Construction of a Rhodobacter sphaeroides Strain That Efficiently Produces Hydrogen Gas from Acetate without Poly(β-Hydroxybutyrate) Accumulation: Insight into the Role of PhaR in Acetate Metabolism

Previously, we observed that the R. sphaeroides 2.4.1-derived strain ΔphaC1, containing a deletion of phaC1 (KEGG locus tag RSP_0382) encoding PHA synthase, exhibited delayed growth under photoheterotrophic growth conditions with acetate and l-glutamate (7), where l-glutamate was supplemented as a poor nitrogen source to derepress genes involved in N₂ fixation. Here, we observed that the ΔphaC1 strain exhibited defects in growth with acetate even under nitrogen-sufficient conditions, where ammonium chloride was supplemented as the sole nitrogen source (Fig. 2A). Under these conditions, PHB accumulation was observed in the wild-type strain, whereas no PHB accumulation was detected in the ΔphaC1 strain (Fig. 2E). We observed that introduction of a plasmid harboring phaC1 (pMGphaC1) restored growth (Fig. 2A) and PHB accumulation (Fig. 2E) of the ΔphaC1 strain under photoheterotrophic growth conditions with acetate. These results indicate that phaC1 is essential for PHB accumulation and normal growth with acetate. The ΔphaC1 strain grew normally with glucose, succinate, and pyruvate (Fig. 3A to C, respectively), while PHB accumulation was not detected in the ΔphaC1 cells grown under these conditions (Fig. 3D to F). These results suggest that PHB accumulation is particularly important for acetate assimilation.

To examine the possibility that the growth defect of the ΔphaC1 strain is due to a diminished ability to harvest light energy, we evaluated the spectral complex formation of the wild-type strain and the ΔphaC1 strain during photoheterotrophic growth with acetate (optical density at 660 nm [OD₆₆₀] ≈ 0.5). As a result, the ΔphaC1 strain produced the spectral complexes during photoheterotrophic growth with acetate, although the amount was less than that of the wild-type strain (Fig. 2D). Subsequently, we evaluated growth of the ΔphaC1 strain with acetate under dark aerobic conditions to examine whether the presence of oxygen supports growth of the ΔphaC1 strain. As a result, the ΔphaC1 strain also exhibited growth defects with acetate under dark aerobic conditions (Fig. 2B). Therefore, the observed severe growth defects were not likely due to an imbalance in the cellular redox state or the diminished ability to harvest light energy.

The phaC2 (KEGG locus tag RSP_1257) present in the genome of R. sphaeroides 2.4.1 shows 37% amino acid sequence identity with phaC1. We also constructed a deletion mutant of phaC2 and evaluated its growth with acetate. Unlike the ΔphaC1 strain, the ΔphaC2 strain exhibited normal growth with acetate (Fig. 2C). Consistent with efficient acetate assimilation, the ΔphaC2 strain accumulated PHB to a level similar to those seen in the wild-type strain (Fig. 2E), indicating that phaC2 is not essential for PHB accumulation under these conditions. In contrast, the ΔphaC12 strain, which harbors deletion of both phaC1 and phaC2, exhibited impaired growth with acetate (Fig. 2C), and PHB accumulation was not detected (Fig. 2E). Introduction of a plasmid harboring phaC1 to the ΔphaC12 strain restored PHB accumulation and growth with acetate. These results confirmed that phaC1, but not phaC2, is essential for PHB accumulation, which is required for efficient acetate assimilation in R. sphaeroides 2.4.1.

To determine the role of phaR in PHB accumulation and growth with acetate in R. sphaeroides 2.4.1, we constructed a deletion mutant of phaR (KEGG locus tag RSP_0380). Similar to the wild-type strain, the ΔphaR strain grew with acetate, while the maximal biomass concentration was 1.3-fold lower than that of the wild-type strain (Fig. 4A). The ΔphaR strain produced 1/3 of the PHB observed in the wild-type strain (Fig. 4C). Introduction of a plasmid harboring phaR (pMGphaR) restored PHB accumulation to the extent seen in the wild-type strain. These results indicate that deletion of phaR reduces PHB accumulation without large effects on acetate assimilation.

PhaR is assumed to act as a transcriptional regulator in response to PHB. To investigate the consequences of phaR deletion on the genetic background of PHB-deficient mutants, we constructed a strain with deletion of both phaC1 and phaR. Surprisingly, the resulting ΔphaC1R strain grew as well as the ΔphaR strain under photoheterotrophic growth conditions with acetate (Fig. 4B), although there was no detectable PHB accumulation (Fig. 4C). Under the conditions used, we further confirmed that deletion of phaC2 in the ΔphaR background had no significant effects on PHB accumulation and growth with acetate. Furthermore, deletion of both phaC1 and phaC2, together with phaR, resulted in growth patterns similar to those of the ΔphaR strain, while no PHB accumulation was detected (Fig. 4C). Introduction of a plasmid harboring phaR (pMGphaR) into the ΔphaC2R strain did not inhibit growth with acetate (Fig. 4B). In contrast, introduction of pMGphaR into the ΔphaC1R and ΔphaC12R strains resulted in almost no growth with acetate until 96 h (Fig. 4B), indicating that the multicopy phaR in PHB-deficient mutants resulted in more severe defects in growth with acetate. It should be noted that they started to grow after 96 h of culture. When these cells were inoculated to fresh medium with acetate, they started to grow within 24 h with growth rates comparable to those of the wild-type strain (see Fig. S1 in the supplemental material), suggestive of suppressor mutations. These results indicate that deletion of phaC1 impairs acetate assimilation when phaR is present in the genome. Given that PHB antagonizes binding of PhaR to DNA (34, 35), genes regulated by PhaR were predicted to be important for acetate assimilation.

We performed transcriptome sequencing (RNA-seq) analysis to predict genes regulated by PhaR. By comparing transcripts of the ΔphaC1 and ΔphaC1R strains grown photoheterotrophically with acetate, we evaluated the effects of phaR deletion in the absence of PHB, which inhibits binding of PhaR to DNA. The matrix of normalized read count data used for differentially expressed gene analysis is shown in Data Set S1. We identified 22 genes whose mRNA levels were upregulated more than 2-fold by deletion of phaR (Table 1). Consistent with the assumption that genes regulated by PhaR are important for acetate assimilation, the expression levels of genes involved in the ethylmalonyl-CoA pathway for acetate assimilation, including ccr, ecm, mcl-2, and mcd (Fig. 1A), were upregulated by deletion of phaR. In agreement with previous studies of R. sphaeroides FJ1 (29, 30), the expression of phaP encoding phasin (Fig. 1B) was upregulated by deletion of phaR. It should be noted that the expression of phaZ encoding intracellular PHB depolymerase was upregulated 1.8-fold (Data set S1), while it was not included in Table 1 since the fold change was less than 2-fold. Although deletion of phaC2 had no significant effects on phenotypes as described earlier, the expression of phaC2 was upregulated by deletion of phaR. In addition, the mRNA levels of genes involved in leucine biosynthesis (leuA, leuC), gluconeogenesis (fbpB), cobalamin biosynthesis (bluB), and heme degradation (bphO) were upregulated. We also observed that the mRNA levels of 133 genes were downregulated by deletion of phaR, (Data set S2), although no PhaR-binding motif was found upstream of these genes, as described below.

PhaR has been reported to bind to a motif sequence, 5′-CTGCN_3-4GCAG-3′, in R. sphaeroides FJ1 (30). We found the PhaR-binding sequence in the upstream region of phaP, phaZ, and phaR (Fig. 5A to C), suggesting that PhaR in R. sphaeroides 2.4.1 recognizes the same motif sequence as that in R. sphaeroides FJ1. Therefore, we searched for the PhaR-binding sequence in the upstream region of each gene whose mRNA levels were upregulated by deletion of phaR. Consequently, we found the PhaR-binding sequence in the intergenic region of ccr and ecm (Fig. 5D) encoding key enzymes for the ethylmalonyl-CoA pathway (Fig. 1A). The PhaR-binding sequence was not found in the upstream region of the other genes whose mRNA levels were upregulated or downregulated by deletion of phaR. Based on the above-described observations, we focused on the role of ccr in the possible PhaR-dependent regulation of acetate metabolism. Notably, introduction of a multicopy plasmid harboring ccr to the ΔphaC1 strain significantly restored growth with acetate (Fig. 5E). These findings suggest that the downregulation of the expression of ccr by PhaR is a major cause of the impaired growth of PHB-deficient mutants with acetate.

Based on the observations described above, deletion of phaC1 together with phaR seemed like a probable means of preventing PHB accumulation with fewer effects on acetate assimilation. Although the wild-type strain does not produce H₂ from acetate in the presence of ammonium, the ΔhupΔcbbP* strain expressing modified nifA from a multicopy plasmid can produce H₂ under the same conditions (8). Using the ΔhupΔcbbP* strain as the reference strain of the constitutive hydrogen producer (CHP), we deleted both phaC1 and phaR with the genetic background of the ΔhupΔcbbP* strain to construct the CHPΔphaC1R strain. For comparison, we also constructed the CHPΔphaC1 and CHPΔphaR strains by deleting phaC1 and phaR, respectively, with the genetic background of the ΔhupΔcbbP* strain. The strains were grown photoheterotrophically with acetate and ammonium chloride in an argon atmosphere, and H₂ production was evaluated. As previously reported. we also added bicarbonate to the culture medium to enhance H₂ production (6). Table 2 summarizes the H₂ production of the CHP strains. As observed in the ΔphaC1 strain, the CHPΔphaC1 strain grew slowly (Fig. 6A), and the amount of H₂ produced was 4.3-fold lower than the H₂ produced by the reference strain ΔhupΔcbbP* during 120 h of culture (Fig. 6B). In agreement with poor growth and H₂ production, more than 60% of acetate remained in the culture medium of the CHPΔphaC1 strain after 120 h (Fig. 6C). These results indicate that deletion of phaC1 is not beneficial for H₂ production from acetate. On the other hand, the CHPΔphaC1R strain harboring deletions in both phaC1 and phaR grew well, similar to the ΔhupΔcbbP* strain, although the maximal biomass concentration was 1.3-fold decreased upon deletion of these genes. These results again demonstrate that deletion of phaC1 impairs growth with acetate, and additional deletion of phaR restores it. Furthermore, the CHPΔphaC1R strain completely consumed acetate within 96 h and produced a 1.4-fold larger amount of H₂ than the ΔhupΔcbbP* strain during 120 h of culture (Fig. 6B). These results indicate that deletion of phaC1 together with phaR is beneficial for H₂ production from acetate. Despite the phaC1-positive genotype, H₂ production, growth, and acetate consumption of the CHPΔphaR strain were similar to those of the CHPΔphaC1R strain.

We assessed PHB accumulation in the CHP strains during H₂ production with acetate to further understand the link between H₂ production and PHB accumulation (Fig. 6D). The reference strain ΔhupΔcbbP* accumulated PHB up to 17.1% of dry cell weight after 72 h and then declined to 11.5% of dry cell weight after 120 h. As phaC1 has been deleted, PHB accumulation was below the detection limit for the CHPΔphaC1 and CHPΔphaC1R strains at any time points. In the CHPΔphaR strain, PHB accumulation was detected in cells cultured for 24 to 72 h; however, PHB was completely degraded after 96 h of culture. This result is consistent with similar H₂ production between the CHPΔphaR and CHPΔphaC1R strains.

We calculated how consumed electrons were utilized in each strain to assess the impact of phaC1 and/or phaR deletion on electron flow (Fig. 6E). In the reference strain ΔhupΔcbbP*, 5% of consumed electrons were utilized to produce PHB. By preventing PHB accumulation by deleting phaC1, H₂ yields were expected to increase to this extent. However, the H₂ yield of the CHPΔphaC1 strain was considerably lower than that of the ΔhupΔcbbP* strain despite no PHB accumulation. Conversely, H₂ yields of the CHPΔphaR and ΔCHPphaC1R strains were higher than expected. These findings imply that PHB accumulation has positive effects on H₂ production through altering the PhaR-dependent regulation. PHB accumulation was not detected in these phaR-negative strains, and biomass formation was decreased compared to the ΔhupΔcbbP* strain.

The bacterial strains and plasmid vectors used in this study are summarized in Table 3. For evaluation of growth with specific carbon sources, R. sphaeroides strains were anaerobically precultured at 30°C with continuous illumination (3,000 K fluorescent lamp, 7000-l×) in a 33-mL rubber stopper-equipped tube containing 10 mL van Niel’s yeast medium (per liter, 10 g yeast extract, 0.5 g MgSO₄ ⋅ 7H₂O, and 1 g K₂HPO₄, pH 7.0) purged with N₂. After 3 days, cells were harvested by centrifugation at 20,400 × g for 1 min at 20°C and resuspended in 0.9% NaCl, followed by inoculation with 10 mL defined mineral medium (7) without sodium bicarbonate. Ammonium chloride (10 mM) was used as the nitrogen source. Carbon sources were added at the following concentrations: 20 mM sodium acetate, 10 mM disodium succinate, 13.3 mM sodium pyruvate, and 6.7 mM glucose. Headspace gas was replaced with N₂ by purging, and the culture tubes were incubated at 30°C with continuous illumination.

To evaluate photofermentative H₂ production with acetate, precultured cells prepared as described above were inoculated into 10 mL defined mineral medium (7) supplemented with 25 mM acetate, 18 mM sodium bicarbonate, and 10 mM ammonium chloride. After headspace gas was replaced with argon by purging, the cells were cultured at 30°C with continuous illumination.

For cultures of R. sphaeroides strains harboring plasmid vectors, kanamycin was added at a final concentration of 20 μg · mL⁻¹. Escherichia coli HST02 was used as a host for gene cloning and plasmid construction, while E. coli S17-1 was used as a donor for conjugation. The E. coli strains were grown in LB medium supplemented with 50 μg · mL⁻¹ kanamycin or 50 μg · mL⁻¹ ampicillin.

For the construction of a plasmid vector for gene deletion of phaC2 (RSP_1257), DNA fragments containing phaC2 were amplified by PCR from the genomic DNA of R. sphaeroides 2.4.1 with the primers RSP_1257_u940_Xba_Fw (5′-GGGTCTAGATGACATTGGGATCGGCATAG-3′) and RSP_1257_d1164__Xba_Rv (5′-GGGTCTAGAAGTTCTCCCGCCTGATTGA-3′) and then subcloned into the XbaI restriction site of the pUC18 vector to generate pUCphaC2. Inverse PCR was performed using a pair of 5′-phosphorylated primers, RSP_1257_1721_Fw (5′-ACCTACATCCTGCAACGCTG-3′) and RSP_1257_30_Rv (5′-ACCTTCCGCATTCCACTTCA-3′), with pUCphaC2 as a template, and the amplified fragments were self-ligated to generate pUCDphaC2. Subsequently, the XbaI-digested fragment of pUCDphaC2 was ligated into the corresponding site of pK19mobsacB to generate pKDphaC2.

To construct a plasmid vector for deletion of phaR (RSP_0380), PCR was performed using the genomic DNA of R. sphaeroides 2.4.1 with the primers RSP_0380_u1041_Hind_Fw (5′-GGGAAGCTTTTCAACGGCTTCCGTCAGTT-3′) and RSP_0380_d1054_Hind_Rv (5′-GGGAAGCTTTGGTCAAGATGCTCTGGTTC-3′) and then subcloned into the HindIII restriction site of the pUC18 vector to generate pUCphaR. Inverse PCR was performed with a pair of 5′-phosphorylated primers, RSP_0380_534_Fw (5′-AAGAAGCTCTCGAAGCTCTGA-3′) and RSP_0380_30_Rv (5′-GATCAGCAACGGCTTGTCAG-3′), with pUCphaR as a template, and the amplified fragments were self-ligated to generate pUCDphaR. The HindIII-digested fragment of pUCDphaR was then ligated into the corresponding site of pK19mobsacB to generate pKDphaR.

To construct a plasmid vector for phaR gene complementation, PCR was performed using the genomic DNA of R. sphaeroides 2.4.1, with primers RSP_0380_EcoRI_Fw (5′-GGGGAATTCTGCCAAGAAAGCCACTGCTC-3′) and RSP_0380_BamHI_Rv (5′-GGGGGATCCTCAGAGCTTCGAGAGCTTCTT-3′). The amplified fragment was digested with EcoRI and BamHI and ligated into the corresponding site of pMG180 to generate pMGphaR.

Gene deletion was performed as previously described (7). Deletion of phaC1, phaC2, and phaR was confirmed by PCR using primers outside of the fragments used for deletion of the corresponding genomic region.

Transformation of R. sphaeroides strains for introduction of a replicating plasmid was carried out using electroporation, as described previously (44).

The wild-type and ΔphaC1 strains were cultured photoheterotrophically in 10 mL minimal medium containing 20 mM acetate until the OD₆₆₀ reached 0.5. Cells were harvested by centrifugation at 5,000 × g at 4°C for 10 min. Cells were resuspended in 0.8 mL of 20 mM potassium phosphate buffer (pH 7.4) containing 1 mM EDTA and disrupted by sonication (20 kHz, 2-s pulse with 10-s intervals for a total of 15 min at 4°C) using a Bioruptor UCD-250 (Cosmo Bio). After centrifugation at 15,400 × g at 4°C for 5 min, the supernatants were diluted to 0.5 mg mL⁻¹ protein. Absorbance spectra (700 to 900 nm) were recorded with a DU800 spectrophotometer (Beckman).

The ΔphaC1 andΔphaC1R strains (three biological replicates for each) were cultured photoheterotrophically in 10 mL mineral medium containing 20 mM acetate until the OD₆₆₀ reached 0.3 to 0.5. RNA was separately isolated from each 10-mL culture as follows. Cells were harvested by centrifugation at 5,000 × g for 10 min at 4°C and treated with RNAprotect bacterial reagent (Qiagen). The resulting pellets were resuspended in 350 μL buffer RA1 (TaKaRa) and disrupted by shaking with glass beads using a multibead shocker MB2000 (Yasui Kikai). After centrifugation at 20,400 × g for 10 min, RNA was purified from the supernatant using a NucleoSpin RNA minikit (TaKaRa). rRNA was removed using the Ribo-Zero rRNA removal kit (Illumina), and cDNA library preparation was done with a NEBNext Ultra directional RNA library prep kit for Illumina (New England Biolabs [NEB]). On a NovaSeq 6000 system (Illumina), cDNA libraries were subjected to 150 × 2-bp paired-end sequencing.

Data quality was confirmed using FastQC (version 0.11.7) (https://www.bioinformatics.babraham.ac.uk/projects/fastqc/). Trimmomatic (version 0.38) was used to trim the sequence data (45). The sequence reads were aligned to the R. sphaeroides 2.4.1 reference genome (ASM1290v2) using HISAT2 (version 2.1.0) (46). Counts per million were calculated using featureCounts (version 1.6.3) (47). Differential gene expression analysis was performed with DESeq2 (48) (version 1.24.0).

Cells were harvested by centrifugation (20,400 × g for 5 min at 4°C) from 1 mL culture and digested in 1 mL My Kitchen bleach (Rocket Soap Co., Ltd.) at 37°C for 2 h. Following centrifugation (20,400 × g for 10 min), the pellets were washed once with 1 mL deionized water, followed by a wash with 1 mL 1:1 ethanol-acetone, and then dried in vacuo. After resuspending the pellets in 1 mL sulfuric acid, the solution was transferred to a glass tube and heated at 100°C for 15 min. PHB was then measured spectrophotometrically as crotonic acid (49) using poly[(R)-3-hydroxybutyric acid] (Sigma-Aldrich) as a reference. Based on the experimentally determined relationship between dry cell weight and OD₆₆₀ values for each strain, the dry cell weight of R. sphaeroides strains was calculated using OD₆₆₀ values.

H₂ was measured using a GC 2014 gas chromatograph (Shimadzu) equipped with a thermal conductivity detector and a molecular sieve-13 × 60/80 (GL Science), as previously described (7). Concentrations of acetate in culture supernatants were determined using a Nexera XR ultra-high-performance liquid chromatograph (Shimadzu) equipped with a Synergi Hydro-RP-HST column (Phenomenex) and a photodiode detector, as previously described (7).