Clostridium acetobutylicum bacterial membrane responses to carbon ion beam irradiation perturbations

staining showed A comparison of differences in fatty acid between 824 and cell under that maintenance the normal physiological characteristics of cell played important the variance analysis. Pairwise comparisons were performed using least signicant difference t-tests. Results with P values of less than 0.05 were considered statistically signicant.


Background
Clostridia spp. are gram-positive bacteria that have important applications in industrial processes [1]. Indeed, in industry, solventogenic Clostridium spp. utilize carbon sources, such as glucose, xylose, mannose, and arabinose, to produce acetone, butanol, and ethanol during anaerobic metabolism. This fermentation process is called acetone-butanol-ethanol (ABE) fermentation [2,3]. The main component, butanol, has the highest commercial value and is used as a multipurpose basic chemical raw material, high-quality renewable transportation fuel, and fuel additive [4]. Clostridium acetobutylicum, as the main strain used in ABE fermentation, expresses amylase and can be directly fermented with sugar and starchy raw materials without prior sacchari cation [1]. Owing to the unique sporulation and solvent production characteristics of C. acetobutylicum and its vital application value, many studies of C. acetobutylicum have been performed in the to evaluate the microbiological morphology, physiology, biochemistry, and molecular biology of the strain. Understanding the mechanisms through which Clostridium spp. adapt to environmental stress is essential for controlling these bacteria and taking advantages of its industrial applications [5].
The metabolic transition of C. acetobutylicum from the acidogenic period to the solventogenic period is an adaptive process resulting from the effects of acid toxicity on the bacterial cells [5]. During the solventogenic process, bacteria can metabolize the products accumulated during the acidogenic period through the reuse route, which causes the pH of the fermentation broth to rise, thereby partially detoxifying the acid [6]. Solvents produced during this process, particularly butanol, have major effects on the growth and metabolism of the bacteria and are therefore key factors limiting the fermentation yield [7,8]. Therefore, butanol is considered the main toxic compound inhibiting the physiological activity of C. acetobutylicum.
Organic solvents rst enter the cell membrane after coming in contact with the cells, and the phospholipid bilayer structure of the membrane is the main target of solvent toxicity [9]. The toxic effects of the solvent on the cell begin with binding of solvents to the phospholipid layer. After solvent accumulates on the cell membrane, membrane permeability to protons and ions increases [9], proton dynamics dissipate [10], energy conduction fails [11,12], and intracellular pH control is suppressed, leading to penetration of intracellular macromolecules (such as RNA, phospholipids, and proteins) [11]. Moreover, solvent accumulation on the cell membrane also affects the activity of the mosaic protein by changing the physical and chemical properties of the membrane [13], such as alterations in the proton-K + pump [11,12], and may alter the membrane structure, resulting in increased membrane uidity [13], reduced microbial metabolism, and decreased growth. Thus, after the solvent enters the membrane, it disrupts the order of the membrane, weakening the function of the membrane as a penetration barrier and protein embedding platform. The increased permeability, uidity, and disorder of the cell membrane make it di cult for microorganisms to resist the toxic effects of solvents, resulting in cell death.
Physiological studies of microorganisms have revealed that solvent toxicity is related to its logPo/w value [11]. The parameter logPo/w refers to the partition coe cient of a given solvent in equimolar octanol and water (indicating the hydrophobicity of the solvent) [14]. The greater the polarity, the smaller the logPo/w value and the greater the solvent toxicity [15]. The logPo/w values of common solvents range from 1 to 4.5, whereas that of butanol is 0.8 [16]. In addition, the solvent relies on its own hydrophobicity to enter and bind to different sites in the bilayer [14]. Binding of the solvent to different positions on the cell membrane can result in varying effects on membrane structure. For example, when a solvent is combined near the polar head, the effects of membrane disorder will be greater than that of the solvent deeply bound to the center of the phospholipid fat chain [17]. The binding positions of alkanols (e.g., butanol, octanol, dodecanol, and myristyl alcohol) on the cell membrane have been examined by Westerman et al. [18] and Pope et al. [19] by X-ray and nuclear magnetic resonance technology. The results indicated that amphiphilic molecules, such as alkanol, position the hydrated part of the molecule close to the polar head of the phospholipid molecule, whereas the fat chain portion is inserted between the fat chains of the phospholipid. Among the four alkanols described above, butanol causes the most serious cell membrane lipid damage in the glycerol skeleton [18]. Indeed, Ezeji et al. [20] found that butanol inhibited the activity of bacteria by destroying the cell membrane and altering its functions, i.e., reducing the substance transfer capacity of the membrane. A series of reliable experimental results showed that there are two systems in C. acetobutylicum, i.e., a speci c self-induced peptide transport system and a signal transduction system; the signal transduction system involves membrane receptor histidine kinase function [21]. Butanol toxicity also results in destruction of the membrane transport system (i.e., the phosphoenolpyruvatecarbohydrate phosphotransferase system), thereby preventing the transport and phosphorylation of glucose and inhibiting the transmembrane transport and assimilation of sugars, amino acids, and other nutrients [22].
In response to butanol toxicity on the cell membrane, a series of studies have been conducted to improve the ABE fermentation and solvent tolerance of C. acetobutylicum. With improvements in molecular biology technology, researchers have attempted metabolic engineering of C. acetobutylicum. For example, Xu et al. [23] knocked out CAC3319 (histidine kinase) in C. acetobutylicum and showed that the tolerance of the strain increased and that butanol production (18.2 g/L) and butanol yield (0.38 g/L/h) were both enhanced. Moreover, Borden et al. [24] screened the resistance genes of C. acetobutylicum by constructing a gene library and showed that recombinant strains carrying genes such as CAC0977, CAC1463, CAC1869, and CAC2495 were enriched. CAC1869 is a transcriptional regulator of the heterogeneous stress element, and its introduction into ATCC 824 increased the butanol tolerance of the recombinant strain by 90%. Mann et al. [25] overexpressed the genes groESL, grpE, and htpG in C. acetobutylicum and showed that 2.0% (v/v) butanol stress for 2 h improved butanol tolerance in engineered strains overexpressing these genes, with survival rates of 45% (groESL), 25% (grpE), and 56% (htpG). In addition to targeted genetic engineering methods, the genes and proteins related to butanol tolerance have been molecularly modi ed to enhance butanol tolerance and survival rates, thereby increasing butanol production. Furthermore, some induced mutations can also be used to obtain butanoltolerant mutants. C. beijerinichii BA101, obtained by chemical mutagenesis in 1991, was used for this application [26]. C. acetobutylicum EA2018 strain was also chemical mutated to become an excellent butanol producer, showing a 10% increase in butanol production compared with the original strain; thus, this strain has been introduced in industrial applications in China [27]. Guo et al. [28] used C. beijerinichii NCIMB 8052 as the starting strain and utilized low-energy ion implantation (N+) for physical mutagenesis to obtain a mutant strain IB4 with high tolerance and high butanol production. Baer et al. obtained a solvent-tolerant mutant strain SA2 through mutagenesis screening and found that the saturation of cell membrane fatty acids was higher than that of the control strain C. acetobutylicum ATCC 824 under various culture conditions between 22 and 37 °C. Moreover, the uidity of the cell membrane remained unchanged under different concentrations of butanol (0-1.5%, v/v) [11].
Compared with traditional mutagenesis technology, as an advanced and e cient mutagenesis technology, heavy ion beam irradiation results in high mutagenesis rates and yields relatively stable mutants, leading to its broad applications as a tool for creating new species. In terms of microbial breeding, approximately 30 genera have been subjected to heavy ion beam irradiation to produce novel strains [29][30][31]. Cadmium resistance of Arthrobacter mutants selected by carbon ion beam irradiation was increased by 2-fold [32], and Lu et al. [33] increased the fermentation rate of Saccharomyces cerevisiae subjected to carbon ion beam irradiation by 25%. Zhou et al. [34] used carbon ion beam irradiation to signi cantly improve the resistance of C. butyricum to butyric acid and the ability to produce butyric. Accordingly, many excellent microbial mutants with applications in industry and research have been bred using heavy ion beam irradiation mutagenesis.
C. acetobutylicum Y217 is a stable and high-yield resistant mutant selected by C. acetobutylicum ATCC 824 as the starting strain through carbon ion beam irradiation. Based on previous studies, Y217 maintains better cell integrity under butanol stress than ATCC 824, potentially because of its high butanol tolerance. In this study, we systematically analyzed and compared the cell morphology, cell surface hydrophobicity, membrane potential, membrane permeability, cell fatty acid composition, and fermentation kinetics of high-butanol tolerant mutant and wild-type strains under butanol stress in order to clarify the membrane physiology and metabolic mechanisms of C. acetobutylicum under butanol stress.

Effects of butanol stress on the growth and surface morphology of mutants
To explore the differences in the surface morphology of Y217 and ATCC 824 cells exposed to increased concentration of butanol, Y217 and ATCC 824 cells were plated on reinforced Clostridium medium (RCM) solid plates and cultured at 37 °C for 48 h. Single colonies were then selected and cultured in fresh medium for 24 h. The resulting seed suspension was inoculated into culture medium containing different concentrations of butanol and then subjected to dry cell weight (DCW) analysis and scanning electron microscopy (SEM). As shown in Fig. 1a, ATCC 824 grew well in medium without butanol, with a lag phase of approximately 4 h; in the presence of butanol, the lag phase increased signi cantly. In cultures of cells incubated with 1.0% (v/v) butanol, the DCW decreased by nearly 35%, and SEM ( Fig. 1b) showed slight folds on the surface of some bacterial cells, which may be a speci c response to butanol stimulation. In 2.0% (v/v) butanol, the cells maintained growth, the butanol concentration continued to increase, and bacterial growth entered a stagnation period. As shown in Fig. 1c, bacteria cultured with 3.0% (v/v) showed severe damage. Figure 2d shows the growth of Y217 at different butanol concentrations. Notably, at butanol concentrations of up to 3% butanol, the growth lag phase of the cells did not differ in the presence or absence of butanol. In the presence of 1.0% or 2.0% (v/v) butanol, Y217 maintained excellent growth, although growth in the presence of 2.0% (v/v) butanol decreased slightly. The surface integrity of Y217 cells was intact in the presence of 1.0% (v/v) butanol. However, in the presence of 3.0% (v/v) butanol stimulation, SEM images showed that cell surface was slightly wrinkled and that integrity was slightly reduced; importantly, the growth of the mutant still met fermentation requirements.

Effects of butanol stress on cell surface hydrophobicity
To evaluate the relationships between changes in cell surface hydrophobicity and butanol tolerance, the microbial adhesion to hydrocarbons (MATH) value of the original strain ATCC 824 and mutant strain Y217 were measured during the logarithmic growth period without butanol and with 2.0% butanol RCM. As shown in Fig. 2, MATH values of ATCC 824 and Y217 were 42.3% and 35.6%, respectively, when butanol stress was absent in the culture environment, indicating moderate hydrophobicity on the cell surface for both strains. When both strains were exposed to butanol environmental stress (2.0% butanol), the hydrophobicity on the cell surface of both cells changed signi cantly, and MATH values increased to 63.5% and 52.7%, respectively. The MATH value of Y217 was lower than that of ATCC 824 under conditions of no butanol addition and 2.0% (v/v) butanol stress, suggesting that the surface hydrophobicity of Y217 cells was lower than that of ATCC 824 cells. Therefore, in a butanol stress environment, Y217 cells could reduce the amount of butanol entering the cell, thereby reducing butanol toxicity in cells. Overall, these ndings demonstrated that reducing the cell surface hydrophobicity could improve the ability of the cells to resist butanol.

Effects of butanol stress on membrane permeability
In order to study the relationships between membrane permeability and butanol tolerance in Y217 and ATCC 824 cells under butanol stress, Y217 and ATCC 824 cells were inoculated into RCM liquid medium and cultured for 24 h. Butanol was then added ( nal concentration: 2.0%), and changes in membrane permeability were determined every 2 h. As shown in Fig. 3a, at 2 h after the addition of butanol, the uorescence intensity of ATCC 824 decreased to 51.7%, indicating that the cell membrane permeability increased by 48.3%. In contrast, for Y217, the uorescence intensity only decreased to 94.6%, and the cell membrane permeability increased by 5.4%. When the cell encounters an adverse environment, stress responses are produced to facilitate adaptation, which may be why uorescence intensity decreased rapidly and then stabilized when butanol was added. After 10 h, the uorescence intensity of ATCC 824 and Y217 cells decreased to 30.5% and 91.2%, respectively, and the permeability of ATCC 824 and Y217 cells increased by 69.5% and 8.8%, respectively. As shown in the Fig. 3b, the conductivity of Y217 remained stable throughout the cultivation cycle, whereas that of ATCC 824 increased with time. The conductivity of ATCC 824 reached a maximum of 1080 µS/cm at 8 h of treatment, and there was a signi cant difference in the conductivity of 623 µS/cm when no butanol was added (p < 0.05). At 6 h, Y217 showed a slight increase in conductivity, although this increase was not signi cant, indicating that butanol did not affect the permeability of this strain.
Nucleic acid molecules show a maximum absorption peak at a wavelength of 260 nm, whereas that of proteins is 280 nm. Thus, we next evaluated the effects of butanol on the absorbance of intracellular biological macromolecules of C. acetobutylicum at 260 and 280 nm (Fig. 3c). Without butanol stimulation, the absorbance values of ATCC 824 bacterial supernatants at 260 and 280 nm were 0.180 ± 0.015 and 0.150 ± 0.02, respectively; those of the Y217 bacterial supernatants were 0.175 ± 0.01 and 0.142 ± 0.02, respectively. After stimulation with butanol, the absorbance of the ATCC 824 supernatant increased as the treatment time increased. After 10 h of butanol treatment, the absorbance values at 260 and 280 nm reached 0.390 ± 0.02 and 0.290 ± 0.015, respectively, indicating that as the butanol treatment time was prolonged, the permeability of the cell membrane of C. acetobutylicum gradually increased, resulting in leakage of intracellular nucleic acids, proteins, and other substances and leading to cell inactivation. However, for Y217, compared with ATCC 824, the amount of macromolecular spillage was small, indicating that relatively stable permeability was maintained in the presence of butanol.

Effects of butanol stress on transmembrane potential
Decreased membrane potential is a characteristic change in the cell membrane. In order to detect the effects of butanol stress on membrane potential, we examined the uorescence intensity of rhodamine 123 (Rh123), a lipophilic cationic uorescent dye that is permeable to membranes and very sensitive to membrane potential, in the two strains. Y217 and ATCC 824 were inoculated into RCM liquid medium and cultured for 24 h. Butanol ( nal concentration: 2.0%) was then added to the fermentation broth, and samples were evaluated every 4 h to determine changes in membrane potential. Flow cytometry was used to detect the intensity of the bound Rh123 to determine the level of membrane potential. Butanol stress in ATCC 824 cells reduced the membrane potential of the cells to a certain extent (Fig. 4), and this different was statistically signi cant (P < 0.01). In contrast, after stimulation with butanol, Y217 showed no signi cant difference in cell membrane potential (Fig. 4). Thus, these ndings demonstrated that the mutant strain Y217 maintained a relatively stable membrane potential under butanol stress, thereby ensuring cell membrane integrity and cell activity.
Effects of butanol stress on membrane fatty acids (FAs) Effects of exogenous butanol on the fermentation kinetics of the mutant Based on the above research on the physiological characteristics of the cell membrane of the mutant strain, we compared the fermentation abilities of the starting strain and the mutant following the addition of 2.0% (v/v) butanol (Fig. 6). The net butanol output of Y217 cells was 7.62 g/L. Additionally, acetone production was 1.8 g/L higher than the control, and ethanol production remained similar in the two strains. During the acid production stage, the amount of acid in ATCC 824 cells was low, and butanol stimulation decreased cell activity during the acid production stage, thereby affecting the progression of the solvent stage. Notably, the nal concentration of butanol in the fermentation broth reached 21.27 g/L for Y217 cells, which was 6.04 g/L higher than that in ATCC 824 cells, indicating that the mutant could still grow normally and participate in the fermentation process to yield butanol when stimulated with 2.0% butanol. The cells showed ultimate butanol tolerance when the butanol concentration exceeded 20 g/L.

Discussion
Heavy ion beam irradiation, which can induce genetic variations in the genome with high mutation rates, is commonly used for mutation breeding. This approach has several advantages in the improvement of biological varieties and has yielded important advancements in the breeding of microorganisms, plants, and other biological varieties. In a previous study, C. acetobutylicum Y217, obtained from 12 C 6+ ion beam irradiation generated by the Heavy Ion Research Facility of Lanzhou (HIRFL), showed resistance to butanol and membrane integrity disruption. Accordingly, in this study, we aimed to explore the mechanisms through which the mutant responded to butanol stress in terms of changes to the membrane.
The relationships between cell morphology and cell tolerance to organic solvents have been studied based on changes in cell size, speci c surface area, and capsules. Capsules have protective effects on cells and can also as a permeability barrier [35,36]. Therefore, the presence of capsules can increase the ability of cells to resist adverse environments. In this study, our SEM results showed that under butanol stress, Y217 cells still exhibited a typical rod-shaped appearance; the cells were also full and intact, without damage. However, after stimulation with butanol, ATCC 824 cells showed folds and depressions on the cell surface. Owing to the uneven extracellular morphology, the cell walls and cell membranes of bacterial cells may have been damaged under butanol pressure. Thus, these ndings indicated that the cell surface barrier played important roles in resisting the impact of organic solvents. According to a previous work [37], when the MATH value is higher than 50%, the cell surface hydrophobicity is strong. In contrast, when the MATH value is between 20% and 50%, the cell surface hydrophobicity is moderate, and when the MATH value is less than 20%, the cell surface becomes hydrophilic. Accordingly, in the context of butanol stress, lower cell surface hydrophobicity was associated with reduced entry of organic solvent in the cells. The mutant strain Y217 could tolerate butanol to a certain extent, possibly because this strain reduced the penetration of butanol into the cell membrane and thereby suppressed the toxic effects of organic solvent.
The cell membrane is a protective barrier for microorganisms. Biological mechanisms actively regulate the dynamic and structural characteristics of the membrane through subtle changes in chemical components and the interactions between components [38]. When microorganisms are in an unfavorable environment or are poisoned by drugs, the protective barrier of the bacteria will be broken, resulting in changes in the uidity and permeability of the cell membrane; large amounts of electrolytes and biomacromolecules can then be released from the cells [39]. Electrolyte leakage can cause changes in the conductivity of the culture uid, and such changes in conductivity and nucleic acid and protein leakage can re ect changes in cell membrane permeability [40]. The conductivity of the solution mainly depends on the number of ions and the speed of movement of the ions [41]. According to the results of our conductivity experiment, after treatment with butanol, intracellular electrolyte leakage occurred, and ion concentration in the supernatant began to increase, resulting in increased conductivity. However, when the ion concentration increases to a certain level, the interaction between ions is enhanced, resulting in reduced speed of ion movement and stable or decreased electrical conductivity. Fluorescein diacetate (FDA) can penetrate the cell membrane to enter the inside of the cell. The esterase produced by living cells can decompose FDA into uorescein to generate uorescence and change its polarity to prevent it from penetrating the cell, allowing the compound to be retained in living cells. When the cell membrane is damaged (but the cell has not died), the permeability of the cell membrane increases, and uorescein ows out of the cytoplasm, resulting in decreased uorescence intensity [42,43]. Therefore, after staining the cells with FDA, the difference in the integrity and permeability of the cell membrane of the original strain ATCC 824 and the mutant Y217 can be determined according to the change in uorescence intensity. In our study, we compared the UV absorbance and conductivity values of the supernatants of the two strains at 260 and 280 nm and showed that as the processing time increased, leakage of intracellular substances, such as nucleic acids and proteins, was lower in Y217 than in ATCC 824. The conductivity of Y217 was relatively stable, whereas the conductivity of ATCC 824 tended to increase rst and then stabilize. Our ndings also showed that butanol could increase the cell membrane permeability of C. acetobutylicum, which may cause cell death, and that Y217, as a butanol-tolerant strain, could resist the damage to cell membrane permeability caused by butanol [40]. When the cells are exposed to butanol stress, butanol can penetrate the cell membrane and enter the interior of the cells, thereby causing changes in the cell membrane. Changes in cell membrane permeability and membrane potential can accurately re ect changes in cell membrane structure [44]. Therefore, FDA and Rh123 can be used to stain the cells and judge the integrity of the cell membrane according to changes in uorescence intensity [45,46]. Our results showed that after addition of butanol, the membrane potential and permeability of ATCC 824 change greatly, indicating that under high butanol stress, ATCC 824 cell membrane damage was greater. In contrast, the membrane potential of Y217 remained stable, indicating that under butanol stress, the Y217 cell membrane showed better integrity. These results are consistent with the results of SEM, indicating that the mutant strain had better ability to resist the impact of butanol. Because the integrity of the cell membrane is essential for maintaining cellular activity, cell death can result when membrane damage is severe. Overall, Y217 showed much better butanol tolerance than ATCC 824.
The uidity of the cell membrane depends on many factors and is a result of interactions of various membrane components. However, its regulation is mainly achieved through changes in fatty acids [47]. Environmental stress causes changes in various characteristics, including increased density of the membrane lipid bilayer. The cell membrane can alter its own composition, called the "compensation effect" [48], to modulate membrane uidity. FAs, as important components of bacterial cell membranes, are highly sensitive to changes in the external environment and are often used as indicators to assess the survival status of microorganisms [49,50]. The membrane lipid structure of the cell membrane is also critical for maintaining the function of the cell membrane and for the growth and reproduction of the cell. The membrane lipid structure is closely related to the FA composition, and changes in the FA composition inevitably affect the relevant membrane protein activity and membrane function [51][52][53]. Generally, the UFA component in the cell membrane is related to the uidity of the cell membrane [54]. If the SFA content is high, the uidity of the cell membrane will be reduced. Changes in cell membrane uidity can alter the functional characteristics of the cell and the induction of apoptosis channels [55]. In our study, the proportion of SFAs in the cell membrane of Y217 was 13% higher than that of the wild-type strain ATCC 824. Compared with UFAs, the chains of SFA are straight and hard, which can increase the rigidity of the cell membrane [56]. When the rigidity of the cell membrane increases, the uidity of the cell membrane will inevitably decrease, thereby reduce the entry of organic solvents into the cell, weakening the adverse effects of solvents, and enhancing the tolerance of the cell to organic solvents [47]. However, these changes in membrane composition in response to stress are rapid responses. The long-term solution is to enhance isomerization, such as altering the composition of the polar head group to enhance phospholipidity agglomeration. Further studies are required to fully elucidate these mechanisms.

Conclusions
In this study, we analyzed and compared the cell morphology, cell surface hydrophobicity, cell membrane permeability, membrane potential, and FA contents of a butanol-tolerant mutant strain and wild-type strain of C. acetobutylicum under butanol stress. Our goals were to elucidate the physiological and metabolic mechanisms related to the adaptation of the C. acetobutylicum cell membrane to a high concentration of butanol and provide a theoretical basis for further improving the ability of C. acetobutylicum to synthesize butanol and tolerate organic solvents. The tolerance of cells to organic solvents is the result of interactions of multiple factors, and it is di cult to explain changes in tolerance based on analysis of only a single factor. Therefore, in the future, systems biology research methods should be used to further explore the solvent resistance of cell and provide a scienti c exploration for the relationships among solvent tolerance, metabolic disturbance law, and butanol production of C. acetobutylicum.

Methods
Strains and cultivation C. acetobutylicum ATCC 824 and Y217 were stored with 40% glycerol at -80 °C in the Biophysics Laboratory of the Institute of Modern Physics at the Chinese Academy of Science, Lanzhou, China. In order to study the relationships of changes in cell membrane permeability, membrane potential, and cell membrane FA contents with butanol tolerance under butanol stress conditions, strains were inoculated into RCM to logarithmic phase (~ 20 h), and butanol was then added to the fermentation broth ( nal concentration: 2.0%). The composition of RCM in distilled water was as follows: beef extract, 10.00 g/L; peptone, 10.00 g/L; yeast extract, 3.00 g/L; glucose, 5.00 g/L; starch, 1.00 g/L; NaCl, 5.00 g/L; sodium acetate, 3.00 g/L; cysteine hydrochloride, 0.50 g/L.

SEM sample preparation
Cells used for SEM were collected from the late logarithmic growth period (~ 24 h), centrifuged at 5000 rpm for 20 min, resuspended, and immobilized with 2.5% glutaraldehyde at 4 °C for 3 h. The cells were centrifuged and washed twice with phosphate buffer, and a series of ethanol concentrations (10%, 30%, 50%, 70%, 90%, and 100%) was used for gradient dehydration (15 min each; 100% dehydration was performed two times). Finally, specimens were dried at room temperature and coated with gold. Samples were observed with an FEI Nova NanoSem 450 (working voltage: 15 kV, ampli cation factor: 80,000).
Detection of cell surface hydrophobicity MATH [37,57] was evaluated as follows. First, cells were cultured at 37 °C for 18-20 h (logarithmic phase) and then transferred to fresh medium (2.0% [v/v] butanol RCM medium and original RCM medium). After 3 h of incubation, cells were collected by centrifugation (4 °C, 5000 rpm), washed with phosphate buffer (pH 7.2), and suspended in the buffer to achieve an OD 600 of 0.6. Next, a 1-mL suspension was removed and added to 300 µL butanol. The mixture was shaken for 2 min and then left to rest for 15 min at room temperature. The butanol phase was removed, and the OD 600 of the water phase was determined. The MATH value was calculated as follows: MATH = ([OD 600 ] pretreatment -[OD 600 ] post-treatment) / (OD 600 ) pretreatment.
Evaluation of membrane permeability and membrane potential FDA ( maximum excitation light wavelength: 490 nm, emission light wavelength: 526 nm; Sigma-Aldrich, St. Louis, MO, USA) was prepared at 5 mg/mL as a stock solution. Rh123 (maximum excitation light wavelength: 507 nm, emission wavelength: 529 nm; Sigma-Aldrich) was prepared at 1 mg/mL as a stock solution. Cells (10 6 -10 7 ) were washed with phosphate buffer and resuspended. The dye was added to the sample solution for appropriate incubation. The uorescence intensity was immediately detected using a multimode reader and ow cytometer. Changes in cell membrane permeability and cell membrane potential in the wild-type and mutant strains in the presence of butanol stress were evaluated by measuring changes in uorescence intensity.

Determination of intracellular biological macromolecular spillage and electrical conductivity
The bacterial suspension of strains stimulated with butanol was centrifuged at 8000 rpm for 15 min. Supernatants were collected, and the ultraviolet absorbance of nucleic acids and proteins was measured at 260 and 280 nm, respectively. Extracellular conductivity was detected with a conductivity meter. Saline was evaluated as a blank. The experiment was repeated three times.

Determination of FA contents
For FA extraction, strains were grown to the logarithmic growth phase and centrifuged at 3000 rpm for 5 min. The supernatants were discoarded, and an inoculating ring was used to pick up 40 mg (wet weight) cells into a clean, dry screw-cap test tube. Saponi cation was then carried out by adding 1.0 mL saponi cation reagent (45 g sodium hydroxide + 150 mL methanol + 150 mL water). The samples were shaken for 5-10 s, placed in boiling water at 95-100 °C for 5 min, cooled at room temperature, incubated in a water bath for 25 min, and then cooled again at room temperature. Methylation was performed by adding 2.0 mL methylation reagent (325 mL of 6 M hydrochloric acid + 275 mL methanol), shaking for 5-10 s, incubating in a water bath at 80 °C for 10 min, and cooling to room temperature with running tap water. For extraction, 1.25 mL of extraction reagent (200 mL n-hexane + 200 mL methyl tert-butyl ether) was added, and samples were gently mixed by rotation for 10 min. The upper organic phase was used for gas phase detection.

Cellular butanol tests and fermentation experiment
For butanol-challenge growth experiments, the seed culture was grown to an OD 600 of 1.0 and inoculated into 10 mL fresh RCM at an inoculation amount of 8%. Cells were then grown anaerobically at 37 °C for 24 h, and samples are colleted every 2 h. Butanol was added at different concentrations in RCM medium (0%, 1%, 2.0%, 2.5%, and 3.0% [v/v]) at an initial pH of 6.5.
The batch fermentation study was carried out in a 250-mL screw capped bottle with 50 mL Clostridium growth medium (CGM). The seed suspension of C. acetobutylicum ATCC 824 or the mutant strain Y217 was anaerobic cultured to the stationary phase at 37 °C in RCM. Screw capped bottles were inoculated with 8% seed culture of the total fermentation culture volume and stationary culture at 37 °C. The composition of CGM in distilled water was as follows: 50 g/L glucose, 2.0 g/L (NH 4 ) 2 SO 4 , 3.0 g/L CaCO 3 , 0.75 g/L K 2 HPO 4 ·3H 2 O, 1.25 g/L KH 2 PO4, 0.71 g/L MgSO 4 ·7H 2 O, 0.01 g/L MnSO 4 ·H 2 O, 0.01 g/L FeSO 4 ·7H 2 O, 1 g/L NaCl, 2 g/L asparagine, and 5 g/L yeast extract. Glucose, FeSO 4 , and asparagine were dissolved independently and added to sterile medium with a 0.22-µm lter. All media were purged with N 2 to remove O 2 to maintain complete anoxic conditions.

Analytical methods
The optical density of the appropriate dilution at 600 nm was detected with an epoch spectrophotometer (Times Legend Bio-Scienti c Co., Ltd., Shanxi, China) to determine bacterial growth. Fluorescence intensity was determined by a multimode reader (In nite 200 PRO; Tecan). The blank was sterile RCM liquid medium. Before analysis of the product concentration, the fermentation broth was analyzed by centrifugation at 8000 rpm for 5 min, and the supernatant was ltered using a a 0.22-µm syringe lter. Solvents (acetone, ethanol, and butanol), acids (acetic acid and butyric acid), and FAs were determined using 456-GC gas chromatography (Bruker, Germany) on a HP-INNOWAX column (30 m × 0.320 mm × 0.50 µm; Agilent J&W, USA) equipped with a ame ionization detector. FA gas chromatographic detection was performed using the heating program of the column oven, as follows: initial temperature, 50 °C; maintained at 50 °C for 1 min; increased at 25 °C/min to 175 °C; increase at 40 °C/min to 230 °C; maintained at 230 °C for 15 min. The injection port temperature was set at 250 °C, the split ratio was 50:1, the detector temperature was 280 °C, the carrier gas was hydrogen (40 mL/min), the air was 400 mL/min (nitrogen, 25 mL/min), the precolumn pressure was 14.00 psi, and the injection volume was 1 µL. Conductivity was detected using a conductivity meter (DDBJ-350; Electronic Scienti c Instrument Co., Ltd., Shanghai, China). Cell membrane potential was measured using a FlowSight ow cytometer (Merck Millipore, Burlington, MA, USA) and analyzed by IDEAS Application software v6.0. For ow cytometry analyses, 20,000 events were recorded per sample.
Statistical analysis SPSS 11.5 software was used for statistical analysis. The experimental data were expressed as means of at least three independent experiments; error bars indicate standard deviations. Means of multiple groups of samples were compared by one-way variance analysis. Pairwise comparisons were performed using least signi cant difference t-tests. Results with P values of less than 0.05 were considered statistically signi cant.   Comparative analysis of membrane potential in the wild-type strain (ATCC 824) and mutant strain (Y217). Histograms (A, B) and signi cance analysis (C, D) are shown. *P < 0.05, **P < 0.01.

Figure 5
Effects of butanol stress on membrane fatty acid composition in ATCC 824 and Y217 strains. Relative content composition of fatty acids (A), ratio of SFAs and UFAs to total FAs (B), ratio of SFAs to UFAs (C), and ratio of MUFAs (D) and polyunsaturated FAs (PUFA) (E) to TFAs. Cn1: n2, where n1 represents the number of carbon atoms, and n2 represents the number of double bonds. *P < 0.05, **P < 0.01.