Obviously, the broth ORP enhances fermentability of the detoxicated SECS by affecting the cells' metabolism. To better understand the working mechanism, genome scale metabolic flux analysis (MFA) was performed to compare the flux profiles of C. acetobutylicum cells under three different culture conditions: the synthetized medium group (SG), detoxicated SECS medium with ORP control at -350mV (OCG) and without ORP control (UCG). The analysis result was given in Fig. 4.
3.3.1 ORP regulation triggers metabolic flux redistribution in C. acetobutylicum ATCC 824
Figure 4 compared the difference of flux distributions at the typical time point of 36 h (cell growth in acidogenesis phase) and 60h (solventogenesis phase) among the three groups. Figure 4A showed the fitted result of biomass of SG with Boltzmann model for the specific growth rate calculation. The other two groups and the detailed calculation process together with the calculation code were given in Appendix E. The genome scale metabolic model used in our study consists of 432 genes, 502 reactions and 479 metabolites. Calculation was carried out using FBA constrained by experimental data (Li et al., 2021b). To assess the diversity among the three groups with regards to metabolic flux distribution, principal component analysis (PCA) was carried out (Fig. 4D). We can see that samples from the three groups can be clearly separated, indicating that the metabolic characteristic differences of C. acetobutylicum ATCC 824 under different fermentation.
Figure 4B showed the topology structure of the metabolic networks, which only preserves the nodes with significant differences calculated by p test lower than 0.05 among the three groups at 36 h. The substrates with the most significant differences are succinyl-CoA (No.62, KEGG: C00091), pyruvate (No.55, KEGG:C00022) and N-Acetyl-L-glutamate 5-semialdehyde (NO. 6, KEGG:C01250). Pyruvate and succinyl-CoA are among the 12 basic biosynthetic precursor compounds that are used to build macromolecules such as nucleic acids and proteins (Orth, 2012). Their differences may be related to cell growth. N-acetyl-l-glutamate 5-semialdehyde is one of the essential precursors for arginine synthesis (Module ID: M00028 in KEGG database). Previous studies have shown that the synthesis of arginine is an energetically expensive process. The cell has to supply high amounts of ATP for this process. Therefore, N-acetyl-l-glutamate 5-semialdehyde might be closed with the differences in the availability of ATP (Korneli et al., 2012; Xia et al., 2015b). However, it is difficult to determine the potential bottlenecks limiting the production based on such information. Hence, it seems highly important to get further insight into the underpinning metabolism. Since ORP directly affects intracellular electron transfer and redox balance involved in intracellular metabolism. The analysis on NADPH, NADH and ATP was carried out
Figure 4C summarized the major redox reactions in acetone-butanol-ethanol fermentation by the bacterium Clostridium (Dai et al., 2021). Figure 4F, Table 3 and Table 4 exhibited the flux distributions among reactions of intracellular force reduction and energy metabolism under the three conditions. In our model, there are 23 reactions involved NADPH, 16 reactions involved NADH and 61 reactions involved ATP. The metabolism of NADPH in the three groups is given in Table 3. At 36h, UCG owned the most active metabolism of NADPH with a total NADPH flux as high as 11.44 mmol/g/h, which is 2.1 and 3.26-fold of SG and OCG. The contributions of each reaction in SG and OCG were similar, except the total flux. In these two groups, respiratory chain was the main source of NADPH, accounting for about 97.7%. Then folate (CA_C2083) and riboflavin synthesis (CA_C0590) account for about 3.09%. However, in the UCG group, the folate synthesis pathway consumed 19.32% of NADPH instead of generation, suggesting the cells in this group consume a large amount of folic acid. This is very reasonable since folic acid is widely involved in the metabolism of cofactors and plays an important role in microbial resistance to stress environment. Compared with 36h, the main source of NADPH at 60h were respiratory chain (EMP19) and TCA cycle (PYR3). Amino acid metabolism (CA_C0510) and carbohydrate metabolism (SULFUR5, CA_C2390) at 60 h tended to decrease. The reaction of fatty acids (CA_C1589, CA_C0764) and COA synthesis (CA_C3254) was significantly enhanced.
Table 3
Simulation of metabolic flux of the NADPH generation/consumption reaction of C. acetobutylicum ATCC 824 strain under the different conditions. The unit of the flux was mmol/g/h. SG: the synthetized medium group, OCG: detoxicated SECS medium with ORP control at -350mV, UCG: detoxicated SECS medium without ORP control (UCG). The abbreviations of the reactions were provided in Appendix A.
Reaction | 36 h | 60 h |
SG | | OGG | | UCG | | SG | | OGG | | UCG |
Flux | Percentage | | Flux | Percentage | | Flux | Percentage | | Flux | Percentage | | Flux | Percentage | | Flux | Percentage |
EMP19 | 5.28 | 97.91 | | 3.43 | 97.82 | | 11.44 | 99.97 | | 9.30 | 94.99 | | 7.95 | 95.68 | | 5.28 | 95.52 |
AMSU8 | -0.03 | -0.63 | | -0.02 | -0.55 | | -0.02 | -0.14 | | 0 | 0 | | 0 | 0 | | 0 | 0 |
NITROGEN6 | -1.95 | -36.19 | | -1.25 | -35.83 | | -3.03 | -26.51 | | 0 | 0 | | 0 | 0 | | 0 | 0 |
SULFUR5 | -0.64 | -11.78 | | -0.45 | -12.72 | | -0.31 | -2.70 | | 0 | 0 | | 0 | 0 | | 0 | 0 |
GST2 | -0.33 | -6.09 | | -0.21 | -5.88 | | -2.42 | -21.16 | | -0.12 | -1.26 | | -0.11 | -1.38 | | -0.08 | -1.43 |
GST3 | -0.23 | -4.34 | | -0.15 | -4.20 | | -2.38 | -20.76 | | -0.05 | -0.56 | | -0.05 | -0.61 | | -0.04 | -0.63 |
VLI3 | -0.07 | -1.35 | | -0.05 | -1.40 | | -0.04 | -0.31 | | -0.04 | -0.39 | | -0.04 | -0.43 | | -0.02 | -0.44 |
VLI7 | -0.28 | -5.12 | | -0.17 | -4.83 | | -0.13 | -1.17 | | -0.15 | -1.49 | | -0.14 | -1.63 | | -0.09 | -1.69 |
LYS2 | -0.09 | -1.75 | | -0.06 | -1.65 | | -0.05 | -0.40 | | -0.05 | -0.51 | | -0.05 | -0.56 | | -0.03 | -0.58 |
PRO4 | -0.08 | -1.41 | | -0.05 | -1.33 | | -0.04 | -0.32 | | 0.01 | 0.12 | | 0.01 | 0.13 | | 0.01 | 0.14 |
PTT4 | -0.17 | -3.18 | | -0.11 | -3.11 | | -0.08 | -0.73 | | -0.09 | -0.93 | | -0.08 | -1.01 | | -0.06 | -1.05 |
UREA3 | -0.1 | -1.82 | | -0.06 | -1.65 | | -0.05 | -0.42 | | 0 | 0 | | 0 | 0 | | 0 | 0 |
PYRM16 | -0.24 | -4.42 | | -0.16 | -4.43 | | -0.12 | -1.01 | | 0 | 0 | | 0 | 0 | | 0 | 0 |
PL7 | -0.2 | -3.76 | | -0.13 | -3.68 | | -0.10 | -0.86 | | 0.10 | 1.03 | | 0.09 | 1.12 | | 0.06 | 1.16 |
FAS3 | -0.06 | -1.1 | | -0.04 | -1.07 | | -0.03 | -0.25 | | -0.22 | -2.29 | | -0.21 | -2.50 | | -0.14 | -2.59 |
FAS4 | -0.49 | -9.16 | | -0.32 | -9.14 | | -0.24 | -2.10 | | -0.04 | -0.45 | | -0.04 | -0.49 | | -0.03 | -0.50 |
FAS5 | -0.08 | -1.42 | | -0.05 | -1.49 | | -0.04 | -0.33 | | -0.02 | -0.17 | | -0.02 | -0.19 | | -0.01 | -0.19 |
FAS6 | -0.04 | -0.72 | | -0.03 | -0.82 | | -0.02 | -0.17 | | -0.16 | -1.66 | | -0.15 | -1.81 | | -0.10 | -1.87 |
FAS7 | -0.29 | -5.33 | | -0.20 | -5.70 | | -0.14 | -1.22 | | -0.01 | -0.12 | | -0.01 | -0.13 | | -0.01 | -0.14 |
PANCOA2 | -0.01 | -0.17 | | -0.01 | -0.27 | | 0.00 | -0.04 | | 0.00 | -0.05 | | 0.00 | -0.06 | | 0.00 | -0.06 |
RIBFLA6 | 0.01 | 0.14 | | 0.01 | 0.28 | | 0.00 | 0.03 | | 0.00 | 0.04 | | 0.00 | 0.05 | | 0.00 | 0.05 |
FOLATE13 | -0.01 | -0.26 | | -0.01 | -0.26 | | -0.01 | -0.06 | | 0 | 0 | | 0 | 0 | | 0 | 0 |
FOLATE18 | 0.11 | 1.95 | | 0.07 | 1.90 | | -2.21 | -19.32 | | 0.01 | 0.12 | | 0.01 | 0.13 | | 0.01 | 0.13 |
FOLATE12 | 0 | 0 | | 0 | 0 | | 0 | 0 | | -0.01 | -0.08 | | -0.01 | -0.08 | | 0.00 | -0.09 |
NITROGEN4 | 0 | 0 | | 0 | 0 | | 0 | 0 | | -0.01 | -0.14 | | -0.01 | -0.15 | | -0.01 | -0.16 |
PRO2 | 0 | 0 | | 0 | 0 | | 0 | 0 | | -0.04 | -0.41 | | -0.04 | -0.45 | | -0.03 | -0.46 |
PYR3 | 0 | 0 | | 0 | 0 | | 0 | 0 | | 0.26 | 2.65 | | 0.24 | 2.90 | | 0.17 | 3.00 |
Table 4
Simulation of metabolic flux of the NADH generation/consumption reaction of C. acetobutylicum ATCC 824 strain under the different conditions. The unit of the flux was mmol/g/h. SG :the synthetized medium group, OCG : detoxicated SECS medium with ORP control at -350mV, UCG: detoxicated SECS medium without ORP control (UCG).
Reaction | 36 h | | 60 h |
SG | | OGG | | UCG | | SG | | OGG | | UCG | |
Flux | Percentage | | Flux | Percentage | | Flux | Percentage | | Flux | Percentage | | Flux | Percentage | | Flux | Percentage | |
EMP10 | 19.01 | 94.47 | | 12.43 | 94.81 | | 14.42 | 78.25 | | 12.35 | 74.56 | | 9.84 | 76.88 | | 6.99 | 75.38 | |
EMP18 | 0.41 | 2.03 | | 0.45 | 3.4 | | 2.48 | 13.44 | | 3.39 | 20.48 | | 2.63 | 20.53 | | 1.94 | 20.88 | |
BUTAN1 | -0.26 | -1.28 | | -1.03 | -8 | | 1.37 | 7.41 | | -0.36 | -2.19 | | -0.35 | -2.75 | | -0.32 | -3.43 | |
BUTAN2 | -0.22 | -1.09 | | -1.01 | -7.81 | | -0.88 | -4.76 | | -0.35 | -2.10 | | -0.84 | -6.53 | | -0.41 | -4.38 | |
BUTAN6 | -4.13 | -20.54 | | -3.11 | -24.08 | | -5.44 | -29.55 | | -3.95 | -23.84 | | -2.83 | -22.13 | | -2.08 | -22.40 | |
BUTAN8 | -8.26 | -41.08 | | -6.21 | -48.17 | | -10.89 | -59.10 | | -6.80 | -41.03 | | -4.70 | -36.72 | | -3.57 | -38.52 | |
BUTAN11 | -3.52 | -17.49 | | -0.7 | -5.44 | | -0.56 | -3.01 | | -0.55 | -3.32 | | -0.48 | -3.78 | | -0.29 | -3.14 | |
BUTAN12 | -3.52 | -17.49 | | -0.7 | -5.44 | | -0.56 | -3.01 | | -0.15 | -0.88 | | -0.63 | -4.96 | | -0.20 | -2.11 | |
TCA2 | -0.07 | -0.33 | | -0.04 | -0.33 | | -0.03 | -0.17 | | 0.16 | 0.94 | | 0.00 | 0.02 | | 0.02 | 0.25 | |
VLI12 | 0.07 | 0.36 | | 0.03 | 0.36 | | 0.03 | 0.19 | | 0.04 | 0.23 | | 0.04 | 0.27 | | 0.02 | 0.26 | |
HIS9 | 0.02 | 0.12 | | 0.02 | 0.12 | | 0.01 | 0.06 | | 0.01 | 0.08 | | 0.01 | 0.09 | | 0.01 | 0.09 | |
HIS10 | 0.02 | 0.12 | | 0.02 | 0.12 | | 0.01 | 0.06 | | 0.00 | 0.00 | | 0.00 | 0.00 | | 0.00 | 0.00 | |
PTT18 | 0.13 | 0.66 | | 0.07 | 0.68 | | 0.06 | 0.35 | | 0.00 | 0.00 | | 0.00 | 0.00 | | 0.00 | 0.00 | |
PUR27 | 0.03 | 0.17 | | 0.02 | 0.17 | | 0.02 | 0.09 | | 0.15 | 0.91 | | 0.02 | 0.15 | | 0.05 | 0.59 | |
PYRM4 | 0.05 | 0.26 | | 0.02 | 0.26 | | 0.03 | 0.14 | | 0.03 | 0.17 | | 0.03 | 0.20 | | 0.02 | 0.19 | |
FOLATE16 | -0.14 | -0.72 | | -0.09 | -0.73 | | -0.07 | -0.38 | | 0.06 | 0.34 | | 0.05 | 0.40 | | 0.04 | 0.38 | |
EMP16 | 0 | 0 | | 0 | 0 | | 0 | 0 | | -9.29 | -56.09 | | -7.99 | -62.41 | | -5.42 | -58.43 | |
GST9 | 0 | 0 | | 0 | 0 | | 0 | 0 | | 0.38 | 2.28 | | 0.18 | 1.41 | | 0.18 | 1.95 | |
LIMPIN3 | 0 | 0 | | 0 | 0 | | 0 | 0 | | 0.00 | 0.03 | | 0.00 | 0.03 | | 0.00 | 0.03 | |
PL6 | 0 | 0 | | 0 | 0 | | 0 | 0 | | -0.11 | -0.65 | | -0.10 | -0.78 | | -0.07 | -0.74 | |
TCA7 | 0 | 0 | | 0 | 0 | | 0 | 0 | | -0.10 | -0.58 | | -0.09 | -0.69 | | -0.06 | -0.66 | |
Table 4 shows the metabolism of NADH in the three groups. At 36h, the NADH fluxes of the three groups were 19.74, 12.66 and 18.43 mmol/g/h respectively. In SG and OCG, the contribution of EMP pathway to NADH was 94%, while that in UCG was only 78.25%. On the contrary, the respiratory chain intensity of the latter group was 6.05 times and 5.51 times that of the first two groups, respectively. It is suggested that the inhibitor decreased the metabolic intensity of EMP. As a compensation mechanism, cells enhanced the metabolic intensity of the respiratory chain. In the UCG, reaction Butan1 (acetaldehyde → acetyl coenzyme A) and reaction Butan8 (acetyl coenzyme A →3-hydroxybutyryl COA) were significantly improved. This change can decrease the synthesis of butyryl COA and ethanol, so as to saving the NADH consumption. At 60H, not surprisingly, SG and OCG owned high flux in almost all the NADH-involved reactions. The total NADH flux of the three groups was 16.57, 12.80 and 9.27 mmol/g/h respectively. This may explain the reason for the high yield of solvents in the first two groups.
Table 5 shows the metabolism of ATP in the three groups. In the genome scale model, there are 61 reactions involved in ATP metabolism, accounting for 12.15% of the total reactions, indicating that ATP metabolism has a very wide impact on cell metabolism. The total ATP flux of the three groups were 44.91, 35.89 and 29.11 mmol/g/h respectively. ATP is mainly used for bacterial synthesis at 36 h, accounting for 42.32% in SG, 42.72% in OCG and 40.16% in UCG of the total ATP respectively (seen Appendix H). Compared with UCG, four reactions in OCG has significantly enhanced, which were FOLATE19(CA_C3201/[EC:6.3.4.3])increased by 159 times, GST4 (CA_C1235/ [EC:2.7.1.39]) increased by 116 times, GST1 (CA_C0278/ [EC:2.7.2.4] or CA_C1810/ [EC:2.7.2.4]) increased by 11 times and TCA1 (CA_C2660/ [EC:6.4.1.1]) increased by 6.24 times. FOLATE19༈CA_C3201/ [EC:6.3.4.3]) is the synthesis of 10-Formyltetrahydrofolate. This substance is the precursor of many cofactors, suggesting, again, ORP induced high-speed synthesis of cofactors of C. acetobutyricum. GST4 and GST1 represent the reactions catalyzed by homoserine kinase [EC:2.7.1.39] and aspartate kinase [EC:2.7.2.4] respectively. They are the key enzymes in aspartate metabolic pathway, which controls the biosynthesis of lysine, methionine, threonine and isoleucine. TCA1 is the reaction catalyzed by pyruvate carboxylase [EC:6.4.1.1]. As the key enzyme of oxaloacetate replenishment pathway in bacteria, it serves as the gate of carbon flow into TCA cycle. In other words, folate, amino acid and TCA cycle of cells were greatly improved under the controlling of oxidoreduction potential. At 60 h the fluxes of four reactions in UCG were at a higher level. it showed 3.23 times higher of PUR17 (CA_C3112/ EC: 2.7.4.3), 1.67 times higher of PANCOA4 (CA_C3200/ [EC:2.7.1.33]), 1.66 times higher of BUTAN10 (CA_C3075/ [EC:2.7.2.7]) and 1.53 times higher of BUTAN4 (CA_C1743/ [EC:2.7.2.1]) of that in OCG. PUR17 [EC: 2.7.4.3] represents the conversion reaction of ATP to ADP, indicating that UCG has higher ADP generation rate. PANCOA4 [EC:2.7.1.33] is a key enzyme that catalyzes COA synthesis. BUTAN4[EC:2.7.2.1] and BUTAN10[EC:2.7.2.7] are the key enzymes that catalyze butyryl phosphate to butyric acid and butanoyl-COA to butyryl phosphate, respectively. These four key enzymes formed a reaction circuit resulting of acid accumulation. The illustration of the metabolic circuit is given in Fig. 4E. The circuit includes three parts: I༉PANCOA4 and PUR17 supply COA and ADP respectively. II) COA and ADP are catalyzed by Butan4 [EC: 2.7.2.1] and butan10 [EC: 2.7.2.7] for butyric acid and acetic acid biosynthesis. III) Under the catalysis of butyryl-CoA-acetoacetate CoA-transferase (EC: 2.8.3.9), acetic acid can capture the CoA group from butanoyl-CoA and convert the latter into butyric acid. The resulted acetoacetyl-CoA can further convert into butyric acid in this circuit. Wang et al. (2011) and Maddox et al. (2000) studied the cause of “acid crash” by adding acid to the culture medium. For the first time, we found a new metabolism circuit that may cause butyric acid accumulation by metabolic pathway analysis method.
3.3.2 ORP regulation changes intercellular redox status
Metabolic flux reflects the instantaneous change in cell metabolism. In order to further confirm the real state of cells, we measured the key metabolites, including ATP concentration, NADH/NAD+ and NADPH/NADP+, within C. acetobutylicum ATCC 824 from the three groups during the solvent-producing phase. As shown in Fig. 5a, all the factors measured in SG kept the highest level among the three groups, followed by those in OCG. Taking the time point of 60 h as example (when the butanol biosynthesis rate was the highest at this point), the ATP concentration in SG is 1.2-fold and 1.5-fold of that in OCG and UCG, respectively; the NADH/NAD + ratio in SG is 2.2-fold of that in OCG and 4.0-fold of that in UCG; and the NADPH/NADP+ ratio in SG is 1.4-fold of that in OCG and 2.3-fold of that in UCG correspondingly. High energy and reducing power availability form one of the main reasons for high butanol production in SG and OCG. Meanwhile, we also detected the activities of key enzymes in the butanol biosynthesis and the result was shown in Fig. 5b. The high activities of butyraldehyde and butanol dehydrogenase in SG explained its high butanol production in the fermentation. Comparing with the ORP uncontrolled group, the activities of butyraldehyde and butanol dehydrogenase in OCG were increased by 2.1-fold and 1.2-fold. Meanwhile, the phosphotransbutyrylase activity was decreased by 29%, indicating that ORP controlling shift more butyryl-CoA towards butanol biosynthesis at the expense of butyrate. It is interesting to note that phosphotransbutyrylase in UCG kept at a stable level, suggesting butyrate was produced continuously during the whole process. This result is quite consistent with our above analysis
3.3.3 ORP regulation changes cell membrane integrity
A major drawback of solvent production by microorganisms is the toxic effect of the alcohols, especially n-butanol, on the cells themselves (Alsaker et al., 2010). It is well known that n-butanol influences the lipid composition, fluidity, and the potential of the membrane and is able to interrupt the phospholipid bilayer of the cell. This toxic effect severely limits the butanol production, even causing cell autolysis (Janssen et al., 2012). Therefore, the ability of cells to withstand the accumulation of toxic products without loss of productivity is a most significant goal (Tomas et al., 2004). Studies showed that the cell's resistance ability depends on a unique mechanism, namely homeoviscous adaptation, which is that clostridia increases saturated fatty acid chain content to modulate their membrane fluidity in the presence of solvents (Alsaker et al., 2010). As key factors for fatty acid and amino acids biosynthesis (Akashi & Gojobori, 2002; Alsaker et al., 2010; Xia et al., 2015a; Xia et al., 2013), it is reasonable to presume that the enhanced ATP, NDPH and NADH availability by ORP controlling can improve the tolerance ability of C. acetobutylicum ATCC 824. To verify our prediction, cell membrane integrity in the three groups were compared and the result was given in Fig. 6.
Fluorescein diacetate (FDA) is a cell-permeant esterase substrate. As Fig. 5a shown, when absorbed, it can be converted into green fluorescent compound "fluorescin" by esterase in the living cell, which can be detected by measuring the fluorescence or absorbance of the sample (Clarke et al., 2001). When the cell membrane was damaged and the permeability increased, FAD will leak from the cell, insulting the decrease of fluorescence intensity. Therefore, it can be used to judge the cell membrane integrity by the change of fluorescence intensity (Chand et al., 1994). In the comparison experiments, butanol was supplemented into the three groups (SG, OCG, and UCG) to keep at a final concentration of 20 g/L and samples were withdrawn very 2 h for cell membrane detection. The result was given in Fig. 6.
We can see that the fluorescence intensity dropped quickly during the first 6 h, with a percentage of 35.1% in SG, 61.7% in OCG and 77.3% in UCG, suggesting there exists differences on the cell membrane integrity among the three groups. Comparing with UCG, the cell membrane integrity was increased by 15.6% in OCG. It also can be found that the decrease of fluorescence intensity trended smaller with processing time increased, especially after 6 h, indicating cell had trigger homeoviscous adaptation to resistant the toxic effect of butanol. Therefore, by experiment, we obtain direct evidence that ORP controlling can increased the cell membrane integrity of C. acetobutylicum. It contributed to the enhancement of butanol production under ORP controlling.