meacham JM, Bose A. 2019. Phototrophic extracellular electron uptake is linked to carbon dioxide fixation in the bacterium Rhodopseudomonas palustris. Nat Commun 10:1355.
41.
McKinlay JB, Oda Y, Rühl M, Posto AL, Sauer U, Harwood CS. 2014. Non-growing Rhodopseudomonas palustris increases the hydrogen gas yield from acetate by shifting from the glyoxylate shunt to the tricarboxylic acid cycle. J Biol Chem 289:1960–1970.
42.
Gosse JL, Engel BJ, Rey FE, Harwood CS, Scriven LE, Flickinger MC. 2007. Hydrogen production by photoreactive nanoporous latex coatings of nongrowing Rhodopseudomonas palustris CGA009. Biotechnol Prog 23:124–130.
43.
Guss AM, Rother M, Zhang JK, Kulkarni G, Metcalf WW. 2008. New methods for tightly regulated gene expression and highly efficient chromosomal integration of cloned genes for Methanosarcina species. Archaea 2:193–203.
44.
van Kessel JC, Marinelli LJ, Hatfull GF. 2008. Recombineering mycobacteria and their phages. Nat Rev Microbiol 6:851–857.
45.
Huff J, Czyz A, Landick R, Niederweis M. 2010. Taking phage integration to the next level as a genetic tool for mycobacteria. Gene 468:8–19.
46.
Huang H, Chai C, Yang S, Jiang W, Gu Y. 2019. Phage serine integrase-mediated genome engineering for efficient expression of chemical biosynthetic pathway in gas-fermenting Clostridium ljungdahlii. Metab Eng 52:293–302.
47.
Baltz RH. 2012. Streptomyces temperate bacteriophage integration systems for stable genetic engineering of actinomycetes (and other organisms). J Ind Microbiol Biotechnol 39:661–672.
48.
Meereboer KW, Misra M, Mohanty AK. 2020. Review of recent advances in the biodegradability of polyhydroxyalkanoate (PHA) bioplastics and their composites. Green Chem 22:5519–5558.
49.
Sabbagh F, Muhamad II. 2017. Production of poly-hydroxyalkanoate as secondary metabolite with main focus on sustainable energy. Renew Sustain Energy Rev 72:95–104.
50.
Karan H, Funk C, Grabert M, Oey M, Hankamer B. 2019. Green bioplastics as part of a circular bioeconomy. Trends Plant Sci 24:237–249.
51.
Bátori V, Åkesson D, Zamani A, Taherzadeh MJ, Sárvári Horváth I. 2018. Anaerobic degradation of bioplastics: a review. Waste Manag 80:406–413.
52.
Sirohi R, Prakash Pandey J, Kumar Gaur V, Gnansounou E, Sindhu R. 2020. Critical overview of biomass feedstocks as sustainable substrates for the production of polyhydroxybutyrate (PHB). Bioresour Technol 311:123536.
53.
Brodin M, Vallejos M, Opedal MT, Area MC, Chinga-Carrasco G. 2017. Lignocellulosics as sustainable resources for production of bioplastics – a review. J Clean Prod 162:646–664.
54.
Kourmentza C, Plácido J, Venetsaneas N, Burniol-Figols A, Varrone C, Gavala HN, Reis MAM. 2017. Recent advances and challenges towards sustainable polyhydroxyalkanoate (PHA) production. Bioengineering (Basel) 4:55.
55.
Maehara A, Doi Y, Nishiyama T, Takagi Y, Ueda S, Nakano H, Yamane T. 2001. PhaR, a protein of unknown function conserved among short-chain-length polyhydroxyalkanoic acids producing bacteria, is a DNA-binding protein and represses Paracoccus denitrificans phaP expression in vitro. FEMS Microbiol Lett 200:9–15.
56.
York GM, Stubbe J, Sinskey AJ. 2002. The Ralstonia eutropha PhaR protein couples synthesis of the PhaP phasin to the presence of polyhydroxybutyrate in cells and promotes polyhydroxybutyrate production. J Bacteriol 184:59–66.
57.
Červenka EJ. 2022. Optimizing PHB bioplastic accumulation in sinorhizobium meliloti grown on CO2-derived formate and bicarbonate. Queen’s University (Canada).
58.
Koch M, Doello S, Gutekunst K, Forchhammer K. 2019. PHB is produced from glycogen turn-over during nitrogen starvation in Synechocystis sp. PCC 6803. IJMS 20:1942.
59.
Wang J, Yu H-Q. 2007. Biosynthesis of polyhydroxybutyrate (PHB) and extracellular polymeric substances (EPS) by Ralstonia eutropha ATCC 17699 in batch cultures. Appl Microbiol Biotechnol 75:871–878.
60.
Li Z, Xin X, Xiong B, Zhao D, Zhang X, Bi C. 2020. Engineering the calvin–benson–bassham cycle and hydrogen utilization pathway of Ralstonia eutropha for improved autotrophic growth and polyhydroxybutyrate production. Microb Cell Fact 19:1–9.
61.
Higuchi-Takeuchi M, Morisaki K, Toyooka K, Numata K. 2016. Synthesis of high-molecular-weight polyhydroxyalkanoates by marine photosynthetic purple bacteria. PLoS One 11:e0160981.
62.
Kamravamanesh D, Pflügl S, Nischkauer W, Limbeck A, Lackner M, Herwig C. 2017. Photosynthetic poly-β-hydroxybutyrate accumulation in unicellular cyanobacterium Synechocystis sp. PCC 6714. AMB Express 7:143.
63.
Quandt J, Hynes MF. 1993. Versatile suicide vectors which allow direct selection for gene replacement in gram-negative bacteria. Gene 127:15–21.
64.
Khan SR, Gaines J, Roop RM II, Farrand SK. 2008. Broad-host-range expression vectors with Tightly regulated promoters and their use to examine the influence of TraR and TraM expression on ti plasmid quorum sensing. Appl Environ Microbiol 74:5053–5062.
65.
Vacani PF, Malek MM, Davey PG. 1993. Cost of gentamicin assays carried out by microbiology laboratories. J Clin Pathol 46:890–895.
66.
Phillips GJ. 1999. New cloning vectors with temperature-sensitive replication. Plasmid 41:78–81.
67.
Nishihata S, Kondo T, Tanaka K, Ishikawa S, Takenaka S, Kang C-M, Yoshida K-I. 2018. Bradyrhizobium diazoefficiens USDA110 PhaR functions for pleiotropic regulation of cellular processes besides PHB accumulation. BMC Microbiol 18:156.
68.
Kobayashi J, Kondo A. 2019. Disruption of poly (3-hydroxyalkanoate) depolymerase gene and overexpression of three poly (3-hydroxybutyrate) biosynthetic genes improve poly (3-hydroxybutyrate) production from nitrogen rich medium by Rhodobacter sphaeroides. Microb Cell Fact 18:40.
69.
Awogbemi O, Kallon DVV. 2022. Application of tubular reactor technologies for the acceleration of biodiesel production. Bioengineering (Basel) 9:347.
70.
Escapa IF, del Cerro C, García JL, Prieto MA. 2013. The role of GlpR repressor in Pseudomonas putida KT2440 growth and PHA production from glycerol. Environ Microbiol 15:93–110.
71.
Majerczak K, Wadkin‐Snaith D, Magueijo V, Mulheran P, Liggat J, Johnston K. 2022. Polyhydroxybutyrate: a review of experimental and simulation studies of the effect of fillers on crystallinity and mechanical properties. Polym Int 71:1398–1408.
72.
Bjørnsen P, Riemann B. 1988. Towards a quantitative stage in the study of microbial processes in pelagic carbon flows. Ergebn Limnol 31:185–193.
73.
Bratbak G, Dundas I. 1984. Bacterial dry matter content and biomass estimations. Appl Environ Microbiol 48:755–757.
74.
Arkin AP, Cottingham RW, Henry CS, Harris NL, Stevens RL, Maslov S, Dehal P, Ware D, Perez F, Canon S, et al. 2018. KBase: the United States department of energy systems biology knowledgebase. Nat Biotechnol 36:566–569.
75.
Wick R. 2024. Quality filtering tool for long reads. Available from: https://github.com/rrwick/Filtlong. Retrieved 11 Apr 2024.
76.
Andrews S. 2010. FastQC: a quality control tool for high throughput sequence data. Babraham Bioinformatics. Available from: http://www. bioinformatics. babraham. ac. uk/projects/fastqc
77.
Li H. 2016. Minimap and miniasm: fast mapping and de novo assembly for noisy long sequences. Bioinformatics 32:2103–2110.
78.
Kolmogorov M, Yuan J, Lin Y, Pevzner PA. 2019. Assembly of long, error-prone reads using repeat graphs. Nat Biotechnol 37:540–546.
79.
Gurevich A, Saveliev V, Vyahhi N, Tesler G. 2013. QUAST: quality assessment tool for genome assemblies. Bioinformatics 29:1072–1075.
80.
Brettin T, Davis JJ, Disz T, Edwards RA, Gerdes S, Olsen GJ, Olson R, Overbeek R, Parrello B, Pusch GD, Shukla M, Thomason JA 3rd, Stevens R, Vonstein V, Wattam AR, Xia F. 2015. RASTtk: a modular and extensible implementation of the RAST algorithm for building custom annotation pipelines and annotating batches of genomes. Sci Rep 5:8365.
81.
Aziz RK, Bartels D, Best AA, DeJongh M, Disz T, Edwards RA, Formsma K, Gerdes S, Glass EM, Kubal M, et al. 2008. The RAST server: rapid annotations using subsystems technology. BMC Genomics 9:1–15.
82.
Overbeek R, Begley T, Butler RM, Choudhuri JV, Chuang H-Y, Cohoon M, de Crécy-Lagard V, Diaz N, Disz T, Edwards R, et al. 2005. The subsystems approach to genome annotation and its use in the project to annotate 1000 genomes. Nucleic Acids Res 33:5691–5702.
83.
Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ. 1990. Basic local alignment search tool. J Mol Biol 215:403–410.
Information & Contributors
Information
Published In
Applied and Environmental Microbiology
Volume 90 • Number 9 • 18 September 2024
eLocator: e01438-24
Editor: Gemma Reguera, Michigan State University, East Lansing, Michigan, USA
PubMed: 39162566
Copyright
© 2024 Ranaivoarisoa et al. This is an open-access article distributed under the terms of the Creative Commons Attribution 4.0 International license.
History
Received: 18 July 2024
Accepted: 30 July 2024
Published online: 20 August 2024
Keywords
Data Availability
All data in the paper will be provided by authors upon request.
Contributors
Editor
Gemma Reguera
Editor
Michigan State University, East Lansing, Michigan, USA
Metrics & Citations
Metrics
Note:
- For recently published articles, the TOTAL download count will appear as zero until a new month starts.
- There is a 3- to 4-day delay in article usage, so article usage will not appear immediately after publication.
- Citation counts come from the Crossref Cited by service.
Citations
Citation
2024.
Overexpression of RuBisCO form I and II genes in Rhodopseudomonas palustris TIE-1 augments polyhydroxyalkanoate production heterotrophically and autotrophically. Appl Environ Microbiol
90:e01438-24.
If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. For an editable text file, please select Medlars format which will download as a .txt file. Simply select your manager software from the list below and click Download.