There are several organisms that can accumulate fumaric acid when given the proper conditions.Rhizopus nigricanswas the first fungus described to produce fumaric acid by Ehrlich (1911). The main organism used in industry and most extensively described in the literature is R. oryzae earlier termed R. arrhizus. A survey of Mucorales
identified in addition toRhizopusalsoMucor,Cunninghamella, andCircinellaspecies as able to accumulate fumaric acid.
Rhizopus oryzaeis a group of fungi with considerable variability described for different isolated species. This has led to an ongoing discussion, with frequent sug- gestions of new species names and with recurrent changes in names. A revision of the genus reported in 1984 used strains of the Dutch CBS culture collection (Schipper 1984). The criteria used were morphological and species delimitation was derived from mating experiments. Abe et al. (2007) suggested to include the organic acid production profile as a criterion for the taxonomy of the genus. Two distinct groups were established based on the analysis of the internal transcribed spacer of rDNA and among others the genes for lactate dehydrogenase B, actin and translation elongation factor-1α. The two acid patterns (fumaric/malic and lactic) correlated with these sequences, and with the presence of 1 or 2 lactate dehydrogenases (ldhAandldhB, see section 15.4). It was suggested to reclassify the fumaric–malic acid producers to Rhizopus delemarand keepR. oryzaefor lactic acid producers. The proposed clas- sification is rarely used in the literature. The synonymy ofR. arrhizusandR. oryzae is well established, and even though the former was described earlier, the preferred name today isR. oryzae, which is used in this chapter.
Different strains ofR. oryzaehave been used for research of fumaric acid produc- tion. The research in the public domain is on a laboratory scale using either shake flasks or laboratory-sized fermentors. The majority of studies have been performed on two different strains, namely ATCC 10260 (NRRL 1526) and ATCC 20344 (NRRL 6400). It is not clear which of these strains is the favored strain on an industrial scale.
It is most probable that mutants with better characteristics will be developed for use in an industrial process. In one study mutants were isolated after treatment with nitrogen ion implantation. Single-layer spore suspensions on plates were implanted by N+beams with a dose of 2×1015ions/cm at 10 keV energy. The isolated mutants had enhanced glucoamylase activity. This allowed for the use of starches as direct carbon source for fumaric acid accumulation with simultaneous saccharification and fermentation resulting in 44.1 g/L fumaric acid from an initial total sugar concentra- tion of corn starch of 100 g/L (Deng et al., 2012). Using femtosecond laser irradiation mutants were isolated that produced fumaric acid to a concentration of 49.4 g/L with a yield of 0.56 g fumaric acid per gram glucose (Yu et al., 2012). This should be compared to the final concentrations reported when fatty acid where added either as Tweens (40 and 80) or corn, soybean, or cottonseed oil with a maximal fumaric acid concentration of 60 g/L with a productivity of 0.63 g/L⋅h in shake flasks (Goldberg and Stieglitz, 1985). In a laboratory size fermentor, 107 g/L of fumaric acid with the highest literature value for productivity of 2 g/L⋅h were obtained by Ng et al.
(1986), as compared to values of 49.4 g/L and 0.59 g/L⋅h, respectively, for the mutant recently described (Yu et al., 2012).
Roa Engel et al. (2008) reviewed data for fumaric acid accumulation by different strains ofRhizopus. The most efficient process was with a rotary biofilm contractor studied by Cao et al. (1996) with volumetric productivities of 4.25 g/L⋅h compared with 2 g/L⋅h for a stirred bioreactor (Ng et al., 1986). The fungus is grown to form a biofilm by self-immobilization on plastic discs, the formed acid is absorbed on a
resin to avoid feedback inhibition, and the organic acid recovery is integrated with the fermentation. The scaleup of such a system would involve novel technological solutions for an integrated process (Fu et al., 2010).
15.2.2 Carbon Sources
The expense of the feedstock for production, especially the carbon source, has a decisive influence on the price of low value–high volume compounds such as fumaric acid. The main production of fumaric acid is by synthetic means based on petroleum- derived maleic anhydride as the raw material. The price of maleic anhydride is two and a half times more expensive than the main fermentation process raw material, glucose (prices in the beginning of 2012).Rhizopus oryzaeis capable of utilizing simple sugars such as glucose; however, sucrose is poorly metabolized byR. oryzae and xylose as carbon source gives low fumaric acid productivities. There are reports that the fungus can also utilize a range of low cost substrates that are less refined than the simple sugars, such as starches, molasses, corn mash, corn steep liquor, potato flour, rice bran, and cassava bagasse.
15.2.3 Solid-State Fermentations
Processes with solid-state fermentations have been investigated using among others corn distillers grain with solubles, where heat or acid hydrolysis treatment of the raw material was required to obtain high concentrations of fumaric acid (West, 2008). The solid-state fermentation is slower than the submerged fermentation, but the capital cost of the former is considerably lower. Cheap substrates such as acid hydrolysate of dairy manure resulted in low yield of 0.15 g/g with a final concentration of 4.9 g/L of fumaric acid (Liao et al., 2008). The use of hydrolysates of lignocellulosic materials have the inherent problem of diauxie (sequential utilization) resulting from the blend of mainly two sugar monomers (glucose and xylose) in hydrolysates. The productivity of fumaric acid formation is so low that the time required for its accumulation puts the economics of the process in question. To circumvent sequential utilization, xylose was used for growth ofR. oryzaeand glucose for biosynthesis of the fumaric acid (Xu et al., 2010). The initial dilute acid hydrolysis of corn straw released mainly xylose, the main component of hemicelluloses, followed by enzymatic digestion to obtain the glucose from the cellulosic fraction of the lignocellulose. The increase in xylose concentration in hydroylsates had a negative effect on the subsequent fumaric acid accumulation stage, probably due to a higher concentration of inhibitors present in these hydrolysates. The maximal productivity obtained was 0.37 g/L⋅h using mild acid hydrolysates for growth and adding pure glucose for the second stage of fumaric acid accumulation with a final concentration of 30 g/L of fumaric acid (Xu et al., 2010).
When using starchy raw material (potato, cassava, and corn) it would be advantageous to have a simultaneous saccharification and fermentation process (SSF), to save on extra enzyme costs. Isolation of strains with glucose insensitive glucoamylase was achieved by isolating glucose analog resistant strains and yields of 0.49 g/g were obtained for SSF of corn starch without any addition of extra enzymes (Deng et al.,
2012). A potentially cheaper substrate is the utilization of cellulose in an effective manner (Xu et al., 2010). Use of the fast-growing evergreen treeEucalyptus globulus as raw material, with limited lignin content and mainly xylose in the hemicellulose fraction, the fumaric acid yield decreased from 0.71 g/g using synthetic media in a fed batch mode to 0.35 g/g when hydrolysates ofE. globuluswere used (Rodriguez- Lopez et al., 2012). With ion-exchange-treated hydrolysates and addition of 15% of the carbon source as glucose it was possible to increase the yield to 0.44 g/g. These results are representative of similar problems encountered in other fermentations of low value/high volume fermentation products, where hydrolysates of lignocellulose used directly as fermentation substrates contain inhibitors that reduce the yields of the desired products (Liu and Blaschek, 2010).
15.2.4 Submerged Fermentation Conditions
Certain fermentation conditions for fumaric acid are required to obtain high pro- ductivity. Among the parameters reported in the literature are pH, morphology, and neutralizing agent. The production media are nitrogen, but not carbon, limited (Gold- berg et al., 2006).
15.2.4.1 Effect of pH Fermentation close to neutral pH during the production phase of the fermentation is ideal for high productivity, yield, and final concentration of fumarate. The disadvantages are the waste salts produced. High base concentra- tions are required to maintain the pH close to neutral. To solubilize the accumulated salt, sulfuric acid is used and the free fumaric acid is obtained with the formation of stoichiometric amounts of gypsum (CaSO4). The possibility to perform the fer- mentation at lower pH would have a great impact on the economics of a large-scale process for fumaric acid fermentations. At lower pH larger fractions of the undisso- ciated form of the acid would be formed with low solubility enabling separation by crystallization without the formation of waste salts (Roa Engel et al., 2011). In shake flasks, pH had strong effect on fumaric acid accumulation. When pH was kept at 5.0, 30 g/L of fumaric acid accumulated compared to 9.4 g/L at pH 3.0. Incubation at lower pH resulted in high accumulation of byproducts like glycerol and ethanol (Roa Engel et al., 2011).
15.2.4.2 Morphology of Fungus The suggested morphology for fumaric acid accumulation usingR. oryzaeis the formation of small pellets (<1 mm in diameter), found to be problematic to achieve by inoculation of spores directly into fermentor vessels (Roa Engel et al., 2011). It is known that fungal morphology is influenced by many factors such as medium composition, inoculums size, pH, medium shear, additives (polymers, surfactants, and chelators), culture temperature, and medium viscosity (Papagianni, 2004). ForR. oryzaein shake flasks pH, shaking frequency, and working volume were the important factors that determined the pellet formation and size (Roa Engel et al., 2011). There is no consensus in the literature on the optimal morphology for fumaric acid formation and neither the exact conditions that will lead to the desired pellet formation. Decrease in pH and higher agitation in large
shake flasks was beneficial for obtaining smaller pellet diameter (Roa Engel et al., 2011), whereas CaCO3addition affected the pellet diameter but not pH in another study (Liao et al., 2008). In a soybean peptone medium with inorganic ions growing R. oryzaeat 30◦C at pH 5, a pellet size of 0.55 mm resulted in the highest fumaric acid accumulation (39.6 g/L in shake flasks, 50 mL in 250 mL flask, and 35.4 g/L in a stirred tank of 7 L both at 200 rpm (Zhou et al., 2011). Transfer of the seed culture in several steps at pH between 2.5 and 4 at 0.5 incremental steps resulted in good pellet formation (up to 1.2 mm in diameter; Fu et al., 2009); however, there is no agreement on how to achieve uniform pellets and different strains and environmental conditions have been used. The challenge is to determine optimal morphology and a way to obtain that morphology for fumaric acid accumulation conditions on an industrial scale.
15.2.4.3 Neutralizing Agent Attempts have been made to use other neutraliz- ing agents than CaCO3, which has been shown to be optimal, with negative results in general (Zhou et al., 2002). CaCO3has several disadvantages, like the low solubility of Ca-fumarate which contributes to the increase in viscosity of the fermentation broth. Sodium salts such as Na2CO3decrease the downstream costs, and thorough cost analysis (ISO 14044, 2006) of the total process is required to determine which neutralizing agent is most economical to use at the industrial level. It should be noted that the solubility of Na-fumarate is 22% w/v, high compared to Ca-fumarate, and its use inhibits fumarate formation (see section 15.5).
15.2.5 Transport of Fumaric Acid
Transport of fumaric acid outside of the fungal cell is poorly understood. It is impor- tant to understand how and at what conditions fumaric acid is transported out of the cell. The ambient pH is one of the main factors found to influence the excretion of organic acids in filamentous fungi and there are different hypothesis on the mech- anisms involved (overflow metabolism hypothesis, charge balance hypothesis, and aggressive acidification hypothesis all detailed in Vrabl et al., 2012).
15.2.6 Production Processes
The biological production of fumaric acid usingR. oryzae was carried out on an industrial scale by Pfizer (see section 15.1.2). With the rapid development of the petrochemical industry, the fermentation route to fumaric acid became uneconom- ical and until today, to the best of our knowledge, all fumaric acid is produced through catalytic oxidation of benzene or butane to maleic anhydride hydrolyzed to maleic acid followed by isomerization to fumaric acid. There is no information available in the public domain to describe current efforts to develop an indus- trial biological process, but there are several research groups that are reporting on attempts to increase fumaric acid accumulation by isolation of high produc- ing mutants as well as using cheaper substrates and also decrease cost and waste material formation of the downstream processes (Huang et al., 2010; Kang et al., 2010). In no case are the newly isolated mutants as effective in production of
fumaric acid as earlier reports using wild-type strains for study of fumaric acid accumulation (Rhodes et al., 1959, 1962; Ng et al., 1986). When corn, at the begin- ning of this century was still at a surplus (before its use as raw material for bio- fuel, ethanol) research was funded for its use as a raw material for fumaric acid production (http://www.reeis.usda.gov/web/crisprojectpages/192861.html, retrieved June 14, 2012)