High ethanol titers from cellulose by using metabolically engineered thermophilic, anaerobic microbes

Abstract

This work describes novel genetic tools for use in Clostridium thermocellum that allow creation of unmarked mutations while using a replicating plasmid. The strategy employed counter-selections developed from the native C. thermocellum hpt gene and the Thermoanaerobacterium saccharolyticum tdk gene and was used to delete the genes for both lactate dehydrogenase (Ldh) and phosphotransacetylase (Pta). The Δldh Δpta mutant was evolved for 2,000 h, resulting in a stable strain with 40:1 ethanol selectivity and a 4.2-fold increase in ethanol yield over the wild-type strain. Ethanol production from cellulose was investigated with an engineered coculture of organic acid-deficient engineered strains of both C. thermocellum and T. saccharolyticum. Fermentation of 92 g/liter Avicel by this coculture resulted in 38 g/liter ethanol, with acetic and lactic acids below detection limits, in 146 h. These results demonstrate that ethanol production by thermophilic, cellulolytic microbes is amenable to substantial improvement by metabolic engineering.

Fig. 1.

Deletion of the C. thermocellum hpt gene. (a) Cartoon schematic illustrating the deletion of hpt. The plasmid used to delete hpt, pMU1657, contained the following: (i) a deletion cassette comprised of hpt 5′ and 3′ flanking DNA; (ii) a cat cassette, with cat driven by C. thermocellum gapDH promoter encoding Tm resistance; (iii) an hpt cassette, with hpt driven by the C. thermocellum cbp promoter encoding sensitivity to AZH. Step 1 depicts plasmid transformation. C. thermocellum transformants were selected using Tm resistance encoded by cat cassette. Step 2 depicts plasmid curing and deletion of chromosomal copy of hpt. Plating on AZH selected for plasmid loss, mediated by the hpt cassette and deletion of chromosomal hpt by homologous recombination. Depiction of chromosomal regions, as labeled on the right, indicate primer binding sites (→) used in diagnostic PCR. Expected amplicon size is shown to the left. (b) DNA gel showing results of diagnostic PCR at the hpt locus after step 2. Gel image contains a 1-kb ladder (lane 1), diagnostic PCR of clones subjected to selection conditions (lanes 2 to 6), no-DNA negative control (lane 7), and C. thermocellum 1313 genomic DNA used as a template to amplify the wild-type hpt locus as a positive control (lane 8).

Fig. 2.

Deletion of the C. thermocellum ldh and pta genes. (a) Cartoon schematic illustrating the deletion of ldh. The plasmid used to delete ldh, pMU1777, contains two major features: (i) tdk cassette, with T. saccharolyticum tdk driven by C. thermocellum cbp promoter, and (ii) an integration cassette, including an internal fragment of the ldh gene, transcriptional fusion of cat and hpt driven by the C. thermocellum gapDH promoter, and 5′ and 3′ flanking DNA labeled as ldh 5′ and ldh 3′. The first step consists of transforming the C. thermocellum Δhpt strain with pMU1777 and selecting on Tm. The second step combines Tm and FUDR, which selects for recombination of the integration cassette onto the chromosome and for loss of the plasmid backbone containing the tdk cassette. This creates a merodiploid (md) strain in which the ldh 5′ flanking DNA is duplicated. The third step utilizes AZH counter-selection to isolate cells that have lost the cat-hpt cassette through recombination between the duplicated ldh 5′ flanking DNA, resulting in deletion of ldh. Chromosomal regions are labeled to the right and indicate primer binding sites (→) used in diagnostic PCR. Expected amplicon size is shown to the left. (b) DNA gel showing results of diagnostic PCR at the ldh locus after step 2. (c) DNA gel showing results of diagnostic PCR at the ldh locus after step 3. Both gel images contain a 1-kb DNA Ladder (lane 1), diagnostic PCR of clones subjected to selection conditions (lanes 2 to 6), no-DNA negative control (lane 7), and Δhpt strain genomic DNA used as template to generate wild-type ldh locus as a positive control (lane 8). (d) Gel image showing molecular confirmation of Δhpt Δpta and Δhpt Δldh Δpta strains. The same selections as those shown in Fig. 2a were performed to delete pta in both the Δhpt and Δhpt Δldh strains. Lanes 2 to 4 show the results of diagnostic PCR at the pta locus for the Δhpt Δpta strain (lane 2), the Δhpt Δldh Δpta strain (lane 3), and the Δhpt strain (lane 4). The expected size of the wild-type pta amplicon is 3.0 kb, while the Δpta amplicon is 2.1 kb. The 1-kb DNA ladder is loaded in lane 1.

Fig. 3.

Fermentation profiles of engineered C. thermocellum monocultures and coculture using both wild-type (wt) or engineered strains of C. thermocellum and T. saccharolyticum. (a) Batch fermentations were maintained at 55°C with initial pH of 7.0 and 19.5 g/liter Avicel. Data represent the averages and standard deviations of duplicate fermentations sampled at 72 h. An evolutionarily (ev) engineered version of the Δhpt Δldh Δpta strain was transferred in batch culture for 2,000 h. (b) Batch fermentations were maintained at 55°C with initial pH of 6.75 and 17.2 g/liter Avicel. Data represent the averages and standard deviations of duplicate fermentations sampled at 120 h. C. thermocellum ev. Δhpt Δldh Δpta and T. saccharolyticum ALK2 comprised the engineered coculture.

Fig. 4.

Percent theoretical maximum (max) ethanol of fermentations performed in this study. Maximum theoretical ethanol yield was based on Avicel conversion and is listed in Tables SA2 and SA3 in the supplemental material. For a given strain or coculture, data represent averaged measured ethanol generated during fermentation divided by the theoretical maximum ethanol. The engineered coculture run in the reactor represents a single experiment.

Fig. 5.

High ethanol titer using the C. thermocellum evolutionarily (ev) engineered version of theΔhpt Δldh Δpta strain and in a T. saccharolyticum ALK2 coculture. A batch reactor with continuous stirring set at 300 rpm with an initial pH of 6.3 and 92.2 g/liter Avicel was maintained at 55°C. Data represent a single fermentation sampled over the course of 146 h.