Engineering Clostridium Tyrobutyricum for High Butanol Production through Induction Expression of Exogenous NADPH-Dependent HBD

Abstract

Clostridium tyrobutyricum Δcat1::adhE2 is a promising cell factory for butanol production because of its robustness, high butanol tolerance, and minimal butyrate production. However, excessive acetate and ethanol production remains a major bottleneck limiting its butanol yield. Coexpressing an exogenous hbd(Ck) encoding the NADPH-dependent 3-hydroxybutyryl-CoA dehydrogenase (HBD) from Clostridium kluyveri with adhE2 could increase the C4 carbon flux, resulting in increased butanol and decreased acetate and ethanol production. However, constitutively overexpressing hbd(Ck) in Δcat1::adhE2 shows little improvement in butanol yield, productivity, and selectivity, which might be caused by redox imbalance and growth inhibition. To alleviate this problem, C. tyrobutyricum MΔcat1::adhE2-Pbgal-hbd(Ck) with a dynamic expression of hbd(Ck) controlled by an inducible promoter was developed. In serum bottle fermentation at 37 °C, when the hbd(Ck) expression was induced at 12 h or in the early exponential phase, butanol production increased ∼20% in yield (from 0.22 to 0.27 g/g glucose), 87.5% in productivity (from 0.16 to 0.30 g/L·h), and 52% in selectivity (from 0.46 to 0.70 g/g total products) compared to the control strain without expressing any hbd(Ck), whereas hbd(Ck) expression induced at 0 or 24 h in MΔcat1::adhE2-Pbgal-hbd(Ck) or constitutively in MΔcat1::adhE2-Pcat1-hbd(Ck) showed significantly lower butanol yield and productivity. At 25 °C, MΔcat1::adhE2-Pbgal-hbd(Ck) with 12 h induction produced the highest butanol titer of 23 g/L with 0.32 g/g yield, 0.16 g/L·h productivity, and 0.83 g/g product selectivity due to much reduced acetate formation. Subsequent scale-up to a stirred-tank bioreactor at 37 °C increased productivity to 0.39 g/L·h while also achieving high butanol titer (21.8 g/L), yield (0.30 g/g), and selectivity (0.67 g/g). The optimized induction timing resulted in a balanced NAD­(P)H pool, effectively channeling substrates toward butanol biosynthesis. It was concluded that the timing for hbd(Ck) expression was critical as it affected glucose catabolism, cell growth, redox balance, and carbon flux distribution. These findings underscore the potential of dynamic metabolic regulation to overcome bottlenecks in biobutanol production, providing a scalable and economically viable bioprocess for industrial application.

1. Introduction

Biobutanol is an important industrial solvent and a potential drop-in biofuel for automobiles. , Traditional solventogenic clostridia naturally produce butanol via the well-known ABE fermentation pathway. − Despite their natural ability to produce solvents, solventogenic clostridia give relatively low butanol yield, productivity, and titer due to the coproduction of acetone and ethanol and poor butanol tolerance, which restrict their industrial applications. − The two-phase ABE fermentation by solventogenic clostridia is also difficult to control for long-term operation in a (semi)­continuous process , because of spontaneous sporulation and culture degeneration. − More recently, Clostridium tyrobutyricum was engineered to overexpress Clostridium acetobutylicum adhE2 encoding the bifunctional aldehyde/alcohol dehydrogenase (AAD) for converting butyryl-CoA to butyraldehyde and then to butanol and has attracted considerable attention as a promising bacterial platform for butanol production due to its high butanol tolerance. − While butyrate and acetate are two major byproducts limiting butanol yield in fermentation by the engineered C. tyrobutyricum overexpressing adhE2, further metabolic engineering and process optimization allowed it to produce butanol at high titer, rate (productivity), and yield (TRY) (15–20 g/L, ∼0.3 g/L·h, and ∼0.3 g/g, respectively) from glucose and biomass hydrolysate sugars. − Furthermore, genome-engineered C. tyrobutyricum Δcat1::adhE2, which was developed using CRISPR-Cas technology for adhE2 insertion on the genome with simultaneous knockout of cat1 encoding the butyryl-CoA/acetate CoA transferase responsible for the biosynthesis of butyric acid, produced butanol at a yield of 0.23 g/g. , Although little butyrate was produced by Δcat1::adhE2, the coproduction of acetate and ethanol in large amounts limited butanol yield to a level far below the theoretical 0.41 g/g glucose. To date, it remains a challenge to engineer C. tyrobutyricum with butanol as the primary or sole product without much coproduction of acetate and butyrate.

In solventogenic clostridia, HBD converts acetoacetyl-CoA to 3-hydroxybutyryl-CoA with NADH as the redox cofactor, which could limit the flux toward C4 metabolites and butanol biosynthesis. Overexpressing an exogenous hbd(Ck) encoding the NADPH-dependent HBD from Clostridium kluyveri was shown to increase butanol biosynthesis in C. acetobutylicum and engineered Clostridium cellulovorans and C. tyrobutyricum overexpressing adhE2. Compared to the strains expressing only adhE2, coexpressing adhE2 and hbd(Ck) in C. tyrobutyricum WT and Ack strains produced significantly more butanol due to increased C4 flux and reducing equivalents. However, overexpressing hbd(Ck) in Δcat1:adhE2 did not show any significant effects on flux redistribution and butanol biosynthesis. The intracellular concentrations of NAD­(P)­H/NAD­(P)⁺ play a critical role in regulating bacterial growth and metabolism. Overexpressing NADPH-dependent HBD in C. tyrobutyricum Δcat1::adhE2 might disrupt its NADH/NAD⁺ homeostasis, resulting in inefficient glycolysis and poor cell growth and metabolic activities. Dynamic metabolic regulation through inducible gene expression systems to control temporal expression of metabolic enzymes and optimize their activity to specific growth stages or fermentation phases may resolve such limitations and alleviate metabolic burden and redox imbalance, resulting in improved product titer, yield, and productivity. ,

In this study, we investigated the use of an inducible promoter Pbgal for the dynamic expression of hbd(Ck) in C. tyrobutyricum MΔcat1::adhE2. The lactose inducible promoter Pbgal was chosen as it has a high dynamic range of 40-fold change in the gene expression level induced with lactose at ∼ 40 mM. Other inducible expression systems, including IPTG and tetracycline inducible promoters, have also been explored for various Clostridium spp., but their compatibility and efficacy with C. tyrobutyricum have not been verified. We aimed to optimize butanol production by controlling the expression of hbd(Ck) in C. tyrobutyricum MΔcat1::adhE2-Pbgal-hbd(Ck) at varying induction times (0, 12, and 24 h) during batch fermentation. Additionally, we evaluated the impacts of methyl viologen (MV) and low-temperature culture for their effects on maximizing metabolic flux toward butanol biosynthesis. Finally, we scaled up the batch fermentation with C. tyrobutyricum MΔcat1::adhE2-Pbgal-hbd(Ck) in a stirred-tank bioreactor with hbd(Ck) expression induced in the early exponential phase, which provided a balanced NAD­(P)H pool to effectively channel substrates toward butanol biosynthesis without impeding cell growth and achieved high butanol titer, productivity, and yield (21.8 g/L, 0.39 g/L·h, and 0.30 g/g, respectively) suitable for industrial biobutanol production. This study contributed to the development of next-generation recombinant clostridia strains for efficient, high-yield, and economically viable biobutanol production.

2. Results

2.1. Inducible Expression of hbd­(Ck) at Different times in Batch Fermentation

Previous studies demonstrated that overexpressing hbd(Ck) increased the carbon flux from acetyl-CoA to butyryl-CoA and butanol production in C. tyrobutyricum strains WT-adhE2, Ack-adhE2 and ΔhydA-adhE2 but not in Δcat1::adhE2. Compared to MΔcat1::adhE2, MΔcat1::adhE2-Pcat1-hbd(Ck) overexpressing hbd(Ck) using the constitutive cat1 promoter showed slightly slower glucose consumption and butanol production with minimal effect on acetate production (Figure A,B). Constitutive expression of NADPH-dependent HBD likely perturbed the redox balance by redirecting NADPH for 3-hydroxybutyryl-CoA formation from biosynthesis/anabolism for cell growth. We thus hypothesized that overexpressing the NADPH-dependent HBD from the beginning of the batch culture might have limited NAD⁺ availability for glycolysis, depleted the NADPH pool needed for cell growth, and caused NADH/NADPH imbalance that resulted in decreased cell growth and butanol production in Δcat1::adhE2-hbd(Ck). To address this issue, we constructed the strain MΔcat1::adhE2-Pbgal-hbd(Ck) with the lactose-inducible promoter Pbgal for controlling the expression of hbd(Ck) and investigated the effect of induction timing of hbd(Ck) expression on butanol production in batch fermentation. As can be seen in Figure C–E, hbd(Ck) induction timing showed significant effects on butanol and acetate production. Induction at 12 h gave the highest butanol production (12.6 g/L) with significantly reduced acetate production (5.1 g/L), while induction at 0 or 24 h showed negligible benefits in reducing acetate or increasing butanol production. Adding lactose at 0 h to induce hbd(Ck) expression immediately after inoculation led to a large amount of acetate produced (11.3 g/L) that surpassed butanol production (10.5 g/L) and resulted in low butanol yield (0.19 ± 0.01 g/g) and productivity (0.18 ± 0.02 g/L·h), which were comparable to or somewhat lower than those from the strains without hbd(Ck) expression or with constitutive hbd(Ck) expression (Figure F). On the other hand, induction at 24 h showed minimal effect on reducing acetate production and produced butanol at a comparable or slightly lower amount than those from Δcat1::adhE2 and Δcat1::adhE2-Pcat1-hbd(Ck). In contrast, induction at 12 h significantly reduced acetate and increased butanol production, resulting in a significantly higher butanol yield (0.23 ± 0.01 g/g) and productivity (0.26 ± 0.01 g/L·h). The impact of the optimized induction time (12 h) for hbd(Ck) expression on flux distribution was also evident from the dramatic increase (∼100%) in the ratios of alcohol to acid [(butanol + ethanol)/(acetate + butyrate)] and C4/C2 [(butanol + butyrate)/(ethanol + acetate)], which increased to ∼2.4 g/g from ∼1.3 g/g (Figure F). Consequently, the product selectivity (butanol/all metabolic products) increased to >0.7 (vs < 0.5 for other conditions). The higher butanol yield, productivity, and selectivity for fermentation with lactose induction at 12 h implied that more NADH was used for butanol biosynthesis (Figure G), which was estimated from the product yields using the stoichiometric equations given in Table S2. It is noted that significant 2,3-BDO production (∼2 g/L) was observed in these batch fermentations. However, the fermentation with hbd(Ck) induction at 12 h produced significantly less amounts of 2,3-BDO, which was eventually converted to butanol by the end of the batch fermentation (Figure D). 2,3-BDO formation could function as an overflow/redox-balancing sink from pyruvate under certain redox and flux states. However, the exact mechanism is not clear and requires further studies.

1.

Batch fermentation kinetics of C. tyrobutyricum MΔcat1::adhE2 (A), MΔcat1::adhE2-Pcat1-hbd(Ck) (B), and MΔcat1::adhE2-Pbgal-hbd(Ck) with hbd expression induced by lactose at 0 h (C), 12 h (D), and 24 h (E), respectively, in serum bottles; the butanol yield, productivity, selectivity, and alcohol/acid and C4/C2 ratios (F); NADH consumption associated with butanol and 2,3-BDO biosynthesis per mol of glucose consumed in the fermentations (G); and the relative transcriptional (mRNA) levels of hbd(Ck) in cells at 36 h in the batch cultures (H). Butanol productivity was based on data at 48 h, selectivity and C4/C2 and alcohol/acid ratios were based on the end points, butanol yield was estimated from the linear regression of butanol vs glucose in the entire batch fermentation, and NADH consumptions (mol/mol glucose consumed) associated with each product biosynthesis was calculated from the product yield using the stoichiometric equations given in Table S2. Relative hbd(Ck) transcription levels were determined in qRT-PCR with typA as the internal reference gene. All experiments were conducted in biological duplicate with the average and standard deviation (error bar) reported in the figures. Trace amounts of ethanol (<0.5 g/L) were also detected in GC analysis but are not included in the fermentation time course figures. Their impacts on the fermentation flux calculation would be small and can be neglected.

Finally, the relative hbd(Ck) transcription levels in cells overexpressing hbd(Ck) (vs control without the overexpressing hbd(Ck) gene) were estimated using qRT-PCR with typA as the internal reference gene. The results confirmed robust hbd(Ck) transcription in cells with 30- to 40-fold higher than the nonexpression control (MΔcat1::adhE2) (Figure H). The Pbgal promoter induced with 40 mM lactose at 0 and 12 h gave mRNA transcription levels similar to that from the constitutive Pcat1 promoter, confirming that the improved phenotype of MΔcat1::adhE2-Pbgal-hbd(Ck) was attributable to optimized induction timing rather than a different hbd(Ck) expression level. It should be noted that Pbgal induction with lactose at 20, 40, and 80 mM, respectively, gave similar fermentation performance, although the hbd(Ck) mRNA expression level increased with increasing lactose concentration (Figure S3). Therefore, 40 mM lactose was used in all of the subsequent experiments in this study.

It can be concluded that optimal induction timing (at 12 h) allowed cells to use NAD⁺ for glycolysis and NADPH for anabolism in the exponential phase and then upon inducer addition to use NADPH for the HBD reaction and NADH for butanol biosynthesis, thus meeting both cell growth and butanol production needs. These findings confirm that fine-tuning the timing of hbd(Ck) expression is critical for balancing metabolic flux and redox cofactors, which in turn minimizes acetate production and results in higher butanol production. Fine-tuning the induction timing and expression levels of exogenous enzymes in engineered pathways has been used to minimize the metabolic burden and relieve the bottlenecks in the reactions, leading to the overproduction of targeted metabolites in recombinant microorganisms. , However, this study, for the first time, demonstrated that optimizing the induction time for hbd(Ck) expression significantly enhanced butanol production in C. tyrobutyricum MΔcat1::adhE2-Pbgal-hbd(Ck), which did not produce butyrate because of the knockout of cat1 encoding for the CoA transferase responsible for the assimilation of acetate and biosynthesis of butyrate in C. tyrobutyricum.

2.2. Effects of hbd­(Ck) Expression on Cell Growth

The intracellular concentrations of NAD­(P)­H/NAD­(P)⁺ play a critical role in regulating bacterial growth and metabolism. NAD⁺ is an essential electron acceptor in glycolysis, and NADH is widely used as an electron donor in metabolism and the electron transport system to generate ATP needed for cell growth and maintenance, whereas NADPH is an electron donor in the biosynthesis of essential biomolecules (fatty acids, amino acids, nucleotides, etc.). We hypothesized that overexpression of NADPH-dependent HBD in C. tyrobutyricum Δcat1::adhE2 might disrupt its NADH/NAD⁺ homeostasis, resulting in inefficient glycolysis and poor cell growth and metabolic activities. The effects of hbd(Ck) expression on the growth kinetics of MΔcat1::adhE2 and its derivative strains expressing hbd(Ck) under Pcat1 or Pbgal control, respectively, were thus investigated in serum tubes with cell density monitored as the optical density (OD₆₀₀) in batch cultures. Figure shows the growth kinetics and compares the specific growth rates for the various strains and conditions studied. In general, cell growth was inhibited by the hbd(Ck) expression with a significantly decreased specific growth rate in the exponential phase and lowered stationary cell density for MΔcat1::adhE2-Pcat1-hbd(Ck), compared to MΔcat1::adhE2, and MΔcat1::adhE2-Pbgal-hbd(Ck), with lactose induction at 0 h, compared to at 12 h or no lactose induction. The results confirmed that optimized induction timing of hbd(Ck) alleviated the adverse effects of hbd(Ck) overexpression on cell growth and was able to maintain stable redox homeostasis in cells, which is essential for efficient production of metabolites such as butanol in industrial fermentation. −

2.

Effects of hbd(Ck) expression on the growth kinetics of MΔcat1::adhE2 and its derivative strains expressing hbd(Ck) under Pcat1 or Pbgal control, respectively. Lactose (40 mM) was added at 0 or 12 h for MΔcat1::adhE2-Pbgal-hbd(Ck) to evaluate the hbd(Ck) effect on growth. (A) Growth curves during 48 h cultivation in 20 mL serum tubes with cell density monitored as OD₆₀₀. (B) Comparison of the specific growth rates in the exponential phase estimated from the OD data in the first 12 h. Cell growth was significantly decreased and reached a lower stationary cell density for MΔcat1::adhE2-Pcat1-hbd(Ck) compared to MΔcat1::adhE2. Similarly, growth was slower for MΔcat1::adhE2-Pbgal-hbd(Ck) with lactose induction at 0 h compared to at 12 h or no lactose induction.

2.3. Effects of MV Addition

As an artificial electron carrier, MV can divert electrons for hydrogen production to NADH regeneration, thereby increasing the intracellular NADH pool for butanol biosynthesis. ,, Therefore, MV addition at various concentrations up to 200 μM was investigated for its effect on butanol production by MΔcat1::adhE2-Pbgal-hbd(Ck) in batch fermentation (Figure ). However, no significant difference in butanol production was observed among all MV concentrations (0, 50, 100, and 200 μM) studied. In general, most of the glucose was rapidly consumed in 48 h, coupled with rapid production of butanol at all MV concentrations. The fermentation ceased at 96 h when the butanol concentration reached ∼17.5 g/L with little glucose left. Although MV addition did decrease acetate production from 5 g/L at MV0 to 3 g/L at MV200, lactate production increased to 2 g/L at MV200. Consequently, butanol yield from glucose remained unchanged at 0.283 ± 0.005 g/g for all MV concentrations studied (Figure E). Nevertheless, the product selectivity increased from 0.70 without MV (MV0) to 0.78 with 100 μM MV. The butanol productivity from these fermentations was 0.32 ± 0.01 g/L/h. About 2.75 mol or 69% of total reducing equivalents (NADH + reduced ferredoxin, Fd_re) generated from each mole of glucose consumed in glycolysis was used in butanol biosynthesis (Figure F). The inducible hbd(Ck) expression at 12 h was effective in providing NADH needed for butanol biosynthesis. Interestingly, butanol productivity in the fermentation increased slightly, instead of decreasing, with increasing MV concentration. These results are different from what have been observed with WT-adhE2, Ack-adhE2, and Δcat1::adhE2 that butanol yield increased while productivity decreased with increasing MV concentration. ,

3.

Batch fermentations by C. tyrobutyricum MΔcat1::adhE2-hbd(Ck) with hbd expression induced by lactose at 12 h in serum bottles with methyl viologen (MV) at various concentrations also added at 12 h: (A) 0 μM MV; (B) 50 μM MV; (C) 100 μM MV; and (D) 200 μM MV. (E) C4/C2 and alcohol/acid ratios, selectivity, and butanol yield and productivity based on the data at 48 h. (F) NADH consumption associated with butanol, lactate, and 2,3-BDO biosynthesis per mol of glucose consumed in the fermentations.

The results imply that inducible expression of hbd(Ck) can be used as an effective strategy for increasing the NADH pool without the MV addition. To boost NADH availability for butanol biosynthesis, prior studies have focused on NADH regeneration strategies, such as adding MV and benzyl viologen, introducing the exogenous formate dehydrogenase and ferredoxin-NAD­(P⁺) oxidoreductase, and integrating bioelectrochemical systems or a photocatalyst to supply electrons to the culture. , Alternatively, NADH availability for butanol biosynthesis can be increased through reducing NADH consumption by replacing the NADH-dependent HBD with NADPH-dependent HBD in the metabolic pathway. However, constitutive expression of NADPH-dependent HBD might cause growth inhibition due to NADH/NADPH imbalance, which was mitigated through the inducible expression of hbd(Ck) at an optimal induction time during batch culture without adding MV. The internal genetic regulation of redox reactions avoids potential toxicity and process complexity and is preferable to chemical redox mediation and bioelectrochemical systems.

2.4. Effects of Low Temperature

Decreasing culture temperature can reduce membrane fluidity and permeability and thus improve the cell’s solvent tolerance. , We evaluated the effects of low temperature (25 °C) on the batch culture of MΔcat1::adhE2-Pbgal-hbd(Ck) with hbd(Ck) induced at 12 h (Figure ). As expected, the low-temperature culture improved butanol production to reach a high titer of 23 g/L at 144 h, accompanied by small amounts of acetate (2.1 g/L) and 2,3-BDO (2.1 g/L; Figure A). Adding 50 μM MV in the low-temperature culture decreased acetate (0.4 g/L) and 2,3-BDO (1.8 g/L) production, but butyrate production increased to 1.5 g/L, while butanol production did not change (Figure B). The production of butyrate in these fermentations suggested the presence of functional thioesterases, such as the one encoded by cat2 (CTK_RS03720), in the genome of C. tyrobutyricum, which remains to be verified. Increasing glucose in the medium to 120 g/L did not raise the final butanol titer above 23 g/L even though the fermentation time was extended to 168 h (Figure C,D). Apparently, 23 g/L butanol was the tolerance limit for C. tyrobutyricum at 25 °C. The low-temperature cultures also gave a much higher butanol yield (0.31 ± 0.01 g/g glucose vs 0.28 ± 0.01 g/g glucose at 37 °C) and product selectivity (∼0.85 vs ∼0.72 at 37 °C) (Figure E). About 3 mol or 75% of the reducing equivalents (NADH + Fd_re) generated from each mole of glucose consumed in glycolysis was used in butanol biosynthesis (Figure F). However, the butanol productivity at 25 °C was only ∼0.16 g/L·h or 50% of that at 37 °C. MFA showed that at 25 °C, 51% of the substrate carbon was retained in butanol and only 2.9% in acetate, whereas at 37 °C, 44% of the substrate carbon was retained in butanol and 3.1% in acetate (see Figure S2).

4.

Batch fermentations by C. tyrobutyricum MΔcat1::adhE2-hbd(Ck) with hbd expression induced by lactose at 12 h in serum bottles at 25 °C: (A) 0 μM MV with initial glucose 70 g/L; (B) 50 μM MV with initial glucose 70 g/L; (C) 0 μM MV with initial glucose 120 g/L; (D) 50 μM MV with initial glucose 120 g/L; (E) C4/C2 and alcohol/acid ratios and butanol yield and productivity; and (F) NADH consumption associated with butanol, butyrate, and 2,3-BDO biosynthesis per mol of glucose consumed in the fermentations.

These results confirm that low-temperature fermentation (25 °C) can effectively increase butanol production to achieve a high titer (23 g/L) and yield (0.31 g/g) due to enhanced butanol tolerance. Lowering the fermentation temperature to 20 °C could further increase the butanol titer to 26 g/L, as previously reported for Δcat1::adhE2,²³ but the productivity would be much lower. It is noted that the butanol yield from Δcat1::adhE2 was much lower at 0.23 g/g due to the excessive amounts of acetate and ethanol coproduced with butanol. Similarly, Clostridium carboxidivorans produced more butanol and ethanol but at significantly reduced rates at 25 °C, while 37 °C provided the optimal conditions for cell growth and organic acids production. Although lowering the fermentation temperature could greatly increase the product titer and yield, the productivity is severely compromised. Cell immobilization or cell recycling in a continuous or sequential batch fermentation process may be used to greatly increase cell density in the bioreactor and thus the reactor productivity, as demonstrated in previous studies. ,,,

2.5. Batch Fermentation in the Bioreactor

To scale up the butanol fermentation for industrial applications, we evaluated MΔcat1::adhE2-Pbgal-hbd(Ck) in a stirred-tank bioreactor operated at 37 °C with pH controlled at 5.5 by using NH₄OH, which allowed us to monitor cell growth by OD measurements that could not be done in serum-bottle fermentations with CaCO₃ for pH buffer. Lactose was added to induce the expression of hbd(Ck) when the cell density or OD reached a predetermined level (0, 2.3, 4.1, 6.5, and 8.3) corresponding to early and mid-exponential phases in the batch culture (Figure ). In general, the fermentation kinetics in glucose consumption and butanol and acetate production were similar to those observed in serum bottles. When lactose was added at the beginning or before inoculation (OD0), there was an apparent lag phase of ∼12 h with negligible cell growth and slow glucose consumption, followed by accelerated glucose consumption and butanol production reaching 18.1 g/L at 72 h (Figure A). The corresponding butanol yield of 0.27 g/g and productivity of 0.28 g/L·h were much higher than those obtained in serum bottles with lactose addition at 0 h. The improvements might be attributed to the NH₄OH used in controlling the bioreactor pH as adequate amounts of ammonia could stimulate cell growth by enhancing the nucleic acid and protein synthesis. , Meanwhile, a slightly delayed expression of hbd(Ck) in the early exponential phase after OD had reached 2.3 allowed cells to grow without an apparent lag phase and increased butanol production to the highest titer, productivity, and yield (21.7 g/L, 0.38 g/L·h, and 0.30 g/g, respectively) at 72 h (Figure B). On the other hand, lactose addition at higher ODs (4.1, 6.5, and 8.3) or in the mid exponential phase led to a progressive decrease in butanol production (Figure C–E) and product selectivity (Figure F). The cultures with delayed hbd (Ck) induction at OD 6.5 and 8.3 also showed significantly slowed glucose consumption or metabolism, probably caused by an imbalance in NADH/NADPH with the overexpression of NADP-dependent HBD. Induction at OD 2.3 gave the highest butanol production in titer, productivity, and yield (Figure F). These results suggested that inducible expression of hbd(Ck) at OD 2.3 or in the early exponential phase provided the optimal balance between cell growth and redox metabolism favoring butanol production in C. tyrobutyricum MΔcat1::adhE2-Pbgal-hbd(Ck).

5.

Batch fermentation kinetics of C. tyrobutyricum MΔcat1::adhE2-hbd(Ck) in bioreactors with hbd expression induced at OD0 (A), OD2.3 (B), OD4.1­(C), OD6.5 (D), and OD8.3 (E) and the corresponding alcohol/acid ratio, selectivity, and butanol yield and productivity (F).

2.6. Intracellular NAD­(P)H and NAD­(P)⁺

In Clostridium, NADH and NADPH serve distinct roles: NAD⁺/NADH drives catabolic energy generation and solvent production, whereas NADP⁺/NADPH primarily supports anabolic processes and biomass growth. ,, The intracellular NADH/NAD⁺ ratio is a key factor affecting cell growth, glycolysis, and metabolic fluxes toward different metabolites. It is thus of high interest to know how dynamic hbd(Ck) expression affected redox metabolism and to elucidate why early exponential phase induction was optimal. We performed MFA and measured intracellular concentrations of NAD­(P)H and NAD­(P)⁺ at 18 and 36 h in cultures with hbd(Ck) expression induced immediately after inoculation (OD 0) and in the early exponential phase (OD 2.3) (Figure ). Figure A shows the carbon flux distributions in key metabolic pathways calculated from the products (butanol, butyrate, acetate, ethanol, and 2,3-BDO) produced from glucose consumed in each of the batch fermentations in the bioreactor. When hbd(Ck) expression was induced in the early exponential phase at OD 2.3, ∼49% of the carbon present in the glucose consumed in the fermentation was channeled toward butanol biosynthesis and resulted in the highest butanol yield of 0.30 g/g, which could be attributed to the higher intracellular NADH/NAD⁺ pool, as evidenced by higher intracellular NADH/NAD⁺ levels at both 18 and 36 h compared with lactose addition at OD 0 (Figure B). This finding supports our hypothesis that the higher NADH availability with the early exponential phase induction gives higher butanol production. It should be noted that the NADH level was much lower, while NAD⁺ was much higher at 36 h than at 18 h, indicating that a significant amount of NADH was used for butanol biosynthesis. Interestingly, NADP⁺ and NADPH remained at a relatively low level throughout the fermentation in both cases, suggesting that the limited NADPH availability might have transiently limited cell growth. When NADPH-dependent HBD was overexpressed, cells must adjust their metabolism to generate the needed NADPH for anabolism to support cell growth.

6.

Effects of lactose addition at different OD levels on carbon flux distributions and intracellular NAD­(P)­H/NAD­(P)⁺ levels in C. tyrobutyricum MΔcat1::adhE2-Pbgal-hbd(Ck). (A) Metabolic flux distribution in key metabolic reactions. The flux values (%) given in the boxes next to the reactions are normalized with glucose (6 carbon) consumed in the fermentation being 100%. (B) Intracellular NAD­(P)­H/NAD­(P)⁺ levels at 18 and 36 h in batch cultures with hbd(Ck) induction at OD0 and OD2.3.

These findings explain why the induction expression of hbd(Ck) at an optimal time is important to butanol production in C. tyrobutyricum MΔcat1::adhE2-Pbgal-hbd(Ck). Introducing the heterologous enzyme (HBD) activity “just-in-time” when the metabolic state shifted from growth to metabolite (butanol) production was critical to the production of the metabolite in fermentation. , Overall, the dynamic expression strategy balanced the intracellular NADH/NAD⁺ and NADPH/NADP⁺ more effectively than the constitutive expression strategy, allowing cells to match cofactor availability with phase-specific metabolic needs.

3. Discussion

Classical ABE fermentation with solventogenic clostridia coproduces butanol, acetone, and ethanol at a 6:3:1 ratio, with butanol production TRY typically at ∼12 g/L, ∼0.35 g/L·h, and ∼0.2 g/g glucose consumed. ,− Extensive engineering of C. acetobutylicum to eliminate byproducts acetone, by knocking out ctfAB (CoA transferase), and butyrate, by knocking out ptb (phosphotransbutyrylase) and buk (butyrate kinase), and to increase NADH availability, by knocking out the redox-sensing transcriptional regulator rex that represses thlA, hbd, crt, and bcd-etfAB in the C4 formation pathway (from acetyl-CoA to butyryl-CoA) and adhE2, greatly increased butanol yield to ∼0.32 g/g or ∼80% of the theoretical maximum butanol yield, but the butanol titer was only ∼12 g/L. Earlier studies also demonstrated that overexpressing thiolase (thlA) improved butanol production in C. acetobutylicum and C. cellulovorans. ,, Expressing a heterologous pathway comprising four genes (atoB from Escherichia coli, hbd from C. kluyveri, crt from C. acetobutylicum, and ter from Treponema denticola) in engineered E. coli produced butanol at 15 g/L with a yield of >70% of the theoretical yield, whereas less than1 g/L butanol was produced by engineered E. coli overexpressing clostridial genes thlA, hbd, crt, and bcd-etfAB. This was because the E. coli thiolase encoded by atoB could tolerate a higher intracellular CoA level and would pull more carbon flux from acetyl-CoA into acetoacetyl-CoA − and the enzyme encoded by ter from T. denticola replacing BCD-ETF complex catalyzed the reaction irreversibly. , These two genes together greatly boosted the metabolic flux from C2 acetyl-CoA to C4 butyryl-CoA and thus enhanced butanol production in engineered E. coli.

Our present study extends previous metabolic engineering efforts in several important ways by achieving a unique balance between cell growth and redox metabolism in MΔcat1::adhE2-Pbgal-hbd(Ck) to attain the highest butanol TRY (>20 g/L, >0.35 g/L·h, >0.30 g/g) and selectivity (butanol/all metabolic products >0.80), which compare favorably with native and engineered microorganisms including E. coli, C. acetobutylicum, and C. tyrobutyricum Δcat1::adhE2 and Δcat1::adhE2-Pcat1-hbd(Ck) (see Table ). The higher butanol titer can be attributed to the high butanol tolerance of C. tyrobutyricum, whereas the higher butanol yield can be attributed to the high product selectivity (>0.8) due to much reduced acetate and ethanol production via the inducible hbd(Ck) expression in C. tyrobutyricum with Δcat1 that could eliminate butyrate biosynthesis. , Moreover, the high butanol titer (21.8 g/L), yield (0.30 g/g), and productivity (0.39 g/L·h) obtained in the stirred-tank bioreactor at 37 °C also compared favorably with those obtained with Ack-adhE2 in repeated batch fermentation with cells immobilized in a fibrous bed bioreactor (FBB) and with the addition of 250 μM MV, which greatly increased butanol production because of the high-cell-density culture environment and the enhancement effect of MV. It is thus possible to further increase butanol production to reach a much higher TRY using MΔcat1::adhE2-Pbgal-hbd(Ck) in FBB fermentation. Compared to industrial ABE fermentation, butanol titer increased >66% (from 12 g/L to >20 g/L), which would reduce energy cost by >40% for butanol separation, and butanol yield increased >50% (from 0.2 g/g to >0.30 g/g), which would reduce feedstock cost by >33%. Therefore, the improvements in TRY with the inducible hbd(Ck) expression in MΔcat1::adhE2-Pbgal-hbd(Ck) should significantly reduce the butanol production cost from the current $4.50 per gallon using the ABE fermentation process to less than $2.50 per gallon, which would be highly competitive to bioethanol and petroleum-derived butanol.

1. Comparison of Butanol Yield, Productivity, Selectivity, and Alcohol/Acid and C4/C2 Ratios in Batch Fermentations with Various C. tyrobutyricum Strains .

strains	temp (°C)	culture system	MV (μM)	butanol yield (g/g)	productivity (g/L·h)	selectivity (g/g)	alcohol/acid (g/g)	C4/C2 (g/g)	ref
Δcat1::adhE2	37	serum bottle	0	0.21	0.16	0.50	1.21	1.21
	37	bioreactor	0	0.20	0.31	0.41	1.27	0.77
	25	bioreactor	0	0.23	0.21	0.47	2.43	1.02
Δcat1::adhE2-Pcat1-hbd (Ck)	37	serum bottle	0	0.21	0.13	0.50	1.32	1.08
	37	serum bottle	50	0.29	0.22	0.73	4.69	3.75
MΔcat1::adhE2	37	serum bottle	0	0.21 ± 0.01	0.19 ± 0.01	0.47 ± 0.01	1.37	0.88	this study
MΔcat1::adhE2-Pcat1-hbd(Ck)	37	serum bottle	0	0.22 ± 0.01	0.17 ± 0.01	0.46 ± 0.01	1.30	0.87
MΔcat1::adhE2-Pbgal-hbd (Ck)	37	serum bottle	0	0.27 ± 0.01	0.30 ± 0.01	0.70 ± 0.01	3.51	3.51	this study
	37	serum bottle	50	0.28 ± 0.01	0.32 ± 0.01	0.74 ± 0.05	3.59	3.59
	37	serum bottle	100	0.28 ± 0.01	0.33 ± 0.01	0.77 ± 0.01	3.73	4.22
	37	serum bottle	200	0.28 ± 0.01	0.33 ± 0.01	0.73 ± 0.08	3.36	3.36
	25	serum bottle	0	0.32 ± 0.01	0.16 ± 0.01	0.83 ± 0.01	10.5	12.0
	25	serum bottle	50	0.31 ± 0.01	0.16 ± 0.01	0.86 ± 0.01	12.9	32.3
	37	bioreactor	0	0.30	0.38	0.67	2.45	2.45
WT-adhE2-hbd (Ck)	37	bioreactor	0	0.22	0.31	0.46	0.91	9.31
Ack-adhE2-hbd (Ck)	37	bioreactor	0	0.25	0.35	0.44	1.26	8.03
Ack-adhE2	37	FBB	250	0.28–0.32	0.18–0.34	0.65–0.69	1.9–2.2	70

^aPbgal promoter was induced with 40 mM lactose at 12 h in batch fermentations with MΔcat1:adhE2-Pbgal-hbd(Ck); FBB: fibrous bed bioreactor.

Cofactor engineering has been widely used as an effective strategy for regulating redox homeostasis, relieving metabolic burden to improve robustness of cells, and increasing bioproduction of chemicals and biofuels in fermentation. − For example, altering the intracellular redox ratio through the expression of a heterologous NADH oxidase in E. coli reduced the NADH/NAD⁺ ratio, which alleviated acetate overflow under high glucose uptake conditions. Also, changing the NADH/NAD⁺ ratio by overexpressing pncB in E. coli redirected the metabolic flux from lactate to ethanol biosynthesis. Furthermore, the nisin-inducible expression of a heterologous NADH oxidase in Lactococcus lactis decreased the intracellular NADH/NAD⁺ ratio and caused a marked redistribution of pyruvate flux away from lactate and toward mixed-acid products. The xylitol biosynthesis from xylose in an engineered E. coli was boosted by dynamically regulating the NADPH-consuming enoyl-ACP reductase and NADPH-producing glucose-6-phosphate dehydrogenase to improve NADPH flux using a two-stage gene circuit. In our study, we also demonstrated the dynamic regulation of NAD­(P)­H/NAD­(P)⁺ through the timely induction of hbd(Ck) expression using an inducible promoter for butanol production by engineered C. tyrobutyricum. Overall, dynamic gene expression is highly effective in balancing intracellular NADH/NAD⁺ and NADPH/NADP⁺ and allows the mutant to match cofactor availability with phase-specific metabolic needs during fermentation. Although gene expression can also be modulated by using constitutive promoters with balanced transcriptional strength, it is cumbersome and labor-intensive to screen promoters with an appropriate strength and building a gene circuit with a biosensor to switch on or off the promoter.

In summary, C. tyrobutyricum MΔcat1::adhE2-Pbgal-hbd(Ck) with inducible expression of hbd(Ck) significantly enhanced butanol production TRY and selectivity. Although the current butanol yield of 0.32 g/g (∼80% of the theoretical value 0.41 g/g) achieved in this study is among the highest ever reported for a biobutanol fermentation process, it may be further increased with rex knockout and expression of heterologous atoB and ter, as demonstrated in recombinant C. acetobutylicum and E. coli. , It should be noted that small but significant amounts of lactate and 2,3-BDO were also produced by MΔcat1::adhE2-Pbgal-hbd(Ck) in fermentation, which not only consumed some NADH but also reduced the butanol yield from glucose. It is thus desirable to knock out genes in the lactate and 2,3-butanediol biosynthesis pathways (see Figure S1) to further increase butanol yield. Meanwhile, protein engineering via active-site mutagenesis can be applied to the enzyme AAD (encoded by adhE2) to increase its activity and selectivity for the conversion of butyryl-CoA to butyraldehyde and butanol, which could be a bottleneck in butanol biosynthesis. Finally, the persistent deficiency in the NADPH supply, as indicated by low intracellular NADPH/NADP⁺ levels, may also limit the fermentation. Engineering NADH-dependent variants of HBD and enhancing NADPH-generating pathways (e.g., the pentose phosphate pathway) could alleviate the redox imbalance caused by the overexpression of NADPH-dependent HBD and allow for constitutive or earlier induction without compromising cell growth and catabolism. Overexpressing a heterologous NAD⁺ kinase, which catalyzes the phosphorylation of NAD⁺ to NADP⁺, may increase the intracellular NADP⁺/NADPH pool and offset the adverse effect of hbd(Ck) overexpression. In addition, the Rex regulator plays a vital role in sensing redox homeostasis and represses bcs operon genes involved in the metabolic reactions from acetyl-CoA to butyryl-CoA when the NAD⁺/NADH ratio is high. These should be considered in the further development of an ideal biobutanol cell factory.

In conclusion, dynamically controlling the expression of hbd(Ck) in C. tyrobutyricum MΔcat1::adhE2-Pbgal-hbd(Ck) via lactose induction in the early exponential phase alleviated redox imbalance and growth inhibition caused by the constitutive expression of NADPH-dependent HBD and gave the highest butanol production TRY (21.8 g/L, 0.39 g/L·h, 0.30 g/g) and selectivity (>0.7) in batch fermentation in a bioreactor operated at 37 °C and pH 5.5. At a lower temperature (25 °C), butanol production in the fermentation increased to 23 g/L and 0.32 g/g with >0.85 product selectivity due to much reduced byproducts formation, but productivity decreased to 0.16 g/L·h. The fermentation using the engineered C. tyrobutyricum strain offers a scalable and economically viable biobutanol production route for industrial application.

4. Materials and Methods

4.1. Strains and Culture Media

E. coli DH5α and CA434 (donor cells in conjugation) were grown at 37 °C in LB medium containing chloramphenicol. Unless otherwise noted, C. tyrobutyricum ATCC 25755 (wild type, WT) and its mutant strains were cultured anaerobically at 37 °C in RCM containing antibiotics. All serum-bottle fermentation kinetic studies were conducted in CGM containing 50–120 g/L glucose and 40 g/L CaCO₃ for buffering the culture at ∼pH 5.5.

4.2. Recombinant Plasmids and Mutant Strains

Table lists the recombinant plasmids and strains used and constructed in this study. The gene knockout plasmid was constructed using an endogenous CRISPR-Cas system, with backbones derived from pMTL83151 and pMTL85151. ,, First, pMTL83151 (or pMTL85151) was linearized. Next, the lactose inducible promoter and terminator (a 30 bp repeat sequence) were amplified and connected by overlapping PCR. We then amplified the 1000 bp homologous arm from C. tyrobutyricum ATCC 25755. These fragments were ligated with the linearized vector to yield pMTL83151Δrease, pMTL85151ΔI, pMTL85151Δupp, and pMTL83153Δcat1::adhE2, respectively. Overexpressing plasmids were built on the pMTL82151 containing the lactose inducible promoter Pbgal.

2. Mutants and Plasmids Used in This Study.

strain/plasmids	description/characteristics
strains
WT	Clostridium tyrobutyricum ATCC 25755
M	C. tyrobutyricum with the inactivation of restriction modification system (Rease; CTK_RS13400), native plasmid (I; pCTK01) and upp (CTK_RS00780) knockouts using the CRSPR-Cas plasmids pMTL83151Δrease, pMTL85151ΔI, and pMTL85151Δupp, respectively
Δcat1::adhE2	WT with the cat1 knockout and adhE2 knock-in on the genome using pMTL82151-Δcat1::adhE2
Δcat1::adhE2-Pcat1-hbd (Ck)	Δcat1::adhE2 transformed with pMTL82151-Pcat1-hbd (Ck) for overexpression of hbd (Ck)
MΔcat1::adhE2	M mutant with the cat1 knockout and adhE2 knock-in on the genome using pMTL83153-Δcat1::adhE2
MΔcat1::adhE2-Pcat1-hbd (Ck)	MΔcat1::adhE2 transformed with pMTL82151-Pcat1-hbd (Ck) for overexpression of hbd (Ck)
MΔcat1::adhE2-Pbgal-hbd (Ck)	MΔcat1::adhE2 transformed with pMTL82151-Pbgal-hbd (Ck) for inducible expression of hbd (Ck)
plasmids
PMTL82151	ColE1 ori, Cm R, pBP1 ori; TraJ
PMTL83151	ColE1 ori, Cm R, pCB102 ori; TraJ
PMTL85151	ColE1 ori, Cm R, pIM13 ori; TraJ
PMTL83151-Pcat1-upp	pMTL83151 derivative containing Pcat1-upp
pMTL83151Δrease	pMTL83151 derivative, targeting restriction modification system subunit
pMTL85151ΔI	pMTL85151 derivative, targeting native plasmid gene
pMTL85151Δupp	pMTL85151 derivative, targeting upp gene
pMTL82151-Δcat1::adhE2	pMTL82151 derivative, targeting cat1 for adhE2 insertion with the promoter Pcat1 for adhE2 expression
pMTL83151-Δcat1::adhE2	pMTL83151 derivative, targeting cat1 for adhE2 insertion with the promoter Pcat1 for adhE2 expression
pMTL82151-Pcat1-hbd (Ck)	pMTL82151 derivative, for hbd (Ck) expression with the promoter Pcat1
pMTL82151-Pbgal-hbd (Ck)	pMTL82151 derivative, for hbd (Ck) expression with the inducible promoter Pbgal

To construct the mutant strains, plasmids were transformed into C. tyrobutyricum via conjugation. , Briefly, the donner cells from the overnight culture of E. coli CA434 and C. tyrobutyricum recipient cells were mixed and incubated on RCM agar plates at 37 °C for 1 day. The cells were then replated on RCM agar plates containing the antibiotics for selection of transformant cells. Positive colonies were confirmed by colony PCR using primers (Table S1) in the flanking regions of homologous arms.

C. tyrobutyricum WT was first modified to create the M strain by knocking out the restriction modification system, native plasmid, and upp gene sequentially, using the CRSPR-Cas plasmids pMTL83151Δrease, pMTL85151ΔI, and pMTL85151Δupp, respectively. CRSPR-Cas genome-editing plasmids were cured or removed using the upp/5-FU counter-selection protocol. Then, the strain MΔcat1::adhE2 was constructed by deleting cat1 and inserting adhE2 in the genome of the M strain. Finally, MΔcat1::adhE2-Pbgal-hbd(Ck) expressing the exogenous genes hbd(Ck), cloned from C. kluyveri, under the inducible promoter Pbgal was constructed by transforming MΔcat1:adhE2 with plasmid pMTL82151-Pbgal-hbd(Ck). Removing the modification system and native plasmid improved the genetic tractability and transformation efficiency for efficient strain construction, whereas upp gene deletion enabled plasmid curing for subsequent genome-engineering cycles. The deletion of cat1 and genomic integration of adhE2 reduced the level of butyrate biosynthesis to almost zero and promoted stable butanol production.

4.3. Batch Fermentation

Unless otherwise noted, batch fermentation was carried out in 100 mL serum bottles each containing 50 mL of CGM with 60 g/L glucose and 40 g/L CaCO₃ to evaluate the effects of hbd(Ck) expression on butanol production. After sparging with N₂ to remove oxygen, autoclaving for sterilization, and cooling to the room temperature, each serum bottle was inoculated at 5% (v/v) with a freshly prepared seed culture and then placed in an static incubator at 37 °C. Lactose at the final concentration of 40 mM, which could induce gene expression at ∼40-fold, was added at 0, 12, or 24 h after inoculation to induce hbd(Ck) expression. Lactose as an inducer was not consumed or used by C. tyrobutyricum in the fermentation. All fermentation conditions were studied in duplicate serum bottles.

Batch fermentation was then studied in 2 L stirred-tank bioreactors. The bioreactor containing 850 mL of CGM was autoclaved at 121 °C and 15 psig for 30 min, mixed with 100 mL of sterilized glucose solution (600 g/L), and sparged with nitrogen gas for 30 min. Then, the reactor was inoculated with a freshly prepared seed culture (50 mL). Lactose (final concentration: 40 mM) was added at the beginning or during the early or mid-exponential phase to induce the expression of hbd(Ck) when the cell density had reached a predetermined level. The bioreactor was operated at 37 °C, 100 rpm agitation, and pH ∼ 5.5 controlled with a 30% (v/v) ammonium solution. Broth samples were taken periodically to monitor cell optical density (OD₆₀₀) and then stored in a freezer for later analysis of glucose and metabolic products, including butanol, butyrate, ethanol, acetate, lactate, and 2,3-butanediol (2,3-BDO).

4.4. Effects of hbd­(Ck) Expression on Cell Growth

The effects of hbd(Ck) expression on the growth of C. tyrobutyricum were studied by comparing the growth kinetics of MΔcat1:adhE2 and its derivative strains expressing hbd(Ck) under the Pcat1 or Pbgal control, respectively. Briefly, cells were cultured in 25 mL serum tubes each containing 15 mL of CGM with 30 g/L glucose as the substrate at 37 °C. For MΔcat1::adhE2-Pbgal-hbd(Ck), lactose (40 mM) was added at 0 or 12 h to induce hbd(Ck) expression. Cell growth was monitored by measuring the optical density at 600 nm (OD₆₀₀) directly in serum tubes by using a Thermo Scientific GENESYS 40 spectrophotometer for the initial 12 h. Cell growth after 12 h was monitored by withdrawing ∼1 mL of the culture sample, which was diluted 5-fold with deionized water for OD₆₀₀ readings. Each condition was studied in two biological replicates.

4.5. Intracellular NAD⁺/NADH and NADP⁺/NADPH

The intracellular NAD⁺/NADH and NADP⁺/NADPH concentrations were assayed using commercial kits #MAK468 and #MAK479, respectively, from Sigma-Aldrich. The assays were based on cycling reactions catalyzed by enzymes (lactate dehydrogenase for NAD⁺/NADH and glucose dehydrogenase for NADP⁺/NADPH). The color intensity of the reduced product of MTT via the formed NAD­(P)H in the reaction was then measured at 565 nm, which was proportional to the NAD­(P)⁺/NAD­(P)H concentration in the sample. Briefly, cells in 1 mL of cell suspension (adjusted to optical density OD₆₀₀ = 2.0) were collected and washed once with ice-cold PBS. The pelleted cells were treated with an extraction buffer, heated at 60 °C for 5 min, mixed with assay buffer (20 μL) and neutralizing buffer (100 μL), vortexed briefly, and then centrifuged at 14,000g for 5 min. 40 μL of the supernatant was transferred into a 96-well plate at room temperature. After addition of 80 μL of the freshly prepared working reagent containing the reaction enzymes, substrate, and MTT, the absorbance at 565 nm (OD₅₆₅) was monitored immediately for 15 min (for NAD⁺/NADH) or 30 min (for NADP⁺/NADPH) using a spectrophotometric plate reader. After subtracting the blank, ΔOD₅₆₅ was used to estimate the concentration (μM) of NAD⁺/NADH or NADP⁺/NADPH present in the sample by comparing to the standard curve (ΔOD₅₆₅ vs NAD­(P)⁺ or NAD­(P)H concentration).

4.6. Quantitative RT-PCR

Quantitative RT-PCR was used to assay the expression levels of hbd(Ck) in cells collected during batch cultures under various conditions. Briefly, 2 mL of culture media containing the suspended cells was collected at 36 h, and cells were harvested by centrifugation. To facilitate cell lysis, cell pellets were resuspended in Buffer Y1 (1 M sorbitol, 0.1 M EDTA, pH 7.4) supplemented with lysozyme (3–5 mg/mL) and 0.1% (v/v) β-mercaptoethanol and incubated at 35 °C with gentle shaking for 30 min. Cells were then pelleted again by centrifugation, the supernatant was removed, and the pellet was washed 1–2 times with Buffer Y1 without β-mercaptoethanol and lysozyme. Total RNA was then extracted using a FastPure cell/tissue total RNA isolation Kit (Qiagen) and reverse-transcribed into cDNA using a HiScript II first Strand cDNA Synthesis Kit (Qiagen). A quantitative RT-PCR (qRT-PCR) assay was conducted using ChamQ Blue Universal SYBR qPCR Master Mix (Vazyme) with the translational GTPase coding gene (typA) as the internal reference gene. The mRNA levels of the hbd(Ck) gene relative to the internal reference gene were calculated and are reported.

4.74.7. Metabolic Flux Analysis

Batch fermentation data were used to estimate product yields, carbon flux distributions, and reducing equivalents (mainly NADH and reduced ferredoxin Fd_re) consumed in the production of various metabolites based on the metabolic pathways (Figure S1) and stoichiometric equations (Table S2) given in the Appendix. Detailed methods can be found elsewhere. ,

4.8. Analytical Methods

The broth optical density at 600 nm (OD₆₀₀) in the bioreactor was measured after proper dilution with deionized water by using a spectrophotometer. Glucose, butyrate, acetate, and lactate were measured by high-performance liquid chromatography (HPLC), whereas butanol, ethanol, and 2,3-BDO were measured by gas chromatography (GC). Details on HPLC and GC methods can be found elsewhere. ,

Glossary

Abbreviations

AAD

aldehyde/alcohol dehydrogenase

ABE

acetone–butanol–ethanol

CGM

clostridial growth medium

HBD

3-hydroxybutyryl-CoA dehydrogenase

MFA

metabolic flux analysis

MV

methyl viologen

PBS

phosphate buffered saline

RCM

reinforced clostridial medium

RT-PCR

reverse-transcription polymerase chain reaction

TRY

titer, rate (productivity), yield

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acssynbio.5c00792.

• PCR primers used for plasmids construction and qPCR, stoichiometric equations for key reactions used in metabolic flux analysis (MFA), metabolic pathways in C. tyrobutyricum with cat1 knockout and adhE2 and hbd(Ck) overexpression, carbon flux distributions in the metabolic pathway for C. tyrobutyricum MΔcat1::adhE2-pbgal-hbd(Ck) in batch fermentations, effects of the lactose induction level on fermentation kinetics, and hbd(Ck) gene expression levels (PDF)

Q.W. and S.T.Y. conceived the study and wrote the manuscript. Q.W. and J.H. did the experiments, analyzed the data, and prepared the figures and tables. J.F., G.W., J.H., and Z.L. assisted in experiments and data analysis. Q.W. and S.T.Y. proofread the manuscript. S.T.Y. secured the funding and directed the project.

The Advanced Research Projects Agency–Energy (ARPA-E) (grant number: DE-AR0001512) The Department of Energy EEREBETO (grant number: DE-EE0010300) The US Department of AgricultureNational Institute of Food & Agriculture (NIFA) (award number: 2022-67021-37606) The National Science Foundation (award number 2328215).

The authors declare no competing financial interest.