Technologies are available which will allow the conversion of lignocellulose into fuel ethanol using genetically engineered bacteria. Assembling these into a cost‐effective process remains a challenge. Our work has focused primarily on the genetic engineering of enteric bacteria using a portable ethanol production pathway. Genes encoding Zymomonas mobilis pyruvate decarboxylase and alcohol dehydrogenase have been integrated into the chromosome of Escherichia coli B to produce strain KO11 for the fermentation of hemicellulose‐derived syrups. This organism can efficiently ferment all hexose and pentose sugars present in the polymers of hemicellulose. Klebsiella oxytoca M5A1 has been genetically engineered in a similar manner to produce strain P2 for ethanol production from cellulose. This organism has the native ability to ferment cellobiose and cellotriose, eliminating the need for one class of cellulase enzymes. The optimal pH for cellulose fermentation with this organism (pH 5.0–5.5) is near that of fungal cellulases. The general approach for the genetic engineering of new biocatalysts has been most successful with enteric bacteria thus far. However, this approach may also prove useful with Gram‐positive bacteria which have other important traits for lignocellulose conversion. Many opportunities remain for further improvements in the biomass to ethanol processes. These include the development of enzyme‐based systems which eliminate the need for dilute acid hydrolysis or other pretreatments, improvements in existing pretreatments for enzymatic hydrolysis, process improvements to increase the effective use of cellulase and hemicellulase enzymes, improvements in rates of ethanol production, decreased nutrient costs, increases in ethanol concentrations achieved in biomass beers, increased resistance of the biocatalysts to lignocellulosic‐derived toxins, etc. To be useful, each of these improvements must result in a decrease in the cost for ethanol production. © 1998 John Wiley & Sons, Inc. Biotechnol Bioeng 58:204–214, 1998.

Cellulolytic strains of Bacillus stearothermophilus were isolated from nature and screened for the presence of activities associated with the degradation of plant cell walls. One isolate (strain XL-65-6) which exhibited strong activities with 4-methylumbelliferyl-4-D-glucopyranoside (MUG) and 4-methylumbelliferyl-p-D-cellobiopyranoside (MUC) was used to construct a gene library in Escherichia coli. Clones degrading these model substrates were found to encode the cellobiose-specific genes of the phosphoenolpyruvate-dependent phosphotransferase system (PTS). Both MUG and MUC activities were present together, and both activities were lost concurrently during subcloning experiments. A functional E. coli ptsI gene was required for MUC and MUG activities (presumably a ptsH gene also). The DNA fragment from B. stearothermophilus contained four open reading frames which appear to form a cel operon. Intergenic stop codons for celA, celB, and celC overlapped the ribosomal binding sites of the respective downstream genes. Frameshift mutations or deletions in celA, celB, and celD were individually shown to result in a loss of MUC and MUG activities. On the basis of amino acid sequence homology and hydropathy plots of translated sequences, celA and celB were identified as encoding PTS enzyme II and celD was identified as encoding PTS enzyme III. These translated sequences were remarkably similar to their respective E. coli homologs for cellobiose transport. No reported sequences exhibited a high level of homology with the celC gene product. The predicted carboxy-terminal region for celC was similar to the corresponding region of E. coli celF, a phospho-,-glucosidase. An incomplete regulatory gene (celR) and proposed promoter sequence were located 5' to the proposed cel operon. A stem-loop resembling a rho-independent terminator was present immediately downstream from celD. These results indicate that B. stearothermophilus XL-65-6 contains a cellobiose-specific PTS for cellobiose uptake. Similar systems may be present in other gram-positive bacteria.

No abstract

Genomic libraries from nine cellobiose-metabolizing bacteria were screened for cellobiose utilization. Positive clones were recovered from six libraries, all of which encode phosphoenolpyruvate:carbohydrate phosphotransferase system (PTS) proteins. Clones from Bacillus subtilis, Butyrivibrio fibrisolvens, and Klebsiella oxytoca allowed the growth of recombinant Escherichia coli in cellobiose-M9 minimal medium. The K. oxytoca clone, pLOI1906, exhibited an unusually broad substrate range (cellobiose, arbutin, salicin, and methylumbelliferyl derivatives of glucose, cellobiose, mannose, and xylose) and was sequenced. The insert in this plasmid encoded the carboxy-terminal region of a putative regulatory protein, cellobiose permease (single polypeptide), and phospho-␤-glucosidase, which appear to form an operon (casRAB). Subclones allowed both casA and casB to be expressed independently, as evidenced by in vitro complementation. An analysis of the translated sequences from the EIIC domains of cellobiose, aryl-␤-glucoside, and other disaccharide permeases allowed the identification of a 50-amino-acid conserved region. A disaccharide consensus sequence is proposed for the most conserved segment (13 amino acids), which may represent part of the EIIC active site for binding and phosphorylation.

No abstract