Skip to main content
NCBI home page
Microbial Biotechnology logoLink to Microbial Biotechnology
. 2021 Jul 21;14(5):2072–2089. doi: 10.1111/1751-7915.13884

Functional dissection and modulation of the BirA protein for improved autotrophic growth of gas‐fermenting Clostridium ljungdahlii

Can Zhang 1,2, Xiaoqun Nie 1, Huan Zhang 1,2, Yuwei Wu 1,2, Huiqi He 1, Chen Yang 1, Weihong Jiang 1,✉, Yang Gu 1,✉
PMCID: PMC8449670  PMID: 34291572

Summary

Gas‐fermenting Clostridium species can convert one‐carbon gases (CO2/CO) into a variety of chemicals and fuels, showing excellent application prospects in green biological manufacturing. The discovery of crucial genes and proteins with novel functions is important for understanding and further optimization of these autotrophic bacteria. Here, we report that the Clostridium ljungdahlii BirA protein (ClBirA) plays a pleiotropic regulator role, which, together with its biotin protein ligase (BPL) activity, enables an effective control of autotrophic growth of C. ljungdahlii. The structural modulation of ClBirA, combined with the in vivo and in vitro analyses, further reveals the action mechanism of ClBirA’s dual roles as well as their interaction in C. ljungdahlii. Importantly, an atypical, flexible architecture of the binding site was found to be employed by ClBirA in the regulation of a lot of essential pathway genes, thereby expanding BirA’s target genes to a broader range in clostridia. Based on these findings, molecular modification of ClBirA was performed, and an improved cellular performance of C. ljungdahlii was achieved in gas fermentation. This work reveals a previously unknown potent role of BirA in gas‐fermenting clostridia, providing new perspective for understanding and engineering these autotrophic bacteria.

The discovery of regulators as new resources for functional studies and strain modifications are promising strategies in industrially important microorganisms, which, however, still remains poorly explored. Here, we identified the ClBirA protein as a determinant regulator controlling the autotrophic growth of C. ljungdahlii, a representative gas‐fermenting Clostridium species, and performed a detailed functional analysis. We revealed the pleiotropic regulatory function of ClBirA as well as the underlying mechanisms, which would contribute to deeper understanding and optimization of this bifunctional protein in C. ljungdahlii.

graphic file with name MBT2-14-2072-g007.jpg

Introduction

Gas‐fermenting Clostridium species, an important class of autotrophic bacteria, can fix and assimilate C1 gases (CO2/CO) via the Wood–Ljungdahl pathway (WLP) to produce multiple organic acids and alcohols, thereby exhibiting huge potential for the sustainable carbon‐neutral production of chemicals and biofuels (Berg et al., 2010). To effectively use C1 gases, these bacteria have evolved their metabolic pathways and regulatory systems, allowing them to adapt to autotrophic growth conditions. For example, a large number of crucial genes involved in carbon fixation and energy supply in gas‐fermenting Clostridium species showed significant transcriptional changes under different fermentation conditions (C1 gases vs. sugars) (Tan et al., 2013; Aklujkar et al., 2017), indicating a unique metabolic regulation model in these autotrophic bacteria in gas fermentation.

To unlock the full potential of gas‐fermenting Clostridium species, the discovery of more crucial regulatory elements and a comprehensive understanding of underlying regulatory mechanism are necessary. However, to date, this aspect remains minimally explored due to the challenges in genetic tools. Specific to transcriptional regulators (TFs), to our knowledge, only a global regulator CcpA and a TetR‐family regulator (CAETHG_0459) have been identified and characterized in these autotrophic bacteria (Lemgruber et al., 2019; Zhang et al., 2020a,2020b). CcpA was found to directly regulate the expression of some WLP genes in Clostridium. ljungdahlii, and the elimination of CcpA’s inhibition on these genes could significantly enhance CO2 utilization (Zhang et al., 2020a,2020b). CAETHG_0459, as a transcription activator in Clostridium autoethanogenum, can directly bind to the core enzyme of RNA polymerase and subsequently activate the expression of multiple WLP genes (Lemgruber et al., 2019). These studies have indicated the importance of TFs in gas‐fermenting clostridia, thereby strongly supporting a continued investigation.

Recently, based on the Pfam database (http://pfam.xfam.org/), we used the similarity search and domain prediction for mining for TFs in C. ljungdahlii. Over 400 TFs were predicted and subsequently subjected to functional screening through gene deletion or transcriptional repression using the CRISPR‐Cas‐based genome editing tools established in our lab (Huang et al., 2016; Zhao et al., 2019). Among these TFs, BirA has been known to be an important bifunctional protein with both the biotin protein ligase (BPL) activity and biotin operon repressor function in some model bacteria, such as Escherichia coli, Bacillus subtilis and Staphylococcus aureus (Weaver et al., 2001; da Costa et al., 2012; Feng et al., 2014). As the ligase, BirA is responsible for the biotinylation of acetyl‐CoA carboxylase (ACC) and pyruvate carboxylase (PYC), thereby significantly influencing the activity of these two enzymes (Li and Sousa, 2012; Peters‐Wendisch et al., 2012); as the biotin operon repressor, BirA regulates biotin synthesis and uptake and thus enables the in vivo biotin homeostasis (Ye et al., 2016; Zhang et al., 2016). Additionally, a recent study revealed that BirA can interact with other regulators to realize a wide regulatory scope in enterohaemorrhagic E. coli (Yang et al., 2015). However, specific to autotrophic bacteria, the function of BirA remains unexplored.

In this study, we identified a functional BirA (ClBirA) protein that greatly affected the performance of C. ljungdahlii in gas fermentation. Next, we dissected the action mechanisms of ClBirA through a combination of omics (ChIP‐seq), biochemical and genetic approaches. The data from these experiments revealed the pleiotropic regulator role of ClBirA as well as the interaction between its dual functions in C. ljungdahlii. Furthermore, a new atypical flexible architecture of ClBirA‐binding sites that are widespread in C. ljungdahlii was found, revealing a wider regulatory scope of the BirA protein in this bacterium. Based on these findings, we achieved an improved autotrophic growth and product synthesis of C. ljungdahlii in gas fermentation through the functional modulation of ClBirA.

Results

ClBirA is crucial for the autotrophic growth of C. ljungdahlii in gas fermentation

As mentioned above, among the TFs that have been examined by gene deletion or transcriptional repression in C. ljungdahlii, ClBirA was found to be crucial for strain growth in gas fermentation. We compared the growth and product synthesis of the birA‐deleted (Clju ∆birA/p) and wild‐type (Clju WT/p) C. ljungdahlii strains in gas fermentation. Here, because BirA has been known to specifically regulate biotin synthesis in bacteria (Weaver et al., 2001; da Costa et al., 2012), the media with and without biotin addition were both adopted.

The results showed that the Clju ∆birA/p strain could not grow in either medium (Fig. 1A and B), and consequently, almost no product (acetic acid or ethanol) was synthesized by this mutant (Fig. 1C and D). However, after birA was reintroduced into the birA‐deleted mutant strain via an expression plasmid, the growth and product synthesis (acetic acid and ethanol) of the resulting strain (Clju ∆birA/pbirA) was restored to the level of the wild‐type strain (Clju WT/p) (Fig. 1). Such a deficiency of the Clju ∆birA/p strain was also found in the fermentation using fructose as the carbon source (Fig. S1). All these findings suggest that the birA gene is essential for the growth and product formation of C. ljungdahlii in fermenting both C1 gases and sugars.

Fig. 1.

Fig. 1

Open in a new tab

The influence of the deletion of the ClBirA‐encoding gene (birA) on the autotrophic growth and product formation of C. ljungdahlii in gas fermentation.

A. The effect of the birA deletion on the growth of C. ljungdahlii grown in the medium without biotin.

B. The effect of the birA deletion on the growth of C. ljungdahlii grown in the medium supplemented with biotin.

C. The effect of the birA deletion on product formation of C. ljungdahlii grown in the medium without biotin.

D. The effect of the birA deletion on product formation of C. ljungdahlii grown in the medium supplemented with biotin. All the fermentations were performed in the modified ATCC medium 1754. Clju WT/p: the WT strain containing an empty plasmid. Clju ∆birA/p: the birA‐deleted strain containing an empty plasmid. Clju ∆birA/pbirA: the birA‐deleted strain complemented with the plasmid‐carried birA expression. Data shown are means ± standard deviations (n = 3) calculated from triplicate individual experiments. Error bars show standard deviations. Statistical analysis was performed by a two‐tailed Student’s t‐test. ***P < 0.001 vs. the Clju WT/p strain.

In addition, we found that extra biotin supplementation in the medium could promote the growth of both the Clju WT/p and Clju ∆birA/pbirA strains, whereas no promotion effect was observed for the Clju ∆birA/p strain (Fig. 1A and B). Therefore, it seems that the deletion of birA greatly impaired biotin metabolism in C. ljungdahlii that may be crucial for cell growth. Furthermore, in light of the BPL activity of BirA, the deletion of birA will destroy the in vivo biotinylation of ACC and PYC, two crucial enzymes for the basic metabolism of cells. The above results also suggest the possibility that ClBirA is involved in the regulation of crucial metabolic processes other than biotin synthesis in C. ljungdahlii. In other words, ClBirA may play a pleiotropic regulatory role in C. ljungdahlii, and the deletion of birA would result in a broad impact on this anaerobic bacterium.

ClBirA mediates a pleiotropic regulation in C. ljungdahlii

To prove the pleiotropic regulator role of ClBirA in C. ljungdahlii, we used the chromatin immunoprecipitation sequencing (ChIP‐Seq) analysis to screen the potential target genes that are directly regulated by ClBirA. The results showed that 24 118 060 and 26 872 496 raw reads were generated from the ChIP sample and mock ChIP sample (negative control, without the addition of antibodies in ChIP) respectively. After removing low‐quality reads, the remaining 22 815 380 and 25 796 238 clean reads were aligned to the reference genome of C. ljungdahlii, yielding 20 640 883 and 24 853 697 uniquely mapped reads (with < 9.5% and 3.7% mismatches), respectively (Table S1), which reached 668 × and 805 × coverage of the C. ljungdahlii reference genome (Table S1). As expected, 439 potential ClBirA‐binding sites (fold enrichment ≥ 1.5) linked with 439 genes were found (Fig. 2A), including bioY, a well‐known target gene directly controlled by BirA in E. coli and B. subtilis (Weaver et al., 2001; Feng et al., 2014). Among these potential ClBirA’s target genes, many were associated with crucial physiological and metabolic processes in C. ljungdahlii (Fig. 2B). All these data strongly suggest that ClBirA has a wide regulatory scope in C. ljungdahlii.

Fig. 2.

Fig. 2

Open in a new tab

ChIP‐seq analysis of ClBirA‐binding sites in the C. ljungdahlii genome.

A. Genome wide mapping of ClBirA‐binding sites. On the y‐axis is the fold enrichment of the peaks.

B. A pie chart presenting the percentage of ClBirA’s targets with functional categories defined in the C. ljungdahlii database.

C. The in vitro confirmation of ClBirA binding to the coding sequence regions of two essential WLP (Wood–Ljungdahl pathway) genes (fdh and meTr). The coding sequence regions of fdh and meTr were used as the DNA probes for EMSAs. Fdh, formate dehydrogenase; Fhs, formyl‐THF synthetase; FolD, Formyl‐THF cyclohydrolase/methylene‐THF dehydrogenase; MetF, methylene‐THF reductase; MeTr, methyltransferase; Codh/Acs, carbon monoxide dehydrogenase/acetyl‐CoA synthase.

To examine the quality of the ChIP‐seq results, we selected 42 potential ClBirA‐binding sites that are associated with the genes responsible for oxidation reduction/energy metabolism and one‐carbon source metabolism for the in vitro EMSA analysis. Finally, 24 sites (associated with 24 genes) were confirmed to be bound by ClBirA (Fig. S2), including a site in the coding sequence region of an essential WLP gene, meTr (Fig. 2C). These data suggest a high credibility of the ChIP‐seq data. The discovery of the ClBirA‐binding to meTr indicates that ClBirA can regulate WLP genes in C. ljungdahlii. To test this anticipation, we examined the interactions between ClBirA to the other WLP genes (cooS1, cooS2, codH, Clju_c06990, Clju_c08930, Clju_c20040, fhs, folD, metF, acsA, acsC, acsD) via EMSAs. As expected, a new ClBirA‐binding site was found in the coding sequence region of fdh (Clju_c06990) (Fig. 2C), although it was not detected in the ChIP‐seq analysis. Therefore, these results suggest that ClBirA can directly regulate carbon fixation and assimilation in C. ljungdahlii.

Next, we employed the in vivo approach to further verify the pleiotropic regulation of ClBirA in C. ljungdahlii. The structure of ClBirA was first evaluated by a homology modelling using the BirA protein from Staphylococcus aureus subsp. aureus ECT‐R 2 as the template. As shown in Fig. 3A, the modelled ribbon structure of ClBirA included a typical N‐terminal DNA‐binding domain and a C‐terminal with enzymatic activity, thus defining ClBirA as a type‐II BirA (Satiaputra et al., 2016). Based on this model, the DNA sequence coding for the N‐terminal 64 amino acid residues (residue 2 to 65) of ClBirA was deleted in the chromosome, yielding a mutant strain (Clju ∆birA‐N) lacking the regulatory function (Fig. S3). By comparing the transcriptional levels of the aforementioned 24 ClBirA’s target genes (Fig. S2) between the Clju ∆birA‐N mutant and WT strain (Clju WT), we found that most genes were up‐regulated in Clju ∆birA‐N (Fig. 3B and C). These data confirmed the pleiotropic regulator role of ClBirA. Moreover, it seems likely that ClBirA mainly acts as a repressor in C. ljungdahlii.

Fig. 3.

Fig. 3

Open in a new tab

Identificaiton of the pleiotropic regulator role of ClBirA in C. ljungdahlii.

A. Homology modelling of ClBirA. A truncation of the N‐terminal DNA‐binding domain of ClBirA (for blocking the regulatory function of ClBirA) was also shown.

B. Transcriptional changes of the genes responsible for oxidation reduction and energy metabolism after the truncation of the N‐terminal DNA‐binding domain of ClBirA.

C. Transcriptional changes of the WLP genes after the truncation of the N‐terminal DNA‐binding domain of ClBirA. The cultivation of C. ljungdahlii cells for comparative transcriptomic analysis was performed in the modified ATCC medium. The data are presented as the means ± standard deviations calculated from two independent experiments. Statistical analysis was performed by a two‐tailed Student’s t‐test. *P < 0.05, **P < 0.01, and ***P < 0.001 vs. Clju WT.

Discovery of an atypical architecture of ClBirA‐binding sites that is wide occurrence in C. ljungdahlii

Interestingly, a further examination of the above 24 binding sites of ClBirA (Fig. S2) revealed that only 13 contained a typical BirA‐binding motif (Rodionov et al., 2002) (Fig. 4A). This finding indicates the existence of atypical motifs for the remaining 11 sites. To verify this hypothesis, two ClBirA‐binding sequences (within the promoter regions of Clju_c04680 and Clju_c23900, respectively) were selected from the above 11 candidates for DNase I footprinting analysis. As expected, an obvious protection region was observed in either of the two DNA probes in the presence of BirA protein (Fig. 4B). A common feature of these two protection regions was the two 5‐nt repeats (quasi‐palindromic sequence) separated by a variable intervening spacer region, namely ‘TTTAC‐N16‐GTTAA’ and ‘TTGTC‐N19‐GGTAA’ (Fig. 4C). The architecture of these two sites, to our knowledge, is different from all the known BirA‐binding motifs, which are normally 14–16 bp long and include several highly conserved nucleotides (Rodionov et al., 2002).

Fig. 4.

Fig. 4

Open in a new tab

Identification of the novel BirA‐binding site consensus in C. ljungdahlii.

A. The genes associated with the typical conserved motif of BirA‐binding sites.

B. DNase I footprinting analysis of the ClBirA binding to Clju_c04680 and Clju_c23900.

C. Mutational analysis of the BirA‐binding sites.

D. EMSAs of the ClBirA binding with various mutated DNA fragments of Clju_c04680 and Clju_c23900.

E. The genes associated with the novel BirA‐binding sites identified in this study. The sequences highlighted with red indicate the putative binding sites. Reactions were performed with 0.04 pmol of Cy5‐labelled probes in the presence of different concentrations of BirA (40, 80 and 120 nM). Free probe is shown by an arrow.

To further confirm the ClBirA binding to these two sites, we separately mutated their palindromic sequences, intervening spacers or flanking regions (Fig. 4D), and the yielding DNA fragments were used as probes for EMSAs. The results showed that the affinity of the probes for ClBirA was almost completely abolished with the mutation of the two palindromic sequences (04680 mut V and 23900 mut V), while the mutation of the other regions remained the affinity to different extents (Fig. 4D). These data verified that the ‘TTTAC‐N16‐GTTAA’ and ‘TTGTC‐N19‐GGTAA’ sequences are the ClBirA‐binding sites, in which the palindromic sequences are crucial for the binding activity. Next, we searched for more ClBirA‐binding sites within the remaining nine candidates (Clju_c02070, Clju_c05330, Clju_c15080, Clju_c16810, Clju_c24130, Clju_c26350, Clju_c29710, Clju_c29780 and Clju_c40120) (Fig. S2) via gene sequence alignment (SnapGene 2.3.2) using a template ‘TTNWC‐N16/19‐GNTAA’ (N represents any bases; W represents A or T; the length of the intervening spacer was 16 or 19 nt). As expected, one and three potential ClBirA‐binding sites were found in the coding region and upstream non‐coding region of their corresponding genes, respectively (Fig. 4E).

To investigate the universality of this atypical motif, we further searched the C. ljungdahlii genome via gene sequence alignment (SnapGene 2.3.2) using the template ‘TTNAC‐N16/19‐GNTAA’. The result revealed total 519 potential ClBirA‐binding sites. This finding suggests that this new binding site architecture with a variable intervening spacer may play a broad role in ClBirA‐mediated metabolic regulation.

Phenotypic changes of C. ljungdahlii caused by functional modulation of ClBirA

Since ClBirA belongs to the type‐II BirA, which has both regulatory function and BPL activity, a derived question is how the two functions of ClBirA influence the performance of C. ljungdahlii in gas fermentation. To answer this question, we constructed two ClBirA variants, in which the N‐terminal DNA‐binding domain and the enzymatically active domain were deleted and inactivated, respectively (Fig. 3A), for phenotypic analysis.

We first purified a ClBirA variant, ClBirAE, lacking the N‐terminal DNA‐binding domain (from the 2nd to 65th amino acid residue). ClBirAE was then tested for its BPL activity. The results showed that the biotinylation degree of the substrate BCCP (the biotin carboxyl carrier protein that is the subunit accB of ACC) by ClBirA and ClBirAE was 96.7% and 97.6%, respectively (Fig. 5A), indicating that the deletion of the N‐terminal portion did not impair the BPL activity of ClBirA. We further constructed the other ClBirA variants without BPL activity by mutating essential amino acid residues in its enzymatically active domain. Based on a homology modelling using the Pyrococcus horikoshii OT3 BirA as the template (Bagautdinov et al., 2008), nine potentially important amino acid residues (Pro176, Asn177, Asp178, Leu190, Glu192, Ile316, Ser317, Gly318, Glu319) of ClBirA were picked out for individual or combined mutations (Fig. 5B), and the yielding mutants were used for BPL activity assays. As shown in Fig. 5C, the ClBirAN177A, ClBirAD178A and ClBirAM (triple mutations, P176A, N177A and D178A) mutants exhibited significantly decreased BPL activity compared to the WT ClBirA, whereas the other two mutants, ClBirAN (double mutations, L190A, E192A) and ClBirAP (quadruple mutations, I316A, S317A, G318A, E319A), had no changes in BPL activity. Therefore, the ClBirAN177A, ClBirAD178A and ClBirAM mutants were adopted for DNA‐binding activity assays (EMSAs) using the promoter region of the bioY gene (an identified direct target of ClBirA, shown in Fig. S2) as the probe. The results showed that only ClBirAN177A still remained the binding activity to bioY (Fig. 5D), indicating that the N177A mutation did not impair the regulatory function of ClBirAN177A. Here, the ClBirAN177A protein was named ClBirAR.

Fig. 5.

Fig. 5

Open in a new tab

Identification of the crucial amino acid residues for the BPL (biotin protein ligase) activity of ClBirA.

A. Detection of the biotinylation of GFP‐BCCP by SDS‐PAGE and Stv‐induced band‐shift analysis.

B. The predicted crucial amino acid residues for the BPL activity of ClBirA.

C. Detection of the GFP‐BCCP biotinylation by ClBirA mutants using SDS–PAGE.

D. The comparison of the binding activities of ClBirAN177A, ClBirAD178A and ClBirAM to the DNA probe of bioY. Biotinylation reactions were performed and processed as described in method. Control reactions were performed in the absence of the ClBirA protein. Fluorescent proteins are analysed by direct UV‐irradiation of the gel and integration of GFP fluorescence.

Next, we began to examine the phenotypic changes of C. ljungdahlii caused by ClBirAE and ClBirAR. These two unifunctional ClBirA variants were transferred into a birA‐deleted C. ljungdahlii chassis (Clju ∆birA), solely or together, via an expression plasmid (Fig. 6A). The resulting three mutant strains, Clju ∆birA‐ClBirAE, Clju ∆birA‐ClBirAR and Clju ∆birA‐ClBirAR+E, were compared to a control strain (Clju ∆birA‐ClBirAWT, the Clju ∆birA strain containing a plasmid‐expressed WT ClBirA) to determine their differences in autotrophic growth in gas fermentation. As shown in Fig. 6B, Clju ∆birA‐ClBirAE and Clju ∆birA‐ClBirR+E exhibited increased and decreased growth rate, respectively, compared to Clju ∆birA‐ClBirAWT, while Clju ∆birA‐ClBirAR lost the growth capability. Obviously, the differences between the growth rate and biomass of Clju ∆birA‐ClBirAR+E and Clju ∆birA‐ClBirAE indicate a negative effect of the pleiotropic regulatory function of ClBirA on C. ljungdahlii.

Fig. 6.

Fig. 6

Open in a new tab

Dissection of the interaction between ClBirA’s dual functions.

A. Construction of the C. ljungdahlii mutants, Clju ∆birA‐ClBirAWT, Clju ∆birA‐ClBirAE, Clju ∆birA‐ClBirAR and Clju ∆birA‐ClBirAR+E.

B. Comparison of the growth and product formation of the Clju ∆birA‐ClBirAWT, Clju ∆birA‐ClBirAE, Clju ∆birA‐ClBirAR and Clju ∆birA‐ClBirAR+E strains in gas fermentation.

C. Real‐time qRT–PCR assay for detecting the transcriptional differences of the fdh, Clju_c26350, and bioY genes between Clju ∆birA‐ClBirAWT and Clju ∆birA‐ClBirAE.

D. The in vivo PYC and ACC activities of Clju ∆birA‐ClBirAWT and Clju ∆birA‐ClBirAE strains. All the fermentations were performed in the modified ATCC medium 1754. Clju ∆birA‐ClBirAWT: the birA‐deleted strain with a plasmid expressing the WT ClBirA; Clju ∆birA‐ClBirAE: the birA‐deleted strain with a plasmid expressing ClBirAE; Clju ∆birA‐ClBirAR: the birA‐deleted strain with a plasmid expressing ClBirAR; Clju ∆birA‐ClBirAR+E: the birA‐deleted strain with a plasmid expressing both ClBirAE and ClBirAR. The data are presented as the means ± standard deviations calculated from two independent experiments. Statistical analysis was performed by a two‐tailed Student’s t‐test. *P < 0.05, **P < 0.01, and ***P < 0.001 vs. Clju ∆birA‐ClBirAWT.

To better understand the improved autotrophic growth of Clju ∆birA‐ClBirAE over Clju ∆birA‐ClBirAWT (Fig. 6B), the regulatory property and BPL activity of ClBirA in these two strains were compared. For the regulatory property analysis, the transcriptional levels of the aforementioned three ClBirA’s target genes, fdh, Clju_c26350 and bioY (Fig. S2 and Fig. 2C), in Clju ∆birA‐ClBirAE and Clju ∆birA‐ClBirAWT, were examined. The results showed that these genes, compared to those in Clju ∆birA‐BirAWT, were significantly up‐regulated at most time points in Clju ∆birA‐BirAE (Fig. 6C). Next, the activities of PYC and ACC, the two known biotinylated enzymes by BirA, in Clju ∆birA‐ClBirAE and Clju ∆birA‐ClBirAWT, were also compared. It was found that, compared to Clju ∆birA‐ClBirAWT, Clju ∆birA‐ClBirAE had much higher PYC activity at 48 h and higher ACC activity at 48 h and 72 h, while no obvious difference was observed at the other time points (Fig. 6D). Therefore, it seems likely that the deletion of the DNA‐binding domain of ClBirA (ClBirAE) destroys its transcriptional repression on target genes, and consequently, ClBirAE can fully function as a BPL to activate PYC and ACC by biotinylation, leading to the improved cell growth of C. ljungdahlii in gas fermentation (Fig. 6B).

Phylogenetic analysis on two types of BirAs in bacteria and archaea

As aforementioned, ClBirA belongs to the type II BirA because it contains both the DNA‐binding and enzymatically active domains (Satiaputra et al., 2016); but its variant, ClBirAE, has lost the regulatory function and, thus, is more similar to the type I BirA. The different performance of C. ljungdahlii strains in gas fermentation caused by ClBirAE and ClBirAR (Fig. 6B) indicates the different roles of the type I and II BirAs in bacteria.

Here, we carried out a maximum likelihood phylogenetic analysis of 70 BirA proteins (28 type I and 42 type II) from 67 bacterial and archaeal strains that have complete genome data in KEGG database. Because type I BirA proteins lack the DNA‐binding domain, the two conservative domains (BPL_ligase domain, PF03099; BPL_ligase C‐terminal domain, PF02237), rather than the whole protein, of these BirAs were used for alignment. The BPL_ligase domains and BPL_ligase C‐terminal domains from five BirAs that have determined structures in the PDB database (PDB number: 3FJP, 2EJ9, 4OP0, 1WNL, 2CGH) were picked out for the amino acid sequence alignment, and the result was used as an initial seed for a further alignment of the other BirA proteins. The results showed that most bacteria have only one type of BirA, in which the type II BirA often occurred in the class Bacilli, Clostridia and γ‐Proteobacteria, while the type I BirA was normally found in the class α‐Proteobacteria, Antibacteria, Thermotogae, Aquificae and Fusobacteria (Fig. 7). Moreover, the BirAs from same species are often clustered in the same clades and subclades (Fig. 7). Interestingly, the type I and type II BirAs were found to coexist in some bacterial hosts, including Clostridium acetobutylicum ATCC 824 (CA_C0589 and CA_C0212), Lactococcus lactis Il1403 (L0191 and L0192) and Francisella novicida U112 (FTN_0811 and FTN_0568).

Fig. 7.

Fig. 7

Open in a new tab

Maximum likelihood phylogenetic tree of BirA homologues in bacteria and archaea. The representative strains of Gram‐positive/negative bacteria and archaea are included. Type I and II BirAs are marked in blue and red respectively.

Function modulation of ClBirA improved autotrophic growth of C. ljungdahlii

The above data not only defined the pleiotropic functions of ClBirA but also provided clues for further strain improvement. It seems likely that a modified ClBirA with only enzymatically active domain is better for C. ljungdahlii, based on the advantage seen in the growth of Clju ∆birA‐ClBirAE over Clju ∆birA‐ClBirAWT in gas fermentation (Fig. 6B).

Thus, we overexpressed the ClBirAE‐coding gene (clbirA E) in the aforementioned Clju ∆birA‐N strain (chromosomal deletion of the N‐terminal DNA‐binding domain of ClBirA) (Fig. 8A); simultaneously, the acc (Clju_c42100‐42140) and pyc (Clju_c37390) genes, coding for the C. ljungdahlii ACC and PYC enzymes, respectively, were also overexpressed separately (Fig. 8A). Such an engineering strategy was expected to specifically block the regulatory function, and simultaneously, reinforce the BPL activity of ClBirA in C. ljungdahlii. The yielding Clju ∆birA‐N‐p (birA E‐acc) and Clju ∆birA‐N‐p (birA E‐pyc) strains were compared to the control strain (C. ljungdahlii Clju WT/p containing an empty plasmid) to determine their difference in growth and product synthesis in gas fermentation. As expected, the growth rate, biomass and final titer of acetic acid (a major product) of Clju ∆birA‐N‐p (birA E‐acc) were much higher than those of Clju WT/p (Fig. 8B and C); in contrast, no obvious phenotypic change was found for Clju ∆birA‐N‐p (birA E‐pyc) (Fig. 8B and C). Besides, the overexpression of the acc or pyc genes without clbirA E in the Clju ∆birA‐N chassis did not cause phenotypic changes (Fig. S4). Collectively, these findings strongly suggest that the reinforcement of the biotinylation of ACC contributes to the performance of C. ljungdahlii in gas fermentation.

Fig. 8.

Fig. 8

Open in a new tab

Improved performance of C. ljungdahlii by the functional modulation of ClBirA.

A. Construction of the Clju ∆birA‐N‐p (birA E ‐acc) and Clju ∆birA‐N‐p (birA E ‐pyc) strains.

B. Comparison of the autotrophic growth of the Clju ∆birA‐N‐p (birA E ‐acc), Clju ∆birA‐N‐p (birA E ‐pyc) and Clju WT/p (control) strains in gas fermentation.

C. Comparison of the production of ethanol and acetic acid of the Clju ∆birA‐N‐p (birA E ‐acc), Clju ∆birA‐N‐p (birA E ‐pyc) and Clju WT/p strains in gas fermentation. All the fermentations were performed in the YT (YTF medium without fructose) medium with the supplement of a mixture of CO–CO2–H2–N2. Clju ∆birA‐N‐p (birA E ‐acc): the Clju ∆birA‐N strain containing the pMTL83151‐birA E ‐acc plasmid. Clju ∆birA‐N‐p (birA E ‐pyc): the Clju ∆birA‐N strain containing the pMTL83151‐birA E ‐pyc plasmid. Data shown are means ± standard deviations (n = 3) calculated from triplicate individual experiments. Error bars show standard deviations. Statistical analysis was performed by a two‐tailed Student’s t‐test. **P < 0.01 and ***P < 0.001 vs. Clju WT/p.

Discussion

Autotrophic bacteria have evolved unique metabolic regulatory approaches for the fixation and conversion of C1 gases. The BirA protein, despite its physiological importance for heterotrophic bacteria, remains unexplored in autotrophic bacteria. This study has elucidated the regulatory scope and action mechanism of ClBirA in the autotrophic gas‐fermenting C. ljungdahlii, providing new insights into the function of this protein in bacteria.

It has been known that biotin synthesis is tightly controlled by BirA in response to extra biotin supply and demand in bacteria (Feng et al., 2014). When the intracellular biotin level exceeds the physiological requirement or biotin‐requiring enzyme supply declines, BirA, together with biotin, will bind directly to the biotin operator and then repress biotin synthesis (Feng et al., 2014). The biotin synthesis normally requires a large amount of energy in bacteria (Satiaputra et al., 2016). In view of that biotin synthesis is a high energy‐consuming process in cells (Feng et al., 2013), such a BirA‐mediated feedback regulation on biotin synthesis may be crucial for the control of energy expenditure in bacteria. Theoretically, this regulatory mechanism is especially important for gas‐fermenting clostridia because the carbon fixation and assimilation in these autotrophic bacteria are a net energy consumption process (Zhang et al., 2020a,2020b), and excessive energy consumption in biotin synthesis will cause metabolic burden. But interestingly, a modified ClBirA with only enzymatically active domain led to a better performance of C. ljungdahlii in gas fermentation (Figs 6 and 8), indicating that an uncontrolled overproduction of biotin is not the key point for the autotrophic growth of C. ljungdahlii. The molecular mechanism of this phenomenon remains to be explored. Taken together, these findings suggest that C. ljungdahlii, as a clostridial chassis for gas fermentation, needs to be further optimized in terms of the in vivo biotin supply.

Additionally, given the fact that ClBirA acts as a pleiotropic regulator in C. ljungdahlii (Figs 2 and 3), the above ClBirA‐mediated feedback regulation in response to the in vivo biotin level is supposed to achieve a large‐scale dynamic metabolic regulation. Of note, ClBirA seems to mainly play a repressor role because the specific deletion of the DNA‐binding domain of ClBirA caused transcriptional up‐regulation of many target genes, whereas only very few were down‐regulated (Fig. 3B and C). This finding offered the possibility for the functional improvement of ClBirA. For example, because ClBirA has been found to directly control the expression of some WLP genes, the functional optimization of this protein may contribute to the metabolism of C1 carbon sources (e.g. CO2, CO and formate) via WLP in C. ljungdahlii. Such an engineering strategy, to our knowledge, has never been reported for strain improvement.

Another interesting finding in this study is that ClBirA employs a ‘flexible’ binding motif (TTNAC…N16/19…GNTAA) in its pleiotropic regulation. To date, most of the reported binding motifs of TFs are reverse palindromes or direct repeats, which are normally between 16 and 18 bp (Wang et al., 2013; Antunes et al., 2016; Bouillaut et al., 2019). Only very TFs have been found capable of recognizing such a ‘flexible’ binding structure. For example, the E. coli cyclic AMP (cAMP) receptor protein involved in regulating carbon metabolism recognizes the binding sites containing a 6‐ or 8‐bp spacer (Barber et al., 1993); the Clostridium acetobutylicum CcpA exerts its global regulation via binding the cre var sites that consists of two inverted repeats separated by a highly variable intervening spacer (Yang et al., 2017a,2017b). Such a variable binding structure is considered to create more flexibility to TFs’ regulation (Yang et al., 2017a,2017b). Therefore, it can be speculated that the binding motif ‘TTNAC…N16/19…GNTAA’ will give the flexibility in both the regulatory scope and strength to ClBirA, thereby enabling diverse regulatory outputs.

In previous studies, researchers have proposed that BirA‐like bifunctional proteins that possess both enzymatic and regulatory activities belong to an intermediate state between specific enzymes and transcription factors in evolution (Commichau and Stulke, 2008). For example, the PutA protein in α‐Proteobacteria (e.g. rhizobia) that appeared relatively early in microbial evolution has only enzymatic activity; but this protein in γ‐Proteobacteria (e.g. Escherichia coli and Salmonella) that appeared later has contained a ribbon–helix–helix (RHH) DNA‐binding domain and thus can play a TF role (Krishnan and Becker, 2005; Liu et al., 2017). Such an evolutionary process may also happen to BirA, namely that type I BirAs appeared relatively early and then became type II BirAs by capturing a DNA‐binding domain during their evolutionary process. As mentioned above, some microbial strains were found to contain two BirA proteins. The study on the two BirAs in Francisella novicida U112 has revealed that one mainly acts as a BPL, while the other plays the regulatory function with very low BPL activity (Feng et al., 2015). Therefore, it seems that functional differentiation has occurred in the BirA proteins in these strains. Additionally, the group II BirA can be found in multiple autotrophic bacteria, such as C. ljungdahlii, Clostridium carboxidivorans and Acetobacterium woodii (Fig. 7), indicating a broad regulatory role of BirA in autotrophs. Considering the pleiotropic regulation of BirA, it is highly possible that BirA controls some crucial metabolic pathways in autotrophic bacteria, thereby playing an important regulatory role in these bacteria.

In summary, our results elucidated the pleiotropic functions of the ClBirA protein in gas‐fermenting C. ljungdahlii, which has renewed our understanding on the role of BirA in bacteria. More importantly, the functional modulation of ClBirA not only revealed the interaction between its dual functions but also achieved the genetic improvement of C. ljungdahlii, thereby showcasing the potential application value of this bifunctional protein in autotrophic bacteria.

Experimental procedures

Bacterial strains, plasmids, media and growth conditions

All strains and plasmids used in this work are listed in Table S2. E. coli Top10 was used for all cloning experiments, and BL21 (DE3) was used for protein expression. E. coli cells were grown in LB (lysogeny broth) medium (Bertani, 2004) supplemented with 12.5 μg ml‐1 chloramphenicol or 50 μg ml‐1 kanamycin antibiotics when needed. C. ljungdahlii DSM 13528 was grown anaerobically at 37°C in the YTF medium (Humphreys et al., 2015) or a modified ATCC medium 1754 (Huang et al., 2016), in which 5 μg ml‐1 of thiamphenicol was added for plasmid stability when needed.

Plasmid construction

The primers used in this work are listed in Table S3.

The pMTLcas‐birA plasmid was constructed as follows: the pMTLcas‐pta plasmid was digested with SalI and XhoI, yielding a linear pMTLcas vector. The sgRNA fragment targeting the birA gene was obtained from the pMTLcas‐pta plasmid by PCR amplification using primers birA‐gRNA‐for/gRNA‐rev. The two homologous arms (HAs) that flank the coding region of birA were obtained by PCR amplification using the C. ljungdahlii genomic DNA as the template and primers birA‐up‐for/birA‐up‐rev and birA‐dn‐for/birA‐dn‐rev and then linked to the sgRNA fragment via overlapping PCR using primers birA‐gRNA‐for/birA‐dn‐rev, yielding a DNA fragment, sgRNA‐HA1. Finally, the linear pMTLcas vector and the sgRNA‐HA1 fragment were assembled in one step using the ClonExpress MultiS One Step Cloning Kit (Vazyme Biotech, Nanjing, China), yielding the pMTLcas‐birA plasmid for the deletion of birA in C. ljungdahlii.

The pMTLcas‐birA‐N plasmid was constructed as follows: The sgRNA fragment targeting the DNA‐binding region (4–195 nt) of birA was obtained by PCR amplification using the pMTLcas‐pta plasmid as the template and primers birA‐gRNA‐for/gRNA‐rev. The two HAs that flank the DNA‐binding region (4–195 nt) of birA were obtained by PCR amplification using the C. ljungdahlii genomic DNA as the template and primers birA‐N‐up‐for/birA‐N‐up‐rev and birA‐N‐dn‐for/birA‐N‐dn‐rev and then linked to the sgRNA fragment via overlapping PCR, yielding the sgRNA‐HA2 fragment. Next, the aforementioned linear pMTLcas vector and the sgRNA‐HA2 fragment were assembled in one step using the ClonExpress MultiS One Step Cloning Kit.

The pMTL83151‐P1440‐birA‐3×FLAG plasmid for the ChIP‐seq experiment was constructed as follows: the DNA fragments of the promoter P1440 and birA were obtained from the C. ljungdahlii genome by PCR amplification using primers 83151‐P1440‐o‐for/P1440‐birA‐o‐rev and birA‐for/birA‐FLAG‐o‐rev respectively. The 3×FLAG fragment was obtained by PCR amplification using primers FLAG‐for/FLAG‐83151‐o‐rev. The P1440‐birA‐3×FLAG fragment was obtained by overlapping PCR using the P1440, birA, 3×FLAG fragments as the templates and primers 83151‐P1440‐o‐for/FLAG‐83151‐o‐rev. Finally, the linear pMTL83151 plasmid and the P1440‐birA‐3×FLAG fragment were assembled in one step using the ClonExpress MultiS One Step Cloning Kit, yielding the pMTL83151‐P1440‐birA‐3×FLAG plasmid.

The construction of the pMTL83151‐birA plasmid was performed as follows: the pMTL83151 plasmid was digested with NdeI and HindIII. The PbirA‐birA fragment was obtained by PCR amplification using primers 83151‐PbirA‐o‐for/birA‐83151‐o‐rev and the C. ljungdahlii genomic DNA as the template. Next, the P‐birA fragment and the linear pMTL83151 were assembled in one step using the ClonExpress MultiS One Step Cloning Kit, yielding the pMTL83151‐birA plasmid. The pMTL83151‐birA N177A and pMTL83151‐birA E plasmids were constructed with the same step except primers used for PCR amplification (83151‐PbirA‐o‐for/177‐1‐rev and 177‐2‐for/birA‐83151‐o‐rev for pMTL83151‐birA N177A; PbirA‐for/PbirA‐rev and birA‐N‐for/birA‐N‐rev for pMTL83151‐birA E).

The pMTL83151‐birA N177A‐birA E plasmid was constructed as follows: the pMTL‐8315‐birA N177A plasmid was digested with NheI and HindIII. The fragments of PbirA and birA E were obtained from the C. ljungdahlii genomic DNA by PCR amplification using primers PbirA‐for/PbirA‐rev and birA‐N‐for/birA‐N‐rev, respectively, and then linked via overlapping PCR using primers PbirA‐for/ birA‐N‐rev, yielding the PbirA‐birA E fragment. The PbirA‐birA E fragment and the linear pMTL83151‐birA N177A were assembled in one step using the ClonExpress MultiS One Step Cloning Kit.

The pMTL83151‐birA E ‐acc plasmid was constructed as follows: the pMTL83151‐birA E plasmid was digested with NheI and HindIII. The DNA fragments of P1440 and acc were obtained by PCR amplifications using the C. ljungdahlii genomic DNA as the template and primers birA E‐P1440‐o‐fw/P1440‐rev and acc‐for/acc‐83151‐o‐rev respectively. Finally, the linear pMTL83151‐birA E was assembled with the DNA fragments of P1440 and acc in one step using the ClonExpress MultiS One Step Cloning Kit. The pMTL83151‐birA E ‐pyc plasmid was constructed with the same steps except that primers pyc‐for/pyc‐83151‐o‐rev were used.

The pUCm‐T‐04680 plasmid was constructed as follows: the DNA fragment of the Clju_c04680 gene was obtained by PCR amplification using the C. ljungdahlii genomic DNA as the template and primers E‐04680‐fw/ E‐04680‐rev and then linked to pUCm‐T using TaqDNA Polymerase, yielding the pUCm‐T‐04680 plasmid. The pUCm‐T‐23900 plasmid was constructed with the same steps, except that primers E‐23900‐fw/ E‐23900‐rev were used for PCR amplification.

Gene expression and deletion in C. ljungdahlii

The plasmids were delivered into C. ljungdahlii by electroporation. The preparation of electrocompetent cells and electroporation procedure were same as previously described (Huang et al., 2016). Gene deletion in C. ljungdahlii was performed using the CRISPR‐Cas9‐based genome editing method, and the procedure was the same as previously described (Huang et al., 2016).

Fermentation

Inoculum preparation of thewild type C. ljungdahlii strain and its derivatives were performed anaerobically in the YTF medium. Fermentations were performed anaerobically in the modified ATCC medium 1754 or YT (YTF medium without fructose), in which 5 μg ml‐1 of thiamphenicol was added when needed. Briefly, 300 μl frozen stock was inoculated into 4 ml liquid YTF medium and incubated anaerobically at 37°C for 24 h. When the optical density (OD600) of grown cells reached 1.0–1.2, 1.5 ml of the grown cells was transferred into 30 ml of the aforementioned modified ATCC medium 1754 or YT medium for gas fermentation. The fermentation was carried out in 125 ml Wheaton serum bottles (Sigma‐Aldrich, Saint Louis, MO, USA) with a headspace of CO‐CO2‐H2‐N2 (56%/20%/9%/15%; pressurized to 0.2 MPa). The gases were added into the headspace (again at 0.2 MPa) for every 24 h.

Protein purification

The genes coding for ClBirA (Clju_c26660) and the derived ClBirA mutants, including ClBirAE, ClBirAN177A, ClBirAD178A, ClBirAM, ClBirAN, ClBirAP and BCCP (Clju_c42140), were obtained by PCR amplification using corresponding primers (Table S3) and the C. ljungdahlii genomic DNA as the template. The DNA fragment of gfp was obtained by PCR amplification using primers gfp‐o‐pET‐28a‐fw/gfp‐o‐bccp‐rev and the pIMP1‐thl9‐gfp plasmid as the template (Yang et al., 2017a,2017b). These genes were inserted into the pET‐28a vector (Invitrogen, Carlsbad, CA, USA), and the yielding plasmids were transferred into E. coli BL21 (DE3) for expression. The subsequent protein purification procedure was same as previously described (Ren et al., 2012).

ChIP‐Seq library construction and sequencing

The pMTL83151‐P1440‐birA‐3×FLAG plasmid carrying the 3×FLAG‐tagged birA was transformed into the Clju ∆birA strain. Cells were cross‐linked with 1% formaldehyde for 10 min at room temperature. The reaction was quenched by adding glycine to a final concentration of 0.125 M. Formaldehyde cross‐linked cells were washed with pre‐chilled phosphate buffer saline and then sonicated to generate 200–500 bp DNA fragments. Immunoprecipitation was performed using monoclonal anti‐3×FLAG antibody (Cat#F1804; Sigma) and protein A magnetic beads (Cat#10002D; Invitrogen) according to the manufacturer’s instructions. Sonicated extracts, without the addition of antibodies, was performed as a negative control. Proteins were then removed by incubation with proteinase K for 2 h at 55°C. The libraries were prepared using the VAHTS Universal DNA Library Prep Kit for Illumina V3 (Catalog NO. ND607; Vazyme) and sequenced using the Novaseq 6000 system (Illumina, San Diego, CA, USA).

Electrophoretic mobility shift assay (EMSA)

The DNA probes for EMSAs were prepared by two‐step PCR amplification. First, the unlabelled DNA fragments were obtained by PCR amplification using the C. ljungdahlii genomic DNA as the template and primers E‐fw/E‐rev that contained a universal sequence (5'‐AGCCAGTGGCGATAAG‐3') at the 5' terminus. Next, the labelled DNA fragments were generated by PCR amplification using a 5'‐Cy5‐labelled universal primer (5'‐AGCCAGTGGCGATAAG‐3') and then recovered by agarose gel electrophoresis using the PCR purification Kit (Cat#AP‐GX‐250; Axygen, Hangzhou, China). The obtained DNA fragments were used as probes for EMSAs, which were carried out as previously described (Ren et al., 2012).

RT‐qPCR

The WT C. ljungdahlii strain and the derived mutants were grown anaerobically at 37°C in the modified ATCC medium 1754 using syngas (CO‐CO2‐H2‐N2). The grown cells were harvested by centrifugation at 8600 g for 10 min. Total RNA was extracted using the kit (Cat#cw0581; CWBIO, Beijing, China) according to the manufacturer’s instructions. Contaminating DNA in the extracted RNA was eliminated by DNase I (TaKaRa, Kyoto, Japan) digestion. The RNA concentration was determined by a NanoDrop spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA). cDNA was obtained by reverse transcription using the PrimeScript RT reagent kit (TaKaRa). The subsequent procedure of qPCR was the same as described previously (Zhang et al., 2018). The rho gene (Clju_c02220) was used as the internal control according to the previous report (Zhao et al., 2019).

Biotinylation activity assay

The biotinylation activity of ClBirA was measured as previously described (Sorenson et al., 2015), using BCCP (an biotin carboxyl carrier protein that is the acetyl‐coenzyme A carboxylase subunit accB) as the substrate. Biotinylation of GFP‐BCCP was performed in 50 µl reactions containing 0.3 µM BirA protein, 0.5 mM biotin, 2.5 mM ATP, 1 mM MgCl2, 30 µM GFP‐BCCP and 25 mM Tris (pH 8.0). The reaction mixtures were incubated for 1 h at room temperature. After incubation, free biotin was removed from the reactions by Ni‐affinity purification using a Pronity IMAC nickel resin column (Cat#732‐6008; Bio‐Rad, Hercules, CA, USA). The resin was then washed twice with 500 µl of 25 mM Tris (pH 8.0) supplemented with 5% glycerol and 150 mM NaCl by centrifugation at 1000 g for 1 min. Fusion proteins were eluted from the column with 50 µl of the same buffer supplemented with 200 mM imidazole and centrifugation at 1000 g for 1 min. Next, 7.5 µl of protein elution and 0.75 µl Stv (streptavidin, 5 mg ml‐1, Invitrogen) were incubated at room temperature for 15 min. The reactions were subjected to SDS–PAGE without a heat‐denaturation step to visualize the protein complexes. Gels were first photographed under a UV transilluminator before Coomassie staining. The fluorescence of GFP–BCCP was analyzed by integration of protein bands (ImageJ; NIH, Bethesda, MD, USA) to determine the extent of biotinylation. Control reactions (without ClBirA) were performed in parallel.

DNase I footprinting assays

To prepare the probes labelled with fluorescent FAM (6‐carboxyfluorescein), the noncoding regions upstream of Clju_c04680 and Clju_c23900 were amplified by PCR from the pUCm‐T‐04680 and pUCm‐T‐23900 plasmids, respectively, using primers M13F/M13R(FAM). The FAM‐labelled probes were purified and quantified with NanoDrop 2000C (Thermo).

The DNase I footprinting assay was performed as follows. First, 250 ng probes were incubated with indicated amounts of purified ClBirA proteins in 40 µl of binding buffer containing 20 mM Tris–HCl (pH 7.9), 1 mM DTT, 10 mM MgCl2, 0.5 mg ml−1 calf BSA and 5% (vol./vol.) glycerol. After incubation at 25°C for 30 min, 10 μl of RNase‐free DNase I buffer, 0.015 unit of DNase I (Promega, Madison, WI, USA) and 100 nmol of fresh CaCl2 were added to digest the DNA probe at 37°C for 1 min. The reaction was stopped by adding 140 µl stop solution (200 mM sodium acetate, 30 mM EDTA and 0.15% SDS). The DNA was recovered by phenol/chloroform extraction and ethanol precipitation and then dissolved in 30 µl MiniQ water. Finally, 1 µl of sample was added to 8.5 µl of HiDi formamide and 0.5 µl of GeneScan‐LIZ600 size standards (Applied Biosystems, Foster City, CA, USA) was sequenced on a 3130xl DNA analyzer (Applied Biosystems). The preparation of the DNA ladder and data analysis was same as previously described (Wang et al., 2012). Genescan results were evaluated with Peak Scanner software v1.0 (Applied Biosystems).

Pyruvate carboxylase and acetyl‐CoA carboxylase activity assays

The C. ljungdahlii strains were grown in the modified ATCC medium 1754 with the supplement of a mixture of CO–CO2–H2–N2. When the the OD600 of grown cells reached 1.0, the cell pellets were harvested by centrifugation (7000 g, 4°C, 10 min), dissolved in the reagents from the following assay kits (as shown below) and vortexed for 1 min for cell lysis. Finally, the cell lysate was centrifuged at 12 000 g for 30 min, and the supernatant was used for PYC and ACC activity assays. The preparation of cell lysis was performed under aerobic conditions.

The PC assay kit (Cat#BC0730; Solarbio, Beijing, China) was used to measure the activity of PYC. The PYC activity was measured following the decrease in absorbance at 340 nm induced by oxidized oxaloacetate (the product from the carboxylation of pyruvate catalyzed by PYC) in the presence of NADH.

The ACC assay kit (Cat#BC0410; Solarbio) was used to measure the activity of ACC. The determination of the ACC was based on the reaction of acetyl‐CoA being catalyzed to malonyl‐CoA and inorganic phosphorus. The inorganic phosphorus was further determined by a colour reaction with molybdenum blue, and the absorbance was read at 660 nm. Standard curve was generated by using known amounts of phosphate in combination with the molybdenum blue reagent as described in the manufacture’s instructions.

Phylogenetic tree construction

Gene annotations were derived from SEED (Overbeek et al., 2005) and KEGG (Kanehisa et al., 2012) database. Comparative genomic analysis was performed to predict genes with unknown functions by using GenomeExplorer (Mironov et al., 2000). Multiple protein alignments were done using MUSCLE (Edgar, 2004) and MUFFT (Katoh and Standley, 2013). The phylogenetic tree was constructed via the maximum likelihood method implemented in PhyML (Guindon et al., 2010) and MEGA (Tamura et al., 2011) and further edited with iTOL (Letunic and Bork, 2011).

Funding Information

This work was supported by the National Key R&D Program of China (2018YFA0901500), the National Natural Science Foundation of China (31970067, 31921006) and Tianjin Synthetic Biotechnology Innovation Capacity Improvement Project (TSBICIP‐KJGG‐016).

Conflict of interest

The authors declare no competing financial interest.

Supporting information

Fig. S1. The influence of the deletion of the ClBirA‐encoding gene (birA) on the growth and product formation of C. ljungdahlii in fermenting fructose. (A) The effect of the birA deletion on the growth of C. ljungdahlii grown in the medium supplemented with biotin. (B) The effect of the birA deletion on product formation of C. ljungdahlii grown in the medium supplemented with biotin. All the fermentations were performed in the modified ATCC medium 1754. Clju WT/p: the wild‐type strain containing an empty plasmid. Clju ∆birA/p: the birA‐deleted strain containing an empty plasmid. Clju ∆birA/pbirA: the birA‐deleted strain complemented with the plasmid‐carried birA expression. Data shown are means ± standard deviations (n = 3) calculated from triplicate individual experiments. Error bars show standard deviations. Statistical analysis was performed by a two‐tailed Student’s t‐test. ***P < 0.001 vs. the Clju WT/p strain.

Fig. S2. EMSAs for determining the ClBirA binding to the 24 potential target genes provided by ChIP‐seq data. The coding sequence regions or promoter regions of genes were used as the DNA probes for EMSAs. The concentration of the BirA protein used in the experiment was 80 nM.

Fig. S3. EMSAs for determining the ClBirAE (lacking the N‐terminal DNA‐binding domain) binding to bioY. The promoter region of the bioY gene was used as the DNA probe. Reactions were performed with 0.04 pmol of Cy5‐labelled probes in the presence of different concentrations of ClBirAE (40, 80, 120 nM).

Fig. S4. The influence of the overexpression of the genes coding for ACC (acetyl‐CoA carboxylase) and PYC (pyruvate carboxylase) on the growth of C. ljungdahlii in gas fermentation. All the fermentations were performed in the YT (YTF medium without fructose) medium with the supplement of a mixture of CO‐CO2‐H2‐N2. Clju ∆birA‐N‐pacc: Clju ∆birA‐N, carriying the pMTL83151‐acc plasmid. Clju ∆birA‐N‐ppyc: Clju ∆birA‐N, carrying the pMTL83151‐pyc plasmid. Data shown are means ± standard deviations (n = 3) calculated from triplicate individual experiments. Error bars show standard deviations. ns: no significance.

Table S1. Statistics for the ChIP‐seq data sets.

Table S2. Strains and plasmids used in this study.

Table S3. Primers used in this study.

Click here for additional data file. (2MB, doc)

Acknowledgements

This work was supported by the National Key R&D Program of China (2018YFA0901500), the National Natural Science Foundation of China (31970067, 31921006) and Tianjin Synthetic Biotechnology Innovation Capacity Improvement Project (TSBICIP‐KJGG‐016).

Microb. Biotechnol. (2021) 14(5), 2072–2089

Contributor Information

Weihong Jiang, Email: wjiang@cemps.ac.cn.

Yang Gu, Email: ygu@cemps.ac.cn.

References

  1. Aklujkar, M., Leang, C., Shrestha, P.M., Shrestha, M., and Lovley, D.R. (2017) Transcriptomic profiles of Clostridium ljungdahlii during lithotrophic growth with syngas or H2 and CO2 compared to organotrophic growth with fructose. Sci Rep‐Uk 7: 13135. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Antunes, A., Golfieri, G., Ferlicca, F., Giuliani, M.M., Scarlato, V., and Delany, I. (2016) HexR controls glucose‐responsive genes and central carbon metabolism in Neisseria meningitidis . J Bacteriol 198: 644–654. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Bagautdinov, B., Matsuura, Y., Bagautdinova, S., and Kunishima, N. (2008) Protein biotinylation visualized by a complex structure of biotin protein ligase with a substrate. J Biol Chem 283: 14739–14750. [DOI] [PubMed] [Google Scholar]
  4. Barber, A.M., Zhurkin, V.B., and Adhya, S. (1993) CRP‐binding sites: evidence for two structural classes with 6‐bp and 8‐bp spacers. Gene 130: 1–8. [DOI] [PubMed] [Google Scholar]
  5. Berg, I.A., Kockelkorn, D., Ramos‐Vera, W.H., Say, R.F., Zarzycki, J., Hügler, M., et al. (2010) Autotrophic carbon fixation in archaea. Nat Rev Microbiol 8: 447–460. [DOI] [PubMed] [Google Scholar]
  6. Bertani, G. (2004) Lysogeny at mid‐twentieth century: P1, P2, and other experimental, systems. J Bacteriol 186: 595–600. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Bouillaut, L., Dubois, T., Francis, M.B., Daou, N., Monot, M., Sorg, J.A., et al. (2019) Role of the global regulator Rex in control of NAD(+)‐regeneration in Clostridioides (Clostridium) difficile . Mol Microbiol 111: 1671–1688. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Commichau, F.M., and Stulke, J. (2008) Trigger enzymes: bifunctional proteins active in metabolism and in controlling gene expression. Mol Microbiol 67: 692–702. [DOI] [PubMed] [Google Scholar]
  9. Edgar, R.C. (2004) MUSCLE: a multiple sequence alignment method with reduced time and space complexity. BMC Bioinformatics 5: 1–19. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Feng, Y.J., Chin, C.Y., Chakravartty, V., Gao, R.S., Crispell, E.K., Weiss, D.S., and Cronan, J.E. (2015) The atypical occurrence of two biotin protein ligases in Francisella novicida is due to distinct roles in virulence and biotin metabolism. MBio 6: e00591–e615. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Feng, Y.J., Napier, B.A., Manandhar, M., Henke, S.K., Weiss, D.S., and Cronan, J.E. (2014) A Francisella virulence factor catalyses an essential reaction of biotin synthesis. Mol Microbiol 91: 300–314. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Feng, Y.J., Zhang, H.M., and Cronan, J.E. (2013) Profligate biotin synthesis in ‐proteobacteria a developing or degenerating regulatory system? Mol Microbiol 88: 77–92. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Guindon, S., Dufayard, J.F., Lefort, V., Anisimova, M., Hordijk, W., and Gascuel, O. (2010) New algorithms and methods to estimate maximum‐likelihood phylogenies: assessing the performance of PhyML 3.0. Syst Biol 59: 307–321. [DOI] [PubMed] [Google Scholar]
  14. Huang, H., Chai, C.S., Li, N., Rowe, P., Minton, N.P., Yang, S., et al. (2016) CRISPR/Cas9‐based efficient genome editing in Clostridium ljungdahlii, an autotrophic gas‐fermenting bacterium. Acs Synth Biol 5: 1355–1361. [DOI] [PubMed] [Google Scholar]
  15. Humphreys, C.M., McLean, S., Schatschneider, S., Millat, T., Henstra, A.M., Annan, F.J., et al. (2015) Whole genome sequence and manual annotation of Clostridium autoethanogenum, an industrially relevant bacterium. BMC Genom 16: 1085. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Kanehisa, M., Goto, S., Sato, Y., Furumichi, M., and Tanabe, M. (2012) KEGG for integration and interpretation of large‐scale molecular data sets. Nucleic Acids Res 40: D109–D114. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Katoh, K., and Standley, D.M. (2013) MAFFT multiple sequence alignment software version 7: improvements in performance and usability. Mol Biol Evol 30: 772–780. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Krishnan, N., and Becker, D.F. (2005) Characterization of a bifunctional PutA homologue from Bradyrhizobium japonicum and identification of an active site residue that modulates proline reduction of the flavin adenine dinucleotide cofactor. Biochemistry 44: 9130–9139. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Lemgruber, R.D.P., Valgepea, K., Garcia, R.A.G., de Bakker, C., Palfreyman, R.W., Tappel, R., et al. (2019) A TetR‐family protein (CAETHG_0459) activates transcription from a new promoter potif associated with essential genes for autotrophic growth in acetogens. Front Microbiol 10: 2549. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Letunic, I., and Bork, P. (2011) Interactive tree of life v2: online annotation and display of phylogenetic trees made easy. Nucleic Acids Res 39: W475–W478. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Li, Y.F., and Sousa, R. (2012) Expression and purification of E. coli BirA biotin ligase for in vitro biotinylation. Protein Expres Purif 82: 162–167. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Liu, L.K., Becker, D.F., and Tanner, J.J. (2017) Structure, function, and mechanism of proline utilization A (PutA). Arch Biochem Biophys 632: 142–157. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Mironov, A.A., Vinokurova, N.P., and Gelfand, M.S. (2000) Software for analysis of bacterial genomes. Mol Biol+ 34: 222–231. [PubMed] [Google Scholar]
  24. Overbeek, R., Begley, T., Butler, R.M., Choudhuri, J.V., Chuang, H.Y., Cohoon, M., 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. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Peters‐Wendisch, P., Stansen, K.C., Gotker, S., and Wendisch, V.F. (2012) Biotin protein ligase from Corynebacterium glutamicum: role for growth and L‐lysine production. Appl Microbiol Biot 93: 2493–2502. [DOI] [PubMed] [Google Scholar]
  26. Ren, C., Gu, Y., Wu, Y., Zhang, W.W., Yang, C., Yang, S., and Jiang, W.H. (2012) Pleiotropic functions of catabolite control protein CcpA in butanol‐producing Clostridium acetobutylicum . BMC Genom 13: 349. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Rodionov, D.A., Mironov, A.A., and Gelfand, M.S. (2002) Conservation of the biotin regulon and the BirA regulatory signal in eubacteria and archaea. Genome Res 12: 1507–1516. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Satiaputra, J., Shearwin, K.E., Booker, G.W., and Polyak, S.W. (2016) Mechanisms of biotin‐regulated gene expression in microbes. Syn Syst Biotechno 1: 17–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Soares da Costa, T.P., Tieu, W., Yap, M.Y., Pendini, N.R., Polyak, S.W., Sejer Pedersen, D., et al. (2012) Selective inhibition of biotin protein ligase from Staphylococcus aureus . J Biol Chem 287: 17823–17832. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Sorenson, A.E., Askin, S.P., and Schaeffer, P.M. (2015) In‐gel detection of biotin–protein conjugates with a green fluorescent streptavidin probe. Anal Methods 7: 2087–2092. [Google Scholar]
  31. Tamura, K., Peterson, D., Peterson, N., Stecher, G., Nei, M., and Kumar, S. (2011) MEGA5: molecular evolutionary genetics analysis using maximum likelihood, evolutionary distance, and maximum parsimony methods. Mol Biol Evol 28: 2731–2739. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Tan, Y., Liu, J.J., Chen, X.H., Zheng, H.J., and Li, F.L. (2013) RNA‐seq‐based comparative transcriptome analysis of the syngas‐utilizing bacterium Clostridium ljungdahlii DSM 13528 grown autotrophically and heterotrophically. Mol Biosyst 9: 2775–2784. [DOI] [PubMed] [Google Scholar]
  33. Wang, R., Mast, Y., Wang, J., Zhang, W.W., Zhao, G.P., Wohlleben, W., et al. (2013) Identification of two‐component system AfsQ1/Q2 regulon and its cross‐regulation with GlnR in Streptomyces coelicolor . Mol Microbiol 87: 30–48. [DOI] [PubMed] [Google Scholar]
  34. Wang, Y., Cen, X.F., Zhao, G.P., and Wang, J. (2012) Characterization of a new GlnR binding box in the promoter of amtB in Streptomyces coelicolor inferred a PhoP/GlnR competitive binding mechanism for transcriptional regulation of amtB . J Bacteriol 194: 5237–5244. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Weaver, L.H., Kwon, K., Beckett, D., and Matthews, B.W. (2001) Corepressor‐induced organization and assembly of the biotin repressor: a model for allosteric activation of a transcriptional regulator. Proc Natl Acad Sci USA 98: 6045–6050. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Yang, B., Feng, L., Wang, F., and Wang, L. (2015) Enterohemorrhagic Escherichia coli senses low biotin status in the large intestine for colonization and infection. Nat Commun 6: 1–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Yang, G.H., Jia, D.C., Jin, L., Jiang, Y.Q., Wang, Y., Jiang, W.H., and Gu, Y. (2017) Rapid generation of universal synthetic promoters for controlled gene expression in both gas‐fermenting and saccharolytic Clostridium species. Acs Synth Biol 6: 1672–1678. [DOI] [PubMed] [Google Scholar]
  38. Yang, Y.P., Zhang, L., Huang, H., Yang, C., Yang, S., Gu, Y., and Jiang, W.H. (2017) A flexible binding site architecture provides new insights into CcpA global regulation in Gram‐positive bacteria. MBio 8: e02004–e2016. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Ye, H.Y., Cai, M.Z., Zhang, H.M., Li, Z.C., Wen, R.H., and Feng, Y.J. (2016) Functional definition of BirA suggests a biotin utilization pathway in the zoonotic pathogen Streptococcus suis . Sci Rep‐Uk 6: 26479. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Zhang, H., Wang, Q., Fisher, D.J., Cai, M., Chakravartty, V., Ye, H., et al. (2016) Deciphering a unique biotin scavenging pathway with redundant genes in the probiotic bacterium Lactococcus lactis . Sci Rep‐Uk 6: 25680. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Zhang, L., Liu, Y.Q., Yang, Y.P., Jiang, W.H., and Gu, Y. (2018) A novel dual‐cre motif enables two‐way autoregulation of CcpA in Clostridium acetobutylicum . Appl Environ Microb 84: e00114–e118. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Zhang, L., Liu, Y.Q., Zhao, R., Zhang, C., Jiang, W.H., and Gu, Y. (2020) Interactive regulation of formate dehydrogenase during CO2 fixation in gas‐fermenting bacteria. MBio 11: e00650–e720. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Zhang, L., Zhao, R., Jia, D.C., Jiang, W.H., and Gu, Y. (2020) Engineering Clostridium ljungdahlii as the gas‐fermenting cell factory for the production of biofuels and biochemicals. Curr Opin Chem Biol 59: 54–61. [DOI] [PubMed] [Google Scholar]
  44. Zhao, R., Liu, Y.Q., Zhang, H., Chai, C.S., Wang, J., Jiang, W.H., and Gu, Y. (2019) CRISPR‐Cas12a‐mediated gene deletion and regulation in Clostridium ljungdahlii and its application in carbon flux redirection in synthesis gas fermentation. Acs Synth Biol 8: 2270–2279. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Fig. S1. The influence of the deletion of the ClBirA‐encoding gene (birA) on the growth and product formation of C. ljungdahlii in fermenting fructose. (A) The effect of the birA deletion on the growth of C. ljungdahlii grown in the medium supplemented with biotin. (B) The effect of the birA deletion on product formation of C. ljungdahlii grown in the medium supplemented with biotin. All the fermentations were performed in the modified ATCC medium 1754. Clju WT/p: the wild‐type strain containing an empty plasmid. Clju ∆birA/p: the birA‐deleted strain containing an empty plasmid. Clju ∆birA/pbirA: the birA‐deleted strain complemented with the plasmid‐carried birA expression. Data shown are means ± standard deviations (n = 3) calculated from triplicate individual experiments. Error bars show standard deviations. Statistical analysis was performed by a two‐tailed Student’s t‐test. ***P < 0.001 vs. the Clju WT/p strain.

Fig. S2. EMSAs for determining the ClBirA binding to the 24 potential target genes provided by ChIP‐seq data. The coding sequence regions or promoter regions of genes were used as the DNA probes for EMSAs. The concentration of the BirA protein used in the experiment was 80 nM.

Fig. S3. EMSAs for determining the ClBirAE (lacking the N‐terminal DNA‐binding domain) binding to bioY. The promoter region of the bioY gene was used as the DNA probe. Reactions were performed with 0.04 pmol of Cy5‐labelled probes in the presence of different concentrations of ClBirAE (40, 80, 120 nM).

Fig. S4. The influence of the overexpression of the genes coding for ACC (acetyl‐CoA carboxylase) and PYC (pyruvate carboxylase) on the growth of C. ljungdahlii in gas fermentation. All the fermentations were performed in the YT (YTF medium without fructose) medium with the supplement of a mixture of CO‐CO2‐H2‐N2. Clju ∆birA‐N‐pacc: Clju ∆birA‐N, carriying the pMTL83151‐acc plasmid. Clju ∆birA‐N‐ppyc: Clju ∆birA‐N, carrying the pMTL83151‐pyc plasmid. Data shown are means ± standard deviations (n = 3) calculated from triplicate individual experiments. Error bars show standard deviations. ns: no significance.

Table S1. Statistics for the ChIP‐seq data sets.

Table S2. Strains and plasmids used in this study.

Table S3. Primers used in this study.

Click here for additional data file. (2MB, doc)

Articles from Microbial Biotechnology are provided here courtesy of Wiley

RESOURCES