Metabolic Engineering for Fumaric and Malic Acids Production

Home , Ammonium fumarate, Maleate isomerase, Maleic acid, Malic acid, Mesaconic acid

Dissertation

Presented in Partial Fulfillment of the Requirement for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University

Baohua Zhang, M.S.

Ohio State Biochemistry Program

The Ohio State University

2012

Dissertation Committee:

Professor Shang-Tian Yang, Advisor

Professor Jeffrey Chalmers

Professor Hua Wang

Baohua Zhang

2012

ABSTRACT

Fumaric acid is a natural organic acid widely found in nature. With a chemical structure of two acid carbonyl groups and a trans-double bond, fumaric acid has extensive applications in the polymer industry, such as in the manufacture of polyesters, resins, plasticizers and miscellaneous applications including lubricating oils, inks, lacquers, styrenebutadiene rubber, etc. It is also used as acidulant in foods and beverages because of its nontoxic feature. Currently, fumaric acid is mainly produced via petrochemical processes with benzene or n-butane as the feedstock. However, with the increasing crude oil prices and concerns about the pollution problems caused by chemical synthesis, a sustainable, bio-based manufacturing process for fumaric acid production has attracted more interests in recent years.

Rhizopus oryzae is a filamentous fungus that has been extensively studied for fumaric acid production. It produces fumaric acid from various carbon sources under aerobic conditions. Malic acid and ethanol are produced as byproducts, and the latter is accumulated mainly when oxygen is limited. Like most organic acid fermentations, the production of fumaric acid is limited by low productivity, yield and final concentration influenced by many factors including the microbial strain used and its morphology, medium composition, and neutralizing agents. Many efforts have been done to improve fumaric acid production through optimization of the fermentation process. The goal of

ii this project was to use metabolic engineering to modify the fumaric acid biosynthesis pathway to increase fumaric acid production in R. oryzae. This is the first attempt to apply genetic engineering strategies to change the metabolic flux towards fumaric acid biosynthesis.

The biosynthesis of fumaric acid in R. oryzae takes place in the cytosol and three enzymes: pyruvate carboxylase (PYC), malate dehydrogenase and fumarase are involved in the reaction. PYC is situated at the branch point of pyruvate metabolism, and its role in affecting fumaric acid and various metabolites accumulation was thus studied in this work. An expression plasmid containing native R. oryzae pyc gene, encoding pyruvate carboxylase, was transformed into the uracil auxotroph R. oryzae M16. Two transformants were obtained: pyc3 and pyc5, and were verified by Southern hybridization.

Compared to the wild type, the PYC activity in the pyc tranformants increased 56%-83%.

However, the pyc transformants grew poorly and had a low fumaric acid yield of less than 0.05 g/g glucose due to the formation of large cell pellets that limited oxygen supply and resulted in the accumulation of ethanol with a high yield of 0.13-0.36 g/g glucose.

An exogenous phosphoenolpyruvate carboxylase (PEPC) was introduced in R. oryzae with the aim to increase CO2 fixation and the carbon flux toward Oxaloacetate.

The pepc gene encoding PEPC was expressed in R. oryzae under the endogenous pgk1 promoter and pdcA terminator. The obtained pepc transformants exhibited significant

PEPC activity of 3-6 mU/mg that was absent in the wild type. Compared to the fermentation kinetics of the wild type, the fumaric acid production by the pepc transformant increased 26% (0.78 g/g glucose vs. 0.62 g/g for the wild type).

iii

Fumarase catalyzing the final step of fumaric acid biosynthesis in R. oryzae was overexpressed to investigate its effects on cell growth and fumaric acid production. Three fumR fragments with different lengths of 5’ and 3’ untranslated regions (UTR) were used to express the fumR gene in R. oryzae. All transformants showed significantly increased fumarase activity during both the seed culture and fermentation stages. However, fumarase overpression yielded more malic acid, instead of fumaric acid in the fermentation. It was attributed to the catalytic prevalence on the direction of fumaric acid to malic acid by the overexpressed fumarase. The results suggested that the overepxressed fumarase by itself was not responsible for the overproduction of fumaric acid in R. oryzae.

Fumaric acid is an intermediate in the succinic acid biosynthesis pathway in E. coli fermentation under anaerobic conditions. The fumarate reductase, encoded by frd, was disrupted in E. coli KJ060, a high succinic acid producer, to study its effect on the metabolic flux distribution. The frd gene was removed from E. coli chromosome through homologous recombination. Under anaerobic conditions, the frd disrupted mutant of E. coli KJ060M produced little succinic acid. However, it did not produce much fumaric acid either. Under aerobic conditions, the mutant produced malic acid as the main product, with a high yield of 0.72 g/g glucose, higher than that (0.65 g/g) produced by the parental strain. A high malic acid concentration of 48.4 g/L was produced in fed-batch fermentation in a 5-liter fermenter with a productivity of 2.38 g/L·h. The mutant thus has the potential for use in industrial production of malic acid, which is widely used as acidulants in foods and beverages.

Dedicated to my parents and husband

ACKNOWLEDGEMENTS

Firstly, I would like to express my special thanks and gratitude to my advisor, Dr.

Shang-Tian Yang, who gave me guidance, encouragement and support throughout my graduate study. I learned a lot from his expertise in science and deep insights in my research.

I would also like to thank Dr. Jeffrey Chalmers and Dr. Hua Wang for their time being on my committee and their valuable advices on my research project.

I want to give my sincere thanks to all the previous and current laboratory members in our research group, especially Dr. Mingrui Yu, Mr. Kun Zhang, and Mr. Zhongqiang

Wang for their helpful suggestions and support.

In addition, I would like to specially thank Dr. Christopher D. Skory from USDA for providing experimental strains, plasmids as well as valuable technical advice and assistance. I would also like to thank Dr. Lonnie O. Ingram from University of Florida for providing E. coli strains.

Financial supports from the United Soybean Board and the Consortium for Plant

Biotechnology Research, Inc. (CPBR) to this research are also acknowledged.

Finally, I wish to thank my parents for all their emotional support and my husband for helping me get through the difficult times. My entire family and all my friends are acknowledged in the depth of my heart.

VITA

1999 - 2003…………………………………...B.S. Biological Science, Nankai University 2003 - 2006…………………………………...... M.S. Microbiology, Nankai University 2006 - 2007……………………..Graduate Fellowship, Ohio State Biochemistry Program The Ohio State University 2007 – 2012………………………………………………….Graduate Research Associate, Chemical & Biomolecular Engineering The Ohio State University

PUBLICATIONS

1. Zhang, B., Skory, C.D., Yang, S.T. Metabolic engineering of Rhizopus oryzae: Effects of overexpressing pyc and pepc genes on fumaric acid biosynthesis from glucose. Metabolic Engineering. 2012. In Press. 2. Zhang, B., Yang, S.T. Metabolic engineering of Rhizopus oryzae: Effects of overexpressing fumR gene on cell growth and fumaric acid biosynthesis from glucose. Process Biochemistry. 2012. Accepted for publication. 3. Yang, S.T., Zhang, K., Zhang, B., Huang, H. Biobased Chemicals - Fumaric Acid. In: Moo-Young M (ed.) Comprehensive Biotechnology, 2nd edition. 2011. pp.163-177. 4. Wang, Z., Feng, S., Huang, Y., Qiao, M., Zhang, B., Xu, H. Prokaryotic expression, purification, and polyclonal antibody production of a hydrophobin from Grifola frondosa. Acta Biochim Biophys Sin (Shanghai). 2010, 42:388-95. 5. Yu, L., Zhang, B., Szilvay, G.R., Sun, R., Jänis, J., Wang, Z., Feng, S, Xu, H., Linder, M.B., Qiao, M. Protein HGFI from the edible mushroom Grifola frondosa is a novel 8 kDa class I hydrophobin that forms rodlets in compressed monolayers. Microbiology. 2008, 154:1677-1685. 6. Yu, L., Shao, B., Zhang, B. Isolation and purification of Trichoderma reesei hydrophobin HFBI. Food and Fermentation Industries. 2005, 31:129-132.

FIELDS OF STUDY

Major Field: Biochemistry Specialty: Biochemical Engineering

vii

Table of Contents

Abstract ...... ii

Dedication………………………………………………………………………………....v

Acknowledgements ...... vi

Vita ...... vii

List of Figures ...... xv

List of Tables ...... xviii

1. Introduction ...... 1

1.1 Objectives ...... 3

1.2 Scope of study ...... 5

1.3 References ...... 8

2. Literature review ...... 14

2.1 Fumaric acid production...... 14

2.1.1 Fumaric acid properties and applications ...... 14

2.1.2 Production of fumaric acid ...... 17

2.1.3 Fumaric acid producing microorganisms ...... 20

2.2 Metabolic engineering and strain development ...... 20

2.2.1 Rhizopus oryzae ...... 20

2.2.2 Metabolic pathway for fumaric acid production in R. oryzae ...... 22

2.2.3 Key enzymes in reductive TCA pathway ...... 24

viii

2.2.4 Metabolic engineering and system biology for strain development ...... 28

2.3 Fermentation process development ...... 33

2.3.1 Medium composition ...... 33

2.3.2 pH effect and neutralizing agents ...... 37

2.3.3 Effects of dissolved oxygen on cell growth and fumaric acid production ...... 39

2.3.4 Morphology control ...... 40

2.4 References ...... 46

3. Effects of overexpressing pyc and pepc genes on fumaric acid biosynthesis from glucose ...... 58

Summary ...... 58

3.1 Introduction ...... 59

3.2 Materials and methods ...... 61

3.2.1 Strains and culture media ...... 61

3.2.2 Construction of plasmids expressing native pyc gene ...... 62

3.2.3 Construction of a plasmid expressing a heterologous pepc gene ...... 62

3.2.4 Transformation ...... 63

3.2.5 Southern hybridization analysis...... 63

3.2.6 Evaluation of genetic stability of transformants ...... 64

3.2.7 Enzyme activity assays ...... 64

3.2.8 Preparation of seed culture for fermentation ...... 65

3.2.9 Fermentation kinetics studies ...... 66

3.2.10 Analytical methods ...... 66

3.2.11 Statistical analysis...... 67

3.3 Results ...... 67

3.3.1 Cloning of pyc and pepc and transformation ...... 67

3.3.2 Southern hybridization analysis...... 68

3.3.3 Genetic stability ...... 69

3.3.4 Enzyme activities ...... 69

3.3.5 Seed culture growth kinetics ...... 70

3.3.6 Fermentation kinetics ...... 71

3.4 Discussion ...... 72

3.5 Conclusions ...... 77

3.6 References ...... 78

4. Effects of overexpressing fumR gene on cell growth and fumaric acid biosynthesis from glucose...... 92

Summary ...... 92

4.1 Introduction ...... 92

4.2 Materials and methods ...... 95

4.2.1 Strains and culture media ...... 95

4.2.2 Construction of plasmids expressing fumR gene ...... 96

4.2.3 Transformation ...... 97

4.2.4 Southern hybridization analysis...... 97

4.2.5 Evaluation of genetic stability of transformants ...... 98

4.2.6 Preparation of seed culture for fermentation ...... 98

4.2.7 Fermentation kinetics studies ...... 99

4.2.8 Fumarase activity assay ...... 99

4.2.9 Analytical methods ...... 100

4.2.10 Statistical analysis...... 101

4.3 Results ...... 101

4.3.1 Cloning of fumR and transformation ...... 101

4.3.2 Southern hybridization analysis...... 102

4.3.3 Genetic stability ...... 102

4.3.4 Seed culture growth kinetics ...... 103

4.3.5 Fermentation kinetics ...... 103

4.3.6 Fumarase activities ...... 105

4.4 Discussion ...... 107

4.5 Conclusions ...... 109

4.6 References ...... 111

5. Metabolic engineering of E. coli for malic acid production ...... 124

Summary ...... 124

5.1 Introduction ...... 125

5.2 Materials and methods ...... 127

5.2.1 Strains and culture media ...... 127

5.2.2 Cloning of frd-disrupted mutant ...... 127

5.2.3 Construction of pTrc99a-fumR expression plasmid ...... 129

5.2.4 Fumarase assay ...... 130

5.2.5 Effect of initial glucose concentration ...... 131

5.2.6 Batch and fed-batch fermentation kinetics ...... 131

5.2.7 Analytical methods ...... 132

5.3 Results ...... 133

5.3.1 Cloning of frd disrupted mutant ...... 133

5.3.2 Overexpression of fumarase ...... 134

5.3.3 Effects of initial glucose concentration on the fermentation kinetics ...... 135

5.3.4 Batch fermentation kinetics in 3-liter fermenter ...... 135

5.3.5 Batch and fed-batch fermentation kinetics in 5-liter fermenter ...... 137

5.4 Discussion ...... 138

5.5 Conclusions ...... 142

5.6 References ...... 144

6. Conclusions and recomendations ...... 161

6.1 Conclusions ...... 161

6.1.1 Metabolic engineering of R. oryzae ...... 161

6.1.2 Fermentation kinetics of R. oryzae transformants ...... 162

6.1.3 Metabolic engineering of E. coli ...... 163

6.1.4 Fermentation kinetics by engineered E. coli strain ...... 164

6.2 Recommendations ...... 165

6.2.1 Metabolic engineering of R. oryzae ...... 165

6.2.2 Genomics and proteomics of R. oryzae ...... 167

6.2.3 Optimization of fumaric acid production by fermentation ...... 167

6.2.4 Metabolic engineering of E. coli ...... 168

Bibliography ...... 170

Appendices ...... 185

Appendix A Medium compositions ...... 185

A.1 Medium compositions for Rhizopus oryzae ...... 186

A.2 Medium compositions for Escherichia coli ...... 186

Appendix B Analytical methods ...... 187

B.1 High performanceliquid chromatography ...... 188

xii

B.2 Protein assay ...... 188

B.3 Enzyme assays ...... 189

B.3.1 Pyruvate carboxylase (PYC) ...... 189

B.3.2 Phosphoenolpyruvate carboxylase (PEPC)...... 189

B.3.3 Fumarase (FUMR) ...... 190

Appendix C Genetic engineering protocols ...... 197

C.1 Preparation of genomic DNA from R. oryzae with QIAGEN genome DNA kit . 198

C.2 Preparation of plasmid DNA with QIAprep Spin Miniprep Kit ...... 199

C.3 PCR amplification of pepc from E. coli genome DNA ...... 199

C.4 Cloning of pepc gene into pGEM-T vector (Promega) ...... 200

C.5 DNA ligation ...... 201

C.6 DNA Transformation in R. oryzae ...... 201

C.6.1 Preparation of tungsten particles ...... 201

C.6.2 DNA coating ...... 202

C.6.3 Transformation ...... 203

C. 7 Southern hybridization by DIG high prime DNA labeling and detection starter kit II ...... 204

C.7.1 DNA labeling ...... 204

C.7.2 DNA transfer and fixation ...... 204

C.7.3 Hybridization ...... 205

C.7.4 Immunological detection ...... 206

C.7.5 Luminescent exposure ...... 207

C.8 Gene disruption through homologous recombination in E. coli ...... 207

C.9 E. coli electrocompetent cell preparation ...... 207

xiii

Appendix D Buffers and reagents ...... 216

xiv

LIST OF FIGURES

Figure 1.1 Metabolic pathways for fumaric acid, lactic acid, and ethanol biosynthesis from glucose in R. oryzae...... 12 Figure 1.2 Research objectives and scope of this study...... 13 Figure 2.1 Molecular structure of fumaric acid...... 57 Figure 2.2 Fumaric acid production via petrochemical route...... 57 Figure 3.1 Metabolic pathways for fumaric acid, lactic acid, and ethanol biosynthesis from glucose in R. oryzae...... 84 Figure 3.2 Plasmid maps of expression vectors pPyrF2.1A, pPyrF2.1A-pyc, pPgk1-Ex, and pPgk1Ex-pepc...... 85 Figure 3.3 Southern hybridization of HpaI-digested DNA from R. oryzae 99880 and 4 transformant isolates...... 86 Figure 3.4 Kinetics of glucose consumption and cell growth in RZ medium for R. oryzae ppc1 transformants from different passage numbers...... 87 Figure 3.5 Comparison of specific enzyme activities in Rhizopus oryzae wild type (WT) and transformants overexpressing pyc, and pepc genes, respectively...... 88 Figure 3.6 Cell morphology of various R. oryzae strains after 24 h incubation in the seed culture medium. Insets show a larger magnification with the scale bar of 100 µm...... 89 Figure 3.7 Batch fermentation kinetics of various R. oryzae strains cultured on glucose- containing production medium at 30 oC at ~pH 5...... 90 Figure 3.8 Comparison of product yields and glucose consumption rate in batch fermentations of R. oryzae wild type (WT) and transformants overexpressing pyc and pepc...... 91

Figure 4.1 Metabolic pathways for fumaric acid, lactic acid, and ethanol biosynthesis from glucose in R. oryzae...... 117 Figure 4.2 Plasmid maps of expression vectors pPyrF2.1A and pPyrF2.1A-fumR...... 118 Figure 4.3 Southern hybridization of HpaI-digested DNA from R. oryzae 99880 and transformant isolate fumR2...... 119 Figure 4.4 Kinetics of glucose consumption and cell growth in RZ medium for R. oryzae transformant fumR2 from different passage numbers. (CDW: cell dry weight) ...... 120 Figure 4.5 Cell morphology, dry weight and mycelial particle number of R. oryzae wild type (WT) and transformant fumR2 after 24 h incubation in the seed culture medium. 121 Figure 4.6 Batch fermentation kinetics of various R. oryzae strains cultured on glucose at 30 oC at ~pH 5...... 122 Figure 4.7 Comparison of specific enzyme activities in wild type (WT) and transformants overexpressing fumR...... 123 Figure 5.1 The anaerobic fermentation pathway showing succinic acid, lactic acid, acetic acid and ethanol biosynthesis from glucose in E. coli...... 153 Figure 5.2 PCR amplification of kan resistance gene and verification of the frd disrupted mutants...... 154 Figure 5.3 Comparison of specific fumarase activites in forward and reverse reactions in E. coli KJ060M (pTrc99a-fumR) with and without IPTG induction...... 155 Figure 5.4 Effect of initial glucose concentration (40 g/L, 60 g/L, 80 g/L, 100 g/L and 120 g/L) on malic acid production...... 156 Figure 5.5 Batch fermentation kinetics of E. coli KJ060, KJ060M and KJ060M (pTrc99a- fumR) in the 3-liter fermenter at 37 ˚C, pH 6.5...... 158 Figure 5.6 Fermentation kinetics of fed-batch fermentation by the mutant E. coli KJ060M in the 3-liter fermenter at 37 ˚C, pH 6.5...... 159 Figure 5.7 Batch and fed-batch fermentation kinetics by the mutant E. coli KJ060M in the 5-liter fermenter at 37 ˚C, pH 6.5...... 160 Figure B.1 HPLC chromatograms for standard samples and sample by R. oryzae fermentation...... 191

xvi

Figure B.2 HPLC chromatograms for standard samples and sample by E. coli fermentation...... 192 Figure B.3 Typical standard curve of protein assay using bovine serum albumin...... 193 Figure B.4 Sample plot of PYC activity determination...... 194 Figure B.5 Sample plot of PEPC activity determination...... 195 Figure B.6 Sample plots of FUMR activity determination...... 196 Figure C.1 Genome DNA of R. oryzae 99-880...... 209 Figure C.2 PCR amplification of pepc from E. coli DH5α genome...... 209 Figure C.3 Cloning of pepc gene into pGEM-T vector...... 210 Figure C.4 Ligation of insert pepc and vector pgk1Ex...... 210 Figure C.5 Biolistic PDS-1000 system main unit...... 211 Figure C.6 R. oryzae Transformants...... 211 Figure C.7 fumR expression plasmid construction and transformation...... 212 Figure C.8 Genome DNA after HpaI digestion...... 213 Figure C.9 Gene disruption strategy in E. coli...... 214 Figure C.10 Plasmids maps of pKD4, pKD46 and pCP20...... 215

xvii

LIST OF TABLES

Table 3.1 Strains and plasmids used in this study...... 81 Table 3.2 PCR primers and conditions used for pyc and pepc gene cloning...... 82 Table 3.3 Cell dry weight, particle number, and particle size of various strains after 24 h incubation in the seed culture medium...... 82 Table 3.4 Comparison of glucose consumption and product yields in shake-flask fermentations by R. oryzae wild type and transformants...... 83 Table 4.1 Strains and plasmids used in this study...... 114 Table 4.2 PCR primers and conditions used for fumR gene cloning. Lower case letters indicate restriction enzyme sites...... 115 Table 4.3 Comparison of glucose consumption and product yields in shake-flask fermentations by R. oryzae wild type and fumR transformants...... 116 Table 5.1 Comparison of malic acid production from various microorganisms...... 147 Table 5.2 Strains and plasmids used in this study...... 148 Table 5.3 PCR primers and conditions used for FRT-flanked kan amplification, fumR amplification and colony verification...... 149 Table 5.4 Malic acid titer, yield and productivity at different initial glucose concentrations...... 150 Table 5.5 Comparison of product concentrations, yields and productivities of the parental strain E. coli KJ060 and the mutants KJ060M and KJ060M (pTrc99a-fumR) in 3-liter fermenter...... 151 Table 5.6 Comparison of glucose consumption rates, product concentrations, yields and productivities of the mutant E. coli KJ060M in batch and fed-batch fermentations in 3- liter and 5-liter fermenters...... 152

xviii

CHAPTER 1

INTRODUCTION

Rhizopus oryzae is a filamentous fungus that is classified under the family

Mucoraceae in the order Mucorales of the phylum Zygomycota (Gryganskyi et al., 2010), and is known as the cause of mucormycosis. R. oryzae is ubiquitous in nature, it can be separated from decaying fruits and vegetables and their seeds, grains, soil and molding bread. R. oryzae strains have a wide range of industrial applications as they can produce various hydrolases, including amylase, proteinases and lipases, and can grow on various carbon sources (Ban et al., 2001; Park et al., 2004; Maas et al., 2006) to produce various industrial products, including ethanol, L-(+)-lactic acid, fumaric acid and lesser extent of

L-(+)-malic acid (Ghosh and Ray, 2011). Fumaric acid is a dicarboxylic acid used extensively in resins, food acidulants, and other applications including oil field fluids, esters etc (Roa Engel et al., 2008). The production of fumaric acid from sustainable biomass has become a promising alternative to the current petroleum chemical route due to public concerns about environmental pollution and the depletion of petroleum resources.

So far as is known, microoganisms capable of producing fumaric acid are confined to filamentous fungi (Foster and Waksman, 1939). R. oryzae strains are most attractive since they produce highest yield and productivity of fumaric acid compared with other

1 fumarate producing species (Gangl et al., 1990; Cao et al., 1996). R. oryzae can be divided into two groups, LA (lactic acid producers) group and FMA (fumaric-malic producers) group according to organic acid production (Abe et al., 2007). These two groups show an obvious phylogenetic distinction based on the analyses of rDNA internal transcribed spacer (ITS), lactate dehydrogenase B (Saito et al., 2004), actin, translation elongation factor-1α and genome-wide amplified fragment length polymorphisms

(AFLP).

The suggested metabolic pathway in R. oryzae is shown in Figure 1.1. In this pathway, the substrate, glucose is reduced to pyruvate through glycolysis. The pyruvate is catalyzed by pyruvate carboxylase with the fixation of CO2 under aerobic conditions

(Overman and Romano, 1969). The carboxylation of pyruvate leads to the formation of oxaloacetic acid. NAD-malate dehydrogenase is responsible for the reduction of oxaloacetate to malate, and fumaric acid is synthesized by the catalysis of fumarase on malate. It has been demonstrated that this C3 plus C1 mechanism is responsible for the fumaric acid accumulation by R. oryzae (Romano et al., 1967). Fumaric acid is accumulated through an exclusively cytosolic pathway (Kenealy et al., 1986; Peleg et al.,

1989). Fumaric acid generated in the citrate cycle is mainly utilized for the biosynthesis during the growth phase but not lead to a significant accumulation (Magnuson and Lasure,

2004). The maximum theoretical yield of fumaric acid from glucose and CO2 (excess) is

2 moles per mole glucose based on the following equation: C6H12O6 + 2 CO2 → 2

C4H4O4 + 2 H2O.

Lactic acid is synthesized from pyruvate by the catalysis of NAD-dependent L- lactate dehydrogenase (Obayashi et al., 1966; Pritchard, 1973). There are two different lactate dehydrogenase genes in R. oryzae, ldhA and ldhB, which provide the genetic basis on the grouping of R. oryzae into Type-I and II (Abe et al., 2003). The type-I strains produce primarily lactic acid due to the possession of both ldhA and ldhB genes. The enzyme LdhA was mainly responsible for conversion of pyruvate to lactic acid. The type-

II strains produce a predominance of fumaric acid and little lactic acid, and only ldhB is detected in this type. It suggests that the enzyme LdhB perhaps plays a role in the utilization of lactic acid (Abe et al., 2003; Skory and Ibrahim, 2007). Ethanol is the other fermentative product mainly synthesized under anaerobic stress through the catalysis of pyruvate decarboxylase (Skory, 2003a) and alcohol dehydrogenase (Skory et al., 1998).

Therefore, aerobic condition is preferred in which available pyruvate is mainly fluxed to fumaric acid synthesis competing with ethanol production.

Submerged fermentation process utilizing either free or immobilized cells is generally applied for fumaric acid production by filamentous fungi (Cao et al., 1997;

Zhou, 1999). Many efforts have been done to improve and optimize the fermentation process including using cheaper biomass feedstock, optimizing medium formulation, developing novel bioreactors, and separation and recovery methods (Cao et al., 1996; Du et al., 1997a; Carta, 1999; Bulut et al., 2009). Fumaric acid production yield and productivity are limited by the morphology control of fungal cells and resultant oxygen limitation in conventional stirred-tank bioreactors (Yang, 2007).

Using metabolic engineering strategy to genetically modify R. oryzae for enhanced fumaric acid production has not been reported yet. This is mainly because of limited genetic engineering tools available for R. oryzae. This fungus is resistant to common antifungal agents used as cloning selection markers (Meussen et al., 2012a). The occurrence of whole genome duplication makes the pathway modifications more difficult

(Ma et al., 2009). Furthermore, DNA introduced into R. oryzae rarely integrates into the genome, and is typically maintained through autonomous extra-chromosomal replication

(Skory, 2005), which hampers gene modifications (eg. gene targeting and gene disruption) of this organism. Several methods such as random mutagenesis (Bai et al., 2004; Ge et al.,

2004), RNA interference (Nakayashiki and Nguyen, 2008; Gheinani et al., 2011) and

Agrobacterium mediated heterologous DNA transformation (Michielse et al., 2004), gene knockout through double cross-over event have been applied to modify the genome of R. oryzae and its transcription (Ibrahim et al., 2010). It is possible to introduce genetic modification by random mutagenesis and heterologous genes by biolistic transformation

(Skory, 2005; Mertens et al., 2006; Meussen et al., 2012b). The bottleneck is caused by the difficulty of genomic integration of vector DNA. Therefore, more research work needs to be done to clarify the mechanism of the DNA integration in R. oryzae.

1.1 Objectives

The overall goal of this research was to employ metabolic engineering techniques to modify the fumaric acid biosynthesis pathway for improved fumaric acid production.

First of all, R. oryzae isolates overexpressing endogenous pyruvate carboxylase gene

(pyc), and heterolgous phosphoenolpyruvate carboxylase gene (pepc) were obtained and characterized in order to evaluate the effects of PYC and PEPC overexpression on the fermentation. The fermentation kinetics in shake flasks was compared between the wild type and pyc and pepc isolates. Secondly, the effect of fumarase (FUMR) overexpression on fumaric acid production was investigated by constructing R. oryzae isolates transformed with plasmids containing endogenous fumR. The reversible reactions of malic acid to fumaric acid catalyzed by fumarase were studied during the growth phase and production phase of fermentation. The third objective of this study was to disrupt fumarate reductase gene (frd) in a succinic acid producing E. coli strain through homologous recombination. The effects of frd knockout on the cellular metabolism and fermentation kinetics of E. coli were evaluated. Figure 1.2 provides an overview of the research objectives, approaches, and the scope of this study, which is described below.

1.2 Scope of study

Task 1: Construction and characterization of R. oryzae isolates overexpressing endogenous pyc gene and herterologous pepc gene derived from E. coli.

The conversion of pyruvate to oxaloacetate catalyzed by PYC is the first step in the pathway leading to fumaric acid biosynthesis. The gene encoding PYC is strictly regulated as it is situated at the branch point of pyruvate metabolism in the cytosol

(Goldberg et al., 2006). Therefore, overexpressing the pyc gene was the first aim to increase fumaric acid production in R. oryzae.

PEPC catalyzes the carboxylation of phosphoenolpyruvate (PEP) to OAA; this enzyme only exists in bacteria and plants (Wohl and Markus, 1972). The introduction of

PEPC into R. oryzae was aimed to increase the carbon flux towards OAA, thus decreasing the pyruvate flux to other competitive pathways. Pepc gene in this study was cloned from E. coli, so the overexpression should be under the endogenous promoter and terminator derived from R. oryzae. The effects of PYC and PEPC overexpression are reported in Chapter 3.

Task 2: Construction and characterization of R. oryzae isolates overexpressing native fumarase encoded by fumR.

Fumarase catalyzes the final step of fumaric acid biosynthesis in R. oryzae. The fumR gene (GenBank accession No. X78576) encoding fumarase was first cloned and sequenced by Fridberg et al. (1995). It also has been reported that the fumR might be involved in the overproduction of fumaric acid by R. oryzae at the transcriptional level

(Friedberg et al., 1995). In this study, endogenous fumR gene was cloned and transformed into R. oryzae to evaluate its potential effects on cell growth and fumaric acid biosynthesis. The results of the fumR overexpression are reported in Chapter 4.

Task 3: Construction and characterization of frd-disrupted E. coli mutant by homologous recombination

Gene disruption has been widely used in the study of protein function and gene expression regulation. The frd encoded enzyme catalyzes the final step of succinic acid

6 biosynthesis, the oxidative conversion of fumarate to succinate in E. coli during anaerobic fermentation (Luna-Chavez et al., 2000). The frd gene on the E. coli chromosome was disrupted through one-step inactivation method using FRT-kan-FRT cassette (Datsenko, 2000). Fermentations were carried out in shake flasks to compare the kinetics of mutant strain and parental strain. Free cell fermentation in a stirred-tank fermenter was performed to evaluate the kinetics of the mutant, and the results are reported in Chapter 5.

1.3 References

Abe, A., Oda, Y., Asano, K., Sone, T., 2007. Rhizopus delemar is the proper name for Rhizopus oryzae fumaric-malic acid producers. Mycologia 99, 714-722.

Abe, A., Sone, T., Sujaya, I.N., Saito, K., Oda, Y., Asano, K., Tomita, F., 2003. rDNA ITS sequence of Rhizopus oryzae: its application to classification and identification of lactic acid producers. Biosci Biotechnol Biochem 67, 1725-1731.

Bai, D.M., Zhao, X.M., Li, X.G., Xu, S.M., 2004. Strain improvement of Rhizopus oryzae for over-production of L-(+)-lactic acid and metabolic flux analysis of mutants. Biochem Eng J 18, 41-48.

Ban, K., Kaieda, M., Matsumoto, T., Kondo, A., Fukuda, H., 2001. Whole cell biocatalyst for biodiesel fuel production utilizing Rhizopus oryzae cells immobilized within biomass support particles. Biochem Eng J 8, 39-43.

Bulut, S., Elibol, M., Ozer, D., 2009. Optimization of process parameters and culture medium for L-(+)-lactic acid production by Rhizopus oryzae. J Chem Eng Jpn 42, 589- 595.

Cao, N., Du, J., Chen, C., Gong, C.S., Tsao, G.T., 1997. Production of fumaric acid by immobilized rhizopus using rotary biofilm contactor. Appl Biochem Biotechnol 63-65, 387-394.

Cao, N.J., Du, J.X., Gong, C.S., Tsao, G.T., 1996. Simultaneous production and recovery of fumaric acid from immobilized Rhizopus oryzae with a rotary biofilm contactor and an adsorption column. Appl Environ Microbiol 62, 2926-2931.

Carta, F.S., Soccol, C.R.,Ramos, L.P., Fontana, J.D., 1999. Production of fumaric acid by fermentation of enzymatic hydrolysates derived from cassava bagasse. Bioresour Technol 68, 23-28.

Datsenko, K.A., 2000. One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products. Proc Natl Acad Sci 97, 6640-6645.

Du, J., Cao, N., Gong, C.S., Tsao, G.T., Yuan, N., 1997. Fumaric acid production in airlift loop reactor with porous sparger. Appl Biochem Biotechnol 63-65, 541-556.

Foster, J.W., Waksman, S.A., 1939. The production of fumaric acid by molds belonging to the genus Rhizopus. J Am Chem Soc 61, 127-135.

Friedberg, D., Peleg, Y., Monsonego, A., Maissi, S., Battat, E., Rokem, J.S., Goldberg, I., 1995. The fumR gene encoding fumarase in the filamentous fungus Rhizopus oryzae: cloning, structure and expression. Gene 165, 139-144.

Gangl, I.C., Weigand, W.A., Keller, F.A., 1990. Economic comparison of calcium fumarate and sodium fumarate production by Rhizopus arrhizus. Appl Biochem Biotechnol 24-25, 663-677.

Ge, C.M., Gu, S.B., Zhou, X.H., Yao, R.M., Pan, R.R., Yu, Z.L., 2004. Breeding of L- (+)-lactic acid producing strain by low-energy ion implantation. J Microbiol Biotech 14, 363-366.

Gheinani, A.H., Jahromi, N.H., Feuk-Lagerstedt, E., Taherzadeh, M.J., 2011. RNA silencing of lactate dehydrogenase gene in Rhizopus oryzae. J RNAi Gene Silencing 14, 363-366.

Ghosh, B., Ray, R.R., 2011. Current commercial perspective of Rhizopus oryzae: a review. J Appl Sci 11, 2470-2486.

Goldberg, I., Rokem, J.S., Pines, O., 2006. Organic acids: old metabolites, new themes. J Chem Technol & Biotechnol 81, 1601-1611.

Gryganskyi, A.P., Lee, S.C., Litvintseva, A.P., Smith, M.E., Bonito, G., Porter, T.M., Anishchenko, I.M., Heitman, J., Vilgalys, R., 2010. Structure, function, and phylogeny of the mating locus in the Rhizopus oryzae complex. PLoS One 5, e15273.

Ibrahim, A.S., Gebremariam, T., Lin, L., Luo, G., Husseiny, M.I., Skory, C.D., Fu, Y., French, S.W., Edwards Jr, J.E., Spellberg, B., 2010. The high affinity iron permease is a key virulence factor required for Rhizopus oryzae pathogenesis. Mol Microbiol 77, 587- 604.

Kenealy, W., Zaady, E., du Preez, J.C., Stieglitz, B., Goldberg, I., 1986. Biochemical aspects of fumaric acid accumulation by Rhizopus arrhizus. Appl Environ Microbiol 52, 128-133.

Luna-Chavez, C., Iverson, T.M., Rees, D.C., Cecchini, G., 2000. Overexpression, purification, and crystallization of the membrane-bound fumarate reductase from Escherichia coli. Protein Expression Purif 19, 188-196.

Ma, L.J., Ibrahim, A.S., Skory, C., Grabherr, M.G., Burger, G., Butler, M., Elias, M., Idnurm, A., Lang, B.F., Sone, T., Abe, A., Calvo, S.E., Corrochano, L.M., Engels, R., Fu, J., Hansberg, W., Kim, J.M., Kodira, C.D., Koehrsen, M.J., Liu, B., Miranda-Saavedra, D., O'Leary, S., Ortiz-Castellanos, L., Poulter, R., Rodriguez-Romero, J., Ruiz-Herrera, J., Shen, Y.Q., Zeng, Q., Galagan, J., Birren, B.W., Cuomo, C.A., Wickes, B.L., 2009.

Genomic analysis of the basal lineage fungus Rhizopus oryzae reveals a whole-genome duplication. PLoS Genetics 5, 1-11.

Maas, R.H., Bakker, R.R., Eggink, G., Weusthuis, R.A., 2006. Lactic acid production from xylose by the fungus Rhizopus oryzae. Appl Microbiol Biotechnol 72, 861-868.

Magnuson, J.K., Lasure, L.L., 2004. Organic acid production by filamentous fungi. In: Tracz JS, Lange L (eds) Advances in fungal biotechnology for industry, agriculture and medicine. Kluwer/Plenum, New York, USA, 307-340.

Mertens, J.A., Skory, C.D., Ibrahim, A.S., 2006. Plasmids for expression of heterologous proteins in Rhizopus oryzae. Arch Microbiol 186, 41-50.

Meussen, B.J., Graaff, L.H., Sanders, J.P.M., Weusthuis, R.A., 2012a. Metabolic engineering of Rhizopus oryzae for the production of platform chemicals. Appl Microbiol Biotechnol 94, 875-886.

Meussen, B.J., Weusthuis, R.A., Sanders, J.P., Graaff, L.H., 2012b. Production of cyanophycin in Rhizopus oryzae through the expression of a cyanophycin synthetase encoding gene. Appl Microbiol Biotechnol 93, 1167-1174.

Michielse, C.B., Hooykaas, P.J.J., Hondel, C.A.M.J.J., Ram, A.F.J., 2004. Agrobacterium-mediated transformation as a tool for functional genomics in fungi. Curr Genet 48, 1-17.

Nakayashiki, H., Nguyen, Q.B., 2008. RNA interference: roles in fungal biology. Curr Opin Microbiol 11, 494-502.

Obayashi, A., Yorifuji, H., Yamagata, T., Ijichi, T., Kanie, M., 1966. Respiration in organic acid forming molds: Part I. Purification of cytochrome c, coenzyme Q and L- lactic dehydrogenase from lactate forming Rhizopus oryzae. Agric Biol Chem 30, 717- 724.

Overman, S.A., Romano, A.H., 1969. Pyruvate carboxylase of Rhizopus nigricans and its role in fumaric acid production. Biochem Biophys Res Commun 37, 457-463.

Park, E.Y., Anh, P.N., Okuda, N., 2004. Bioconversion of waste office paper to L(+)- lactic acid by the filamentous fungus Rhizopus oryzae. Bioresour Technol 93, 77-83.

Peleg, Y., Battat, E., Scrutton, M.C., Goldberg, I., 1989. Isoenzyme pattern and subcellular localization of enzymes involved in fumaric acid accumulation by Rhizopus oryzae. Appl Microbiol Biotechnol 32, 334-339.

Pritchard, G.G., 1973. Factors affecting the activity and synthesis of NAD dependent lactate dehydrogenase in Rhizopus oryzae. J Gen Microbiol 78, 125-137.

Roa Engel, C.A., Straathof, A.J.J., Zijlmans, T.W., Gulik, W.M., Wielen, L.A.M., 2008. Fumaric acid production by fermentation. Appl Microbiol Biotechnol 78, 379-389.

Romano, A.H., Bright, M.M., Scott, W.E., 1967. Mechanism of fumaric acid accumulation in Rhizopus nigricans. J Biotechnol 93, 600-604.

Saito, K., Saito, A., Ohnishi, M., Oda, Y., 2004. Genetic diversity in Rhizopus oryzae strains as revealed by the sequence of lactate dehydrogenase genes. Arch Microbiol 182, 30-36.

Skory, C.D., 2003. Induction of Rhizopus oryzae pyruvate decarboxylase genes. Curr Microbiol 47, 59-64.

Skory, C.D., 2005. Inhibition of non-homologous end joining and integration of DNA upon transformation of Rhizopus oryzae. Mol Genet Genomics 274, 373-383.

Skory, C.D., Freer, S.N., Bothast, R.J., 1998. Production of L-lactic acid by Rhizopus oryzae under oxygen limiting conditions. Biotechnol Lett 20, 191-194.

Skory, C.D., Ibrahim, A.S., 2007. Native and modified lactate dehydrogenase expression in a fumaric acid producing isolate Rhizopus oryzae 99-880. Curr Genetics 52, 23-33.

Wohl, R.C., Markus, G., 1972. Phosphoenolpyruvate carboxylase of Escherichia coli. Purification and some properties. J Biol Chem 247, 5785-5792.

Yang, S.T., 2007. Bioprocessing – from biotechnology to biorefinery. In: Yang ST (eds.) Bioprocessing for value-added products from renewable resources – new technologies and applications. Elsevier.

Zhou, Y., 1999. Fumaric acid fermentation by Rhizopus oryzae in submerged systems. PhD thesis. Purdue University.

Glucose

PEP NADH+ H+ a Lactate Pyruvate Oxaloacetate f NADH+ H+ ATP CO2 d b CO2 Malate TCA Acetaldehyde

+ c e NADH+ H Mitochondria Ethanol Fumarate

Figure 1.1 Metabolic pathways for fumaric acid, lactic acid, and ethanol biosynthesis from glucose in R. oryzae. a. pyruvate carboxylase; b. malate dehydrogenase; c. fumarase; d. pyruvate decarboxylase; e. alcohol dehydrogenase; f. lactate dehydrogenase.

Objective Improved fumaric acid production by metabolically engineered R. oryzae

Effect of pyc and pepc Effect of fumR Effect of frd gene overexpression overexpression disruption in E. coli  PCR amplification of  PCR amplification of  Construct FRT-kan- pyc and pepc fumR FRT cassette  Construct the plasmid  Construct the plasmid  Develop frd disrupted  Biolistic transformation  Biolistic transformation mutant  Characterize the R.  Characterize the R. Characterize the oryzae isolates oryzae isolates mutant  Compare fermentation  Compare fermentation  Fermentation kinetics kinetics

Figure 1.2 Research objectives and scope of this study.

CHAPTER 2

LITERATURE REVIEW

2.1 Fumaric acid production

2.1.1 Fumaric acid properties and applications

Fumaric acid is a natural organic acid which is firstly found from the plant Fumaria officinalis. Therefore, fumaric acid is named after the genus fumaria, it is also known as

(E)-2-butenedioic acid, boletic acid, lichenic acid, allomaleic acid, and trans-1,2- ethylene-dicarboxylic acid (Zhou, 1999). Fumaric acid is also a key intermediate in the tricarboxylic acid cycle. Therefore, fumaric acid exists in a small amount in many microorganisms, plants and mammals.

Fumaric acid (HOOCCH=CHCOOH) has a chemical structure of two acid carbonyl groups and a trans-double bond (Figure 2.1). It has a molecular weight of 116.07 g/mol with pKa values of 3.03 and 4.44 at 25˚C. Pure fumaric acid has an appearance of white

3 crystalline solid with no odor and a very slight acid taste. It has a density of 1.635 kg/m and a melting point of 286˚C. If heated cautiously at 200˚C, fumaric acid will sublime without decomposition (Yang et al., 2011). At higher temperatures, it will form traces of maleic anhydride. No anhydride of fumaric acid exists.

Fumaric acid has a solubility of 0.63% in water at 25˚C, the solubility is increased with the increasing temperature of water. The solubilities of fumaric acid in organic

14 solvents are also very low: 0.02% in chloroform, 5.44% in 95% ethanol, 0.56% in ethyl ether, 0.003% in benzene, 0.027% in carbon tetrachloride at 25˚C and 1.29% in acetone at 20˚C. Different fumarate salts have different solubilities in water. The solubilities of calcium salt are as low as 1.22%, whereas sodium salt is much more soluble with the solubility of 22% at 20˚C.

Fumaric acid is an important specialty chemical and has extensive applications in various industries, such as the manufacture of paper resins, food and beverages, unsaturated polyesters, alkyd resins, plasticizers and miscellaneous applications including lubricating oil, inks, lacquers, carboxylating agent for styrenebutadiene rubber, personal care additives (Yang, 2007; Roa Engel et al., 2008). Fumaric acid has also been used for the synthesis of succinic acid, maleic acid.

The characteristics of trans-double bond and two carbonyl groups have made fumaric acid the largest use on the industry of polymerization (Otsu et al., 1984; Lee et al., 2004). Both fumaric acid and maleic anhydride are currently used in the synthesis of these resins, and maleic anhydride has relatively lower price as the raw material.

However, fumaric acid offers better characteristics when substituting maleic anhydride in the formulations of alkyd resins, including greater hardness in the polymer structure, nontoxic nature and higher durability, which make fumaric acid resins superior to other resins within the same cost range. Recently, fumaric acid derivatives of oligomeric esters become more attractive in the field of biodegradable polymer networks since these compounds and the degradation products can be expected to be nontoxic and biocompatible. The special functionalities of fumaric acid could allow crosslinking by

UV radical polymerization and hence provide a novel type of polymer networks potentially applied in tissue engineering and biomedical research (Grijpma et al., 2005).

Nontoxic feature approved the wide application of fumaric acid in food and beverage products since 1946. When used as a food additive, the hydrophobic nature of fumaric acid results in persistent, long lasting sourness and flavor impact. A much lower amount of fumaric acid can be used because of its high rate of sourness therefore reducing cost of many foods and beverage products. In contrast to other organic acids, fumaric acid is not hygroscopic and can be used to keep moisture from hardening food powders such as baking powder, cake mixes, etc. Since fumaric acid has good solubility in hot water, it is often used in dry products intended for hot preparation. Currently, fumaric acid is used in wheat and corn tortillas, sour dough and rye breads, refrigerated biscuit doughs, fruit juice and nutraceutical drinks, gelatin desserts, gelling aids, pie fillings and wine (Yang et al., 2011).

Fumaric acid has proven to be a particularly effective additive in animal feed. It acts as acidulant which can replace antibiotics to minimize the presence of antibiotic-resistant pathogens in meat and milk products. The inclusion of fumaric acid and the resultant adjustment of the pH value demonstrate improved animal health and weight gain by controlling microbial populations in animal digestive systems and in feed. Another potential application is that fumaric acid can be used as supplement in cattle feed. It can reduce up to 70% methane emissions of cattle, as farm animals are responsible for 14% of the methane emission caused by human activity (McGinn et al., 2004). Fumaric acid can improve the feed conversion ratio in broiler chickens because of its ability to lower

16 the pH of the crop and gizzard contents (Skinner et al., 1991), therefore reducing the load of pathogenic bacteria which favor higher pHs in the digestive system.

Fumaric acid plays an essential role in metabolism and is a naturally occurring compound which can be produced by the human body. Esters of fumaric acid (FAEs) were introduced since 1959 to treat patients with psoriasis who cannot naturally produce a sufficient quantity of this important compound. Various FAEs with different molecular formula, such as monoethylfumarate (MEF), monomethylfumarate (MMF), diethylfumarate (DEF) and dimethylfumarate (DMF) have been used in the treatment of psoriasis in European countries for over 30 years (Nieboer et al., 1989; Mrowietz et al.,

1999; Moharregh-Khiabani et al., 2009). In pharmaceutical company, fumaric acid is mainly used on the synthesis of ferrous fumarate which is used to treat iron deficiency anemia (a lack of red blood cells caused by having too little iron in the body). It has the advantage of high iron content, resistance to oxidation and little side effect (Zlotkin et al.,

2001).

2.1.2 Production of fumaric acid

Fumaric acid can be manufactured via petrochemical route and by fermentation. In the early 1940s fumaric acid was produced by microbial fermentation using filamentous fungi Rhizopus arrhizus by Pfizer in a commercialized scale, the production capacity reached about 4000 tons per year (Goldberg et al., 2006). Nevertheless, the fermentation process was discontinued and chemical synthesis became a prevailing process because of its economic advantage. However, with the increasing oil prices and more concerns about

17 the environmental pollutions accompanied with chemical petroleum based process, biobased fumaric acid fermentation as a sustainable alternative has attracted more and more interests.

2.1.2.1 Chemical synthesis

The production capacity of fumaric acid through chemical synthesis has reached

90,000 tons/year (Roa Engel et al., 2008). Currently, fumaric acid is mainly produced through a cis-trans isomerization of maleic acid, which is derived from hydrolysis of maleic anhydride, the latter being industrially produced by oxidation of benzene or other aromatic compounds. Due to rising benzene prices, n-butane has become more important feedstock in most maleic anhydride plants.

The reaction of n-butane oxidation to maleic anhydride is: 2C4H10 + 7O2 →

2C4H2O3 + 8H2O (Lorences et al., 2003). Maleic anhydride is absorbed by organic acid in the reaction gas with a high recovery rate of 98%. Then the pure maleic anhydride is separated from the solvent through fractional distillation and hydrolyzed into maleic acid.

The hydrolysis reaction of maleic anhydride to maleic acid is: C4H2O3 + H2O → C4H4O4

(Roa Engel et al., 2008).

The reaction process of maleic acid to fumaric acid is fulfilled by various types of catalysts: mineral acids, peroxy compounds with bromides and bromates, and sulfur containing compounds such as thiourea and its derivatives (Morgan and Friedmann, 1938;

Bachmann and Scott, 1948). Figure 2.2 illustrates a simplified fumaric acid chemical synthesis flowchart. Fumaric acid with a high purity is obtained by cooling the reaction

18 mixture and separating, washing, and drying the crystal. Several decolorizing and filtration techniques are applied to make high grade fumaric acid from impure maleic liquors.

2.1.2.2 Enzymatic catalysis

The isomerization of maleic acid to fumaric acid by chemical catalysts is restricted by reaction equilibrium, and conversion yield was affected by the byproducts formation under the high temperature environment (Meek, 1975). Bioconversion was introduced as an alternative method to efficiently produce fumaric from maleic acid.

It is known that maleate isomerase isomerizes maleic acid to fumaric acid.

Microorganisms producing maleate isomerase are Pseudomonas species, Alcaligenes faecalis IB-14, Pseudomonas fluolescens ATCC 23728, and Arthrobacter species strain

TPU 5446 (Otsuka, 1961; Takamura et al., 1969; Kato et al., 1995). However, details of maleate isomerase such as molecular weight, subunit structure, amino acid sequence, gene sequence have not been verified.

It has been reported that maleate isomerases from these microorganisms were unstable even at a moderate temperature. Bacillus stearothermophilus, Bacillus brevis, and Bacillus sp. MI-105 (Goto M. et al., 1998) have been found to have thermo-stable maleate isomerases which can improve fumaric acid production process. Instead of using cell-free enzyme catalysis, whole cells catalysis is a better option in industrial process when considering the simplified process and low production cost. Pseudomonas alcaligenes strain XD-1 has been selected as the best strain with respect to its conversion

19 of maleic acid to fumaric acid. Fumaric acid productivity of the strain reached 41.9 g/L at

6 h of incubation under the optimal conditions (Nakajima-Kambe et al., 1997). Ichikawa inactivated fumarase by heating the cells of Pseudomonas alcaligenes strain XD-1 in the presence of 30 mM CaCl2 at 70 ˚C for 1 h, the maleate isomerase still remained active.

The heat-treated cells produced fumaric acid of 57 g/L in a 4 h reaction period and fumaric acid yield was improved to 95% from maleic acid (Sosaku Ichikawa, 2003).

2.1.3 Fumaric acid producing microorganisms

Insofar as is known, microbial production of fumaric acid is mainly confined to filamentous fungi. Ehrlich first identified fumaric acid production in Rhizopus nigricans in 1911. Foster and Waksman then screened 41 strains from eight genera of Mucorales and identified Rhizopus, Mucor, Cunninghamella, and Circinella as fumarate producers

(Foster and Waksman, 1939). Several other fungal cultures outside the order Mucorales, including Penicillium griseo-fulvum, Aspergillus glaucus and Caldariomyces fumago, are also able to produce fumaric acid. Among them, several Rhizopus species (nigricans, arrhizus, oryzae, and formosa) were identified as the best fumarate producers. However, only R. arrhizus and R. oryzae have been extensively studied for their fumaric acid production potential (Rhodes et al., 1959).

2.2 Metabolic engineering and strain development

2.2.1 Rhizopus oryzae

Rhizopus spp., which belong to the class of zygomycetes, have had extensive industrial applications for thousands years. They can produce various hydrolases, including amylases, proteinases and lipases, and can degrade starch and starch-based materials to produce various industrial products, including alcohol, lactic acid, and fumaric acid (Amedioha, 1993; Bakir et al., 2001; Ban et al., 2001; Taherzadeh et al.,

2003; Karmakar and Ray, 2010; Ghosh and Ray, 2011; Rani and Ghosh, 2011). Their cell walls also contain large amounts of chitin and chitosan, which can be valuable byproducts. R. oryzae and R. arrhizus are the two most studied species for fumaric acid production. However, not all strains within these species can be used to produce fumaric acid. In general, R. arrhizus NRRL 2582 gave the highest product titer and yield (Rhodes et al., 1962), whereas R. oryzae ATCC 20344 gave the highest productivity (Cao et al.,

1996). In submerged cultures under aerobic conditions, some strains of R. oryzae produce fumaric acid while others produce L-lactic acid as the main metabolic product. Under anaerobic conditions, these strains may produce more ethanol, instead of lactic or fumaric acid. The difference in the major metabolites produced by different strains and under different fermentation conditions can be attributed to the varying activities of key enzymes, such as lactate dehydrogenase, in the metabolic pathway that is discussed in the following sections. The fumaric acid producing ability of the heterothallic fungi R. nigricans was also found to be related to its sexuality: the male race (+) can produce fumaric acid in high yields, whereas the female race (-) seldom produces any fumaric acid (Foster and Waksman, 1939; Gryganskyi et al., 2010).

The R. oryzae can be separated into two groups, Type-I and II, based on genetic differences. Type-I strains produce primarily lactic acid, and contain two lactate dehydrogenase genes, ldhA and ldhB. While Type-II strains produce predominately fumaric acid and only have ldhB gene (Abe et al., 2003; Saito et al., 2004). Recently, R. oryzae has been proposed to be divided into two groups, LA (lactic acid producers) and

FMA (fumaric-malic acid producers) according to organic acid production. The phylogenetic distinction of this grouping was confirmed based on the analyses of rDNA

ITS, lactate dehydrogenase B, actin, translation elongation factor-1α and genome-wide

AFLP (Abe et al., 2007).

2.2.2 Metabolic pathway for fumaric acid production in R. oryzae

Although fumaric acid is an important intermediate in the tricarboxylic acid (TCA) cycle present in most aerobic organisms, its production by Rhizopus is not through this pathway. Initially, a C-2 plus C-2 condensation similar to the reactions in the glyoxylate bypass was hypothesized for fumaric acid biosynthesis (Foster and Carson, 1949).

However, later experiments showed a high fumaric acid molar yield of 140%, exceeding the theoretical molar yield (100%) for the postulated pathway, and thus disapproved this hypothesis (Romano et al., 1967). Also, the high glucose concentration essential for fumaric acid production inhibited the activity of the key enzyme, isocitrateglyoxylatelyase, of the glyoxylate pathway. Later, a C-3 plus C-1 mechanism with CO2 fixation was proposed as the pathway for fumaric acid biosynthesis, which was supported by the discovery that R. oryzae harbored cytosolic pyruvate carboxylase,

22 malate dehydrogenase and fumarase. The presence of this reductive TCA pathway (rTCA) was further confirmed in 13C nuclear magnetic resonance and enzymatic activity studies

(Kenealy et al., 1986). This mechanism was also supported by the finding that the addition of cycloheximide (as inhibitor of the cytosolic fumarase) caused a large decrease in fumaric acid production.

Figure 1.1 shows the metabolic pathways in R. oryzae (Romano et al., 1967; Wright et al., 1996). The main pathway for fumaric acid biosynthesis is rTCA, which is located in the cytosol and involves three reactions starting from pyruvate. The first reaction is catalyzed by pyruvate carboxylase (Overman and Romano, 1969), which is ATP- dependent condensation of pyruvate and CO2 to form oxaloacetic acid (OAA). The product OAA is then converted to malate by malate dehydrogenase and then to fumarate by fumarase in cytosol. Fumaric acid is also an intermediate metabolite in the TCA cycle.

However, fumarate generated in the TCA cycle is mainly utilized for the biosynthesis of cell constituents and cannot be accumulated in a significant amount during active cell growth. Therefore, the production of fumarate from glucose by Rhizopus is believed to operate entirely through the cytosolic rTCA pathway with a high theoretical molar yield of 200% (Kenealy et al., 1986). The rTCA pathway can balance the excessive ATP and

NADH and adsorb CO2 produced in the oxidative TCA cycle. However, the two pathways compete for the sole pyruvate carbon flux. Under aerobic conditions, C4 (TCA cycle) intermediates can be formed in cytosol for biosynthesis during the growth phase.

When nitrogen is limited and cell growth stops, the continuing metabolism of glucose and

CO2 fixation lead to the overproduction of fumaric acid at a theoretical yield of 129%

(w/w) from glucose. Meanwhile, the TCA cycle must remain active in order to provide energy for cellular maintenance and material transportation. Therefore, the actual fumaric acid yield in the fermentation is usually much lower than the theoretical yield. In addition, there are other competing pathways that lead to other products such as lactic acid and ethanol. Ethanol production usually occurs under anaerobic conditions or when oxygen is limited in the culture, and thus can be decreased by providing sufficient oxygen to the culture. Depending on the strain and culture conditions, additional byproducts including malic acid, citric acid, and succinic acid could also be formed in relatively small amounts.

2.2.3 Key enzymes in reductive TCA pathway

The key enzymes in rTCA pathway are pyruvate carboxylase, malate dehydrogenase and fumarase. Research concerning these enzymes has focused on their phenotypic structural and functional properties, and regulation by the addition of effectors. However, knowledge concerning the expressions and regulations of these enzymes in R. oryzae is lacking as compared to the wealth of biochemical and molecular information available from yeast (Goldberg et al., 2006). The majorities of the published work focused on a particular part, such as fumarase and its encoding gene, which does not provide a complete picture of the regulatory network in R. oryzae.

Pyruvate carboxylase (EC 6.4.1.1) catalyzes the carboxylation of pyruvate to oxaloacetate, which is the precursor for the biosynthesis of many C4 intermediates and is used in gluconeogenesis, biosynthesis of amino acids and fat metabolism (Jitrapakdee and Wallace, 1999). Pyruvate carboxylase belongs to the family of biotin-dependent

24 carboxylase and is composed of four identical subunits (~130 kDa each) organized as a tetramer (Attwood, 1995). It is present in many organisms including bacteria, fungi, plants and animals. Pyruvate carboxylase is situated in mitochondria in most eukaryotic organisms. However, in some filamentous fungi, such as Aspergillus nidulans,

Aspergillus terreus and R. oryzae, and in the yeast S. cerevisiae, pyruvate carboxylase is localized exclusively in the cytosol and lacks the peptide for mitochondrial targeting

(Osmani and Scrutton, 1985; Goldberg et al., 2006). The cytosolic localization seems to afford these microorganisms the ability to produce large amounts of fumaric acid

(Goldberg et al., 2006). Biotin is an important regulator of pyruvate carboxylase activity.

Long-term regulation of this enzyme in yeast can be realized by controlling the availability of biotin (Acar, 2004). The implication of pyruvate carboxylase in the production of fumaric acid by R. arrhizus has also been supported by using biotin as the effecter. Additionally, acetyl-CoA and L-aspartate could be used as the activator and inhibitor, respectively, for short term regulation (Acar, 2004).

Malate dehydrogenase (EC 1.1.1.37) catalyzes the reversible conversion of oxaloacetate to L-malate with the participation of NAD(H). Malate dehydrogenase is a multimeric enzyme consisting of identical subunits often arranged as either a dimer or a tetramer. The molecular weight of each subunit is about 30~40 kDa. Each subunit performs its catalytic reaction independently with no proof of cooperativity between catalytic sites. Malate dehydrogenase can be found in many organisms including bacteria, fungi, plants and animals. Malate dehydrogenase has several isoenzymes located in different subcellular organelles, including mitochondria, chloroplasts, glyoxysomes and

25 peroxysomes, and is involved in several metabolisms including TCA cycle and glyoxylate cycle. The known inhibitors of this enzyme include thyroxine, iodine, cyanide and chlorothricin, while phosphate, arsenate, and zinc ions are activators for this enzyme

(Peleg et al., 1989).

Fumarase (EC 4.2.1.2) catalyzes the reversible dehydration of L-malate to fumaric acid. Fumarase plays a key role in the TCA cycle and is extensively distributed in microorganisms, plants and animals. Fumarase can be divided into Class I and Class II with distinct properties (Woods et al., 1988). Much research has been carried out on the characterization of fumarase in microorganisms. Class I fumarases in E. coli (FumA and

FumB) (Guest et al., 1985), Euglena gracilis (Shibata et al., 1985), Bacillus stearothermophilus (Reaney et al., 1993) and a syntrophic propionate-oxidizing bacterium strain MPOB (Van Kuijk et al., 1998) have been identified and characterized, while Class II fumarases have been found and characterized in E. coli (FumC),

Sulfolobus solfataricu, Thermus thermophilus, S. cerevisiae, R. oryzae, Pseudomonas putida, Bacillus subtilis, and mammalian cells (Weaver et al., 1995; Mizobata et al., 1998;

Gerbod et al., 2001; Yang et al., 2004).

The subcellular distribution of fumarase has been the research interest for a long time. Though the existence of mitochondrial and cytosolic fumarases has been found in many organisms including S. cerevisiae, R. oryzae, human and mouse cells, and pig and rat liver tissues, the mechanisms of the allocation and formation of the full enzyme are still under investigation and many details remain unknown (Goldberg et al., 2006). Much work has been done centering on the subcellular distribution of fumarase in S. cerevisiae

(Pines et al., 1996). There were some controversies concerning the number of the encoding genes, the number of translational product and the size of the two isoenzymes.

Recently, it was confirmed that S. cerevisiae has only one FUM1 gene encoding fumarase and only one single translational product, which is targeted and processed in mitochondria before distribution between the cytosol and mitochondria (Suzuki et al.,

1992; Stein et al., 1994). It was proposed that a subset of the processed fumarase molecules is fully imported into the matrix, whereas the majority (70%) is partially translocated, so that their amino termini become accessible to mitochondrial matrix peptidase. These latter molecules are released back into the cytosol as soluble enzyme by retrograde movement through the translocation pore.

In R. oryzae, fumarase is a dual localized protein distributed in mitochondria and cytosol (Yogev et al., 2011; Yogev and Pines, 2011). For the cytoplasmic fumarase, the main direction of the catalyzed reaction is leading to the accumulation of fumaric acid, which is opposite to that catalyzed by the mitochondrial fumarase. A plausible explanation is that R. oryzae harbors two genes encoding these two fumarases (Friedberg et al., 1995). After inoculation of R. oryzae into the acid production medium, the induction of fumarase with unique characteristics prompts the conversion from malic acid to fumaric acid. A higher intracellular fumaric acid concentration of over 2 mmol/L could fully inhibit the reverse reaction to malic acid, which might contribute to the continued production and accumulation of fumaric acid in the fermentation. The fumR gene encoding fumarase in R. oryzae has been cloned and analyzed for its sequence and expression (Friedberg et al., 1995). Results showed that fumarase in R. oryzae is encoded

27 by gene fumR consisting of 2019 nt. The enzyme belongs to Class-II fumarase and is composed of 494 amino acids with the molecular mass of 50 kDa. It was also found that transcriptional regulation of fumarase might be involved in the production of fumaric acid under stress conditions.

Song et al. (2011) cloned 1440 bp R. oryzae fumarase fumR gene (GenBank accession number X78576.1) and expressed in E. coli BL21 (DE3). The FUMR protein was purified and its enzymatic properties were characterized. The fumR gene had a deletion of 15-amino acid sequence in the N-terminal region comparing with the fumR gene cloned by Friedberg (Friedberg et al., 1995). The result showed that the FUMR activity was optimal at 30˚C and pH7.2. Mg2+ had a slight inhibition and Ca2+ had s small stimulatory effect of FUMR activity. The Km for L-malic acid and fumaric acid were

0.46 mM and 3.07 mM, respectively. The activity of FUMR catalyzing the conversion of fumarate to L-malate was fully inhibited by 2 mM fumaric acid, which indicated that overexpression of FUMR could improve fumaric acid production in R. oryzae.

The studies on the catalytic properties of fumarase indicated that the reverse reaction of fumaric acid to malic acid is favored in the cellular metabolism since the ΔG0´ of the conversion of malic acid to fumaric acid is 3.6 KJ/mol (Gajewski et al., 1985). And the fumarase properties are similar to that in E. coli, the ΔG0´ of the conversion of fumarate to malate is -1.3 kcal/mol (Henry et al., 2006).

2.2.4 Metabolic engineering and system biology for strain development

The maximal theoretical yield is 2 mol of fumaric acid per mole of glucose consumed or 1.23 g/g in the reductive TCA pathway in a nongrowth situation. However, the actual fermentation yield is usually much less than 1.0 g/g and the productivity is also low compared to other organic acids, including lactic acid and citric acid, produced by filamentous fungi. Strain development through metabolic engineering offers a promising approach to increase fumaric acid production in filamentous fungi, although very little has been done in this area due to the difficulty in cloning, the complexity of the metabolic pathway and lack of knowledge of the regulatory network in these organisms (Skory,

2002, 2004b, 2005; Meussen et al., 2012a). Metabolic engineering of R. oryzae to overproduce fumaric acid has not been reported yet. However, Agrobacterium mediated

DNA transfer systems (Michielse et al., 2004) and several expression vectors have been constructed and can be used to express heterologous proteins in R. oryzae. For example, green fluorescent protein has been successfully expressed under the promoters of pdcA, amyA and pgk1 in R. oryzae (Mertens et al., 2006). Cyanophycin originated from microorganisms such as cyanobateria was successfully produced by R. oryzae 99-880 under the endogenous pdcA promoter and terminator (Meussen et al., 2012b).

Up to now, there is no available antifungal resistance marker to use on the selection of the plasmid transformed R. oryzae strains. Therefore, auxotrophic strains of R. oryzae are generated by N-methyl-N′-nitro-N-nitrosoguanidine and used as target organism. pyrG and pyrF which are encoded by orotidine-5′-monophosphate gene and orotate phosphoribosyltransferase gene respectively are mutagenized and complemented by the introduction of plasmid DNA (Skory, 2004a; Skory and Ibrahim, 2007). Biolistic

29 transformation is an efficient method to transform constructed plasmids into R. oryzae

(Klein et al., 1987; Gonzalez-Hernandez et al., 1997). The transformed plasmids hardly integrate into the genome chromosome, only replicate in a concatenated structure with a high molecular weight (>23 kb) outside the chromosome (Skory, 2004a; Skory and

Ibrahim, 2007).

In addition, the genome of Rhizopus oryzae 99-880 has been sequenced through the

Fungal Genome Initiative at the Whitehead Center for Genome Research at MIT, which should facilitate genetic manipulation and metabolic engineering of R. oryzae. The genome is 45.3 Mb in size, has 13,895 protein-coding genes, and contains abundant transposable elements. The order and genomic arrangement of the duplicated gene pairs and their common phylogenetic origin indicates that an ancestral whole-genome duplication event has occurred (Ma et al., 2009). The phenomena of genome duplication can make some difficulties on the genetic modification especially gene knockout in R. oryzae.

One way to increase fumaric acid production is to reduce the formation of byproducts, such as lactic acid and ethanol that are also produced in the cytosol along with fumaric acid. The synthesis of ethanol tends to occur primarily under anaerobic stress, presumably through the catalysis of pyruvate decarboxylase and alcohol dehydrogenase (ADH) (Skory, 2003a). R. oryzae mutants with reduced ADH activities obtained after mutagenesis with UV and nitrosoguanidine and selection on allyl alcohol containing plates produced 21.1% more fumaric acid and 83.7% less ethanol as compared to the wild type parental strains (Fu et al., 2010a). It is clear that the competitive ethanol

30 production needs to be disrupted in order to divert more pyruvate away from ethanol and increase fumaric acid production. However, the major obstruction in this approach is that gene disruption by homologous recombination is not applicable in R. oryzae as this fungus rarely integrates DNA used for transformation, but instead replicate plasmid autonomously in high molecular weight concatenated structures (Van Heeswijck, 1986).

The only one successful example is that the high-affinity iron permease gene (ftr1) was disrupted through double cross-over homologous recombination in an auxotrophic mutant derived from R. oryzae. But a homokaryotic null allele could not be segregated because of the multinucleated property of R. oryzae (Ibrahim et al., 2010).

The conversion of pyruvate to lactic acid is known to proceed by way of NAD- dependent L-lactate dehydrogenase (Skory, 2000a). The gene encoding lactate dehydrogenase was amplified from R. oryzae using degenerate primers by RT-PCR

(Hakki and Akkaya, 2001). Some genetic manipulations have been developed to overproduce lactic acid through metabolically engineered R. oryzae mutants (Bai et al.,

2004; Skory et al., 2009). Skory reported that a R. oryzae mutant overexpressing lactate dehydrogenase yielded higher levels of lactic acid from glucose (Skory et al., 1998;

Skory, 2004a). Lactate dehydrogenase gene ldhA has also been expressed in S. cerevisiae under the yeast endogenous promoter and terminator, lactic acid yielded 40% more than the strain without ldhA expression (Skory, 2003b). The production of lactic acid can be detected by transforming ldhB gene in fumaric acid producing strain R. oryzae 99-880

(Skory and Ibrahim, 2007). Similar approach can be used to overexpress key genes in the rTCA pathway for the overproduction of fumaric acid.

RNA interference (RNAi) with direct introduction of double-stranded RNA to trigger the degradation of a homologous sequence of mRNA, has been used for fungi to downregulate the gene expression since stable hairpin RNA (hpRNA) expression plasmids have been constructed (Goldoni et al., 2004). The genome sequence of R. oryzae 99-880 has been analyzed for the availability of proteins required by RNAi machinery (Nakayashiki and Nguyen, 2008). RNA silencing has been applied in R. oryzae to suppress the ldhA gene expression. An 85.7% reduction of lactic acid production can be achieved by short (25nt) synthetic siRNAs targeting the ldhA gene, and yield of ethanol was increased 15.4% (Gheinani et al., 2011).

As discussed before, the first reaction in the rTCA pathway is catalyzed by pyruvate carboxylase, which is localized exclusively in cytosol in R. oryzae and Saccharomyces cerevisiae (Osmani and Scrutton, 1985). Overexpressing this enzyme could increase CO2 fixation and oxaloacetate production from pyruvate, and thus may increase fumaric acid production. R. oryzae contains both cytosolic and mitochondrial isozymes of NADP- malate dehydrogenase and NAD-malate dehydrogenase (Osmani and Scrutton, 1985).

However, no change was observed in the isoenzyme pattern of malate dehydrogenase during fumaric acid production (Peleg et al., 1989). The gene fumR encoding R. oryzae fumarase has been cloned and analyzed for its structure and expression. fumR encodes a single transcript, but exists in both cytosolic and mitochondrial fraction (Peleg et al.,

1989). The level of fumR RNA increased in cells producing fumaric acid under stress conditions. This increased transcript of fumR leads to the cytosolic form of fumarase

(Friedberg et al., 1995). Thus, overexpressing fumR gene under a constitutive promoter may increase fumaric acid production, although this strategy has not been tested.

It should be noted that the complex regulatory networks involving redox balance,

ATP generation, and cell growth are largely unknown in R. oryzae. Moreover, the roles of enzymes in the rTCA pathway need to be further investigated in order to find out an efficient metabolic engineering strategy. For efficient strain improvement, both metabolic engineering and evolutionary engineering approaches should be applied, and aided with the knowledge obtained via “omic” technologies and systems biology, which analyzes cellular behavior on a global scale (Lee et al., 2005).

2.3 Fermentation process development

2.3.1 Medium composition

Glucose is used as the most common carbon source during the fermentation of fumaric acid producing strains. Fumaric acid was also produced from xylose by immobilized R. arrhizus cells in shake flasks, but the productivity was as low as 0.087 g/

L·h (Kautola, 1989). Although low cost carbon sources derived from starch-based materials such as corn steep liquor, molasses, potato, corn flour, and cassava using

Rhizopus spp. fermentation were investigated (Moresi, 1991, 1992; Carta, 1999; West,

2008; Kang et al., 2010; Xu et al., 2010). The strain R. formosa MUCL 28422 was selected as the best fumaric acid producer, yielding 21.3 g/L in a media containing enzymatic hydrolysis of cassava bagasse as the sole carbon source by submerged fermentation (Carta, 1999). Fumaric acid production by R. arrhizus from potato flour was

33 studied and a productivity of 0.42 g/L·h was achieved (Moresi, 1991). Moresi screened substrates including cassava, corn and potato flour and fumaric acid yield and productivity reached 0.6 g/g and 0.5 g/L·h using corn starch as carbon substrate (Moresi,

1992). Three strains of R. oryzae were screened for their ability to produce fumaric acid on untreated and treated corn distillers’ grains with soluble. R. oryzae ATCC 52918 had a highest fumaric acid productivity of 0.14 g/ L·h with 1.0% sulfuric acid hydrolyzed grains through solid-state fermentation (West, 2008).

Lignocellulosic material dairy manure can release carbohydrate after hydrolysis by which the fumaric acid yield reached 0.31 g/g using R. oryzae ATCC 20344 during pellet fermentation (Liao et al., 2008). However, the processes of acid pretreatment and hydrolysis of cellulose and hemicellulose in the plant biomass are complicate and costly, and some toxic compounds are released which severely inhibit cell growth and fermentation. R. nigricans can produce fumaric acid with a titer of 33.1 g/L using apple juice which contains 5% reducing sugar (Podgorska et al., 2004).

The nutritional and physical requirements of R. arrihzus to optimally improve fumaric acid biosynthesis have been studied (Rhodes et al., 1959). Nitrogen limitation is an essential requirement for filamentous fungi to accumulate high concentration of fumaric acid. It is because that when the medium contains sufficient nitrogen source, most of the carbon source will be consumed for biomass instead of fumaric acid biosynthesis (Bulut et al., 2009). In fungal fermentation, only when the nitrogen source is fully depleted, fumaric acid can be accumulated and secreted into the medium (Goldberg et al., 2006). It has been reported that C: N ratios ranging from 120:1 to 150:1 will

34 convert 60% to 70% of the glucose to fumaric acid (w/w) in submerged fermentation

(Magnuson and Lasure, 2004). A high fumaric acid yield of 85% on glucose was obtained using an initial C: N molar ratio of 200:1 for R. arrhizus (Roa Engel et al.,

2008). The effects of carbon-nitrogen ratio on the fumaric acid production were investigated in R. oryzae. The production of fumaric acid increased from 14.4 to 40.3 g/L with the urea concentration decreased from 2.0 to 0.1 g/L (Ding et al., 2011).

Fumaric acid production varies with the selection of nitrogen sources. Suitable nitrogen sources include organic and inorganic sources such as urea, ammonium chloride, ammonium sulfate, ammonium acetate, ammonium nitrate, ammonium biphosphate, asparagine and protein hydrolysate (Ling and Ng, 1989). Organic nitrogen source, such as yeast extract mainly promotes cell growth with little fumaric acid accumulation.

Inorganic source, eg. (NH4)2SO4 and urea, is superior for fumaric acid production. To obtain optimal fumaric acid yield, available nitrogen in the fermentation medium should be limited to a range of 0.14 to 0.42 g/L (Ling and Ng, 1989). Phosphorus limitation is another alternative when nitrogen limitation is unavailable in certain circumstances

(Riscaldati et al., 2000).

Rhizopus has very little nutritional demands. Only substrates and inorganic salts can give a satisfactory fermentation performance. The addition of some trace metals may significantly enhance fumaric acid accumulation. Zn2+ and Mg2+ at a concentration of 10 and 30 ppm have simulatory effect on the fumaric acid production. Phosphorus at 200 ppm is beneficial for overproduction of fumaric acid (Rhodes et al., 1959). Cu2+ may inhibit fumaric acid production at a concentration higher than 1 ppm. In addition, trace

35 metals in the growth medium also affect the morphology of mycelia growth. The optimal medium formulation containing 500 ppm Mg2+, 4 ppm Zn2+ and 100 ppb Fe2+ can promote small (1 mm) spherical pellets formation, which is beneficial to fumaric acid production (Zhou, 1999).

CO2 is involved in the metabolic conversion of pyruvate to oxaloacetate by pyruvate carboxylase. CaCO3 is added as a neutralizing agent as well as another carbon source except glucose (Rhodes et al., 1962; Ling and Ng, 1989). With the sufficient supply of

CO2 via reductive pyruvate carboxylation, the maximal theoretical yield of fumaric acid is 2 mole per mole of glucose. If no CO2 or carbonate is added, CO2 will be totally derived from the catabolism of glucose via citrate cycle. In this case, the maximal theoretical yield should be 1.5 mole of fumaric acid per mole of glucose (Roa Engel et al.,

2008).

Other nutrients like vitamins also play an important role in fumaric acid biosynthesis.

It has been reported that biotin is the activator of pyruvate carboxylase, and it is necessary for cell growth and metabolism of fats and amino acids. Riboflavin, also known as vitamin B2, is the central component of cofactors flavin adenine dinucleotide

(FAD) and flavin mononucleotide (FMN), which significantly affect energy metabolism.

Up to now, few studies have been done to investigate the effect of individual vitamins on the fumaric acid production. Methanol has shown to enhance citric acid production via its regulation on the citrate cycle in Aspergillus niger (Maddox et al., 1986). With respect to

Rhizopus species, the same mechanism could benefit the optimal fumaric acid production with methanol addition.

2.3.2 pH effect and neutralizing agents

The presence of a neutralizing agent to continuously maintain the pH of the fermentation process is particularly important for optimal yield of fumaric acid. The produced fumaric acid will accumulate protons in the fermentation medium which decrease the pH (Roa Engel et al., 2008). At low pH, excreted fumaric acid will diffuse back to the cytoplasm of the fungal cells and decreases the intracellular pH, cell metabolism will be impaired and fumaric acid biosynthesis is terminated.

CaCO3 is the most common used neutralizing agent by which highest fumaric acid yield and productivity were obtained (Zhou et al., 2002) . The low solubility of CaCO3 in the fermentation broth eliminates the negative impact of Ca2+ to the cell metabolism. At the same time, CO2 is supplied as the co-substrate for the fumaric acid biosynthesis.

However, the low solubility of CaCO3 also causes the high viscosity of the fermentation broth and cell interaction with the precipitated product, which negatively affects the mixing and oxygen transfer in the fermentation. And the solubility of products calcium fumarate and fumarate are only 21 g/L and 7 g/L, sulfuric acid is required to dissolve and recover fumaric acid from the fermentation broth.

Other neutralizing agents like Ca(OH)2, Na2CO3, NaHCO3 and (NH4)2CO3 have also been used for fumaric acid fermentation pH control (Gangl et al., 1990; Federici et al.,

1993; Zhou et al., 2002). Federici et al. (1993) used a combination of CaCO3 and

KOH/K2CO3 as the neutralizing agent for the conversion of glucose syrup to fumaric acid and comparable fumaric acid yield was obtained. Gang et al. (1990) used Na2CO3 to

37 neutralize the fumaric acid produced in order to avoid the complicated product recovery process.

In general, fumaric acid yield and productivity are lower when comparing with

+ CaCO3 as neutralizing agent. It maybe because that a high concentration of Na could negatively affect cell metabolism. In addition, sodium fumarate has a relatively higher solubility in water that could cause product inhibition and reduce fumaric acid production.

However, using NaHCO3 and Na2CO3 is still attractive since the high solubility of sodium fumarate leads to cheaper downstream processing. Heating is not required as compared with CaCO3. Moreover, the NaHCO3 alternative has the advantage of cell reuse for the next batch fermentation (Gangl et al., 1990; Zhou et al., 2002). These advantages may offset the disadvantages of lower fumaric acid yield and productivity.

Riscaldati et al. used (NH4)2CO3 as a neutralizing agent to directly produce ammonium fumarate in the condition of controlled mycelial growth by phosphorous limitation (Riscaldati et al., 2000). The fumaric acid yield and productivity were not comparable to those when CaCO3 was used. But the greater solubility of ammonium fumarate appeared to have no product inhibition effect.

Fermentation process without pH neutralizing agents has been proposed with the consideration of preventing product inhibition. One resolution is to employ simultaneous fermentation-separation process to continuously remove fumaric acid during fermentation

(Cao et al., 1996). The prevention of production inhibition is supposed to greatly enhance the process economics. Another possible alternative is to genetically manipulate the organisms to improve their acid tolerance (Maris et al., 2004) or export the fumaric acid

38 faster out of the plasma membrane. It should be noted that CO2 supply must be taken into consideration without carbonate neutralizing agents (Roa Engel et al., 2011).

2.3.3 Effects of dissolved oxygen on cell growth and fumaric acid production

Fumaric acid is produced by Rhizopus species in acerobic process whereas dissolved oxygen plays a key role in the fungal growth and fumaric acid yield. Since the solubility of oxygen in water is very low, only 40 mg/L at 25˚C. A submerged fermentation with continuous aeration and agitations is generally used to achieve a good oxygen transfer rate. Oxygen transfers from the air bubble into the fermentation broth, and dissolved oxygen transfers through the solution to the surface of the fungal cell and diffuses into the cell. In stirred-tank fermenters, higher agitation speeds could better disperse air bubbles, break them into smaller ones and increase the contact of liquid and air, and thereby increase oxygen transfer rate (Thongchul, 2005).

CaCO3 as a usual neutralizing agent in the fermentation process could severely affect oxygen limitation. The low solubility of CaCO3 causes the viscosity of the fermentation broth which deteriorates the oxygen transfer. Another situation that affects oxygen transfer rate it the fungal morphology. Rhizopus species tend to grow into mycelial mats or clumps, which will anchor onto the walls, stirrer, propellers and probes of the reactor (Tsao et al., 1999). This will result in the interference of oxygen and therefore enhanced ethanol production at the expense of fumaric acid accumulation. One way to avoid this problem is to control the growth of the fungal cells to obtain small compact mycelia pellets and then subculture into the non-growth fermentation medium.

The high fumaic acid production of 130 g/L was achieved by Rhizopus in a cultured medium with controlled dissolved oxygen levels in which the dissolved oxygen concentration is maintained between about 80 and 100% for the cell-growth phase and about 30-80% for the acid-production phase (Ling and Ng, 1989). Higher dissolved oxygen level is required in cell growth phase to accumulate enough biomass for subsequent acid production. During acid production phase, the metabolism of glucose is regulated towards reaction of reductive pyuvate carboxylation rather than with oxygen under conditions of limited oxygen availability (Ling and Ng, 1989). Therefore, a two- stage fermentation process with different dissolved oxygen control strategy is required for enhanced fumaric acid production. It has been reported that a high fumaric acid production of 56.2 g/L and high yield of 0.54 g/g on glucose were obtained by applying the dissolved oxygen concentration strategy of 80% in the first 18 h and then 30% afterwards. The fumaric acid productivity increased 37% with the increase of dissolved oxygen level from 30% to 80% (Fu et al., 2010b).

2.3.4 Morphology control

When filamentous fungi are grown in submerged culture, different types of morphology are formed under different fermentation process. The type of growth is generally classified into two groups: individual filamentous mycelia and pellet form (Cui et al., 1998b). Filamentous form consists of entanglements (Riley et al., 2000) and mycelia clumps which are differentiated based on the hyphal loops. Freely dispersed mycelia can grow in nutritious culture medium with sufficient dissolved oxygen and

40 substrate concentration, and all the hyphae are exposed to the medium and thus capable of contributing to growth. With respect to the pellet form, the growth rate was lowered due to oxygen starvation in the center of the pellets.

The filamentous morphology leads to highly viscous and pseudoplastic fermentation broth, negatively affects mass transfer and reduces the homogeneity in fermenters. In general, highly branched fungal mycelia are hard to operate in conventional stirred-tank bioreactors due to the difficulty in mixing and aeration caused by morphology. On the other hand, growing filamentous fungi in pellet form could lower the broth viscosity, increase the homogeneity, and improve the mass transfer in fermenters. Therefore power input in mixing and aeration can also be greatly reduced. However, the pellet form causes the problem of nutrient transport into the inner pellet. Metabolism at the pellet center is reduced by insufficient nutrient and oxygen supply, which results in decreased fumaric acid production rate.

Fungal morphology plays an important role in the metabolism during the fumaric acid fermentation process, and the morphology control is highly desired in industrial fermentations. Numerous factors influence the fungi growing into loose mycelia or pellets. They include the strain, medium composition, pH, temperature, mechanical forces, inoculums size and the dissolved oxygen and carbon dioxide (Cui et al., 1998a).

Physical factors include agitation systems, rheology and culture modes, eg., batch, fed- batch or continuous (Papagianni, 2004).

2.3.4.1 Pellet formation

Pellet formation can be classified into two types: coagulating type and noncoagulating type (Zhou, 1999). The coagulation type defines the process that spores aggregate together, or with other small agglomerates to form pellets. Whereas the noncoagulating pellet is formed through the process that one spore germinates to one pellet. Depending on the strain and culture conditions, the structure of pellets varies considerably from fluffy loose pellets, compact smooth pellets to hollow smooth pellets

(Zhou, 1999).

Fermentation process employing the fungal morphology of pellets is highly desirable since the pellet form exhibits low viscosities and ease of agitation and aeration

(Du et al., 1997a; Zhou et al., 2000; Chotisubha-anandha et al., 2011). However, the central area of larger pellets undergoes autolysis due to nutrient limitation, which can have a significantly negative effect on both cellular metabolism and fumaric acid biosynthesis (Papagianni, 2004). Therefore, small pellets with the diameter less than 1 mm are preferred more than the larger ones during the development of fungal fermentation.

A number of factors affecting the acquisition of uniform pellets of a desired size have been extensively studied (Liao et al., 2007a; Liao et al., 2007b; Liu et al., 2008).

These factors can be summarized as follows: inoculum size, type and age of strains, genetic factors, medium composition, surfactants, shear forces, temperature and pressure, medium viscosity, etc (Papagianni, 2004). It has been reported in P. Chrysogenum that pellet formation occurs by the agglomeration of dispersed mycelia (Nielsen et al., 1995).

The size of pellets depends on the inoculated spore concentration. At low concentrations,

42 agglomeration of hyphal elements was limited which results in the formation of small pellets. At higher spore concentrations, agglomeration was severed and large pellets were formed. It has been reported that pH of the culture medium also strongly influences the pellet formation. In general, the tendency of pellet formation increases with the increase of pH value of the culture medium. In the study of R. oryzae in a stirred tank reactor, freely dispersed hyphal elements were formed at pH value between 3.0 and 3.5. At pH value above 4.0, pellet was observed and it was caused by agglomeration of the spores.

At pH higher than 6.0, only pellets were obtained and the pellet size increased with increasing pH value (Papagianni, 2004). Different nitrogen source contributes to the formation of pellet morphology. In R. nigricans, increased nitrogen content led to larger and denser pellet formation, while at lower nitrogen levels, small and fluffy pellets and even clumps tended to form (Znidarsic et al., 2000).

2.3.4.2 Cell immobilization

Fermentation of filamentous fungi in a stirred tank bioreactor is usually troublesome because of the diversified fungal morphology which in turn affects fumaric acid production. Besides growing the spores to small, uniform and dispersed pellets, immobilization process generally leads to a dramatic decrease in broth viscosity, enhanced nutrient and oxygen transfer, and repeated batch and continuous process are feasible and easy to operate.

Various immobilization techniques have been innovated and applied in submerged filamentous fermentation. Fungal mycelia can be immobilized on solid carriers through

43 entrapment and adsorption. Calcium alginate beads and polyurethane foam are usually employed as supports for mycelia to grown in the inner structure of these porous materials. Entrapment is often used to immobilize Rhizopus cells for fumaric acid production in most of the earlier studies. The solid carriers for entrapment included polyurethane foam cubes, cork pieces, perlite and alumina, Ca-alginate and polyurethane sponge (Buzzini et al., 1995; Petruccioli et al., 1996; Ganguly et al., 2007). Adsorption is another way of immobilization by which cells are attached to the surface of solid carriers through physical or chemical interaction. Compared to entrapment, adsorption has become more attractive lately due to its advantages of simple operation and lower cost.

Rhizopus mycelia can also be immobilized on the plastic surface of the plates and form the biofilm during the growth stage. A rotary biofilm contactor (RBC) has been introduced by Cao et al. to produce fumaric acid by R. oryzae ATCC 20344 (Cao et al.,

1997). The six plastic discs with the total surface area of 750 cm2 acted as supports for the growth of filamentous fungi. The discs were mounted on a horizontal shaft and placed in the fermenter containing growth medium. After the immobilization, the broth in the reactor was changed from growth to the nitrogen depleted fermentation medium. During the production phase, the shaft rotated in the liquid medium and slowly exposed the biofilm to the gas space and liquid medium for oxygen and nutrient uptake. The fumaric acid production rate of 3.78 g/L·h and yield of 0.75 g/g on glucose were obtained. With the combination of an adsorption column for fumaric acid in-situ separation and recovery, the yield of fumaric acid reached over 90% of the theoretical maximum yield, 0.85 g/g glucose consumed (Cao et al., 1996). The biofilm in the RBC can be used in continuous

44 process since the fungal mycelia are remained biologically active with nitrogen-rich medium.

A fibrous-bed bioreactor (RFB) was developed for fungal cell immobilization for lactic acid production (Tay and Yang, 2002). A rotating fibrous bed bioreactor was modified by affixing a perforated stainless steel cylinder mounted with a 100% cotton cloth to the agitation shaft. The spores were inoculated into the growth culture and cells were immobilized on the fibrous matrix after 48 h (Thongchul, 2005). Cell immobilization on the fibrous matrix is very effective since no mycelia found in the elements of the bioreactor and cells can be easily reused for lactic acid production periodically after replenishing the production medium.

2.4 References

Abe, A., Oda, Y., Asano, K., Sone, T., 2007. Rhizopus delemar is the proper name for Rhizopus oryzae fumaric-malic acid producers. Mycologia 99, 714-722.

Acar, S., 2004. Biochemical and genetics studies on the pyruvate branch point enzymes of Rhizopus oryzae. The Middle East Technical University.

Amedioha, A.C., 1993. Production of cellolytic enzymes by Rhizopus oryzae in culture and Rhizopus-infected tissues of potato tubers. Mycologia 85, 574-578.

Attwood, P.V., 1995. The structure and the mechanism of action of pyruvate carboxylase. Int J Biochem Cell Biol 27, 231-249.

Bachmann, W.E., Scott, L.B., 1948. The reaction of anthracene with maleic and fumaric acid and their derivatives and with citraconic anhydride and mesaconic acid. J Am Chem Soc 70, 1458-1461.

Bai, D.M., Zhao, X.M., Li, X.G., Xu, S.M., 2004. Strain improvement of Rhizopus oryzae for over-production of L-(+)-lactic acid and metabolic flux analysis of mutants. Biochem Eng J 18, 41-48.

Bakir, U., Yavascaoglu, S., Guvenc, F., Ersayin, A., 2001. An endo-β-1,4- xylanase from Rhizopus oryzae: production, partial purification and biochemical characterization. Enzyme Microb Techn 29, 328-334.

Bulut, S., Elibol, M., Ozer, D., 2009. Optimization of process parameters and culture medium for L-(+)-lactic acid production by Rhizopus oryzae. J Chem Eng Jpn 42, 589- 595.

Buzzini, P., Gobbetti, M., Rossi, J., Ribaldi, M., 1995. Comparison in different unconventional supports for the immobilization of Rhizopus arrhizus cells to produce fumaric acid on grape must. Ann Microbiol Enzymol 43, 53-60.

Cao, N., Du, J., Chen, C., Gong, C.S., Tsao, G.T., 1997. Production of fumaric acid by immobilized rhizopus using rotary biofilm contactor. Appl Biochem Biotechnol 63-65, 387-394.

Carta, F.S., Soccol, C.R.,Ramos, L.P., Fontana, J.D., 1999. Production of fumaric acid by fermentation of enzymatic hydrolysates derived from cassava bagasse. Bioresour Technol 68, 23-28.

Chotisubha-anandha, N., Thitiprasert, S., Tolieng, V., Thongchul, N., 2011. Improved oxygen transfer and increased L-lactic acid production by morphology control of Rhizopus oryzae in a static bed bioreactor. Bioprocess Biosyst Eng 34, 163-172.

Cui, Y.Q., Okkerse, W.J., van der Lans, R.G.J.M., Luyben, K.C.A.M., 1998a. Modeling and measurements of fungal growth and morphology in submerged fermentations. Biotechnol Bioeng 60, 216-229.

Cui, Y.Q., van der Lans, R.G.J.M., Luyben, K.C.A.M., 1998b. Effects of dissolved oxygen tension and mechanical forces on fungal morphology in submerged fermentation. Biotechnol Bioeng 57.

Ding, Y., Li, S., Dou, C., Yu, Y., Huang, H., 2011. Production of fumaric acid by Rhizopus oryzae: role of carbon–nitrogen ratio. Appl Biochem Biotechnol 164, 1461- 1467.

Du, J., Cao, N., Gong, C.S., Tsao, G.T., Yuan, N., 1997. Fumaric acid production in airlift loop reactor with porous sparger. Appl Biochem Biotechnol 63-65, 541-556.

Federici, F., Moresi, M., Parente, E., Petruccioli, M., Piccioni, P., 1993. Effect of stirring rate and neutralizing agent on fumaric acid production by Rhizopus arrhizus. Ital J Food Sci 4, 387-396.

Foster, J.W., Carson, S.F., 1949. Aerobic formation of fumaric acid in the mold Rhizopus nigricans, synthesis by direct C2 condensation. Proc Natl Acad Sci U S A 35, 663-672.

Foster, J.W., Waksman, S.A., 1939. The production of fumaric acid by molds belonging to the genus Rhizopus. J Am Chem Soc 61, 127-135.

Fu, Y., Xu, Q., Li, S., Chen, Y., Huang, H., 2010a. Strain improvement of Rhizopus oryzae for over-production of fumaric acid by reducing ethanol synthesis pathway. Korean J Chem Eng 27, 183-186.

Fu, Y.Q., Li, S., Chen, Y., Xu, Q., Huang, H., Sheng, X.Y., 2010b. Enhancement of fumaric acid production by Rhizopus oryzae using a two-stage dissolved oxygen control strategy. Appl Biochem Biotechnol 162, 1031-1038.

Gajewski, E., Goldberg, R.N., Steckler, D.K., 1985. Thermodynamics of the conversion of fumarate to L-(-)-malate. Biophys Chem 22, 187-195.

Gangl, I.C., Weigand, W.A., Keller, F.A., 1990. Economic comparison of calcium fumarate and sodium fumarate production by Rhizopus arrhizus. Appl Biochem Biotechnol 24-25, 663-677.

Ganguly, R., Dwivedi, P., Singh, R.P., 2007. Production of lactic acid with loofa sponge immobilized Rhizopus oryzae RBU2-10. Bioresour Technol 98, 1246-1251.

Gerbod, D., Edgcomb, V.P., Noel, C., Vanacova, S., Wintjens, R., Tachezy, J., Sogin, M.L., Viscogliosi, E., 2001. Phylogenetic relationships of class II fumarase genes from Trichomonad species. Mol Biol Evol 18, 1574-1584.

Gheinani, A.H., Jahromi, N.H., Feuk-Lagerstedt, E., Taherzadeh, M.J., 2011. RNA silencing of lactate dehydrogenase gene in Rhizopus oryzae. J RNAi Gene Silencing 14, 363-366.

Ghosh, B., Ray, R.R., 2011. Current commercial perspective of Rhizopus oryzae: a review. J Appl Sci 11, 2470-2486.

Goldberg, I., Rokem, J.S., Pines, O., 2006. Organic acids: old metabolites, new themes. J Chem Technol & Biotechnol 81, 1601-1611.

Goldoni, M., Azzalin, G., Macino, G., Cogoni, C., 2004. Efficient gene silencing by expression of double stranded RNA in Neurospora crassa. Fungal Genet Biol 41, 1016- 1024.

Gonzalez-Hernandez, G.A., Herrera-Estrella, L., Rocha-Ramirez, V., Roncero, M.I.G., Gutierrez-Corona, J.F., 1997. Biolistic transformation of Mucor circinelloides. Mycol Res 101, 953-956.

Goto M., Nara T., Tokumaru I., Fugono N., Uchida Y., Terasawa M., H., Y., 1998. Method of producing fumaric acid. US5783428.

Grijpma, D.W., Hou, Q., Feijen, J., 2005. Preparation of biodegradable networks by photo-crosslinking lactide, ε-caprolactone and trimethylene carbonate-based oligomers functionalized with fumaric acid monoethyl ester. Biomaterials 26, 2795-2802.

Guest, J.R., Miles, J.S., Roberts, R.E., Woods, S.A., 1985. The fumarase genes of Escherichia coli: location of the fumB gene and discovery of a new gene (fumC). J Gen Microbiol 131, 2971-2984.

Hakki, E.E., Akkaya, M.S., 2001. RT-PCR amplification of a Rhizopus oryzae lactate dehydrogenase gene fragment. Enzyme Microb Technol 28, 259-264.

Henry, C.S., Jankowski, M.D., Broadbelt, L.J., Hatzimanikatis, V., 2006. Genome-scale thermodynamic analysis of Escherichia coli metabolism. Biophys J 90, 1453-1461.

Jitrapakdee, S., Wallace, J.C., 1999. Structure, function and regulation of pyruvate carboxylase. Biochem J 340 ( Pt 1), 1-16.

Kang, S.W., Lee, H., Kim, D., Lee, D., Kim, S., Chun, G.-T., Lee, J., Kim, S.W., Park, C., 2010. Strain development and medium optimization for fumaric acid production. Biotechnol Bioprocess Eng 15, 761-769.

Karmakar, M., Ray, R.R., 2010. Extra cellular endoglucanase production by Rhizopus oryzae in solid and liquid state fermentation of agro wastes. Asian J Biotechnol 52, 27-36.

Kato, Y., Yamagishi, J., Asano, Y., 1995. Maleate cis-trans Isomerase from Arthrobacter sp. TPU 5446. J Ferment Bioeng 80, 610-612.

Kautola, H., Linko, Y., 1989. Fumaric acid production from xylose by immobilized Rhizopus arrhizus cells. Appl Microbiol Biotechnol 31, 448-452.

Kenealy, W., Zaady, E., du Preez, J.C., Stieglitz, B., Goldberg, I., 1986. Biochemical aspects of fumaric acid accumulation by Rhizopus arrhizus. Appl Environ Microbiol 52, 128-133.

Klein, T.M., Wolf, E.D., Wu, R., Sanford, J.C., 1987. High-velocity microprojectiles for delivering nucleic acids into living cells. Nature 326, 70-73.

Lee, S.Y., Hong, S.H., Lee, S.H., Park, S.J., 2004. Fermentative production of chemicals that can be used for polymer synthesis. Macromol Biosci 4, 157-164.

Lee, S.Y., Lee, D.Y., Kim, T.Y., 2005. Systems biotechnology for strain improvement. Trends in Biotechnol 23, 349-358.

Liao, W., Liu, Y., Chen, S., 2007a. Studying pellet formation of a filamentous fungus Rhizopus oryzae to enhance organic acid production. Appl Biochem Biotechnol 137-140, 689-701.

Liao, W., Liu, Y., Frear, C., Chen, S., 2007b. A new approach of pellet formation of a filamentous fungus -Rhizopus oryzae. Bioresour Technol 98, 3415-3423.

Liao, W., Liu, Y., Frear, C., Chen, S., 2008. Co-production of fumaric acid and chitin from a nitrogen-rich lignocellulosic material – dairy manure – using a pelletized filamentous fungus Rhizopus oryzae ATCC 20344. Bioresource Technology 99, 5859- 5866.

Ling, L.B., Ng, T.K., 1989. Fermentation process for carboxylic acids.

Liu, Y., Liao, W., Chen, S., 2008. Study of pellet formation of filamentous fungi Rhizopus oryzae using a multiple logistic regression model. Biotechnol Bioeng 99, 117- 128.

Lorences, M.J., Patience, G.S., Díez, F.V., Coca, J., 2003. Butane oxidation to maleic anhydride: kinetic modeling and byproducts. Ind Eng Chem Res 42, 6730-6742.

Maddox, I.S., Hossain, M., Brooks, J.D., 1986. The effect of methanol on citric acid production from galactose by A. niger. Appl Microbiol Biotechnol 23, 203-205.

Maris, A.J.A.v., Konings, W.N., Dijken, J.P.v., Pronk, J.T., 2004. Microbial export of lactic and 3-hydroxypropanoic acid: implications for industrial fermentation processes. Metab Eng 6, 245-255.

McGinn, S.M., Beauchemin, K.A., Coates, T., Colombatto, D., 2004. Methane emissions from beef cattle: Effects of monensin, sunflower oil, enzymes, yeast, and fumaric acid. J Anim Sci 82, 3346-3356.

Meek, J.S., 1975. The determination of a mechanism of isomerization of maleic acid to fumaric acid. J Chem Educ 52.

Mertens, J.A., Skory, C.D., Ibrahim, A.S., 2006. Plasmids for expression of heterologous proteins in Rhizopus oryzae. Arch Microbiol 186, 41-50.

Meussen, B.J., Graaff, L.H., Sanders, J.P.M., Weusthuis, R.A., 2012a. Metabolic engineering of Rhizopus oryzae for the production of platform chemicals. Appl Microbiol Biotechnol 94, 875-886.

Michielse, C.B., Hooykaas, P.J.J., Hondel, C.A.M.J.J., Ram, A.F.J., 2004. Agrobacterium-mediated transformation as a tool for functional genomics in fungi. Curr Genet 48, 1-17.

Mizobata, T., Fujioka, T., Yamasaki, F., Hidaka, M., Nagai, J., Kawata, Y., 1998. Purification and characterization of a thermostable class II fumarase from Thermus thermophilus. Arch Biochem Biophys 355, 49-55.

Moharregh-Khiabani, D., Linker, R.A., Gold, R., Stangel, M., 2009. Fumaric acid and its esters: an emerging treatment for multiple sclerosis. Curr Neuropharmacol 7, 60-64.

Moresi, M., Parente, E., Petruccioli, M., Federici, F., 1991. Optimization of fumaric acid production from potato flour by Rhizopus arrhizus. Appl Microbiol Biotechnol 36, 35-39.

Moresi, M., Parente, E., Petruccioli, M., Federici, F., 1992. Fumaric acid production from hydrolysates of starch-based substrates. J. Chem. Tech. Biotechnol. 54, 283-290.

Morgan, E.J., Friedmann, E., 1938. Interaction of maleic acid with thiol compounds. Biochem J 32, 733-742.

Mrowietz, U., Christophers, E., Altmeyer, P., 1999. Treatment of severe psoriasis with fumaric acid esters: scientific background and guidelines for therapeutic use. The German Fumaric Acid Ester Consensus Conference. Br J Dermatol 141, 424-429.

Nakajima-Kambe, T., Nozue, T., Mukouyama, M., Nakahara, T., 1997. Bioconversion of maleic acid to fumaric acid by Pseudomonas alcaligenes strain XD-1. J Ferment Bioeng 84, 165-168.

Nakayashiki, H., Nguyen, Q.B., 2008. RNA interference: roles in fungal biology. Curr Opin Microbiol 11, 494-502.

Nieboer, C., de Hoop, D., van Loenen, A.C., Langendijk, P.N., van Dijk, E., 1989. Systemic therapy with fumaric acid derivates: new possibilities in the treatment of psoriasis. J Am Acad Dermatol 20, 601-608.

Nielsen, J., Johansen, C.L., Jacobsen, M., Krabben, P., Villaden, J., 1995. Pellet formation and fragmentation in submerged culture of Penicillium chrysogenum and its relation to penicillin production. . Biotechnol Prog 11, 93-98.

Osmani, S.A., Scrutton, M.C., 1985. The sub-cellular localization and regulatory properties of pyruvate carboxylase from Rhizopus arrhizus. Eur J Biochem 147, 119-128.

Otsu, T., Yasuhara, T., Kohei, S., Mori, S., 1984. Radical high polymerization of di-tert- butyl fumarate and novel synthesis of high molecular weight poly(fumaric acid) from its polymer. Polymer Bulletin 12, 449-456.

Otsuka, K., 1961. Cis-trans isomerase; isomerization from maleic acid to fumaric acid. Agric Biol Chem 25, 726-730.

Overman, S.A., Romano, A.H., 1969. Pyruvate carboxylase of Rhizopus nigricans and its role in fumaric acid production. Biochem Biophys Res Commun 37, 457-463.

Papagianni, M., 2004. Fungal morphology and metabolite production in submerged mycelial processes. Biotechnol Adv 22, 189-259.

Petruccioli, M., Angiani, E., Federici, F., 1996. Semi continuous fumaric acid production by Rhizopus arrhizus immobilized in polyurethane sponge. Process Biochem 31, 463-469.

Pines, O., Even-Ram, S., Elnathan, N., Battat, E., Aharonov, O., Gibson, D., Goldberg, I., 1996. The cytosolic pathway of L-malic acid synthesis in Saccharomyces cerevisiae: the role of fumarase. Appl Microbiol Biotechnol 46, 393-399.

Podgorska, E., Kaspizak, M., Szwajgier, D., 2004. Fumaric acid production by Rhizopus nigricans and Rhizopus oryzae using apple juice. Pol J Food Nutr Sci 13, 47-50.

Rani, R., Ghosh, S., 2011. Production of phytase under solid-state fermentation using Rhizopus oryzae: novel strain improvement approach and studies on purification and characterization. Bioresour Technol 102, 10641-10649.

Reaney, S.K., Bungard, S.J., Guest, J.R., 1993. Molecular and enzymological evidence for two classes of fumarase in Bacillus stearothermophilus (var. non-diastaticus). J Gen Microbiol 139, 403-416.

Rhodes, R.A., Lagoda, A.A., Jackson, R.W., Misenhei, T.J., Smith, M.L., Anderson, R.F., 1962. Production of fumaric acid in 20 liter fermentors. Appl Microbiol 10, 9-15.

Rhodes, R.A., Moyer, A.J., Smith, M.L., Kelley, S.E., 1959. Production of fumaric acid by Rhizopus arrhizus. Appl Microbiol Biotechnol 7, 74-80.

Riley, G.L., Tucker, K.G., Paul, G.C., Thomas, C.R., 2000. Effect of biomass concentration and mycelial morphology of fermentation broth rheology. Biotechnol Bioeng 68, 160-172.

Riscaldati, E., Moresi, M., Federici, F., Petruccioli, M., 2000. Direct ammonium fumarate production by Rhizopus arrhizus under phosphorous limitation. Biotechnol Lett 22, 1043- 1047.

Roa Engel, C.A., Straathof, A.J.J., Zijlmans, T.W., Gulik, W.M., Wielen, L.A.M., 2008. Fumaric acid production by fermentation. Appl Microbiol Biotechnol 78, 379-389.

Roa Engel, C.A., Van Gulik, W.M., Marang, L., van der Wielen, L.A.M., Straathof, A.J.J., 2011. Development of a low pH fermentation strategy for fumaric acid production by Rhizopus oryzae. Enzyme and Microb Technol 48, 39-47.

Romano, A.H., Bright, M.M., Scott, W.E., 1967. Mechanism of fumaric acid accumulation in Rhizopus nigricans. J Biotechnol 93, 600-604.

Saito, K., Saito, A., Ohnishi, M., Oda, Y., 2004. Genetic diversity in Rhizopus oryzae strains as revealed by the sequence of lactate dehydrogenase genes. Arch Microbiol 182, 30-36.

Shibata, H., Gardiner, W.E., Schwartzbach, S.D., 1985. Purification, characterization, and immunological properties of fumarase from Euglena gracilis var. bacillaris. J Bacteriol 164, 762-768.

Skinner, J.T., Izat, A.L., Waldroup, P.W., 1991. Research note: fumaric acid enhances performance of broiler chickens. Poult Sci 70, 1444-1447.

Skory, C.D., 2000. Isolation and expression of lactate dehydrogenase genes from Rhizopus oryzae. Appl Environ Microbiol 182, 30-36.

Skory, C.D., 2002. Homologous recombination and double-strand break repair in the transformation of Rhizopus oryzae. Mol Genet Genomics 268, 397-406.

Skory, C.D., 2003a. Induction of Rhizopus oryzae pyruvate decarboxylase genes. Curr Microbiol 47, 59-64.

Skory, C.D., 2003b. Lactic acid production by Saccharomyces cerevisiae expressing a Rhizopus oryzae lactate dehydrogenase gene. J Ind Microbiol Biot 30, 22-27.

Skory, C.D., 2004a. Lactic acid production by Rhizopus oryzae transformants with modified lactate dehydrogenase activity. Appl Microbiol Biotechnol 64, 237-242.

Skory, C.D., 2004b. Repair of plasmid DNA used for transformation of Rhizopus oryzae by gene conversion. Curr Genet 45, 302-310.

Skory, C.D., 2005. Inhibition of non-homologous end joining and integration of DNA upon transformation of Rhizopus oryzae. Mol Genet Genomics 274, 373-383.

Skory, C.D., Freer, S.N., Bothast, R.J., 1998. Production of L-lactic acid by Rhizopus oryzae under oxygen limiting conditions. Biotechnol Lett 20, 191-194.

Skory, C.D., Ibrahim, A.S., 2007. Native and modified lactate dehydrogenase expression in a fumaric acid producing isolate Rhizopus oryzae 99-880. Curr Genetics 52, 23-33.

Skory, C.D., Mertens, J.A., Rich, J.O., 2009. Inhibition of Rhizopus lactate dehydrogenase by fructose 1,6-bisphosphate. Enzyme Microb Technol 44, 242-247.

Song, P., Li, S., Ding, Y., Xu, Q., Huang, H., 2011. Expression and characterization of fumarase (FUMR) from Rhizopus oryzae. Fungal Biol 115, 49-53.

Sosaku Ichikawa, T.I., Seigo Sato, Tadaatsu Nakahara, Sukekuni Mukataka, 2003. Improvement of production rate and yield of fumaric acid from maleic acid by heat treatment of Pseudomonas alcaligenes strain XD-1. Biochemical Engineering Journal, 7- 13.

Stein, I., Peleg, Y., Even-Ram, S., Pines, O., 1994. The single translation product of the FUM1 gene (fumarase) is processed in mitochondria before being distributed between the cytosol and mitochondria in Saccharomyces cerevisiae. Mol Cell Biol 14, 4770-4778.

Suzuki, T., Yoshida, T., Tuboi, S., 1992. Evidence that rat liver mitochondrial and cytosolic fumarases are synthesized from one species of mRNA by alternative translational initiation at two in-phase AUG codons. Eur J Biochem 207, 767-772.

Taherzadeh, M.J., Fox, M., Hjorth, H., Edebo, L., 2003. Production of mycelium biomass and ethanol from paper pulp sulfite liquor by Rhizopus oryzae. Bioresour Technol 88, 167-177.

Takamura, Y., Takamura, T., Soejima, M., Uemura, T., 1969. Studies on the induced synthesis of maleate cis-trans isomerase by malonate (III). Purification and properties of maleate cis-trans isomerase induced by malonate. Agric Biol Chem 33, 718-728.

Tay, A., Yang, S.T., 2002. Production of L(+)-lactic acid from glucose and starch by immobilized cells of Rhizopus oryzae in a rotating fibrous bed bioreactor. Biotechnol Bioeng 80, 1-12.

Thongchul, N., 2005. Lactic acid production by immobilized Rhizopus oryzae in a rotating fibrous bed bioreactor. The Ohio State University.

Tsao, G.T., Cao, N.J., Du, J., Gong, C.S., 1999. Production of multifunctional organic acids from renewable resources. Adv Biochem Eng/Biotechnol 65, 243-280.

Van Heeswijck, R., 1986. Autonomous replication of plasmids in Mucor transformants. Carlsberg Res Commun 51, 433-443.

Van Kuijk, B.L., Schlosser, E., Stams, A.J., 1998. Investigation of the fumarate metabolism of the syntrophic propionate-oxidizing bacterium strain MPOB. Arch Microbiol 169, 346-352.

Weaver, T.M., Levitt, D.G., Donnelly, M.I., Stevens, P.P., Banaszak, L.J., 1995. The multisubunit active site of fumarase C from Escherichia coli. Nat Struct Biol 2, 654-662.

West, T.P., 2008. Fumaric acid production by Rhizopus oryzae on corn distillers' grains with solubles. Res J Microbiol 3, 35-40.

Woods, S.A., Schwartzbach, S.D., Guest, J.R., 1988. Two biochemically distinct classes of fumarase in Escherichia coli. Biochim Biophys Acta 954, 14-26.

Wright, B.E., Longacre, A., Reimers, J., 1996. Models of metabolism in Rhizopus oryzae. J Theor Biol 182.

Xu, Q., Li, S., Fu, Y., Tai, C., Huang, H., 2010. Two-stage utilization of corn straw by Rhizopus oryzae for fumaric acid production. Bioresour Technol 101, 6262-6264.

Yang, J., Wang, Y., Woolridge, E.M., Arora, V., Petsko, G.A., Kozarich, J.W., Ringe, D., 2004. Crystal structure of 3-carboxy-cis,cis-muconate lactonizing enzyme from Pseudomonas putida, a fumarase class II type cycloisomerase: enzyme evolution in parallel pathways. Biochemistry 43, 10424-10434.

Yang, S.T., Zhang, K., Zhang, B., Huang, H., 2011. Biobased chemicals - fumaric acid. In: Moo-Young M (ed.) Comprehensive Biotechnology, 2nd edition.

Yogev, O., Naamati, A., Pines, O., 2011. Fumarase: a paradigm of dual targeting and dual localized functions. FEBS J 278, 4230-4242.

Yogev, O., Pines, O., 2011. Dual targeting of mitochondrial proteins: Mechanism, regulation and function. Biochim Biophys Acta (BBA) - Biomembr 1808, 1012-1020.

Zhou, Y., 1999. Fumaric acid fermentation by Rhizopus oryzae in submerged systems. PhD thesis. Purdue University.

Zhou, Y., Du, J., Tsao, G.T., 2000. Mycelial pellet formation by Rhizopus oryzae ATCC 20344. Appl Biochem Biotechnol 84-86, 779-789.

Zhou, Y., Du, J., Tsao, G.T., 2002. Comparison of fumaric acid production by Rhizopus oryzae using different neutralizing agents. Bioprocess Biosyst Eng 25, 179-181.

Zlotkin, S., Arthur, P., Antwi, K.Y., Yeung, G., 2001. Treatment of anemia with microencapsulated ferrous fumarate plus ascorbic acid supplied as sprinkles to complementary (weaning) foods. Am J Clin Nutr 74, 791-795.

Znidarsic, P., Komel, R., Pavko, A., 2000. Influence of some environmental factors on Rhizopus nigricans submerged growth in the form of pellets. World J Microbiol Biotechnol 16, 589-593.

Figure 2.1 Molecular structure of fumaric acid.

O H O 2 2 Hydrolysis n-Butane Maleic anhydride Maleic acid

cis- trans Isomerization

Crude Decolorization Centrifugation fumaric acid

Filtration Crystallization Drying

Fumaric acid cystal

Figure 2.2 Fumaric acid production via petrochemical route.

CHAPTER 3

EFFECTS OF OVEREXPRESSING pyc AND pepc GENES ON FUMARIC ACID

BIOSYNTHESIS FROM GLUCOSE

Summary

Fumaric acid, a dicarboxylic acid used as a food acidulant and in manufacturing synthetic resins, can be produced from glucose in fermentation by Rhizopus oryzae.

However, the fumaric acid yield is limited by the co-production of ethanol and other byproducts. To increase fumaric acid production, overexpressing endogenous pyruvate carboxylase (PYC) and exogenous phosphoenolpyruvate carboxylase (PEPC) to increase the carbon flux toward oxaloacetate were investigated. Compared to the wild type, the

PYC activity in the pyc transformants increased 56%‒83%, whereas pepc transformants exhibited significant PEPC activity (3‒6 mU/mg) that was absent in the wild type.

Fumaric acid production by the pepc transformant increased 26% (0.78 g/g glucose vs.

0.62 g/g for the wild type). However, the pyc transformants grew poorly and had low fumaric acid yields (<0.05 g/g glucose) due to the formation of large cell pellets that limited oxygen supply and resulted in the accumulation of ethanol with a high yield of

0.13-0.36 g/g glucose. This study is the first attempt to use metabolic engineering to

58 modify the fumaric acid biosynthesis pathway to increase fumaric acid production in R. oryzae.

3.1 Introduction

Fumaric acid (HOOCCH=CHCOOH), an organic acid with a trans-double bond and two carboxylic acid groups, has many industrial applications, mainly as a chemical feedstock for the manufacturing of synthetic resins, biodegradable polymers, and plasticizers (Yang et al., 2011). It is also widely used as an acidulant in foods and beverages, and miscellaneous industrial products, including lubricating oils, inks, lacquers, and carboxylating agents for styrenebutadiene rubber (Cao et al., 1996; Roa

Engel et al., 2008). Currently, fumaric acid is primarily produced through the petrochemical route, including the catalytic oxidation of hydrocarbons (e.g., benzene, n- butane or n-butane/n-butene mixture) to maleic anhydride, hydrolysis of maleic anhydride to maleic acid, and finally cis-trans isomerization of maleic acid to fumaric acid (Lohbeck et al., 1990). However, the escalating prices of crude oils and petroleum products, and concerns of depleting oils and environmental pollution caused by petroleum refinery have generated renewed interest in the production of biobased fumaric acid by filamentous fungal fermentation using Rhizopus species (Yang, 2007).

Fumaric acid is an intermediate metabolite in the TCA cycle. However, fumarate generated in the TCA cycle is mainly used for the biosynthesis of cell constituents and cannot be accumulated in a significant amount during active cell growth. The biosynthesis of fumaric acid in Rhizopus mainly takes place in cytosol and involves three

59 enzymes from pyruvate: pyruvate carboxylase (PYC), which catalyzes the ATP- dependent condensation of pyruvate and CO2 to form oxaloacetic acid (OAA), malate dehydrogenase (MDH), which converts OAA to malate, and fumarase, which converts malate to fumarate (Osmani and Scrutton, 1985; Kenealy et al., 1986; Peleg et al., 1989).

Figure 3.1 shows the metabolic pathways in R. oryzae. In general, glucose is first converted to pyruvate, which is then converted to lactic acid, fumaric acid, ethanol, and cell biomass (through TCA cycle), depending on the growth conditions. Fumaric acid and lactic acid are the main metabolites produced respectively under aerobic conditions by

FMA (fumaric-malic acids) and LA (lactic acid) producing strains (Saito et al., 2004;

Abe et al., 2007), whereas ethanol is produced when oxygen is limited (Magnuson and

Lasure, 2004).

The goal of this study was to apply metabolic engineering strategies to control and maximize the metabolic flux towards fumaric acid production (Curran and Alper, 2012).

We focused on the overexpression of the endogenous pyc gene, which is involved in the conversion of pyruvate to OAA, the first reaction step in the fumaric acid biosynthesis pathway, and an exogenous pepc gene encoding PEPC, which naturally is not present in

R. oryzae but can catalyze the production of OAA from PEP with CO2 fixation, in R. oryzae. Our hypothesis was that overexpressing PYC or PEPC could increase CO2 fixation and the carbon flux toward OAA, and thus potentially could increase fumaric acid production from glucose. R. oryzae M16 was chosen as the host organism in this study. It is a uracil auxotroph with a single-nucleotide mutation on the region of orotate phosphoribosyl transferase gene, pyrF, which can be complemented with the plasmid

60 pPyrF2.1A containing the native pyrF gene (Skory and Ibrahim, 2007). Overexpressing pyc and pepc genes using pPyrF2.1A in R. oryzae M16 and their effects on cell growth and fumaric acid production in shake-flask fermentation were studied and the results are reported in this chapter.

3.2 Materials and Methods

3.2.1 Strains and culture media

R. oryzae M16, a uracil auxotrophic transformant of R. oryzae 99880, was used as the parental strain of various transformant strains (see Table 3.1) developed in this work.

The uracil auxotroph has a mutation in pyrF gene encoding OMP pyrophosphorylase

(Skory and Ibrahim, 2007). The DNA isolated from R. oryzae NRRL 6400 was used as template to obtain the pyc gene, while E. coli DH5α (Invitrogen, Carlsbad, CA) was used to obtain the pepc gene and in the preparation of all recombinant plasmids. Unless otherwise noted, R. oryzae was cultured at 35 oC in a medium containing 50 g/L glucose and 2.5 g/L yeast extract. The RZ minimal medium containing 100 g/L glucose, 2 g/L

(NH4)2SO4, 0.5 g/L KH2PO4, 0.25 g/L MgSO4, 2.2 mg/L ZnSO4·7H2O, 0.5 mg/L

MnCl2·4H2O, and 0.5 % Trypticase peptone was used in the selection of R. oryzae transformants (Skory, 2000a). E. coli was grown at 37 ºC in LB with ampicillin (50

µg/mL) or on solid LB plates (15 g/L agar, 100 µg/mL ampicillin). All strains were in the same class of fumarate-producing R. oryzae. Table 3.1 lists the strains and plasmids with their relevant characteristics used in this study. Figure 3.2 shows detailed maps of the expression plasmids used in this work.

3.2.2 Construction of plasmids expressing native pyc gene

The DNA of R. oryzae NRRL 6400 was extracted by using Qiagen DNeasy plant mini kit (Qiagen, Valencia, CA) and used as template to amplify pyc genes. One 5.2 kb pyc fragment, which includes 3.7 kb native pyc coding region with 4 introns (GenBank:

HM130700) was amplified using PCR primers listed in Table 3.2. The PCR reaction mixture (50 µl) was prepared from 5 µl 10× High Fidelity PCR buffer, 1 µl 10 mM dNTP mixture, 2 µl 50 mM MgSO4, 1 µl primer mix (10 µM each), 1 µl template DNA and 0.2

µl Platinum Taq high fidelity. The PCR was performed for 30 thermal cycles under the following conditions: initial denaturation at 94 oC for 30 seconds; annealing at oC for 30 seconds; extension at 68 oC for 2.5–5.5 min (see Table 3.2). The PCR amplification products containing pyc gene were purified using QIAquick gel extraction kit (Qiagen, Valencia, CA) and cloned into the pGEMT vector (Promega, Madison, WI).

The resulting plasmids were digested with NotI and ligated with NotI linearized plasmid pPyrF2.1A, which contained the pyrF gene as the selection marker. The resulting expression vectors containing the pyc gene, designated as pPyrF2.1A-pyc, were confirmed by sequencing the two ends of the pyc fragment in the constructed plasmids.

3.2.3 Construction of a plasmid expressing a heterologous pepc gene

For construction of the pepc expression plasmid, the 2.6 kb pepc gene was PCR- amplified from E. coli DH5α genomic DNA using the primers shown in Table 3.2. The

PCR fragment was cloned into the pGEMT vector first. The pPgk1-Ex vector (Mertens et al., 2006) was modified by replacing its pyrG gene with the pyrF gene from plasmid

62 pPyrF2.1A, digested with XmaI and XhoI. Then, the pepc gene was cloned into the modified pPgk1-Ex PyrF vector, using the XbaI and NheI restriction sites. The resulting expression vector was designated as pPgk1Ex-pepc.

3.2.4 Transformation

The expression plasmids for pyc and pepc genes were transformed into R. oryzae

M16 spores using microprojectile particle bombardment (PDS-1000/He system, BioRad

Laboratories, Hercules, CA) following the procedures described by Skory (Skory, 2002).

Ungerminated spores were transformed directly on RZ medium plates. Approximately

5‒7 days after bombardment, spores were collected from the plates, diluted in sterile water, and replated to obtain single-spore isolates, which were tested for their genetic stability, enzyme activities, and fermentation kinetics.

3.2.5 Southern hybridization analysis

Spores of R. oryzae 99880 and transformant isolates were inoculated at 1 × 105 spores/ml in growth media containing 50 g/L glucose and 2.5 g/L yeast extract. After 16 h cultivation at 30 oC, 200 rpm, the mycelia were collected, washed with distilled water, and filtered. The genomic DNA was extracted using OmniPrep kit (G-Biosciences, St.

Louis, MO) and digested with HpaI. Southern analyses were performed using DIG high prime DNA labeling and detection starter kit II (Roche, Mannheim, Germany). DIG labeled Lambda DNA, cleaved with HindIII was used as the molecular weight marker on the gel. The pyrF probe used for hybridization was an internal 582-bp PCR fragment

63 obtained with the primers: P1 5’-TGCACTTGCCAATGATGTCTTA-3’ and P2 5’-

CAAAGCCAATTCAGCCTCAAATG-3’. The detection of hybridization blots was performed with Kodak Biomax XAR film (Kodak, Cedex, France) according to the manufacturer’s recommendations.

3.2.6 Evaluation of genetic stability of transformants

The genetic stability of the transformant strains was tested by culturing them on nonselective Potato Dextrose Agar (PDA, Difco, BD, Franklin Lakes, NJ) plates with consecutive passages for 10 days. Briefly, the spores from the frozen glycerol stock were inoculated onto the plates and incubated at 35 oC. The mycelia grown on the plates were consecutively transferred to new PDA plates at 24 h intervals for a total of 10 passages.

The spores, which formed after culturing for additional 23 days, from the first, fifth and tenth passages were collected from the respective plates, washed with sterile water, and inoculated into RZ medium in Erlenmeyer shake-flasks to test their ability to grow in the selective medium. About 1×106 spores were inoculated into each flask containing 50 mL

RZ medium (with 30 g/L calcium carbonate for pH control) and incubated for 4 days at

30 oC with agitation at 150 rpm. Cell growth and glucose consumption were monitored by measuring cell dry weight and glucose concentration at 24 h intervals. Each spore sample was tested in duplicate flasks.

3.2.7 Enzyme activity assays

R. oryzae wild type and transformants were cultivated in medium containing 50 g/L glucose and 2.5 g/L yeast extract. After 24 h, mycelia were collected, washed and suspended in the buffer used in enzyme activity assays (see below). Cells were disrupted by treating for 2 min with Mini-beadbeater-16 (Biospec, Bartlesville, OK). The cell lysate was centrifuged at 13,000 rpm, 4 °C for 10 min, and the supernatant was collected for enzyme assays.

PYC activity was measured in a solution containing 0.05 M NaHCO3, 0.005 M

MgCl2, 0.075 mM acetyl CoA, 0.01 M pyruvate, 0.0025 M ATP, 0.2 mM 5,5'-dithio- bis(2-nitrobenzoic acid) and 200 U/ml citrate synthase (Payne and Morris, 1969). The reaction was initiated by adding a cell extract and the increase in absorbance of 5-thio-2- nitrobenzoate at 412 nm was measured. The extinction coefficient of reduced DNTB at

412 nm is 13.6 mM-1cm-1. One unit of PYC activity corresponds to the formation of 1

µmol of oxaloacetate per minute at 30 °C and pH 8.0.

PEPC activity was determined in a reaction mixture containing 0.1 M Tris-HCl

buffer at pH 8.0, 0.01 M MgCl2, 2.5 mM phosphoenolpyruvic acid, 0.2 mM NADH, 0.01

M NaHCO3 and 5 units of malate dehydrogenase (Payne and Morris, 1969). The reaction was initiated by adding aliquots of protein extracts and measuring the decrease in absorbance of NADH at 340 nm. The extinction coefficient of NADH is 6.22 mM-1cm-1.

One PEPC activity unit was defined as the amount of enzyme that oxidizes 1 µmol of

NADH per minute at 25°C and pH 8.0.

3.2.8 Preparation of seed culture for fermentation

Unless otherwise noted, seed cultures for fermentation kinetics studies were prepared in 250-ml Erlenmeyer shake-flasks. The spores of various strains were collected from the agar plates and 4.5 ×106 spores were inoculated into each flask containing 50 ml of a medium containing 10 g/L glucose and 10 ml soybean meal hydrolysate, which was obtained by treating soybean meal with 2.5% (v/w) HCl at 121 oC for 30 min to hydrolyze its proteins and polysaccharides. The initial pH of the medium was adjusted to

3.0 with the addition of NaOH. The seed cultures were collected after 24 h incubation at

37 °C with 200 rpm agitation in an orbital incubator-shaker (Lab-Line 3527, Cole-Parmer,

Vernon Hills, IL).

3.2.9 Fermentation kinetics studies

Unless otherwise noted, each flask containing 30 ml of the fermentation medium (85 g/L glucose, 0.6 g/L KH2PO4, 0.5 g/L MgSO4, 0.018 g/L ZnSO4, 0.0005 g/L FeSO4) was inoculated with 10 ml of a seed culture and 1.5 g of CaCO3 to maintain the fermentation pH at ~5.0. Some batch fermentations also included 0.5 g/L yeast extract in the medium to evaluate its effects on the fermentation. All fermentations were carried out at 30 °C with 200 rpm agitation for 96 h, with periodically sampling for analysis of glucose and fermentation products. Duplicate or triplicate flasks were used for each strain studied.

3.2.10 Analytical methods

Cell dry weight of the fungal mycelia or pellets in the seed culture was measured after dissolving any residual CaCO3 with 0.7N HCl, filtering through filter paper, washing, and drying at 100 oC for overnight. The number of mycelial clumps or pellets was also counted under a microscope (Olympus IX70 Fluorescence Microscope, Center

Valley, PA).

A high performance liquid chromatography (HPLC) system was used to analyze the organic compounds, including glucose, malic acid, fumaric acid, lactic acid and ethanol in the fermentation broth. The fermentation samples (1 ml each) were centrifuged at

13,000×g for 10 min in a microcentrifuge. The supernatant was collected for HPLC analysis using an organic acid analysis column (Bio-Rad HPX-87H) at 45°C. The HPLC system (Shimadzu Scientific Instruments, Columbia, MD) consisted of an automatic injector (SIL-10Ai), a pump (LC-10Ai), a column oven (CTO-10A), and a refractive index detector (RID-10A). The eluent was 0.01 N H2SO4 at a flow rate of 0.6 ml/min.

3.2.11 Statistical analysis

Unless otherwise noted, all experiments were replicated, and the average values with standard errors are reported. Student’s t-test analysis of the data was performed using

JMP with p = 0.05 as the threshold for significant difference.

3.3 Results

3.3.1 Cloning of pyc and pepc and transformation

The 5.2 kb pyc fragment was amplified from R. oryzae genomic DNA and cloned into the expression vector pPyrF2.1A. This fragment contained the structural gene as well as 500 bp upstream and 1 kb downstream regions of ORF. Thus, the expression of these genes in the transformants was controlled by their respective endogenous promoters. The

E. coli DH5α pepc gene (2.6 kb) was cloned into the expression vector pPgk1Ex, which contained endogenous pgk1 promoter and pdcA terminator. This pgk1 promoter has been used to constitutively express exogenous genes in Rhizopus (Mertens et al., 2006). These plasmids were transformed into R. oryzae M16 spores on minimal RZ medium agar plates using microprojectile bombardment. Only transformants carrying plasmids containing the pyrF gene could grow on the minimal RZ medium (Skory, 2004b). After culturing for 5-7 days, two isolates of the transformants with pyc gene (designated as pyc3 and pyc5), and two isolates of the pepc transformants (designated as ppc1 and ppc3) were obtained from the resulting colonies on the agar plates.

3.3.2 Southern hybridization analysis

HpaI was used to digest fungal genome DNA but not the transformed plasmids. As can be seen in the Southern blots (Figure 3.3), HpaI digested PyrF fragment in the genome DNA was located at 5.5 kb, which was the only band found for the wild type

(Lane 6). However, all transformant isolates also had a second band at 23.1 kb in addition to the 5.5 kb band. The results confirmed that all transformants contained a second PyrF gene in plasmids replicating in a high-molecular-weight structure and none of the transformations resulted in chromosomal integration. As reported by Skory (Skory, 2002),

DNA transformed as circular plasmids was maintained extrachromosomally in a high- molecular-weight (>23 kb) concatenated form. Integration event could happen only when a linearized plasmid was transformed.

3.3.3 Genetic stability

Spores from the first, tenth passages on the non-selective PDA plates were tested for their ability to grow in the selective RZ medium. Spores without the selective marker gene pyrF on the recombinant plasmid would not be able to survive in minimal RZ medium because of uracil deficiency. Figure 3.4 shows the growth kinetics for the ppc1 transformant. In general, no significant difference in growth rate and biomass production was observed among the spores obtained from different passage numbers, suggesting that the ppc1 transformant was genetically stable, even after culturing for over 10 days on

PDA plates without the selective pressure. Similar results were also found with the pyc5 transformant.

3.3.4 Enzyme activities

Figure 3.5 shows the specific enzyme activities for pyruvate carboxylase (PYC) and phosphoenolpyruvate carboxylase (PEPC) in R. oryzae wild type (99880) and transformants. Compared to the wild type, the specific PYC activity in the pyc transformants increased 56%83% (0.028 U/mg in pyc5 and 0.033 U/mg in pyc3, vs.

0.018 U/mg in wild type; p < 0.05). The increased PYC activities can be attributed to the higher copy number of pyc gene in the corresponding transformant strains. Similarly,

69 pepc transformants exhibited significant PEPC activity (0.0036 U/mg) that was absent in the wild type, as pepc gene naturally only exists in bacteria and plants. This was the first study demonstrating that E. coli pepc gene can be expressed with significant activity in R. oryzae.

3.3.5 Seed culture growth kinetics

24 h were examined and compared for their cell morphology (Figure 3.6), cell dry weight, and mycelial clump or pellet number and size (Table 3.3). In general, the wild type and ppc1 transformant reached a similar dry weight after 24 h cultivation, suggesting they all grew at a similar growth rate in the seed culture medium. In contrast, the transformant pyc5 had a significantly lower cell dry weight at 24 h. In addition, different strains exhibited distinctive morphologies. The wild type grew into loose mycelial clumps of less than 100150 m and ppc1 grew into small, spherical pellets of ~150 m in diameter, whereas the pyc5 formed large, compact pellets of 250500 m (Figure 3.5). The large difference in the measured particle number among these different strains can be attributed to variations in cell morphology and particle size, which might have also contributed to the different growth rates (cell dry weights). Oxygen limitation in the large cell pellets formed by the pyc5 transformant could inhibit its growth. It is clear that overexpressing

70 pyc and pepc affected cell growth and morphology, although the exact mechanisms involved are not known and need further investigation.

3.3.6 Fermentation kinetics

To evaluate the effects of gene overexpression on fumaric acid biosynthesis in R. oryzae, batch fermentations with wild type and various transformant strains were studied in shake-flasks. The glucose consumption rates and product yields from these fermentations are summarized and compared in Table 3.4. Figure 3.7 shows typical fermentation kinetics of wild type and selected transformant strains cultured in the glucose medium with 0.5 g/L yeast extract. As expected, the wild-type strain produced fumaric acid as the main product with malic acid as a byproduct and only a little of ethanol (Figure 3.7A). Compared to the wild type, pyc5 transformant produced more malic acid and ethanol, but only a little of fumaric acid (Figure 3.7B). Similar results were found with the pyc3 transformant strains (see Table 3.4). In general, the pyc transformants grew poorly and had much lower glucose consumption and little fumaric acid production (<0.05 g/g glucose), presumably due to the large cell pellets that limited oxygen supply in the fermentation, resulting in the accumulation of ethanol with high yields (0.13 g/g to 0.36 g/g glucose). This is because the biosynthesis of fumaric acid in R. oryzae requires high oxygen, whereas oxygen limitation shifts the fermentation to the anaerobic ethanol production pathway (Skory, 2003b; Magnuson and Lasure, 2004).

The expression of heterologous PEPC in R. oryzae significantly increased fumaric acid production (Figure 3.7C). Compared to the wild type strain 99880, fumaric acid

71 yield from the pepc transformants increased 26% (0.78 g/g glucose vs. 0.62 g/g) for fermentation with yeast extract and 11%20% (0.710.77 g/g vs. 0.64 g/g) for fermentation without yeast extract (see Table 3.4). PEPC catalyzes the biosynthesis of

OAA from PEP with the fixation of one CO2 (see Figure 3.1). The increased fumaric acid production thus can be attributed to increased availability of OAA, which is the substrate of the fumaric acid biosynthesis pathway and increasing its availability should drive more carbon flux towards the formation of fumaric acid as the final product.

It should be noted that the wild type strain 99880, instead of its auxotrophic transformant strain M16, was used as the control in this study because both the seed culture and fermentation media without yeast extract were deficient in uracil. However, the addition of yeast extract in the fermentation medium did not significantly change the product yields, although the additional nutrients present in the yeast extract increased the glucose consumption rate of all strains tested except for the pyc transformant. Figure 3.8 summarizes and compares the glucose consumption rates and product yields from various strains in fermentations with yeast extract. Clearly, expressing pepc was beneficial to fumaric acid production, whereas overexpressing pyc had negative effect on fumaric acid biosynthesis in R. oryzae. The reasons for these observed effects are discussed in the following section.

3.4 Discussion

Fumaric acid is an intermediate in the TCA cycle found in almost all organisms.

However, in filamentous fungi, the accumulation of fumaric acid involves a C3 plus C1

72 pathway (Wright et al., 1996), and fumaric acid is accumulated in R. oryzae mainly via a cytosolic pathway that converts pyruvate to fumaric acid by the combined activities of pyruvate carboxylase, malate dehydrogenase, and fumarase under aerobic conditions

(Kenealy et al., 1986). Pyruvate carboxylase is a biotin-dependent tetrameric enzyme that catalyzes the ATP-dependent condensation of pyruvate and CO2 to form oxaloacetic acid

(Kenealy et al., 1986). In general, pyruvate carboxylase in eukaryotic cells is localized in mitochondria. However, in R. oryzae, this enzyme is situated exclusively in the cytosol

(Osmani and Scrutton, 1985). The cytosolic localization appears to be important for the ability of these organisms to accumulate high concentrations of organic acids. The gene encoding pyruvate carboxylase should be strictly regulated since it is situated at the branch point of pyruvate metabolism in the cytosol (Goldberg et al., 2006).

This study aimed at increasing fumaric acid biosynthesis in R. oryzae. We demonstrated the feasibility of engineering R. oryzae for increased fumaric acid production from glucose by increasing the carbon flux toward OAA. To our best knowledge, there has been no published research related to improved fumaric acid production through metabolic engineering. This is mainly because of limited genetic engineering tools available for R. oryzae, which is resistant to common antifungal agents

(e.g., hygromycin B, G418, glufosinate ammonium) that are used as the cloning selection markers (Suarez and Eslava, 1988; D'Halluin et al., 1992; Sugui et al., 2005).

Furthermore, DNA introduced into Rhizopus is typically maintained through autonomous extra-chromosomal replication (Skory, 2005), which hampers gene modifications (e.g., gene targeting and gene replacement) of this organism. Nevertheless, using the uracil

73 auxotroph M16, we were able to select transformants with recombinant plasmids expressing both the endogenous pyc and exogenous pepc genes.

Pyruvate is the metabolite at the branch point of the metabolic pathways leading to various end products, including fumaric acid, lactic acid, and ethanol (see Figure 3.1).

The conversion of pyruvate to OAA catalyzed by PYC is the first step in the pathway leading to fumaric acid biosynthesis. Therefore, we first aimed at overexpressing the endogenous pyc gene to increase fumaric acid production in R. oryzae. However, our results showed that the overexpression of PYC led to poor cell growth and increased malic acid production with a dramatic reduction in fumaric acid yield, which is contrary to our original expectation. The poor cell growth with large pellets of pyc transformants was probably due to the shortage of ATP caused by increased PYC catalyzed reaction

(Blankschien et al., 2010). Although there would be no net ATP consumption from PEP to pyruvate and then from pyruvate to OAA, overexpressing PYC might change the flux distribution for pyruvate between cytosol and mitochondria (Osmani and Scrutton, 1985;

Goldberg et al., 2006). The latter is important for the TCA cycle and additional ATP generation needed in supporting cell growth. Because the pyc gene is on the plasmid, its overexpression due to multiple copies of the plasmids could cause imbalanced pyruvate flux distribution and thus impair cell growth, as observed in this study. The imbalanced pyruvate flux distribution could have also contributed to the shift from fumarate production to the accumulation of malic acid in the pyc transformants, although the exact mechanism is not clear. Consequently, growth was reduced and fumaric acid production

74 was inhibited in the pyc transformants, with dramatically increased ethanol production due to oxygen limitation caused by the large cell pellets.

The overexpression of PEPC has been shown to increase the carbon flux towards

OAA, resulting in more succinate and less acetate production in E. coli (Gokarn et al.,

2001). Pyruvate kinase defect in Corynebacterium glutamicum led to a significant increase in PEPC activity and therefore an enhanced overall flux from PEP to oxaloacetic acid (Sawada et al., 2010). PEPC catalyzes the direct conversion of PEP to OAA, thus reducing the carbon flux towards pyruvate and other pathways. Unlike PYC, no ATP is required in the reaction catalyzed by PEPC and overexpressing the heterologous pepc gene would not affect the pyruvate flux distribution between cytosol and mitochondria.

Therefore, expressing pepc in R. oryzae did not impose any stress on cell growth and was able to increase fumaric acid biosynthesis because more carbon flux was diverted toward

OAA. In our study, PEPC expression using the plasmid pPgk1Ex-pepc led to ~20% increase in fumaric acid production by R. oryzae.

As can be seen in Figure 3.7, malic acid was a major byproduct in the fumaric acid fermentation, suggesting fumaric acid biosynthesis is also limited by fumarase, which catalyzes the reversible hydration of fumaric acid to L-malic acid. It has been reported that fumR encodes a single transcript, but exists in both cytosol and mitochondria (Peleg et al., 1989). The genome sequence of R. oryzae 99880 also clearly shows that there is only one fumarase gene, which is likely co-localized between the cytoplasm and mitochondria (Yogev et al., 2011; Yogev and Pines, 2011). The level of fumR RNA increased in cells producing fumaric acid under stress conditions (Friedberg et al., 1995).

However, the fumR RNA is abundant during nonproduction phase, indicating that additional mechanisms and/or limiting steps are involved in the enhanced biosynthesis of fumaric acid. Recently, the cDNA of fumR from R. oryzae ATCC 20344 was cloned in E. coli and the expressed fumarase protein was purified and analyzed for its structure and enzyme properties by Song et al. (2011). Their purified fumarase (GenBank accession number GU013473) exhibited the characteristics of cytosolic fumarase, and had a deletion of 15-amino acid sequence from the N-terminal region of the fumarase previously reported (GenBank accession number X78576), suggesting that the modified fumarase might be responsible for the accumulation of fumaric acid in R. oryzae. Thus, it would be interesting to overexpress this fumR gene and investigate its effect on fumaric acid biosynthesis in R. oryzae. However, overexpressing the truncated, cytosolic form of

R. oryzae fumarase in Aspergillus niger led to the conversion of fumarate to malate and resulted in increased citrate production (De Jongh and Nielsen, 2008).

Lactate and ethanol are two other fermentation products derived from pyruvate.

Conversion of pyruvate to lactic acid is accomplished by a NAD-dependent L-lactate dehydrogenase (Obayashi et al., 1966; Skory, 2000a; Skory, 2004a) that is absent in this fumaric acid producing strain (Abe et al., 2007), so no lactic acid was produced by the strains used in this study. Under anaerobic stress, ethanol biosynthesis becomes the main pathway, which is catalyzed through pyruvate decarboxylase and alcohol dehydrogenase

(Skory, 2003a). In fumaric acid fermentation, even though small pellets were produced in the seed culture and then used in the subsequent fermentation to avoid oxygen limitation, there were still significant amounts of ethanol produced by both the wild-type and pepc

76 transformant strains. The disruption of the pyruvate decarboxylase genes thus could be an effective way to block the competitive ethanol formation since more pyruvate and NADH can be preserved for fumaric acid biosynthesis. However, gene disruption in Rhizopus remains a major challenge as there is no effective genetic engineering technique to do it.

3.5 Conclusions

In conclusion, this is the first study using metabolic engineering to modify the fumaric acid biosynthesis pathway in order to increase fumaric acid production in R. oryzae. We successfully demonstrated that the transformation of circular plasmids containing pyc gene can lead to increased PYC activities in R. oryzae, and that the exogenous gene pepc can also be expressed with the endogenous pgk1 promoter in R. oryzae, resulting in increased fumaric acid production from glucose in batch fermentation.

Even though the gene modification in R. oryzae is still limited by immature genetic engineering techniques for Rhizopus, successful endogenous and exogenous gene overexpressions illustrated in this work can pave the way for further engineering this filamentous fungus for industrial production of fumaric acid from renewable carbohydrates.

3.6 References

Abe, A., Oda, Y., Asano, K., Sone, T., 2007. Rhizopus delemar is the proper name for Rhizopus oryzae fumaric-malic acid producers. Mycologia 99, 714-722.

Blankschien, M.D., Clomburg, J.M., Gonzalez, R., 2010. Metabolic engineering of Escherichia coli for the production of succinate from glycerol. Metab Eng 12, 409-419.

Curran, K.A., Alper, H.S., 2012. Expanding the chemical palate of cells by combining systems biology and metabolic engineering. Metab Eng 14, 289-297.

D'Halluin, K., Bonne, E., Bossut, M., De Beuckeleer, M., Leemans, J., 1992. Transgenic maize plants by tissue electroporation. Plant Cell 4, 1495-1505.

De Jongh, W.A., Nielsen, J., 2008. Enhanced citrate production through gene insertion in Aspergillus niger. Metab Eng 10, 87-96.

Gokarn, R.R., Evans, J.D., Walker, J.R., Martin, S.A., Eiteman, M.A., Altman, E., 2001. The physiological effects and metabolic alterations caused by the expression of Rhizobium etli pyruvate carboxylase in Escherichia coli. Appl Microbiol Biotechnol 56, 188-195.

Goldberg, I., Rokem, J.S., Pines, O., 2006. Organic acids: old metabolites, new themes. J Chem Technol & Biotechnol 81, 1601-1611.

Kenealy, W., Zaady, E., du Preez, J.C., Stieglitz, B., Goldberg, I., 1986. Biochemical aspects of fumaric acid accumulation by Rhizopus arrhizus. Appl Environ Microbiol 52, 128-133.

Lohbeck, K., Haferkorn, H., Fuhrmann, W., Fedtke, N., 1990. Maleic and fumaric Acids. Ullmann’s Encyclopedia of Industrial Chemistry, VCH, Weinheim, Germany.

Mertens, J.A., Skory, C.D., Ibrahim, A.S., 2006. Plasmids for expression of heterologous proteins in Rhizopus oryzae. Arch Microbiol 186, 41-50.

Osmani, S.A., Scrutton, M.C., 1985. The sub-cellular localization and regulatory properties of pyruvate carboxylase from Rhizopus arrhizus. Eur J Biochem 147, 119-128.

Payne, J., Morris, J.G., 1969. Pyruvate carboxylase in Rhodopseudomonas spheroides. J Gen Microbiol 59, 97-101.

Roa Engel, C.A., Straathof, A.J.J., Zijlmans, T.W., Gulik, W.M., Wielen, L.A.M., 2008. Fumaric acid production by fermentation. Appl Microbiol Biotechnol 78, 379-389.

Saito, K., Saito, A., Ohnishi, M., Oda, Y., 2004. Genetic diversity in Rhizopus oryzae strains as revealed by the sequence of lactate dehydrogenase genes. Arch Microbiol 182, 30-36.

Sawada, K., Zen-in, S., Wada, M., Yokota, A., 2010. Metabolic changes in a pyruvate kinase gene deletion mutant of Corynebacterium glutamicum ATCC 13032. Metab Eng 12, 401-407.