Today, the most commonly used biofuels are bioethanol, generated from sugar — and starch-based processes, and biodiesel, generated from animal fats or vegetable oils. As of 2005, worldwide production capacity for bioethanol fuel was about 45 million Lyear-1 [12]. Global capacity for biodiesel is much lower at about 4 millionL [13-16], although certain countries (notably Ger­many) are investing in expanded capacity for this fuel [14]. The installed capacity for both fuels is rising dramatically in the face of high oil prices; biodiesel production has risen by an average of 50% annually between 2000 and 2005, while about 15% annual growth has been observed in bioethanol production over the same period. While biodiesel is increasing in importance, it is clear that bioethanol will remain the dominant biofuel for some years to come.

The simplest way to generate bioethanol is to use yeast to ferment hex — ose sugars such as glucose, which can be obtained directly from agricultural crops such as sugarcane or sugar beet. In Brazil, the sugar-based industry currently has the capacity to produce almost half of the world’s bioethanol supply, or about 17 billion Lyear-1 [12]. Another source of the sugars re­quired for fermentation is starch, produced in corn, wheat, and other cereal crops. Starch must be broken down through acid or enzymatic hydrolysis in order to release glucose, which can then be fermented to bioethanol [6]. Both sugar and starch-based processes are employed in Europe, with France (629 millionL) and Spain (520 million L) currently leading production [12]. In North America, corn (or maize) is currently the dominant biomass source for the bioethanol industry, due in part to the high proportion of starch found in its kernels and its high yield per hectare in comparison to other cereal crops. Corn, like sugarcane or switchgrass, is a C4 plant, which can utilize an extra carbon molecule in the photosynthetic process as compared to wheat or trees, which are C3 plants. Warmer growing conditions found across the USA favor C4 plants, while cooler regions (including the Canadian prairie) are well-suited to C3 plant production. Comparatively, C4 plants have relatively high water efficiency, while C3 plants have the ability to increase photosynthetic activity in the presence of elevated CO2 levels. Thus, growing conditions in any given year will determine optimal bioethanol feedstocks for specific regions [17].

The USA has a bioethanol production capacity of over 18 billion L [18], while Canada’s bioethanol production capacity is currently about 245 million L but expected to grow to more than 1 bill L by 2008 [15]. Various other coun­tries around the world have increased bioethanol production significantly since the mid-1990s. The dominant emerging bioethanol producers include China, which is home to Jilin Fuel Alcohol, the world’s largest corn-based bioethanol plant with a current capacity in excess of 350 million L year-1. The development of biofuel capacity over the past quarter century may be seen in Fig. 1.

As bioethanol is the most dominant biofuel found today, it is useful to look at the policies that supported development of this fuel in different jurisdic­tions around the world, and to evaluate the impact that different policies may have on creating increases in production capacity.

2.1

There has recently been considerable interest in the redirection of metabolism in bacteria such as E. coli for the overproduction of specific metabolites and higher value products. At a commercial level, the large-scale production by Tate & Lyle/Dupont of 1,3-propandiol using a highly engineered strain of E. coli is indicative of an increasing trend towards such bio-based pro­cesses. The fast specific rates of sugar uptake by Z. mobilis, its highly efficient metabolism for a specific product (ethanol), and its relatively small genome size (facilitating genetic manipulation) may make it an ideal candidate for producing other metabolites via its genetic engineering.

As shown in Fig. 7, Z. mobilis has an incomplete TCA cycle and the po­tential exists via “knock out” mutation to redirect metabolism away from end-products such as lactate and ethanol, towards higher value products like succinic acid. As reported recently by Kim et al. [42], succinic acid overproducing Z. mobilis strains have been developed by disruption of the genes for pyruvate decarboxylase (pdc) and lactate dehydrogenase (ldh). Such strains can produce relatively high concentrations of succinic acid at yields of 1.73 mole/mole glucose (86% theoretical). The yield was reported to be more than 30% greater when compared to those of other succinic acid — producing bacteria such as Actinobacillus succinogenes and Mannheimia suc- ciniciproducens (about 1.34 mole/mole glucose). These strains of Z. mobilis were also reported to exhibit higher overall rates of succinic acid production (1.62 gL-1 h-1) under Na bicarbonate supplemented conditions compared to those of other succinic acid producing bacteria (1.35 gL-1 h-1).

Entner-Doudoroff Pathway

f

I

2-keto-3-deoxy-6-P-gluconate

4.3

4.1

Cost Effective Growth Media

In order for ethanol production to be commercially feasible, the growth media cost should be kept at a minimum. In addition to engineering strains to require less nutritional supplementation, the design of simpler, and therefore cheaper, growth media is important for the expansion ofbioethanol production.

AM1 [76] and NBS mineral salts media [77] are two simple mineral salts media developed in our laboratory. Both have been shown to support high levels of cell growth and ethanol production. AM1 is a derivative of NBS, with a 65% reduction in salts. With low total alkali (4.5 mM) and total salts (4.2 gL-1), AM1 was able to support production of ethanol from xylose and lactate from glucose with average productivities of 18 -19 mmol L-1 h-1.

OUM1 medium contains corn steep liquor, mineral salts, and urea as sole nitrogen source; K. oxytoca BW21 produced over 40 gL-1 ethanol (0.47 g ethanol per gram glucose) in this medium within 48 h [57]. The use of urea as sole nitrogen source has the benefit of cost reduction while also reducing media acidification [78].

On-site preparation of crude yeast autolysate from spent yeast offers poten­tial synergy between grain-based and lignocellulosic processes. Preparation of this autolysate, optimization of the resulting media, and ethanol production by KO11 were demonstrated by [31], with ethanol yields comparable to LB.

K. oxytoca is able to utilize urea as sole nitrogen source, where urea has roughly half the cost of ammonium on an equivalent nitrogen basis. Addition­ally, because urea metabolism does not contribute to media acidification [79], the use of urea reduces the cost of pH control. With the goal of reducing the nutrient cost of K. oxytoca-based ethanol production, optimized urea medium (OUM1) was developed. In addition to containing urea as the sole nitrogen source, OUM1 contains corn steep liquor, mineral salts, and glucose [57].

4.2

The effluent from bioethanol production still contains a large amount of or­ganics that are not composed of carbohydrates. Anaerobic digestion (AD) has been used for a long time to treat organic waste streams with a high concen­tration of organic matter. The benefits of anaerobic treatment are stabilization of the waste stream, the high reduction of organic matter, and the production of methane, which can be used as energy source [25]. This gives an overall positive energy balance of the waste treatment process compared to aerobic waste treatment. The income from the methane produced after bioethanol production constitutes a value corresponding to a lowering of the ethanol production price by 34%.

The effluent from the fermentation step of bioethanol production con­tains low-molecular weight lignin degradation products primarily generated during the physical-chemical pretreatment. These aromatic compounds are generally difficult to degrade under anaerobic conditions. Furthermore, a re­peated reuse of the process water has the potential to cause a build up of these fermentation inhibitors. It is therefore important to achieve an anaero­bic purification technology that is able to remove these compounds from the process water. Experiments in our laboratory have shown that all problematic organic components can be removed in the anaerobic step. The low hydraulic retention times and the removal of organics are of great importance, looking at the overall process feasibility [26].

The Maxifuel concept has been implemented at pilot scale at the Techni­cal University of Denmark, DTU (Fig. 5) and the concept is planned to go into demonstration phase in 2008.

The plant is dimensioned to convert 150 kg dry biomass/day, and con­sists of 17 tanks (fermentation, reactors, and holding tanks). The ethanol fermentation takes place in two 2700 L fermenters. The plant includes all pro­cesses from straw to ethanol, and was brought into operation in the autumn of 2006.

4.6

P. L. Rogers1 (И) • Y. J. Jeon1 • K. J. Lee2 • H. G. Lawford3

School of Biotechnology and Biomolecular Sciences, UNSW, 2052 Sydney, Australia p. rogers@unsw. edu. au

2School of Biological Sciences, Seoul National University, 151-742 Seoul, Korea 3Department of Biochemistry, University of Toronto, Toronto Ont., M5S 1A8, Canada

1 Introduction……………………………………………………………………………………………… 264

2 Development of Recombinant Strains of Z. Mobilis…………………………………………. 265

2.1 Increased Substrate Range Through Expression

of a Single Heterologous Gene……………………………………………………………….. 265

2.2 Strain Construction for Utilization of C5 Sugars………………………………………….. 266

2.3 NMR Analysis of Metabolic Characteristics of Recombinant Strains…. 269

2.4 Kinetic Characteristics of Recombinant Strains…………………………………………….. 269

2.5 Kinetic Model Development……………………………………………………………………….. 273

2.6 Effect of Inhibitors in Lignocellulosic Hydrolysates………………………………………. 274

2.7 Application to Industrial Raw Materials……………………………………………………… 275

3 Genome Sequence of Z. Mobilis…………………………………………………………………….. 278

4 Applications for Higher Value Products……………………………………………………… 278

4.1 Metabolites and Related Products………………………………………………………………. 278

4.2 Metabolic Engineering for Organic Acids and TCA Cycle Intermediates. . 279

4.3 Enzyme Based Biotransformations……………………………………………………………… 281

4.3.1 Sorbitol/Gluconate Production…………………………………………………………………… 281

4.3.2 Pharmaceutical Intermediates and Fine Chemicals………………………………………. 282

5 Discussion and Conclusions……………………………………………………………………….. 283

References…………………………………………………………………………………………………….. 286

Abstract High oil prices, increasing focus on renewable carbohydrate-based feedstocks for fuels and chemicals, and the recent publication of its genome sequence, have provided continuing stimulus for studies on Zymomonas mobilis. However, despite its apparent advantages of higher yields and faster specific rates when compared to yeasts, no com­mercial scale fermentations currently exist which use Z. mobilis for the manufacture of fuel ethanol. This may change with the recent announcement of a Dupont/Broin part­nership to develop a process for conversion of lignocellulosic residues, such as corn stover, to fuel ethanol using recombinant strains of Z. mobilis. The research leading to the construction of these strains, and their fermentation characteristics, are described in the present review. The review also addresses opportunities offered by Z. mobilis for higher value products through its metabolic engineering and use of specific high activity enzymes.

Keywords Ethanol production • Glycose/Xylose fermentations • Higher value products • Lignocellulosics • Metabolic engineering • Zymomonas mobilis

1

Introduction

Zymomonas mobilis has attracted considerable interest over the past decades as a result of its unique metabolism and ability to rapidly and efficiently pro­duce ethanol from simple sugars. An early paper by Millis [1] characterized the role which Zymomonas sp. play in causing cider sickness and a compre­hensive review by Swings and DeLey [2] provided much of the background for the subsequent stimulus in research activity in the early 1980s which fol­lowed the first of the “oil price shocks”. Further reviews over the ensuing decades [3-9] included extensive data on genetic and kinetic characteriza­tion of strains of Zymomonas mobilis capable of growing on an increasingly wide range of sugars. In a fine example of metabolic (pathway) engineering, recombinant strains of Z. mobilis were reported in 1995/6 from the National Renewable Energy Laboratory (NREL) Golden, CO, USA, that were capable of the efficient conversion to ethanol of the C5 sugars, xylose and arabinose present in lignocellulosic hydrolysates [10,11]. Most recently, the reporting of the complete genome sequence of Z. mobilis ZM4 (ATCC 31821) [12] has opened up further potential for strain enhancement and for its use for higher value products.

Table 1 provides an outline of the key research milestones which have oc­curred for Z. mobilis over the past three decades with the present review focusing particularly on those developments which have been reported over the past 5-10 years.

2

The oldest example of widespread biofuel development is found in Brazil, which produces bioethanol from sugar- or starch-based material in the form of sugarcane and sugarcane residues. Because of Brazil’s optimal climate, two seasons of sugarcane growth can be achieved, adding greatly to the poten­tial production of both sugar and bioethanol products. In response to the first oil crisis of the 1970s, Brazil invested heavily in fuel alcohol primarily as a means of increasing fuel security and saving foreign currency on petroleum purchases. The original policy choice was to create direct funding sources to create biofuel capacity. In 1975, a diversification program for the sugar in­dustry called Proalcool was created with large public and private investments supported by the World Bank, allowing expansion of the sugarcane plantation area and construction of alcohol distilleries, either autonomous or attached to existing sugar plants [19].

The second group of policies introduced in Brazil provided a subsidy for bioethanol use. Two related financing schemes were organized to guarantee fuel sale price; the FUPA program guaranteed US $ 0.12 L-1 for E22 (a blend of 22% ethanol in gasoline), while the FUP program provided US$ 0.15 L-1 for E100 (or pure, anhydrous ethanol) fuel. By 1996/97, the total subsidy de­livered via these programs reached about US $ 2 billion year-1 [19].

The presence of a renewable fuel standard and of strong subsidies to E100 production, combined with the second oil shock of the early 1980s, resulted in the successful adaptation of engines to E100 fuel use. By 1984, E100 vehicles accounted for 94.4% of domestic automobile manufacturers’ production, and in 1988 participation in the E100 program reached 63% of total vehicle use in the country [20]. The upward trend ended, however, when high global sugar prices led to a crash in availability of fuel alcohol, resulting in a consumer shift away from E100 vehicles.

From 1989 to 1996, the sugar export market was very strong, and thus the cost of sugar to the bioethanol industry soared and fuel bioethanol short­ages resulted. In response, the Brazilian government made a failed attempt to restrict sugar exports, and then announced that the fuel market would be deregulated as of 1997. While deregulation began with E100 fuels, subsi­dies for blended fuels remained in place for an additional period, which had the effect of increasing overall alcohol production at the time. When price controls on E22 were removed in 1999, however, the prices for bioethanol collapsed [19].

Faced with an excess of bioethanol and collapsed prices at home, major producer groups joined together to form Brasil Alcool SA in March 1999, and made the decision to export excess bioethanol at any price. Later that year, a mechanism to create a monopoly on fuel bioethanol named Bolsa Brasileira de Alcool Ltda was created by the founders of Brasil Alcool. This monopoly drove a dramatic increase in bioethanol export prices for a period after its inception, with prices doubling within a year [20]. Since 1999, the total production of bioethanol in Brazil has risen; this trend has been driven by the expansion of export markets for bioethanol, rising world prices for oil, and an increase in domestic oil supply. The Brazilian industry today follows a simple biorefinery model, where the production of a combination of prod­ucts, including refined sugar, bioethanol, and energy from the combustion of sugarcane residues (bagasse) improves both economic and environmen­tal performance. Brazil controls more than 75% of the world’s export market, with primary exports going to the USA, Europe, Korea, and Japan; Brazil’s es­timated total exports will be approximately 3.1 billion L in 2006 [12]. Many countries that lack significant biomass resources, such as Japan, have made Brazilian bioethanol a part of their renewable fuel strategies.

Brazil’s domestic market still utilizes the single largest portion of fuel bioethanol capacity in the country. The presence of a Renewable Fuel Stan­dard means that all Brazilian gasoline has a legal alcohol content requirement that has ranged between 20% and 25% (currently 23%, as of 20 November 2006) [21]. Most vehicles are being run on E20 or E22, but sales of flex-fuel vehicles capable of operating on E85 blends are strong. Brazil has developed a unique distribution infrastructure for this fuel, with a network of more that 25 000 gas stations with E20 pumps.

Today, Brazil remains a dominant bioethanol producer and the single larg­est exporter of this fuel, with shipments expected to hit a record 3 billion L in the 2006-07 harvest. Rising demand for bioethanol — in part caused by poli­cies in other countries — has created an impetus for new product capacity. Recently, it was reported that UNICA plans to open 77 new bioethanol plants by 2013, adding to the existing 248 plants. When complete, this will raise the country’s production capacity to about 35.7 billion L [21].

2.2

4.3.1

Sorbitol/Gluconate Production

The production of sorbitol by Z. mobilis when grown on sucrose or a mixture of glucose and fructose has been reported earlier by several groups [94,95]. In subsequent studies on the mechanism of sorbitol production, an enzyme complex was identified by Leigh et al. [96] which was capable of oxidizing glucose to gluconic acid concomitant with the reduction of fructose to sor­bitol. This enzyme was described as a glucose-fructose oxidoreductase with a tightly coupled (non-dialyzable) co-factor identified as NADP [97]. The mechanism for sorbitol/gluconic acid production and the associated enzymes are shown in Fig. 8 with the pathway from gluconate to ethanol not being functional if cells of Z. mobilis are fully permeabilized. As shown in Fig. 8, the possibility exists also of producing a mixture of sorbitol and gluconolactone if gluconolactonase activity is deleted.

Kinetic studies have been reported for a 60% sugar solution (300 gL-1 glucose and 300 g L-1 fructose) using toluene-treated permeabilized cells of Z. mobilis in which a sorbitol concentration of 290 g L-1 and a gluconic acid concentration of 283 gL-1 were achieved after 15 h in a batch process [98]. A continuous process with immobilized cells was developed with only a small loss of enzyme activity (less than 5%) evident after 120 h. With a strongly basic anion exchange resin and a buffer system at pH = 9.0, good separation of sorbitol and gluconic acid was achieved. Subsequent studies using immo-

4.3.2

In order to attain the desired high product titers, biocatalysts must be sup­plied with high levels of sugars. These high sugar levels in turn create osmotic stress, which is compounded by the desire to use simple mineral salts media. Osmolytes such as trehalose, betaine, proline, and glutamate help bacteria maintain appropriate cell turgor and volume despite changes in extracellu­lar osmolality; osmolyte uptake and synthesis are reviewed in [80]. Increased activity of the native trehalose synthesis pathway elevated the growth rate of E. coli W3110 in the presence of various osmotic stress agents [81], and betaine supplementation increased the production of D-lactic acid by E. coli SZ132 in NBS mineral salts media [42]. A combination of betaine supple­mentation and elevated trehalose synthesis increased the tolerance of W3110 to xylose, glucose, sodium lactate, and sodium chloride more than betaine supplementation or elevated trehalose synthesis alone [82].

As described above, the poor performance of ethanologenic E. coli strain KO11 in minimal media has been attributed to NADH-mediated inhibition of citrate synthase, limiting the availability of glutamate, a protective os — molyte [33]. Additionally, the increased performance of LY01 relative to KO11 can be partly attributed to increased osmolyte production and uptake [83]. NMR analysis confirmed that intracellular pools of glutamate, trehalose, and betaine are very low in KO11 during anaerobic growth relative to aerobic growth in the same medium [84]. Growth and ethanol production of KO11 was increased by supplementation with various osmolytes, demonstrating that the glutamate limitation is related to osmotic stress, not to a specific metabolic demand for glutamate [84].

4.3

A further development of biorefineries is the use of hybrid techniques com­bining biological conversion with catalytic downstream processing (Wester — mann P, J0rgensen B, Lange L, Ahring BK, Christensen CH (2007) Int J Hydro­gen Energy, accepted for publication). For instance, highly efficient autother­mal reformers capable of converting 1 mol ethanol to 5 mol hydrogen have recently been demonstrated [27]. Since 2 mol of ethanol can be achieved for each sugar molecule, the hydrogen yield of this two-step process is 83% of the theoretical maximum, compared to the 10-20% achieved by direct hydrogen fermentation. Hydrogen produced in the thermophilic ethanol fermentation process described above would add to this yield, approaching the theoretical maximum yield of 12 mol hydrogen/mol monosaccharide.

Hydrogen has been suggested as a future energy carrier to succeed the fossil fuel era [28]. The introduction of downstream catalytic conversion of biofuels leaves the possibility of combining a less complex fuel handling technology (ethanol instead of hydrogen) for transportation purposes with all the benefits of the fuel cell technology. In the transition period before a hydrogen-based energy economy has been realized, a gradual change to the use of renewable energy can be facilitated by the use of catalytically con­verted biofuels in existing internal combustion engines. Although ethanol in even high ethanol:gasoline mixtures can be used for ground transportation with few modifications of the engines, biogasoline produced by catalytic con­version of methane and bioethanol will have potential use as a high energy alternative for aviation and air transport. If these transportation means are sustained in the future, the availability of safe liquid fuels with high energy content storable under ambient conditions is a prerequisite.

4.7

2.1

Increased Substrate Range Through Expression of a Single Heterologous Gene

One of the possible disadvantages of Z. mobilis is that it has a limited carbon substrate range as it can only use the simple C6 sugars glucose, fructose and sucrose. As a result early studies on its genetic manipulation focused on ex­tending its substrate range for ethanol production. Skotnicki et al. [18] first reported high frequency conjugal transfer of plasmids from Escherichia coli and Pseudomonas aeruginosa, and this was followed by expression of the lac Z gene and production of ^-galactosidase in strains of Z. mobilis [21,22]. How­ever, the strain ZM6100 (RP1:Tn 951) derived from this work was shown to progressively lose all plasmid markers in batch culture under non-selective conditions. Subsequently a new strain, ZM6306, was developed in contin­uous culture which showed 100% stability for all plasmid markers when grown without selection pressure. Synthesis of ^-galactosidase was induced in continuous culture by addition of lactose resulting in increased ethanol production and unutilized galactose [23].

Further studies to extend the substrate range were reported which involved the cloning and expression of a в-glucosidase gene from Xanthomonas al — bineans [24] and a-glucosidase gene from a Bacillus sp [44], however enzyme expression levels were low.

2.2