Determination of the contribution of the CBB operons and hydrogenases to autotrophic growth of R. eutrophic H16
The CBB cycle is an expensive metabolic pathway that consumes large amounts of energy and reducing equivalents. The production of 1 mol of 3-PG from 3 mol of CO2 requires 9 mol of ATP and 6 mol of NADPH. In R. eutropha H16 cells, the hydrogenase systems are employed to provide both the energy and the reducing equivalents for the CBB cycle [4]. The hydrogenases of R. eutropha H16 are insensitive to oxygen, which is rather unusual. They catalyze the oxidation of molecular hydrogen into protons and electrons, which are then transferred to membrane-bound or cytoplasmic electron carriers with oxygen as the terminal electron acceptor [26]. This process either generates the proton electromotive force for ATP production [27], or is used for the regeneration NADH through the cytoplasmic electron transport chain [28]. There are few studies on the three main hydrogenases of the autotrophic system of R. eutropha H16, i.e. SH, MBH and RH, and their individual contribution to carbon fixation and autotrophic growth is unknown. It was reported that the CBB operon on chromosome 2 of R. eutropha H16 contains 13 CBB coding sequences, along with a cbbR gene on the negative strand, while the CBB operon on the megaplasmid of R. eutropha H16 consists of 12 CBB coding sequences with a deficient cbbR gene. The CbbR expressed from chromosome 2 was considered to control the expression of both CBB operons [19]. Here, we estigated which hydrogenase or CBB operon is more important for autotrophic growth by constructing corresponding knockout strains.
The knockout strains and strains carrying overexpression plasmids were subjected to autotrophic fermentation in minimal medium supplemented with a gas mixture comprising H2, CO2, and O2 at a volume ratio of 7:1:1. Under such conditions, CO2 was the only carbon source for the synthesis of cellular building blocks, and the cell growth efficiency was assumed to be directly correlated with the carbon fixation efficiency. The strains are summarized in supplement Table 1, and their autotrophic growth phenotype is illustrated in Figure 1. The results indicated that a double knockout of both MBH and SH hydrogenases completely eliminated the autotrophic growth capacity of R. eutropha H6, while a single deletion of either MBH or SH only partly affected the autotrophic growth. Based on the growth profile, the deletion of SH had a more significant impact (Figure 1A). Thus, while both hydrogenases were functional during autotrophic growth of R. eutropha H6 and contributed to the cell metabolism, the results indicated that SH probably played a more significant role.
In the case of the CBB enzymes, both single deletion of RuC (RuBisCO operon on Chromosome 2) or RuP (RuBisCO operon on the megaplasmid) affected the autotrophic growth efficiency, and the growth decrease was similar for both operons (Figure 1B), which indicated that both CBB operons were active and contributed almost equally to the carbon fixation process.
To determine if there was a possible polar effect of the knockout experiments, complementation experiments were performed to see if we could restore the autotrophic growth capacity of the double knock out strains of RuBisCO or hydrogenases via plasmid-based expression of the knocked-out genes. The complemented strain H16ΔRuPΔRuC(pRub_R) recovered its autotrophic growth ability, indicating that there was no polar effect of the RuBisCO knockouts (Figure 1C). However, the complementation of the two hydrogenases in strain H16ΔMBHΔSH was not successful. Considering the difficulties of a manipulating large number of genes in R. eutropha, the failed complementation experiment does not necessarily indicate a polar effect for the knockout.
In summary, we determined that both the SH and MBH hydrogenases contribute to the autotrophic growth of R. eutropha, and both CBB operons are active in the carbon fixation process.
Engineering the CBB cycle for improved autotrophic growth of R. eutropha
The autotrophic metabolism of R. eutropha H16 constitutes of the CBB cycle and hydrogenases. In the CBB cycle, RuBisCO catalyzes the carboxylation reaction converting ribulose-1,5-diphosphate and CO2 to generate 2 molecules of 3-phosphoglyceric acid for the synthesis of organic carbon compounds. The efficiency of RuBisCO is low and it is considered the speed-limiting step of the CBB cycle (Figure 2) [9]. Therefore, we intended to improve the carbon fixation efficiency of R. eutropha H16 by increase the efficiency of its RuBisCO enzyme. The RuBisCO with the highest reported efficiency is that from the cyanobacterium Synechococcus sp. PCC 7002 [8]. Due to the complex structure of RuBisCO, its heterologous folding and maturation might not be ideal and requires chaperones or accessory proteins. Previous studies have shown that the GroES/EL chaperone system of E. coli is of great significance for the folding of heterologous RuBisCO proteins [29]. Therefore, in this study we attempted to overexpress the endogenous RuBisCO from R. eutropha or the heterologous cyanobacterial RuBisCO, coupled with various chaperone systems to find an optimal strategy for increasing the CO2 fixation capacity.
Overexpression plasmids with different combinations of RuBisCO genes and chaperone systems were constructed based on the pBBR1-MCS multiple-copy vector as listed in Table 1. The strains carrying these plasmids were subjected for autotrophic gas fermentation in minimal medium supplemented with the gas mixture as described above. However, we found that most of the engineered strains had decreased growth compared with the control strain carrying an rfp expression plasmid with the same vector backbone (Fig. 3). The strains with decreased growth included those overexpressing only RuBisCO genes, RuBisCO genes together with the E. coli chaperone genes groES/groEL, and the endogenous R. eutropha RuBisCO together with endogenous groES/groEL. Only the strain H16(pRub_cyano, pGroESL_R), which overexpresses the Synechococcus sp. PCC 7002 RuBisCO genes together with the endogenous chaperone genes groES/groEL showed an increased growth phenotype. Its OD600 after 72 hours of growth was 89.15% higher than that of the control strain (Fig. 3E).
These results indicated that the assembly and maturation of a functional cyanobacterial RuBisCO in R. eutropha was successfully achieved with the assistance of overexpressed endogenous GroES/EL chaperons, as well as confirming the feasibility of increasing the carbon fixation efficiency of R. eutropha using heterologous RuBisCO enzymes.
Engineering of the hydrogenase module and CBB cycle for improved autotrophic growth of R. eutropha H16
To engineer the hydrogenase systems, the genes encoding each hydrogenase were overexpressed using the same plasmids that were constructed for complementation in the knockout experiments, and the engineered strains were analyzed for their autotrophic growth phenotype. As illustrated in Fig. 4, while overexpression of the hoxABCJ genes encoding the RH hydrogenase decreased the growth (Fig. 4A), while overexpressing the PHG064 and PHG065 genes encoding Hy4 had no effect. By contrast, expression of either hoxFUYHI encoding SH or hoxKGZ encoding MBH had a positive effect on the autotrophic growth. Compared with the control, after 96h of autotrophic growth, the OD600 of the SH overexpression strain C5(pRH_R) and MBH overexpression strain C5(pMBH_R) increased 13.8 and 58.71%, respectively.
Since the plasmid expression system has limitations for more complex engineering, we intended to engineer the hydrogenase systems by modification the R. eutropha genome. The BBaJ_23100, BBaJ_23109 and BBaJ_23119 promoters from E. coli were selected (supplement Table 3) to modulate the expression of the MBH and SH hydrogenase operons. The corresponding strains were constructed by replacing the original promoters with the BBaJ promoters. Because the R. eutropha C5 strain we constructed is simpler to transform [23, 24], we decided to use it instead of H16 for these experiments. As shown in Figure 5A, while the promoters BBaJ_23100 and BBaJ_23109 failed to increase the autotrophic growth, BBaJ_23119 with a stronger efficiency was able to increase the growth when inserted instead of the original MBH promoter. This effective regulator was introduced to increase the expression of the SH operon to obtain the strain C5-sh19. Subsequently, strain C5-shmbh19 in which both the MBH and SH operons were upregulated by BBaJ_23119 were constructed. Both strains C5-sh19 and C5-sh-mbh19 were found to have increased autotrophic growth compared with the parent strain, and C5-shmbh19 had a slightly higher growth than C5-sh19.
In this part, we modulated the expression of both SH and MBH by both plasmid-based overexpression and chromosomal promoter modulation. The results suggested that increased expression of both hydrogenase operons benefited the autotrophic growth of R. eutropha (Fig. 5B).
Engineering of both the CBB module and the hydrogenase module improved the autotrophic growth and PHB production of R. eutropha
The genomic engineering of the hydrogenase operons provided a basis for engineering a strain in which both the CBB module and the hydrogenase module were engineered. By transforming strain C5-sh-mbh19 with the plasmids pRub_cyano and pGroESL_R, the strain C5-sh-mbh19(pRub_cyano, pGroESL_R) was obtained. Combining all the findings of this research, this strain may represent the most deeply engineered R. eutropha strain to date in terms of its autotrophic metabolism.
R.eutropha can accumulate polyhydroxybutyrate (PHB) [30], which is a potential bioplastic material of great interest, in intracellular granules. Therefore, both the autotrophic growth status and PHB production capacity of C5-sh-mbh19(pRub_cyano, pGroESL_R) were evaluated. As illustrated in Figure 6, the engineered strain displayed a substantial improvement in both the growth phenotype and PHB production, with increases of 93.4% and 74.7% in 96 h respectively compared to the parent strain. C5-mbh-sh19(pRub_cyano, pGroESL_R) produced 0.34 g/g PHB, which was a significant improvement over the 0.17 g/g produced by H16(pBAD-RFP).