Systematic Adaptation of Bacillus licheniformis to 2-Phenylethanol Stress

ABSTRACT

The compound 2-phenylethanol (2-PE) is a bulk flavor and fragrance with a rose-like aroma that can be produced by microbial cell factories, but its cellular toxicity inhibits cellular growth and limits strain performance. Specifically, the microbe Bacillus licheniformis has shown a strong tolerance to 2-PE. Understanding these tolerance mechanisms is crucial for achieving the hyperproduction of 2-PE. In this report, the mechanisms of B. licheniformis DW2 resistance to 2-PE were studied by multi-omics technology coupled with physiological and molecular biological approaches. 2-PE induced reactive oxygen species formation and affected nucleic acid, ribosome, and cell wall synthesis. To manage 2-PE stress, the antioxidant and global stress response systems were activated; the repair system of proteins and homeostasis of the ion and osmotic were initiated. Furthermore, the tricarboxylic acid cycle and NADPH synthesis pathways were upregulated; correspondingly, scanning electron microscopy revealed that cell morphology was changed. These results provide deeper insights into the adaptive mechanisms of B. licheniformis to 2-PE and highlight the potential targets for genetic manipulation to enhance 2-PE resistance.

IMPORTANCE The ability to tolerate organic solvents is essential for bacteria producing these chemicals with high titer, yield, and productivity. As exemplified by 2-PE, bioproduction of 2-PE represents a promising alternative to chemical synthesis and plant extraction approaches, but its toxicity hinders successful large-scale microbial production. Here, a multi-omics approach is employed to systematically study the mechanisms of B. licheniformis DW2 resistance to 2-PE. As a 2-PE-tolerant strain, B. licheniformis displays multifactorial mechanisms of 2-PE tolerance, including activating global stress response and repair systems, increasing NADPH supply, changing cell morphology and membrane composition, and remodeling metabolic pathways. The current work yields novel insights into the mechanisms of B. licheniformis resistance to 2-PE. This knowledge can also be used as a clue for improving bacterial performances to achieve industrial-scale production of 2-PE and potentially applied to the production of other relevant organic solvents, such as tyrosol and hydroxytyrosol.

INTRODUCTION

The compound 2-phenylethanol (2-PE) is an aromatic chemical with a profound rose-like odor, which is synthesized by some flowers and plants, especially roses. As a kind of bulk flavor or fragrance, 2-PE has been widely employed in the food, cosmetics, and perfumery industries (1, 2). Furthermore, 2-PE is used as an antiseptic due to its antibacterial property and as a substrate for the production of 2-phenethyl acetate (2, 3). The global demand for 2-PE has continued to mount and exceeded 10,000 tons in 2010 (4). The 2-PE market has exceeded US$255 million in 2021 and is estimated to grow at over 5.5% compound annual growth rate (CAGR) between 2022 and 2028. Currently, chemical synthesis and plant extraction are the main methods for commercial production of 2-PE. However, these traditional methods cannot meet the growing demand for natural 2-PE (1). Microbial synthesis of 2-PE is an appealing alternative due to its green, sustainable, and natural features.

Fermentative production of 2-PE has been achieved using contemporary biotechnological approaches. Many microorganisms, such as Bacillus licheniformis, Saccharomyces cerevisiae, Yarrowia lipolytica, Kluyveromyces marxianus, Proteus mirabilis, and Escherichia coli have been successfully engineered to produce 2-PE (2, 5–7). For example, a high-yielding 2-PE production (6.24 g/L) was achieved using a genetically engineered B. licheniformis strain via engineering the Ehrlich pathway, increasing the precursor supply, and blocking by-product synthesis (5).

However, the bioproduction of organic solvent is limited using the microbial fermentation method due to its toxicity to the microbial cells. Many researchers have reported that cell growth is inhibited by a low concentration of 2-PE (2 to 4 g/L) (8–10). Although 2-PE can be removed from fermentation broth using in situ product recovery (ISPR) methods, these processes increase the purification cost (1, 3). Thus, gaining insights into the mechanism of 2-PE tolerance may help to improve the strains’ performance.

Several studies have attempted to uncover the inhibitory mechanism of 2-PE to microbial cells (11–13). It has been reported that 2-PE changes membrane permeability, affects nutrient uptake, and damages the respiratory chain (11). In addition, as an antiseptic, 2-PE inhibits the synthesis of DNA, RNA, and protein (12). A transcriptomics study has been performed to compare the gene expression levels in Penicillium italicum with or without 2-PE stress. The transcriptomics results indicate that amino acid and protein biosynthesis, cell proliferation, and cell death associated pathways are significantly modulated (13). However, to our knowledge, these studies have only explained the inhibitory mechanism of 2-PE at the physiological and transcriptional levels. The tolerance mechanisms to 2-PE in 2-PE-tolerant strains are not yet systematically elucidated.

The well-characterized B. licheniformis DW2 has been engineered to produce various antimicrobial peptides, enzymes, and compounds (5, 14, 15). B. licheniformis DW2 exhibits strong tolerance to 2-PE stress. The maximum tolerance concentration of DW2 for 2-PE is 5.0 g/L, which is higher than other species (16). However, its tolerance mechanisms remain unclear. The aim of this study is to understand the mechanisms of 2-PE resistance of B. licheniformis using a multi-omics approach. Firstly, the protein expression levels of B. licheniformis cells in the presence and absence of 2-PE are compared using a quantitative label-free proteomics approach. Furthermore, the metabolic and lipid alterations are investigated by metabolomic and lipidomic profiling analyses. The results of multi-omics analyses are confirmed by extensive physiological and molecular biological experiments. Our results will provide insights into the mechanisms of B. licheniformis DW2 resistance to 2-PE and identify potential targets that can be manipulated to improve strain tolerance to 2-PE.

RESULTS AND DISCUSSION

Overview of proteome profile of B. licheniformis in response to 2-PE.

In our previous study, B. licheniformis DW2 is shown to be capable of tolerating high concentrations of 2-PE (16). Thus, B. licheniformis DW2 is a good representative to investigate the microbial resistance mechanisms to 2-PE. To minimize cell death while observing changes in protein abundance, 3 g/L of 2-PE (the 2-PE concentration corresponding to a 50% growth decrease) was chosen for the treatment of B. licheniformis cells prior to proteomics analysis in this study. A sequential windowed acquisition of all theoretical fragment ion mass spectra (SWATH-MS)-based quantitative proteomics approach was employed to investigate the global protein alterations of B. licheniformis in the presence and absence of 2-PE (Fig. 1A). SWATH-MS is an advanced data-independent acquisition (DIA) proteomics method with deep coverage capability and consistent, accurate quantitation features (17). Herein, a total of 1,112 proteins were quantitated, corresponding to 26% coverage of the B. licheniformis proteome. Figure 1B shows the heatmap with enriched KEGG pathways highlighting the proportion of significant changes. As shown in Table S1 in the supplemental material, 377 differentially expressed proteins (DEPs), including 149 upregulated proteins and 228 downregulated proteins, were observed.

FIG 1.

Overview of proteome profile in B. licheniformis DW2 in response to 2-PE. (A) Workflow of the multi-omics assay in B. licheniformis DW2 with or without 2-PE stress. (B) Hierarchical clustering analyses of differentially expressed proteins.

To understand the functional information of the altered proteins, gene ontology (GO) enrichment analyses were carried out using the Gene Ontology database. The results indicated that the DEPs were mainly distributed in various biological processes and comprised nearly all subcellular localizations (see Fig. S1 in the supplemental material). We further analyzed the KEGG pathways of DEPs upon 2-PE stress. We found that DEPs were mainly involved in key metabolic pathways, including the biosynthesis of secondary metabolites, microbial metabolism in diverse environments, biosynthesis of antibiotics, carbon metabolism, tricarboxylic acid cycle (TCA cycle), amino acid, ribosome, purine, and pyrimidine metabolism (see Fig. S2 in the supplemental material).

To validate the proteomic data, reverse transcriptase quantitative PCR (RT-qPCR) analysis was performed (see Fig. S3 in the supplemental material). The results uncovered upregulation of the gene expression of antioxidant and global stress response systems-associated proteins (msrA, trxA, trxB, tpx, yugJ, sodA, dhaS, sigB, and groS) (Fig. S3A). The gene expression of the fatty acid biosynthesis-associated protein (fabD), the sugar catabolism-associated proteins (mdxF, sacB), and the ribosome synthesis protein (rplL) were downregulated (Fig. S3B). These results were well correlated with the results of the proteomic analysis.

The protein-protein interaction (PPI) networks were analyzed using STRING v11.0 to evaluate the molecular interaction of DEPs of B. licheniformis (see Fig. S4 and Fig. S5 in the supplemental material). PPI results revealed that 2-PE invoked the following responses: (i) activation of the global stress response system, (ii) changes in central carbon metabolism, (iii) maintainance of metal ion homeostasis, (iv) modulation of carbohydrate and amino acid metabolism; and (v) remodeling cell wall and membrane. Therefore, these responses were further examined by metabolomic, lipidomic, physiological, and molecular studies.

Global stress response and antioxidant systems are activated in response to 2-PE.

Environmental stimuli and stress generally induce the generation of reactive oxygen species (ROS) (18). In this study, the ROS levels in B. licheniformis cells under 2-PE stress were measured using a 2′,7′-dichlorofluorescin diacetate (DCFH-DA) fluorescence probe. As shown in Fig. 2A, the ROS level in the 2-PE-stressed group was significantly higher than that of the control group. Many studies have illustrated that higher levels of ROS could cause oxidative damage, consequently leading to the damage of numerous cellular components, such as DNA, proteins, and membrane lipids (19). This result suggested that 2-PE might cause oxidative stress in B. licheniformis cells by increasing ROS levels.

FIG 2.

Physiological analysis and ROS levels in response to 2-PE stress. (A) Effects of 2-PE stress on reactive oxygen species (ROS) production in B. licheniformis cells (arbitrary units). (B) Changes in abundances of enzymes involved in redox homeostasis by 2-PE stress. (C) Bacterial growths (OD₆₀₀) in minimal medium (MM) with 3 g/L 2-PE at 10 h after inoculation of the control strain DW2/pHY300 and the single-gene-overexpression strains. (D) Bacterial growths (OD₆₀₀) in MM with 3 g/L 2-PE at 10 h after inoculation of the control strain DW2/pHY300 and the groESL-overexpression strain. Data are represented as means from three biological replicates, and error bars represent the standard deviations. *, P < 0.05; **, P < 0.01; ***, P < 0.001.

The proteomic data showed that the abundances of proteins associated with the thioredoxin system were significantly increased under the 2-PE stress conditions (Fig. 2B). Multiple oxidoreductases, such as NADH-dependent nitro/flavin oxidoreductases (NfrA), iron-containing alcohol dehydrogenases (UniProt accession number Q65MF1, YkwC, YugJ, YtkK), and aldehyde dehydrogenases (GabD and DhaS), were upregulated by 2-PE stress. Previous studies have shown that the upregulation of these oxidoreductases could increase resistance to benznidazole and hypochlorite (20, 21). Upregulated expression levels of methionine-(R)-sulfoxide reductase (MsrB) and peroxidase (UniProt accession number Q65DQ3) were observed. Moreover, the transcriptional levels of thioredoxin (trxA), peroxiredoxin (tpx), and thioredoxin reductase (trxB) were upregulated by 2-PE stress (Fig. S3A). Thioredoxin, peroxiredoxin, and thioredoxin reductase have been reported to protect the bacteria against oxidative damage (22, 23). Moreover, the abundances of proteins (YceC, YceD, YceE, YfkM) involved in general stress response, activated by sigma factor SigB, were also upregulated in response to 2-PE stress.

The antioxidant system relies on NADPH to remove the oxidative damage (Fig. 2B). The abundance of the protein (NadE) associated with NAD⁺ and NADP⁺ synthesis was upregulated by 2-PE stress, suggesting that the supply of NAD⁺ and NADP⁺ might be improved. An untargeted metabolomics approach was applied to explore the metabolic responses to 2-PE stress (see Fig. S6A in the supplemental material). A total of 105 metabolites were identified, and 22 exhibited a significant increase in abundance in response to 2-PE, while 51 were downregulated (Fig. S6B). These differential metabolites were involved in a wide range of pathways, including amino acid, carbohydrate, lipid, nucleotide, and central carbon metabolic pathways (Fig. S6C and D). Metabolomic results indicated that nicotinate, nicotinamide, nicotinamide mononucleotide, and NADP⁺ in the 2-PE stress group were upregulated compared to the control group (Fig. 2B). These results indicated that intracellular NADP⁺ supply was improved in response to 2-PE stress.

In this study, the growths of recombinant strains with overexpression in trxB, tpx, yugJ, and sigB were significantly improved when compared to the control strain's growth in response to 2-PE stress, while no difference was found with nfrA and sodA overexpression (Fig. 2C). In contrast, overexpression of msrAB showed a significant decrease in cell yield during growth with 2-PE compared to the control strain. Collectively, these results indicated that a multilayered antioxidant system in B. licheniformis DW2 cells was activated by 2-PE stress to reduce oxidative damage.

To promote survival during 2-PE stress, B. licheniformis DW2 cells seemed to enhance the machinery of protein homeostasis. Some proteins associated with the heat shock proteins (HSPs) and proteolytic enzymes were upregulated in response to 2-PE stress. These HSPs included UniProt accession number Q65NH0, GroS, GroL, DnaK, and HtpG (Table S1). The proteolytic enzymes YhfE, Vpr, IspA, PrfA, AmpS, YkvY, BprB, Ggt, PepA, PepF, PatA, HtrA, HtrB, ClpP, ClpC, and McsB were also upregulated. As the gatekeepers, the HSPs can prevent protein misfolding by assisting in proper protein folding (18, 22, 24). It is well-known that the increased expression of chaperones and proteases facilitates protein folding as well as avoids damage to proteins in response to a wide variety of stressors and stimuli, such as solvents or toxic compounds (22, 25). To validate these observations, the groESL was overexpressed in the DW2 strain. The groESL overexpression strain showed increased tolerance to 2-PE compared to the control strain (Fig. 2D). Taken together, upregulation of proteases and chaperones might help maintain the protein quality and increase the tolerance of cells to 2-PE stress.

Central carbon metabolism is remodeled under the 2-PE stress conditions.

NADPH is an important reducing metabolite to remove oxidative damage. NADPH could be produced by glucose-6-phosphate dehydrogenase (Zwf), 6-phosphogluconate dehydrogenase (Gnd), and isocitrate dehydrogenase (Icd) in B. licheniformis. The proteomic results showed that Zwf and Icd were upregulated (Fig. 3A). The icd overexpression strain exhibited higher tolerance to 2-PE than the control strain (Fig. 3B). In addition to Icd, other proteins involved in the TCA cycle, such as citrate synthase (CitZ), 2-oxoglutarate dehydrogenase (OdhAB), succinate-CoA ligase (SucCD), succinate dehydrogenase (SdhB), and malate dehydrogenase (MDH) were also upregulated by 2-PE stress (Fig. 3A). Similarly, the TCA cycle was activated as an adaptive mechanism to stress conditions in many bacteria (26, 27). In this study, 2-PE stress might cause osmotic shock to the bacterial cells, which was consistent with the results of butanol and isobutanol stress (22). The abundance of succinate semialdehyde dehydrogenase GabD with 2-PE stress was increased by 2.7-fold compared to that of the control strain. The strain with overexpression of gabD led to a 30% increase in cell density in the presence of 2-PE compared to that of the control strain (Fig. 3B), which represented the most significant effect shown among all overexpression strains used in this study.

FIG 3.

Modulating the central carbon metabolism pathways of B. licheniformis. (A) Multistep reactions are indicated with multiple arrows. Upregulated proteins and metabolites are shown in red, and downregulated proteins and metabolites are shown in green. (B) Comparison of levels of bacterial growth (OD₆₀₀) in MM plus 3 g/L 2-PE at 10 h after inoculation of the control strain DW2/pHY300 and the single-gene-overexpression strains. *, P < 0.05; **, P < 0.01; ***, P < 0.001. (C) Bacterial growth (OD₆₀₀) of the control strain DW9i/pHY-300 and citrate synthase inhibited strains grown in the presence of 3 g/L 2-PE.

The citrate synthase CitZ showed an increase in abundance (1.8-fold) in response to 2-PE, while the abundances of the citrate synthase CitA and MmgD were significantly downregulated. CitZ has been shown to act as a contingency enzyme for CitA and is repressed by the actions of CcpA and CcpC. Upregulation of CitZ might prevent oxidative damage to CitA and MmgD in the presence of 2-PE. To study the functions of the three citrate synthases, three knockdown strains were constructed. The citZ knockdown strain has a significant decrease in 2-PE tolerance compared to that of the control strain, while the cell growth of citA and mmgD knockdown strains showed no significant changes (Fig. 3C). These results indicated that upregulation of CitZ might be a crucial physiological response to 2-PE stress.

Metabolomic data displayed that the levels of phosphoenolpyruvate (PEP) and pyruvic acid in the 2-PE stress group were reduced (Fig. 3A). The abundances of pyruvate dehydrogenase and citrate synthase were increased by 2-PE stress. These results indicated that there was an increase in carbon flux distributed into the TCA cycle. An accelerated TCA cycle and acetyl coenzyme A (acetyl-CoA) accumulation might result in more biosynthesis of NADH and GTP.

Expression levels of transporters are increased under the 2-PE stress conditions.

The proteomic data showed that some membrane transport proteins were upregulated in the presence of 2-PE. For example, the abundance of drug efflux pump protein LmrB, and ATP binding protein YlmA were significantly upregulated by 2-PE stress (Fig. 4A). LmrB is a drug efflux pump protein related to tolerance to lincomycin and puromycin (28). In this study, the lmrB gene knockdown strain showed an increased sensitivity to 2-PE compared to that of the control strain (Fig. 4B), indicating LmrB might be a candidate for 2-PE efflux. The oligopeptide ABC transporters (OppA, OppB, OppC, and OppF) and dipeptide ABC transporters (DppA and DppE) were clearly upregulated by 2-PE stress (Fig. 4A). The oligopeptide ABC transporters play an important role in cell-cell communication of Gram-positive bacteria (29). The oppABCD knockdown strain shows a significant decrease in 2-PE tolerance compared to that of the control strain (Fig. 4B), suggesting that it might play an important role in regulating signal transduction in response to 2-PE. Manganese is an essential cofactor for superoxide dismutase SodA. MntAB transporter, a conserved manganese transport system, plays a vital role in promoting cell growth in response to Mn²⁺. The proteomic data showed that MntA was significantly increased in abundance in response to 2-PE (Fig. 4A).

FIG 4.

(A) Changes of transports in B. licheniformis DW2 in response to 2-PE stress. (B) Bacterial growths (OD₆₀₀) in MM with 3 g/L 2-PE addition after inoculation of the control strain DW9i/pHY-300 and mntB, lrmB, and oppABCD knockdown strains.

2-PE caused protein and metabolic changes in carbohydrate and amino acid metabolism pathways.

Several carbohydrate pathways were disturbed by 2-PE stress. Proteomic data showed that several proteins involved in maltodextrin uptake, i.e., MdxG, MdxF, MdxE, and MsmX, were 3.567-, 2.46-, 2.33-, and 2.25-fold less abundant, respectively, in B. licheniformis cells grown with 2-PE (see Fig. S7A in the supplemental material). The sugar phosphotransferase systems (lactose/cellobiose, sucrose transports) in B. licheniformis were downregulated in response to 2-PE, which was similar to vanillin stress (30). These sugar catabolism pathways were downregulated (such as glycoside hydrolases, xylA, sacA, sacB, mtnK, and mtnA) (Fig. S7A). Several ABC transports, including UniProt accession number BL00257, UniProt accession number BL03201, BceA, YvrC, ArtM, and ExpZ, were downregulated (Fig. S7A). Interestingly, metabolomic results showed that the contents of stachyose, lactose, trehalose, and raffinose were increased under the 2-PE stress conditions. The combined data suggested that these sugar catabolism pathways were less active under the 2-PE stress conditions. In the present study, other transporters were observed to be significantly downregulated under 2-PE stress, including methionine ABC transporter and import proteins and other transporters (Table S1). The decrease in the abundance of transporters might be reasonably regarded as an energy conservation strategy to meet the increased energy demands for 2-PE resistance.

The enzymes and metabolites present in the cysteine and methionine metabolism pathways were compared. Cysteine synthases (CysK and YtkP) related to the biosynthesis of cysteine were significantly more highly abundant under the 2-PE stress conditions (Fig. S7B). Cysteine is involved in essential cellular functions like the assembly of iron-sulfur clusters. The upregulation of cysteine synthesis might increase the supply of iron-sulfur cluster proteins to combat oxidative damage. Thus, the upregulation of cysteine biosynthesis may be a self-detoxification mechanism of DW2 against 2-PE stress. However, we observed a reduction of proteins associated with the methionine metabolism pathway (Table S1). Among them, MetH, YxjG, YrhB, MetE, YjcJ, YitJ, YjcI, and MtnW related to the methionine synthesis and salvage and MetQ and MetN related to the methionine uptake system were found to be reduced. In agreement with these observations, l-methionine was in low abundance under 2-PE stress conditions (Fig. S7B). The results showed that the increased pool sizes of cysteine and reduced pool sizes of methionine were caused by 2-PE stress.

To better understand the regulated metabolic pathways, these altered proteins and metabolites were analyzed by integrated analysis. The integrated proteomic and metabolomic results showed that most proteins and metabolites involved in purine and pyrimidine biosynthesis were less abundant under 2-PE stress (Fig. 5). Moreover, a decrease of 1.5- to 28.0-fold was observed in the abundance of ribosomal protein subunits under 2-PE stress (Table S1). The levels of aminoacyl-tRNA biosynthesis-related metabolites such as l-valine, l-isoleucine, l-methionine, l-proline, and l-histidine were also decreased (Fig. 5). These results are in agreement with previously reported physiological and transcriptome results (12, 13), which showed that 2-PE inhibited macromolecular synthesis, including DNA, RNA, and protein synthesis.

FIG 5.

Changes in nucleic acid and protein synthesis pathways in response to 2-PE stress. Multistep reactions are indicated with multiple arrows. Red indicates upregulation, and green indicates downregulation.

Changing cell morphology.

Many studies have reported that the cell morphology of the microbe is changed in response to solvent stress (31, 32). In this study, the cell morphology of the DW2 strain under the 2-PE stress conditions was examined by the scanning electron microscopy (SEM) assay. It was noteworthy that the width of cells exposed to 2-PE (1.74 ± 0.24 μm) was longer than that of the nonstressed cells (1.21 ± 0.15 μm) (Fig. 6A), which was similar to that of butanol stress (32).

FIG 6.

(A) Representative SEM images of B. licheniformis DW2 cells cultured without or with 3 g/L of 2-PE. Images shown are at ×15,000 magnification (scale bar is 1 μm). (B) Changes in fatty acid and membrane phospholipid compositions in response to 2-PE stress. Red indicates upregulation, and green indicates downregulation.

Metabolic perturbations in peptidoglycan and lipid synthesis.

The bacterial cell wall is regarded as the first line of defense against solvent stress, and the peptidoglycan determines the mechanical strength and permeability of cell walls (33). The two-component system CssS/CssR contributes to coping with cell wall stresses caused by environmental stress (34). Upregulation of CssS/CssR by 2-PE stress suggests that they are indicators of a cell wall stress response. The proteomic analysis showed that the abundance of proteins involved in the cell wall integrity pathway, including YhcJ, YtpR, TagD, TagF, MurA2, MurC, and MurA1, displayed a significant decrease response to 2-PE stress (Table S1). DltA and DltD involved in the incorporation of d-alanine into lipoteichoic acid (LTA) were downregulated under 2-PE stress (Table S1). A decrease in d-alanine substitutions on LTA can decrease the positive surface charge of the cell wall by binding metal ions, which results in peptidoglycan damage (35). The metabolites involved in peptidoglycan syntheses, such as N-acetylneuraminate, N-acetyl-galactosamine, and glucosamine, were also downregulated in response to 2-PE stress (see Table S2 in the supplemental material). Altogether, the downregulation of proteins related to cell wall integrity implied that cell wall stress was caused by 2-PE stress.

Organic solvents can enter the cell membrane to disrupt its integrity, increasing the permeability, concomitantly resulting in cell membrane damage (31). Considering that lipids remodeling was a part of resistance mechanisms, an untargeted lipidomic analysis was performed to compare the lipid changes between the 2-PE stress group and the control group. A total of 49 varied lipids from 5 different classes were observed (see Table S3 in the supplemental material). Our lipidomic results showed that fatty acid contents in the 2-PE stress group were decreased compared to the control group (Table S3). In agreement with lipidomic analysis results, the proteomic data also showed that proteins related to fatty acids biosynthesis (AcpP, FabG, FabI, FabHA, FabF, and FabD) were downregulated by 2-PE stress (Fig. 6B).

Importantly, the distribution of the membrane phospholipid head groups was changed to adapt to the 2-PE stress. The abundances of phosphatidylethanolamine (PE), cardiolipin (CL), and phosphatidylglycerol (PG) in the 2-PE stressed cells were higher compared to those of the nonstressed cells (Fig. 6B) (Table S3). Altering cell membrane lipid composition is an adaptive response to solvent stress (31, 36–38). Thus, these findings indicate that B. licheniformis DW2 regulates membrane phospholipid headgroup distribution to improve 2-PE tolerance. As such, it can be hypothesized that one of the mechanisms of B. licheniformis DW2 resistance to 2-PE is through improving the cell membrane integrity via remodeling membrane phospholipids composition.

To better understand the tolerance mechanisms of B. licheniformis DW2 in response to 2-PE, a response network of B. licheniformis under 2-PE stress was proposed in Fig. 7. When 2-PE entered cells and induced ROS formation, there was primary damage to DNA, RNA, ribosome, and cell wall and perturbation of metabolism. Antioxidant and global response systems were activated for global cellular regulation to alleviate oxidative damage. Synthesis of proteins with the protein protective mechanism was initiated, and the production of NADPH was increased. The 2-PE stress stimulated the TCA cycle pathway to increase NADH and GTP supplies. To block 2-PE access to cytoplasm, the distribution of the membrane phospholipid head groups was changed. Some transporters were downregulated to reduce energy consumption to meet the increased energy demands for 2-PE resistance.

FIG 7.

Schematic representation of the systemic changes in B. licheniformis in response to 2-PE stress. Multistep reactions are indicated with multiple arrows. Red indicates upregulation, and green indicates downregulation.

Conclusions.

In this study, the global response of B. licheniformis to 2-PE stress at the protein and metabolic levels was systematically investigated using a multi-omics approach. To resist this solvent stress, DW2 strain employs multifactorial strategies, including activating global stress responses, strengthening its antioxidant system, trigging energy conservation, and modulating glycerophospholipid and central carbon metabolism. The results of physiological and genetic engineering experiments were in line with those of the multi-omics analyses. Collectively, our results provide insights into the tolerance mechanisms of B. licheniformis to 2-PE and identify potential targets that can be manipulated to improve strain resistance to relevant organic solvents.

MATERIALS AND METHODS

Bacterial culture and growth conditions.

All strains used in this work are presented in Table 1, and the primers are summarized in Table 2. E. coli DH5α was used for plasmid storage and propagation, and B. licheniformis DW2 and DW9i (39) were used as the initial strains to construct mutants. Gene overexpression and knockdown vectors were derived from the plasmids pHY300PLK and pGRNA, respectively. Strains were grown in lysogeny broth (LB) medium, with 20 μg/L tetracycline/kanamycin added as necessary for plasmid selection.

TABLE 1.

Strains and plasmids in this study

Strain or plasmid	Characteristic(s)	Source
Strain
B. licheniformis DW2	B. licheniformis DW2 (CCTCC M2011344)	Laboratory stock
DW2/pHY-300	DW2 harboring pHY300PLK	This work
DW2/pHY-trxB	DW2 harboring pHY-trxB	This work
DW2/pHY-tpx	DW2 harboring pHY-tpx	This work
DW2/pHY-yugJ	DW2 harboring pHY-yugJ	This work
DW2/pHY-nfrA	DW2 harboring pHY-nfrA	This work
DW2/pHY-msrAB	DW2 harboring pHY-msrAB	This work
DW2/pHY-sigB	DW2 harboring pHY-sigB	This work
DW2/pHY-sodA	DW2 harboring pHY-sodA	This work
DW2/pHY-groESL	DW2 harboring pHY-groESL	This work
DW2/pHY-gabD	DW2 harboring pHY-gabD	This work
DW2/pHY-icd	DW2 harboring pHY-icd	This work
DW2/pHY-zwf	DW2 harboring pHY-zwf	This work
DW2/pHY-oxdC	DW2 harboring pHY-oxdC	This work
DW9i/pHY-300	DW9i harboring pHY300PLK	This work
DW9i/pHYi-mmgD	DW9i harboring pHYi-mmgD	This work
DW9i/pHYi-citA	DW9i harboring pHYi-citA	This work
DW9i/pHYi-citZ	DW9i harboring pHYi-citZ	This work
DW9i/pHYi-mntB	DW9i harboring pHYi-mntB	This work
DW9i/pHYi-oppA	DW9i harboring pHYi-oppA	This work
DW9i/pHYi-lmrB	DW9i harboring pHYi-lmrB	This work
E. coli DH5a	F– Φ80d/lacZΔM15, Δ(lacZYA-argF)U169, recA1, endA1, hsdR17(rK−, mK+), phoA, supE44, λ–, thi−1, gyrA96, relA1	Laboratory stock
Plasmid
pHY300PLK	E. coli-B. licheniformis shuttle vector, Apr (E. coli), Tetr (E. coli and B. licheniformis)	Laboratory stock
pHY-kivD	pHY300PLK carrying kivD gene with P43 promoter
pHY-trxB	pHY300PLK carrying trxB gene	This work
pHY-tpx	pHY300PLK carrying tpx gene	This work
pHY-yugJ	pHY300PLK carrying yugJ gene	This work
pHY-nfrA	pHY300PLK carrying nfrA gene	This work
pHY-msrAB	pHY300PLK carrying msrAB gene	This work
pHY-sigB	pHY300PLK carrying sigB gene	This work
pHY-sodA	pHY300PLK carrying sodA gene	This work
pHY-groESL	pHY300PLK carrying groESL gene	This work
pHY-gabD	pHY300PLK carrying gabD gene	This work
pHY-icd	pHY300PLK carrying icd gene	This work
pHY-zwf	pHY300PLK carrying zwf gene	This work
pHY-oxdC	pHY300PLK carrying oxdC gene	This work
pHYi-mmgD	Derived from pGRNA, with the sgRNA cassette targeting the mmgD ORFa	This work
pHYi-citA	Derived from pGRNA, with the sgRNA cassette targeting the citA ORF	This work
pHYi-citZ	Derived from pGRNA, with the sgRNA cassette targeting the citZ ORF	This work
pHYi-mntB	Derived from pGRNA, with the sgRNA cassette targeting the mntB ORF	This work
pHYi-oppA	Derived from pGRNA, with the sgRNA cassette targeting the oppABCD operon	This work
pHYi-lmrB	Derived from pGRNA, with the sgRNA cassette targeting the lmrB ORF	This work

^aORF, open reading frame.

TABLE 2.

Primers used in this study

Primer	5′→3′ sequence
300-T5-F	TGATCCTTCCTCCTTTAGATCTGCT
300-T5-R	AAGAGCAGAGAGGACGGATTTCCTG
msrAB-F	CTAAAGGAGGAAGGATCAATGGCTGAAAAACGAGAA
msrAB-R	TCCGTCCTCTCTGCTCTTCTATTTTTCCCCTTCAAA
tpx-F	CTAAAGGAGGAAGGATCAATGATGGCTTCAATTACA
tpx-R	TCCGTCCTCTCTGCTCTTTTACTTAACGAGTGATTT
trxB-F	CTAAAGGAGGAAGGATCAATGTCAGAAGAAAAAATG
trxB-R	TCCGTCCTCTCTGCTCTTTTATTTTACAGCCTTTTC
dhaS-F	CTAAAGGAGGAAGGATCAATGACAAACATGAGTTCA
dhaS-R	TCCGTCCTCTCTGCTCTTTTATTCGCCTGTATGAAT
nfrA-F	CTAAAGGAGGAAGGATCAATGAATAAAACGATTGAA
nfrA-R	TCCGTCCTCTCTGCTCTTTTATCTTTTGTTAAAACC
yugJ-F	CTAAAGGAGGAAGGATCAATGGATAATTTTACATAT
yugJ-R	TCCGTCCTCTCTGCTCTTTTATAAAGAAGCCTTCAA
sigB-F	CTAAAGGAGGAAGGATCAATGGCACAACCGTCCAAA
sigB-R	TCCGTCCTCTCTGCTCTTTCATGGTTTAAGCTCCAT
gabD-F	CTAAAGGAGGAAGGATCAATGACCAGCATGCTAGAA
gabD-R	TCCGTCCTCTCTGCTCTTTTATTCATCCAAACCGATG
oxdC-F	CTAAAGGAGGAAGGATCAATGGAAAAAAACAAATCA
oxdC-R	TCCGTCCTCTCTGCTCTTGAAATTTGATAAGAAATAA
sodA-F	CTAAAGGAGGAAGGATCAATGATGGCTTACAAACTTCC
sodA-R	TCCGTCCTCTCTGCTCTTTTATTTTGCTTCGCTGTAT
groESL-F	CTAAAGGAGGAAGGATCATTGTTAAAGCCATTAGGTG
groESL-R	TCCGTCCTCTCTGCTCTTTTACATCATGCCGCCCATA
icd-F	CTAAAGGAGGAAGGATCAATGTTTTTCTATATTGAA
icd-R	TCCGTCCTCTCTGCTCTTTTAAGACATATTTTTGATAAG
zwf-F	CTAAAGGAGGAAGGATCATTGAAAAAAGATCAAATGG
zwf-R	TCCGTCCTCTCTGCTCTTTTAAAGCGGCCACCAATGA
oppA-g-F	AGCCGAAGCCGCAGGCGCTGGTTTTAGAGCTAGAAATA
oppA-g-R	CAGCGCCTGCGGCTTCGGCTAATGGTACCGCTATCACT
lmrB-g-F	TAAAATGGCGATAATCATGAGTTTTAGAGCTAGAAATAG
lmrB-g-R	TCATGATTATCGCCATTTTAAATGGTACCGCTATCACT
mntB-g-F	TAAATCTTTTACAGTAATCGGTTTTAGAGCTAGAAATAG
mntB-g-R	CGATTACTGTAAAAGATTTAAATGGTACCGCTATCACT
citZ-g-F	TAGTGGCGACTACCCCTTCAGTTTTAGAGCTAGAAATAG
citZ-g-R	TGAAGGGGTAGTCGCCACTAAATGGTACCGCTATCACT
citA-g-F	TTCCCTCTTCACCGTCAATGGTTTTAGAGCTAGAAATAG
citA-g-R	CATTGACGGTGAAGAGGGAAAATGGTACCGCTATCACT
mmgD-g-F	TTCCGCTTCCGTATCTAAGAGTTTTAGAGCTAGAAATAG
mmgD-g-R	TCTTAGATACGGAAGCGGAAAATGGTACCGCTATCACT
pHY-F	GTTTATTATCCATACCCTTAC
pHY-R	CAGATTTCGTGATGCTTGTC
Used in RT-qPCR
rpsE-RT-F	ATCGGACGTTTCGGTGCAGG
rpsE-RT-R	AACGTCCTCAGCGCGTTTCA
msrA-RT-F	GAAGCGGTGCAAATCACGTT
msrA-RT-R	AATGTCGGTCGCGATTGGAT
trxA-RT-F	GAAGGCGTAGTTCTTGCCGA
trxA-RT-F	TCAACCACTTCGCCGTCTTT
trxB-RT-F	GTTTGTGACGGCGCATTCTT
trxB-RT-R	ACTCTTCTTCTTCGCCCGTG
tpx-RT-F	CCGGTGACACTGGTAGGTTC
tpx-RT-R	TTCAATTCCGTTTGCTCCGC
nfrA-RT-F	TGACCGAAGAAGAGGTTCGC
nfrA-RT-R	AGGTCTCCGTCCCTTCAAGA
yugJ-RT-F	ATGTGCGAATCCCTGCTCAA
yugJ-RT-R	TCAAAAACGCGTACGGCAAG
sodA-RT-F	TTTCTCCAAACGGAGGAGGC
sodA-RT-R	TATGCGTGCTCCCATACGTC
dhaS-RT-F	TGAAACGCTCGCCACCTTAT
dhaS-RT-R	GTTATTTTCGTCGCCCAGCC
sigB-RT-F	CCGTCAACAAACAAGTCGCA
sigB-RT-R	GCGGACATTTCCAGGCATTC
groS-RT-F	GCCATTAGGTGATCGCGTTG
groS-RT-R	AAATGTCGCTTTCACGCAAGA
purF-RT-F	AAACCGCTATGTCGGCAGAA
purF-RT-R	CGCGATCAGTTCTTCATGCG
mdxF-RT-F	TTGTTCTCGGCATCGTGACA
mdxF-RT-R	AAACCGAGCCACGTCTGAAT
sacB-RT-F	TGATGAAATCGAACGGGCCA
sacB-RT-R	CGTGATGACGACTTTGTCGC
citA-RT-F	TTGCGTTCAAGCCGACAATG
citA-RT-R	TTGCCGAAAAGGTTGAAGCG
fabD-RT-F	AAGGTTCGCAGCATATCGGC
fabD-RT-R	TACCAATGCGCTGTACTCGC
aroA-RT-F	CAGCAAAGCACTGCTCGTTT
aroA-RT-R	TGTACGGGGTTTGAAGGCTC
glpK-RT-F	TGATGAACACCGGCGAGAAA
glpK-RT-R	CCGGGTCAGTCCGAATACAG
atpD-RT-F	GGTGTTACACGCGGAATGGA
atpD-RT-R	AAGACCGATTTTTCCGCCTT
rplL-RT-F	TGATCTTGTTCTTGCTGGTGC
rplL-RT-R	ACTTAACTTCTACAGAAGCGCC
metH-RT-F	TCGTCAAACAGTTCGGAGGG
metH-RT-R	TTTGATGAGGCGGATGCCTT

B. licheniformis DW2 cells were inoculated into 5 mL of LB broth and cultivated at 37°C for 16 h. The inoculum was then added into the glucose minimal medium (MM) (40) containing 10 g/L glucose, 3.0 g/L KH₂PO₄, 8.5 g/L Na₂HPO₄·2H₂O, 1.0 g/L NH₄Cl, 21 mg/L citric acid monohydrate, 0.5 g/L NaCl, 246 mg/L MgSO₄·7H₂O, 14.7 mg/L CaCl₂·2H₂O, 13.5 mg/L FeCl₃·6H₂O, 1.7 mg/L ZnCl₂, 1 mg/L MnCl₂·4H₂O, 0.6 mg/L CoCl₂·6H₂O, 0.6 mg/L Na₂MoO₄·7H₂O, and 0.43 mg/L CuCl₂·2H₂O.

Experimental design.

B. licheniformis DW2 cells were inoculated in 50 mL minimal medium without 2-PE (control group) and with 3 g/L 2-PE supplement (stress group). The stress samples were harvested at the onset of the stationary phase, corresponding to 10 h of cultivation. The control samples were collected at the same growth phase (10 h). All samples were collected in triplicates. The collected cells were washed twice with precooled phosphate-buffered saline (PBS). Cell pellets were stored at −80°C for the subsequent multi-omics studies (Fig. 1A).

Protein extraction and digestion.

Cell pellets were suspended using 100 mM triethylammonium bicarbonate (TEAB) buffer, pH 8.0 (Thermo Scientific, USA). Cell lysis was carried out by adding lysozyme (1 mg/mL) and protease inhibitor cocktail (Sigma-Aldrich), followed by the supplement of 400 μL of lysis buffer (100 mM TEAB buffer, 6.25% SDS, protease inhibitor cocktail, 10 μg/mL DNase, and 10 μg/mL RNase). After sonicating, the protein concentration was measured by the bicinchoninic acid (BCA) assay.

A total of 100 μg proteins were reduced and alkylated with dithiothreitol (DTT) and iodoacetamide, respectively. The alkylated protein was buffer exchanged, and digested with 2 μg mass spectrometry grade trypsin (Promega) at 47°C for 1 h on a suspension trapping (S-TRAP) spin cartridge (Protifi, Farmingdale, NY). Peptides were eluted and dried at a SpeedVac vacuum centrifuge according to the manufacturer’s protocol and resuspended in 0.5% formic acid (FA).

liquid chromatography-tandem mass spectrometry acquisition and data analysis.

The Evotips preloaded with 1 μg peptide were automatically picked up by the Evosep One LC system (Denmark), which sequentially elutes the trapped peptides from the Evotips using a gradient from pump A (0.1% FA in water) and B (0.1% FA in acetonitrile). Evosep One also generates an offset gradient using pump C (0.1% FA in water) and D (0.1% FA in acetonitrile) to subsequently elute the peptides from an analytical C₁₈ column (100 μm × 150 mm, 3 μm) at 1 μL/min in a 44-minute gradient with a default program. Mass spectrometry (MS) and tandem mass spectrometry (MS/MS) were performed on a TripleTOF 6600 MS system (SCIEX).

In order to construct the B. licheniformis spectral library for SWATH-MS analysis following our previous method (41), data-dependent acquisition MS (DDA-MS) was employed, and the mass spectrometer was set in top 50 selected MS1 precursor ions across a mass range of 400 to 1,500 m/z, followed by MS/MS scan using the 50 ms accumulation time per spectrum. For SWATH-MS acquisition experiments, a 250 ms survey scan (MS1 from 350 to 1,250 m/z) with 100 variable window was used, and MS2 spectra in high-sensitivity mode from 100 to 1,500 m/z for 30 ms. The total cycle time is ~3.3 s.

The DDA-MS data were analyzed using the ProteinPilot software v.5.0 (AB SCIEX, Canada). Data were searched against the UniProt database of B. licheniformis ATCC 14580 (UP000000606). The parameters were set as previously described (42). The PeakView v.2.2 (AB SCIEX, Canada) was employed to analyze SWATH acquisition data using the reference spectral library with the reported parameters (41).

To identify differentially expressed proteins (DEPs), the MarkerView 1.3 software was used to analyze the SWATH-MS data. The protein peaks were normalized to the total peak area for each run. DEPs were considered by selecting quantitation ratios corresponding to the cutoff values of 1.35 and P values of <0.05. Gene ontology (GO) enrichment analysis was performed to analyze the functional categories of DEPs, and KEGG pathway enrichment analysis of DEPs was carried out using the KOBAS website (http://kobas.cbi.pku.edu.cn/).

Metabolome analysis.

The intracellular metabolites were extracted using a cold 75% (vol/vol) ethanol solution and freezing with liquid nitrogen according to the method reported by Yuan et al. (43). The intracellular metabolites were analyzed using an ultra-high performance liquid chromatography (UPLC) system (Acquity; Waters) coupled to a mass spectrometer (Q Exactive MS; Thermo Fisher Scientific). The parameters were set as previously described (44).

The raw MS data were processed by XCMS. The mass peaks were identified using the in-house database and the Kyoto Encyclopedia of Genes and Genomes (KEGG) (https://www.genome.jp/kegg) database. Then, the metabolites of interest were confirmed by MS2 spectra comparison. The identified metabolites were analyzed using multivariate (SIMCA-P+ software, version 13.0.3; Umetrics, Sweden) (44). The varied metabolites between the 2-PE-free and 2-PE-stressed cells were selected based on ratios of >1.5 and P values of <0.05 for the Student’s t test.

Untargeted lipidomics analyses.

Cell pellets were collected at the stationary phase, and the total lipids were extracted from the cell pellets and measured by LC-MS as reported previously (45). Untargeted lipidomics analyses were performed using a UPLC Shimadzu 30A system (Kyoto, Japan) coupled to a TripleTOF 6600 mass spectrometer (Q-TOF MS; Sciex, Singapore) and a C₁₈ column (100 × 2.1 mm, 2.6 μm; Phenomenex Kinetex). The parameters of UPLC and mass spectrometer were set as previously described (46, 47).

The lipidomics data were processed and analyzed by MS-DIAL software (version 4.24), PeakView software (version 2.2), MasterView (version 2.0), and MultiQuant (version 3.0.3) (AB Sciex, Toronto, ON, Canada). The parameters of these software were set as previously described (46, 47). The upregulated proteins related to metabolites and downregulated proteins related to metabolites were summarized in Table S4 in the supplemental material.

Quantitative real-time PCR analysis.

To confirm the proteomics data, the gene expression levels of the antioxidant and global stress response systems, fatty acids biosynthesis, the sugar catabolism-associated proteins, and ribosome synthesis protein were analyzed by quantitative PCR (see Table S5 in the supplemental material). The cells were harvested at the onset of the stationary phase. An RNA extraction kit (APExBIO, China) was employed to extract the total RNA (14). A reverse transcription kit was used to obtain the cDNA. Next, a CFX96 real-time PCR system (Bio-Rad, Hercules, CA, USA) was employed for quantitative PCR analysis. Data were normalized by the reference gene rpsE using the 2^−ΔΔCT method. The relative gene expression analysis was performed in triplicate.

Measurement of ROS and cell morphology.

ROS levels were analyzed by the fluorescent probe, 2′,7′-dichlorofluorescin diacetate (DCFH-DA) kit according to the kit's instructions (Beyotime Co., China). In brief, B. licheniformis cells (optical density at 600 nm [OD₆₀₀] = 0.5) was collected at the onset of the stationary phase and then washed three times with the precooled PBS. The pellets were resuspended by 10 mmol/LDCFH-DA and incubated at 37°C for 30 min in the dark. After washing thrice with the precooled PBS, the ROS contents were measured using a microplate reader (Molecular Devices Ltd, Shanghai) with an excitation wavelength of 485 nm and an emission wavelength of 525 nm.

A high-resolution scanning electron microscopy (HR-SEM) was performed to measure the cell morphology as previously described (32). In brief, B. licheniformis cells (OD₆₀₀ = 2) were collected and washed with 0.1 M sodium cacodylate buffer (pH 7.2) and fixed with 2.0% glutaraldehyde for 2 h. Subsequently, samples were treated with osmium tetroxide in the dark. The treated samples were processed by dehydrating the cells using a graded ethanol series (25%, 50%, 75%, 95%, and 100% ethanol), followed by drying and gold-mounting and coating using Quorum Technologies SC7640 Sputter Coater. Finally, the cells were observed using the Zeiss Supra 55 scanning electron microscope.

Gene expression.

groESL fragment was amplified using DW2 genomic DNA as a template and inserted into linearized pHY-kivD using the ClonExpress one-step cloning kit (Vazyme Biotech Co., Ltd, Nanjing, China) (16). The recombinant plasmid pHY-groESL was confirmed by DNA sequencing and electroporated into B. licheniformis DW2 strain to obtain groES and groEL expression strain. Other gene (gabD, oxdC, icd, zwf, trxB, tpx, yugJ, nfrA, msrAB, sigB, and soda) overexpression strains were constructed using the same method.

Gene knockdown using CRISPRi method.

Gene knockdown was performed via the CRISPR interference (CRISPRi) approach (39). The linearized fragment of pGRNA was obtained by inverse PCR using pGRNA as the template, and the linearized pGRNA was cyclized by Gibson assembly to obtain the recombinant plasmid pHYi-citZ. Other CRISPRi plasmids harboring single guide RNAs (sgRNAs) (citA, mmgD, lrmB, oppA, and mntB) targeting single genes were constructed using a similar method. The guide RNA (gRNA) expression plasmids were electroporated into B. licheniformis DW9i strain, and pGRNA was used as the control.

Data availability.

The data sets used and/or analyzed during the current study are available from the corresponding author upon reasonable request.