Improvement of cellulose catabolism in Clostridium cellulolyticum by sporulation abolishment and carbon alleviation

Abstract

Background: Clostridium cellulolyticum can degrade lignocellulosic biomass, and ferment the soluble sugars to produce valuable chemicals such as lactate, acetate, ethanol and hydrogen. However, the cellulose utilization efficiency of C. cellulolyticum still remains very low, impeding its application in consolidated bioprocessing for biofuels production. In this study, two metabolic engineering strategies were exploited to improve cellulose utilization efficiency, including sporulation abolishment and carbon overload alleviation.

Results: The spo0A gene at locus Ccel_1894, which encodes a master sporulation regulator was inactivated. The spo0A mutant abolished the sporulation ability. In a high concentration of cellulose (50 g/l), the performance of the spo0A mutant increased dramatically in terms of maximum growth, final concentrations of three major metabolic products, and cellulose catabolism. The microarray and gas chromatography-mass spectrometry (GC-MS) analyses showed that the valine, leucine and isoleucine biosynthesis pathways were up-regulated in the spo0A mutant. Based on this information, a partial isobutanol producing pathway modified from valine biosynthesis was introduced into C. cellulolyticum strains to further increase cellulose consumption by alleviating excessive carbon load. The introduction of this synthetic pathway to the wild-type strain improved cellulose consumption from 17.6 g/l to 28.7 g/l with a production of 0.42 g/l isobutanol in the 50 g/l cellulose medium. However, the spo0A mutant strain did not appreciably benefit from introduction of this synthetic pathway and the cellulose utilization efficiency did not further increase. A technical highlight in this study was that an in vivo promoter strength evaluation protocol was developed using anaerobic fluorescent protein and flow cytometry for C. cellulolyticum.

Conclusions: In this study, we inactivated the spo0A gene and introduced a heterologous synthetic pathway to manipulate the stress response to heavy carbon load and accumulation of metabolic products. These findings provide new perspectives to enhance the ability of cellulolytic bacteria to produce biofuels and biocommodities with high efficiency and at low cost directly from lignocellulosic biomass.

Figure 1

Generation and characterization of the spo0A mutant. (A) Confirmation of pure culture of the spo0A mutant by PCR, using different combinations of primers. Amplification of intron-spo0A junction regions using one primer in the genome and the other in the intron (Spo0AF/intronF1 and intronR1/Spo0AR) resulted in bands from the Ccel_1894 mutant (lanes 1 and 2), but not in wild-type (WT) cells (lanes 4 and 5). In PCR reactions using Spo0AF/Spo0AR primers, the mutant showed a single band (lane 3), which was 915 bp larger than the single band (lane 6) in WT cells, confirming the expected intron insertion. (B) Southern blot with biotin-labeled intron-specific probes showing a single intron insertion band for spo0A mutant (lane 3), and no band for WT (lane 2). The intron donor plasmid was used as a positive control (lane 1). (C) Agar plate showing that the colonial morphology of spo0A mutant (right) was flatter and more translucent compared to WT (left). (D) Comparison of cellulose fermentation broth showing that spo0A mutant culture produced a yellow-green pigment (right), which was not observed for WT (left).

Figure 2

Growth curves and maximum production of lactate, acetate and ethanol by C. cellulolyticum strains. (A) Growth measured by optical density at 600 nm (OD₆₀₀) in 5 g/l cellobiose. Wild-type (WT) and spo0A mutant reached a similar maximum growth, and spo0A mutant had a slightly lagged log phase. (B) Growth measured by total cellular protein in 10 g/l cellulose. WT and spo0A mutant reached a similar maximum growth, and spo0A mutant had a more obvious lagged log phase. (C and D) Final metabolic product concentrations measured for WT and spo0A mutant in 5 g/l cellobiose and 10 g/l cellulose, respectively. In both of the carbon sources, the WT and spo0A mutant produced similar amounts of lactate, acetate and ethanol. The error bars represent standard deviations of measurements from three replicate cultures.

Figure 3

Total cellulose utilization by wild-type (WT) and spo0A mutant grown in 50 g/l and 10 g/l cellulose, respectively. The fermentation time was 256 hours with 50 g/l and 156 hours with 10 g/l cellulose, respectively. The error bars represent standard deviations of measurements from three replicate cultures.

Figure 4

Production profiles of lactate, acetate and ethanol and growth curves of C. cellulolyticum strains in 50 g/l cellulose. All of the three major metabolic products, lactate (A), acetate (B), and ethanol (C) were increased in spo0A mutant, with the highest increase in ethanol production. The major metabolic product of both strains was lactate. The growth of spo0A mutant (D) was increased 53% in the total cellular biomass, but the total cellular biomass of spo0A mutant dropped more rapidly than WT after reaching the peak growth. The error bars represent standard deviations of measurements from three replicate cultures.

Figure 5

Gene expression levels by quantitative real-time PCR. Relative expression level of the selected genes Ccel_1736, Ccel_0127, Ccel_0128, Ccel_3435, and Ccel_0592 in wild-type (WT) and spo0A mutant grown on 10 g/l cellulose (A) and 5 g/l cellobiose (B) were compared by normalizing with the expression of the calibrator gene recA. The means and standard deviations were calculated from the values of three biological replicates. *Significant difference between WT and spo0A mutant (P <0.05, Student t-test).

Figure 6

Flow cytometry data with fluorescent detector channel, FL1-A. In the cellobiose medium, left upper chart showed an obvious shift of pLyc017 transformants from negative control (NC) with fluorescence intensity, but not for pLyc027 transformants, corresponding to more than 10-fold fluorescent signal intensity increase in the middle upper chart for pLyc017 and only a slightly increase for pLyc027. In the cellulose medium, pLyc017 transformants showed an obviously shifted peak from NC, and pLyc027 transformants also showed a shifted peak (left lower chart). The data also demonstrated that the cipP promoter had different strength in cellobiose and cellulose media, and even with cellulose as the carbon source it was not as strong as the ferrodoxin promoter. The error bars of mean and median FL1-A values represent standard deviations of measurements from three replicate cultures.

Figure 7

Microscopic images of cells grown in cellobiose (CB) and cellulose (CU) media. pLyc017 transformants showed strong fluorescence in both CB and CU media, whereas pLyc027 transformants showed weak fluorescence in CB medium and moderate fluorescence in CU medium.

Figure 8

Final production of lactate, acetate, ethanol and isobutanol and cellulose utilization by C. cellulolyticum strains transformed with pLyc025. The metabolic productivity of the transformants was compared in 5 g/l cellobiose (A), 10 g/l cellulose (B) and 50 g/l cellulose (C), respectively. The cellulose utilization at the two different concentrations was also comared (D). The error bars represent standard deviations of measurements from three replicate cultures.