Tolerance to furfural-induced stress is associated with pentose phosphate pathway genes ZWF1, GND1, RPE1, and TKL1 in Saccharomyces cerevisiae.

Engineering yeast to be more tolerant to fermentation inhibitors, furfural and 5-hydroxymethylfurfural (HMF), will lead to more efficient lignocellulose to ethanol bioconversion. To identify target genes involved in furfural tolerance, a Saccharomyces cerevisiae gene disruption library was screened for mutants with growth deficiencies in the presence of furfural. It was hypothesized that overexpression of these genes would provide a growth benefit in the presence of furfural. Sixty two mutants were identified whose corresponding genes function in a wide spectrum of physiological pathways, suggesting that furfural tolerance is a complex process. We focused on four mutants, zwf1, gnd1, rpe1, and tkl1, which represent genes encoding pentose phosphate pathway (PPP) enzymes. At various concentrations of furfural and HMF, a clear association with higher sensitivity to these inhibitors was demonstrated in these mutants. PPP mutants were inefficient at reducing furfural to the less toxic furfuryl alcohol, which we propose is a result of an overall decreased abundance of reducing equivalents or to NADPH's role in stress tolerance. Overexpression of ZWF1 in S. cerevisiae allowed growth at furfural concentrations that are normally toxic. These results demonstrate a strong relationship between PPP genes and furfural tolerance and provide additional putative target genes involved in furfural tolerance.

Isolation of microorganisms for biological detoxification of lignocellulosic hydrolysates.

Acid pretreatment of lignocellulosic biomass releases furan and phenolic compounds, which are toxic to microorganisms used for subsequent fermentation. In this study, we isolated new microorganisms for depletion of inhibitors in lignocellulosic acid hydrolysates. A sequential enrichment strategy was used to isolate microorganisms from soil. Selection was carried out in a defined mineral medium containing a mixture of ferulic acid (5 mM), 5-hydroxymethylfurfural (5-HMF, 15 mM), and furfural (20 mM) as the carbon and energy sources, followed by an additional transfer into a corn stover hydrolysate (CSH) prepared using dilute acid. Subsequently, based on stable growth on these substrates, six isolates--including five bacteria related to Methylobacterium extorquens, Pseudomonas sp, Flavobacterium indologenes, Acinetobacter sp., Arthrobacter aurescens, and one fungus, Coniochaeta ligniaria--were chosen. All six isolates depleted toxic compounds from defined medium, but only C. ligniaria C8 (NRRL 30616) was effective at eliminating furfural and 5-HMF from CSH. C. ligniaria NRRL 30616 may be useful in developing a bioprocess for inhibitor abatement in the conversion of lignocellulosic biomass to fuels and chemicals.

Fermentation of sugar mixtures using Escherichia coli catabolite repression mutants engineered for production of L-lactic acid.

Conversion of lignocellulose to lactic acid requires strains capable of fermenting sugar mixtures of glucose and xylose. Recombinant Escherichia coli strains were engineered to selectively produce L-lactic acid and then used to ferment sugar mixtures. Three of these strains were catabolite repression mutants (ptsG(-)) that have the ability to simultaneously ferment glucose and xylose. The best results were obtained for ptsG(-) strain FBR19. FBR19 cultures had a yield of 0.77 (g lactic acid/g added sugar) when used to ferment a 100 g/l total equal mixture of glucose and xylose. The strain also consumed 75% of the xylose. In comparison, the ptsG(+) strains had yields of 0.47-0.48 g/g and consumed 18-22% of the xylose. FBR19 was subsequently used to ferment a variety of glucose (0-40 g/l) and xylose (40 g/l) mixtures. The lactic acid yields ranged from 0.74 to 1.00 g/g. Further experiments were conducted to discover the mechanism leading to the poor yields for ptsG(+) strains. Xylose isomerase (XI) activity, a marker for induction of xylose metabolism, was monitored for FBR19 and a ptsG(+) control during fermentations of a sugar mixture. Crude protein extracts prepared from FBR19 had 10-12 times the specific XI activity of comparable samples from ptsG(+) strains. Therefore, higher expression of xylose metabolic genes in the ptsG(-) strain may be responsible for superior conversion of xylose to product compared to the ptsG(+) fermentations.

Recombinant Escherichia coli engineered for production of L-lactic acid from hexose and pentose sugars.

Recombinant Escherichia coli have been constructed for the conversion of glucose as well as pentose sugars into L-lactic acid. The strains carry the lactate dehydrogenase gene from Streptococcus bovis on a low copy number plasmid for production of L-lactate. Three E. coli strains were transformed with the plasmid for producing L-lactic acid. Strains FBR9 and FBR11 were serially transferred 10 times in anaerobic cultures in sugar-limited medium containing glucose or xylose without selective antibiotic. An average of 96% of both FBR9 and FBR11 cells maintained pVALDH1 in anaerobic cultures. The fermentation performances of FBR9, FBR10, and FBR11 were compared in pH-controlled batch fermentations with medium containing 10% w/v glucose. Fermentation results were superior for FBR11, an E. coli B strain, compared to those observed for FBR9 or FBR10. FBR11 exhausted the glucose within 30 h, and the maximum lactic acid concentration (7.32% w/v) was 93% of the theoretical maximum. The other side-products detected were cell mass and succinic acid (0.5 g/l).

Use of catabolite repression mutants for fermentation of sugar mixtures to ethanol.

Use of agricultural biomass, other than corn-starch, to produce fuel ethanol requires a microorganism that can ferment the mixture of sugars derived from hemicellulose. Escherichia coli metabolizes a wide range of substrates and has been engineered to produce ethanol in high yield from sugar mixtures. E. coli metabolizes glucose in preference to other sugars and, as a result, utilization of the pentoses in hemicellulose-derived sugar mixtures is delayed and may be incomplete. Residual sugar lowers the ethanol yield and is problematic for downstream processing of fermentation products. Therefore, a catabolite repression mutant that simultaneously utilizes glucose and pentoses would be useful for fermentation of complex substrate mixtures. We constructed ethanologenic E. coli strains with a glucose phosphotransferase (ptsG) mutation and used the mutants to ferment glucose, arabinose, and xylose, singly and in mixtures, to ethanol. Yields were 87-94% of theoretical for both the wild type and mutants, but the mutants had an altered pattern of mixed sugar utilization. Phosphotransferase mutants metabolized the pentoses simultaneously with glucose, rather than sequentially. Based upon fermentations of sugar mixtures, a catabolite-repression mutant of ethanologenic E. coli is expected to provide more efficient fermentation of hemicellulose hydrolysates by allowing direct utilization of pentoses.

BenR, a XylS homologue, regulates three different pathways of aromatic acid degradation in Pseudomonas putida.

Pseudomonas putida converts benzoate to catechol using two enzymes that are encoded on the chromosome and whose expression is induced by benzoate. Benzoate also binds to the regulator XylS to induce expression of the TOL (toluene degradation) plasmid-encoded meta pathway operon for benzoate and methylbenzoate degradation. Finally, benzoate represses the ability of P. putida to transport 4-hydroxybenzoate (4-HBA) by preventing transcription of pcaK, the gene encoding the 4-HBA permease. Here we identified a gene, benR, as a regulator of benzoate, methylbenzoate, and 4-HBA degradation genes. A benR mutant isolated by random transposon mutagenesis was unable to grow on benzoate. The deduced amino acid sequence of BenR showed high similarity (62% identity) to the sequence of XylS, a member of the AraC family of regulators. An additional seven genes located adjacent to benR were inferred to be involved in benzoate degradation based on their deduced amino acid sequences. The benABC genes likely encode benzoate dioxygenase, and benD likely encodes 2-hydro-1,2-dihydroxybenzoate dehydrogenase. benK and benF were assigned functions as a benzoate permease and porin, respectively. The possible function of a final gene, benE, is not known. benR activated expression of a benA-lacZ reporter fusion in response to benzoate. It also activated expression of a meta cleavage operon promoter-lacZ fusion inserted in an E. coli chromosome. Third, benR was required for benzoate-mediated repression of pcaK-lacZ fusion expression. The benA promoter region contains a direct repeat sequence that matches the XylS binding site previously defined for the meta cleavage operon promoter. It is likely that BenR binds to the promoter region of chromosomal benzoate degradation genes and plasmid-encoded methylbenzoate degradation genes to activate gene expression in response to benzoate. The action of BenR in repressing 4-HBA uptake is probably indirect.

Development of new ethanologenic Escherichia coli strains for fermentation of lignocellulosic biomass.

Two new ethanologenic strains (FBR4 and FBR5) of Escherichia coli were constructed and used to ferment corn fiber hydrolysate. The strains carry the plasmid pLOI297, which contains the genes from Zymomonas mobilis necessary for efficiently converting pyruvate into ethanol. Both strains selectively maintained the plasmid when grown anaerobically. Each culture was serially transferred 10 times in anaerobic culture with sugar-limited medium containing xylose, but no selective antibiotic. An average of 93 and 95% of the FBR4 and FBR5 cells, respectively, maintained pLOI297 in anaerobic culture. The fermentation performances of the repeatedly transferred cultures were compared with those of cultures freshly revived from stock in pH-controlled batch fermentations with 10% (w/v) xylose. Fermentation results were similar for all the cultures. Fermentations were completed within 60 h and ethanol yields were 86-92% of theoretical. Maximal ethanol concentrations were 3.9-4.2% (w/v). The strains were also tested for their ability to ferment corn fiber hydrolysate, which contained 8.5% (w/v) total sugars (2.0% arabinose, 2.8% glucose, and 3.7% xylose). E. coli FBR5 produced more ethanol than FBR4 from the corn fiber hydrolysate. E. coli FBR5 fermented all but 0.4% (w/v) of the available sugar, whereas strain FBR4 left 1.6% unconsumed. The fermentation with FBR5 was completed within 55 h and yielded 0.46 g of ethanol/g of available sugar, 90% of the maximum obtainable.

An aerotaxis transducer gene from Pseudomonas putida.

An aerotaxis gene, aer, was cloned from Pseudomonas putida PRS2000. A P. putida aer mutant displayed an altered aerotactic response in a capillary assay. Wild-type P. putida clustered at the air/liquid interface. In contrast, the aer mutant did not cluster at the interface, but instead formed a diffuse band at a distance from the meniscus. Wild-type aer, provided in trans, complemented the aer mutant to an aerotactic response that was stronger than wild-type. The P. putida Aer sequence is similar over its entire length to the aerotaxis (energy taxis) signal transducer protein, Aer, of Escherichia coli. The amino-terminus is similar to redox-sensing regulatory proteins, and the carboxy-terminus contains the highly conserved domain present in chemotactic transducers.

Fermentations with new recombinant organisms.

United States fuel ethanol production in 1998 exceeded the record production of 1.4 billion gallons set in 1995. Most of this ethanol was produced from over 550 million bushels of corn. Expanding fuel ethanol production will require developing lower-cost feedstocks, and only lignocellulosic feedstocks are available in sufficient quantities to substitute for corn starch. Major technical hurdles to converting lignocellulose to ethanol include the lack of low-cost efficient enzymes for saccharification of biomass to fermentable sugars and the development of microorganisms for the fermentation of these mixed sugars. To date, the most successful research approaches to develop novel biocatalysts that will efficiently ferment mixed sugar syrups include isolation of novel yeasts that ferment xylose, genetic engineering of Escherichia coli and other gram negative bacteria for ethanol production, and genetic engineering of Saccharoymces cerevisiae and Zymomonas mobilis for pentose utilization. We have evaluated the fermentation of corn fiber hydrolyzates by the various strains developed. E. coli K011, E. coli SL40, E. coli FBR3, Zymomonas CP4 (pZB5), and Saccharomyces 1400 (pLNH32) fermented corn fiber hydrolyzates to ethanol in the range of 21-34 g/L with yields ranging from 0.41 to 0.50 g of ethanol per gram of sugar consumed. Progress with new recombinant microorganisms has been rapid and will continue with the eventual development of organisms suitable for commercial ethanol production. Each research approach holds considerable promise, with the possibility existing that different "industrially hardened" strains may find separate applications in the fermentation of specific feedstocks.

benK encodes a hydrophobic permease-like protein involved in benzoate degradation by Acinetobacter sp. strain ADP1.

The chromosomal benK gene was identified within a supraoperonic gene cluster involved in benzoate degradation by Acinetobacter sp. strain ADP1, and benK was expressed in response to a benzoate metabolite, cis,cis-muconate. The disruption of benK reduced benzoate uptake and impaired the use of benzoate or benzaldehyde as the carbon source. BenK was homologous to several aromatic compound transporters.

PcaK, a high-affinity permease for the aromatic compounds 4-hydroxybenzoate and protocatechuate from Pseudomonas putida.

PcaK is a transporter and chemoreceptor protein from Pseudomonas putida that is encoded as part of the beta-ketoadipate pathway regulon for aromatic acid degradation. When expressed in Escherichia coli, PcaK was localized to the membrane and catalyzed the accumulation of two aromatic substrates, 4-hydroxybenzoate and protocatechuate, against a concentration gradient. Benzoate inhibited 4-hydroxybenzoate uptake but was not a substrate for PcaK-catalyzed transport. A P. putida pcaK mutant was defective in its ability to accumulate micromolar amounts of 4-hydroxybenzoate and protocatechuate. The mutant was also impaired in growth on millimolar concentrations of these aromatic acids. In contrast, the pcaK mutant grew at wild-type rates on benzoate. The Vmax for uptake of 4-hydroxybenzoate was at least 25 nmol/min/mg of protein, and the Km was 6 microM. PcaK-mediated transport is energized by the proton motive force. These results show that although aromatic acids in the undissociated (uncharged) form can diffuse across bacterial membranes, high-specificity active transport systems probably also contribute to the ability of bacteria to grow on the micromolar concentrations of these compounds that are typically present in soil. A variety of aromatic molecules, including naturally occurring lignin derivatives and xenobiotics, are metabolized by bacteria and may be substrates for transport proteins. The characterization of PcaK provides a foundation for understanding active transport as a critical step in the metabolism of aromatic carbon sources.

Repression of 4-hydroxybenzoate transport and degradation by benzoate: a new layer of regulatory control in the Pseudomonas putida beta-ketoadipate pathway.

Pseudomonas putida PRS2000 degrades the aromatic acids benzoate and 4-hydroxybenzoate via two parallel sequences of reactions that converge at beta-ketoadipate, a derivative of which is cleaved to form tricarboxylic acid cycle intermediates. Structural genes (pca genes) required for the complete degradation of 4-hydroxybenzoate via the protocatechuate branch of the beta-ketoadipate pathway have been characterized, and a specific transport system for 4-hydroxybenzoate has recently been described. To better understand how P. putida coordinates the processes of 4-hydroxybenzoate transport and metabolism to achieve complete degradation, the regulation of pcaK, the 4-hydroxybenzoate transport gene, and that of pcaF, a gene required for both benzoate and 4-hydroxybenzoate degradation, were compared. Primer extension analysis and lacZ fusions showed that pcaK and pcaF, which are adjacent on the chromosome, are transcribed independently. PcaR, a transcriptional activator of several genes of the beta-ketoadipate pathway, is required for expression of both pcaF and pcaK, and the pathway intermediate beta-ketoadipate induces both genes. In addition to these expected regulatory elements, expression of pcaK, but not pcaF, is repressed by benzoate. This previously unrecognized layer of regulatory control in the beta-ketoadipate pathway appears to extend to the first two steps of 4-hydroxybenzoate degradation, since levels of 4-hydroxybenzoate hydroxylase and protocatechuate 3,4-dioxygenase activities were also depressed when cells were grown on a mixture of 4-hydroxybenzoate and benzoate. The apparent consequence of benzoate repression is that cells degrade benzoate in preference to 4-hydroxybenzoate. These findings indicate that 4-hydroxybenzoate transport is an integral feature of the beta-ketoadipate pathway in P. putida and that transport plays a role in establishing the preferential degradation of benzoate over 4-hydroxybenzoate. These results also demonstrate that there is communication between the two branches of the beta-ketoadipate pathway.

Identification of the pcaRKF gene cluster from Pseudomonas putida: involvement in chemotaxis, biodegradation, and transport of 4-hydroxybenzoate.

Pseudomonas putida PRS2000 is chemotactic to 4-hydroxybenzoate and other aromatic acids. This behavioral response is induced when cells are grown on 4-hydroxybenzoate or benzoate, compounds that are degraded via the beta-ketoadipate pathway. Isolation of a transposon mutant defective in 4-hydroxybenzoate chemotaxis allowed identification of a new gene cluster designated pcaRKF. DNA sequencing, mutational analysis, and complementation studies revealed that pcaR encodes a regulatory protein required for induction of at least four of the enzymes of the beta-ketoadipate pathway and that pcaF encodes beta-ketoadipyl-coenzyme A thiolase, the last enzyme in the pathway. The third gene, pcaK, encodes a transporter for 4-hydroxybenzoate, and this protein is also required for chemotaxis to aromatic acids. The predicted PcaK protein is 47 kDa in size, with a deduced amino acid sequence indicative of membership in the major facilitator superfamily of transport proteins. The protein, expressed in Escherichia coli, catalyzed 4-hydroxybenzoate transport. In addition, whole cells of P. putida pcaK mutants accumulated 4-hydroxybenzoate at reduced rates compared with that in wild-type cells. The pcaK mutation did not impair growth at the expense of 4-hydroxybenzoate under most conditions; however, mutant cells grew somewhat more slowly than the wild type on 4-hydroxybenzoate at a high pH. The finding that 4-hydroxybenzoate chemotaxis can be disrupted without an accompanying effect on metabolism indicates that this chemotactic response is receptor mediated. It remains to be determined, however, whether PcaK itself is a chemoreceptor for 4-hydroxybenzoate or whether it plays an indirect role in chemotaxis. These findings indicate that aromatic acid detection and transport are integral features of aromatic degradation pathways.