NOVEL INSIGHTS INTO SUGAR AND SUCCINATE METABOLISM OF
ACTINOBACILLUS SUCCINOGENES FROM STRAINS EVOLVED FOR IMPROVED
GROWTH ON LIGNOCELLULOSE HYDROLYSATE SUGARS
By
Nikolas Robin McPherson

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
Genetics–Doctor of Philosophy
2017

PUBLIC ABSTRACT
NOVEL INSIGHTS INTO SUGAR AND SUCCINATE METABOLISM OF
ACTINOBACILLUS SUCCINOGENES FROM STRAINS EVOLVED FOR IMPROVED
GROWTH ON LIGNOCELLULOSE HYDROLYSATE SUGARS
By
Nikolas Robin McPherson
A wide variety of important chemicals are currently produced from petroleum, using very
well-refined processes and a large industrial infrastructure. However, petroleum processing has a
number of hazardous and otherwise negative impacts on the environment, as well as on human
health. Additionally, the supply and price of oil into the future are uncertain. Thus, renewable
substrates can help extend the current supply of oil and eventually replace it. Succinate is a
specialty chemical currently produced from maleic anhydride from petroleum processing that
can also be produced from renewable substrates by microorganisms. Sustainable production of
succinate by microbial catalysis is an attractive alternative to petrochemical succinate because it
could use the existing industrial infrastructure and be scaled up to meet the demands of this $15
billion commodity chemical market. Lignocellulose, the polymeric material that makes the stiff
cell walls of plants, is a major potential feedstock source for conversion to succinate. Primarily
made of sugars, lignocellulose from agricultural waste or from bioenergy crops could become a
viable option for succinate production using organisms able to grow on its sugar components. Of
particular interest is the bacterium Actinobacillus succinogenes, which is among the best natural
succinate producers in terms of production rate and yield. It also grows on a wide variety of
carbohydrates, including the major sugars that make up lignocellulose.
A. succinogenes grows well on glucose, the most common sugar in lignocellulose, but
does not grow as quickly with the other lignocellulosic sugars. I have evolved strains of A.

succinogenes for faster growth on xylose, the second most common sugar in lignocellulose and
second most abundant sugar on the planet, as well as arabinose and galactose. Further, I evolved
strains for optimal growth and succinate production from lignocellulose hydrolysate, the sugar
and byproduct solution that results from treating lignocellulose to release its sugars. Many of the
evolved strains of A. succinogenes produce more succinate than the parental strain, even though
the evolution process only selected for growth. Genomic analyses in the new strains identified
mutations that shed light on the physiological traits that could have resulted in faster growth.
RNA sequencing of the xylose-evolved strains identified changes in transcription that provided
additional insights into xylose fermentation and growth in these strains. For example, many
genes encoding the enzymes responsible for converting sugars to succinate in the xylose-evolved
strains were upregulated. By contrast, genes that encode enzymes that redistribute carbon
molecules toward byproducts were downregulated. I also evolved a strain that can grow on
galactose, a sugar that the parental strain cannot use as sole carbon and energy source. The final
evolved strain also grew faster on xylose, arabinose, and lignocellulose hydrolysates. I also
engineered the parental A. succinogenes strain with a mutation from a xylose-evolved strain.
This strain produced 40% more succinate than the parental A. succinogenes strain, although its
growth rate was less than half the growth rate of the parental strain.
In summary, I have generated several strains of A. succinogenes for improved growth and
succinate production from lignocellulosic substrates. I have also identified and characterized
mutations responsible for some of the phenotypes and have provided novel insights into the
factors that control fast growth and succinate production. This knowledge sets the foundation for
further genetic improvements in A. succinogenes and other bacteria needed to efficiently convert
renewable substrates into succinate.

ABSTRACT
NOVEL INSIGHTS INTO SUGAR AND SUCCINATE METABOLISM OF
ACTINOBACILLUS SUCCINOGENES FROM STRAINS EVOLVED FOR IMPROVED
GROWTH ON LIGNOCELLULOSE HYDROLYSATE SUGARS
By
Nikolas Robin McPherson
A wide variety of industrially vital chemicals are currently produced from petroleum,
using very well-refined processes and a large industrial infrastructure. However, petroleum
processing has a number of hazardous and otherwise negative impacts on the environment, as
well as on human health. The supply and price of oil into the future are uncertain as well, and
supplementing oil with feedstocks from renewable sources can help extend the current supply of
oil and eventually replace it. Succinate is a specialty chemical currently produced from maleic
anhydride from petroleum processing. If bio-based succinate could compete with the cost of
maleic anhydride, it could replace maleic anhydride in a $15 billion commodity chemical
market, taking advantage of the existing chemical production infrastructure. A major potential
feedstock source for conversion to succinate is lignocellulose from agricultural waste or from
bioenergy crops. The bacterium Actinobacillus succinogenes is one of the best natural succinate
producers and it grows on a wide variety of carbohydrates, including the major sugars in
lignocellulose. A. succinogenes grows well on glucose, the most common sugar in lignocellulose,
but does not grow as quickly on other lignocellulosic sugars.
I have evolved strains of A. succinogenes to grow faster on xylose, the second most
common lignocellulosic sugar, as well as on arabinose, galactose, and lignocellulose
hydrolysates. Many of the evolved strains produce more succinate than the parental strain as
well, even though the evolution process did not specifically select for succinate production. The

evolved strains were resequenced to identify the mutations accumulated during evolution. RNA
sequencing of the xylose-evolved strains helped identify changes in transcript levels and was
used to refine our conclusions about the xylose-evolved (X) strains. I discovered that the genes
that encode many glycolytic enzymes were upregulated in at least one X strain, several genes
encoding succinate production enzymes were upregulated, while genes that encode enzymes that
redirect fluxes from the succinate pathway to other fermentation products were downregulated.
During the directed evolution process, I obtained a strain of A. succinogenes that can grow on
galactose, a sugar that the base strain cannot use. The final evolved strain grew faster than the
wild-type strain on xylose, arabinose, and lignocellulose hydrolysate, and could grow on
galactose. I determined that A. succinogenes will co-consume glucose and xylose, but that xylose
represses arabinose consumption. After directed evolution, though, arabinose represses xylose
consumption. Finally, I used multiplex transformation to introduce mutations from the evolved
strains into the wild-type strain. The first strain produced, using the xylose symporter mutation
from a xylose-evolved strain, produced 40% more succinate than wild-type A. succinogenes,
even though it grew at less than half the speed.
In summary, I have evolved a set of A. succinogenes strains that grow faster on
lignocellulose sugars and some have a higher succinate yield, I know the location and nature of
their mutations and have RNA sequencing data for the xylose-evolved strains. I have conducted
numerous additional experiments to characterize sugar consumption in A. succinogenes and what
causes the evolved strains to be able to grow faster and produce more succinate. My results lay
solid groundwork for future work with A. succinogenes and other bacteria being grown on sugars
and synthesizing succinate.

ACKNOWLEDGEMENTS

The research presented in this dissertation is the result of years of trial and error and has
been helped along by a great many people at Michigan State University. Members of the Vieille
lab, including Maris Laivenieks, Dr. Justin Beauchamp, Dr. Bryan Schindler, and Rajasi Joshi,
have been an immense help, both personally and professionally. Dr. Bryan Schindler trained me
to work with Actinobacillus succinogenes and started the evolution on xylose in addition to being
a great graduate student mentor for my first several years. Rajasi Joshi is the other member of the
Vieille lab to work with A. succinogenes and the brainstorming and experimental help that she
provided were absolutely invaluable to the work presented here. Additionally, her work with
natural transformation in A. succinogenes gave me the means to engineer the multiplex natural
transformation strains and finally directly test small mutations. I was fortunate to work with and
train a number of gifted undergraduates as well, include Jake Gibson, Maeva Bottex, and Vlad
Jdanov. All of them were adept at performing experiments, preparing media, and other tasks in
the laboratory.
I am grateful to the previous director of the Genetics program, Dr. Barb Sears, for helping
me get a successful start at graduate school, and the current director, Dr. Cathy Ernst, for her
help securing funding at the end of my graduate career. Jeannine Lee, the previous Graduate
Secretary of the Genetics program, also deserves thanks for her monumental assistance along
almost the entirety of my PhD.
I must also thank my committee members, Drs. Gemma Reguera, Jonathan Walton, and
Yair Shachar-Hill, for their expertise and assistance. Their input was vital at several key points
during my career and helped guide this project’s trajectory.

iv

My friends at Michigan State University were constants who helped me weather the years
of experiments, both successes and failures, and deserve sincere thanks. I thank my family,
including my parents, my brother, and sister-in-law, for their support and positive outlooks
through the years. I also thank Matt Chansler for his tireless devotion and for being a sounding
board for ideas.
I thank Drs. Terri McElhinny and Jeanette McGuire for the opportunity to develop as a
teacher while I was their teaching assistant. Dr. Terri McElhinny especially has taken me under
her wing and taught me much more than I otherwise would have learned about building a course,
curriculum, teaching, problem-solving, and interacting with students.
Finally, and most importantly, I must thank my mentor, Dr. Claire Vieille. Her unending
support, guidance, and expertise were a driving force behind the research presented here. I am
and will always be impressed by her ability, drive, and devotion to research science. I am
grateful to have had the opportunity to work in her laboratory, and have developed tremendously
as a researcher because of her. I know I would not have had the success I did without her
contributions and willingness to give input, and I cannot thank her enough.

v

TABLE OF CONTENTS

LIST OF TABLES ......................................................................................................................... ix
LIST OF FIGURES ........................................................................................................................ x
Chapter 1: Prospects for Actinobacillus succinogenes as a biocatalyst for conversion of
lignocellulosic sugars to succinate .................................................................................................. 1
1.1 Succinate as an industrial chemical ........................................................................ 2
1.2 Succinate-producing microbes of interest ............................................................... 5
1.2.1 Anaerobiospirillum succiniciproducens ........................................................ 6
1.2.2 Rumen bacteria .............................................................................................. 6
1.2.2.1 Actinobacillus succinogenes ............................................................. 7
1.2.2.2 Mannheimia succiniciproducens ..................................................... 12
1.2.3 Escherichia coli ........................................................................................... 13
1.2.4 Corynebacterium glutamicum ...................................................................... 16
1.2.5 Saccharomyces cerevisiae ........................................................................... 17
1.3 Succinate exporters ............................................................................................... 18
1.4 Lignocellulosic feedstocks .................................................................................... 20
1.4.1 Lignocellulose background .......................................................................... 21
1.4.2 Pretreatment methods and byproducts ......................................................... 22
1.4.3 Plant sources of lignocellulosic material ..................................................... 22
1.5 Succinate production from lignocellulose hydrolysates ....................................... 23
1.6 Introduction to project........................................................................................... 24
APPENDIX ................................................................................................................. 26
REFERENCES ........................................................................................................... 32
Chapter 2: Regulation of D-xylose utilization in Actinobacillus succinogenes and evolution for
fast growth .................................................................................................................................... 42
2.1 Abstract ................................................................................................................. 43
2.2. Introduction .......................................................................................................... 44
2.3 Materials and Methods .......................................................................................... 45
2.3.1 Bacterial strains and media .......................................................................... 45
2.3.2. A. succinogenes evolution for fast growth on xylose ................................. 46
2.3.3. Growth curves and fermentation balance analyses ..................................... 47
2.3.4. Plasmids and DNA manipulations .............................................................. 47
2.3.5. Multiplex natural transformation ................................................................ 48
2.3.6. Strain resequencing ..................................................................................... 49
2.3.7. Xylose isomerase assays ............................................................................. 49
2.3.8. Internal xylose concentration measurement ................................................ 50
2.3.9. RNA purification and RT-qPCR................................................................. 51
2.3.10. RNA sequencing ....................................................................................... 53
2.3.11. Determination of fusidic acid minimum inhibitory concentration (MIC) 54
2.4. Results .................................................................................................................. 54
2.4.1. Growth rates and fermentation balances of the xylose-evolved strains ...... 54

vi

2.4.2. Mutations in the xylose-evolved strains ..................................................... 55
2.4.3. Transcription of the xylose utilization gene cluster .................................... 56
2.4.4. Transcript levels and succinate production ................................................. 58
2.4.5. Xylose isomerase activity, internal xylose concentration, and xylose uptake
rate......................................................................................................................... 61
2.4.6. Mutations in fusA and translation elongation rates in the evolved strains .. 62
2.4.7. Reintroduction of individual mutations into wild-type 130Z ..................... 63
2.5. Discussion ............................................................................................................ 64
2.6. Acknowledgements .............................................................................................. 67
APPENDIX ................................................................................................................. 69
REFERENCES ........................................................................................................... 87
Chapter 3: An Actinobacillus succinogenes strain evolved for fast growth on lignocellulosic
sugars is able to grow on galactose ............................................................................................... 91
3.1 Abstract ................................................................................................................. 92
3.2 Introduction ........................................................................................................... 93
3.3. Materials and Methods ......................................................................................... 94
3.3.1. Strains and media ........................................................................................ 94
3.3.2. RNA purification and sequencing ............................................................... 95
3.3.3. Growth curves and fermentation balance analyses ..................................... 95
3.3.4. Strain evolution ........................................................................................... 95
3.3.5. Sequencing of the evolved strains............................................................... 96
3.3.6. Transmembrane helix prediction ................................................................ 97
3.4 Results ................................................................................................................... 97
3.4.1 Glucose, mannose, fructose, and xylose assimilation pathways .................. 97
3.4.2 Sugar co-utilization ...................................................................................... 99
3.4.3 Growth rates of evolved strains ................................................................. 100
3.4.4 Fermentation balances of the evolved strains ............................................ 101
3.4.5 Mutations in the evolved strains ................................................................ 101
3.4.6 Galactose assimilation ............................................................................... 102
3.5 Discussion ........................................................................................................... 104
APPENDIX ............................................................................................................... 109
REFERENCES ......................................................................................................... 125
Chapter 4: Development of a markerless knockout method for Actinobacillus ......................... 128
succinogenes ............................................................................................................................... 128
Chapter 5: Conclusions and Future Directions ........................................................................... 130
5.1 Introduction ......................................................................................................... 131
5.2 Evolution of A. succinogenes for growth on lignocellulose hydrolysate sugars 131
5.2.1 Xylose ........................................................................................................ 131
5.2.2 Galactose .................................................................................................... 132
5.2.3 Other sugars and AFEX corn stover hydrolysate ...................................... 133
5.2.4 Discussion of the evolution experiment protocol ...................................... 134
5.3 Future work ......................................................................................................... 135
5.3.1 X strain testing and combination with other engineering .......................... 135

vii

5.3.2 Further work with A, G, C, and H strains .................................................. 135
5.3.3. Continuing development of genetic tools and other methods for A.
succinogenes ....................................................................................................... 136

viii

LIST OF TABLES

Table 1.1. Succinate-producing bacterial and yeast strains .......................................................... 30
Table 2.1. Plasmids used in this study .......................................................................................... 70
Table 2.2. Primers for PCR and RT-qPCR ................................................................................... 71
Table 2.3. Fermentation balances of wild-type, evolved, and mutant A. succinogenes strains. ... 72
Table 2.4. Mutations identified in strains X1, X2, and X3. .......................................................... 73
Table 2.5. Fold changes of xylR, xylF, xylA, and xylE transcripts relative to wild type strain 130Z
in glucose- and xylose-grown cultures of strains X1, X2, and X3 calculated from RT-qPCR data
....................................................................................................................................................... 75
Table 2.6. Fold changes in transcripts of the xylose assimilation genes in the glucose- and
xylose-grown X strains compared to the parental strain, 130Z, calculated from RNAseq data. .. 76
Table S2.1. Fold changes in transcripts levels of central metabolism genes. ............................... 82
Table S2.2. Xylose isomerase activity in the wild-type and the X strains.................................... 84
Table S2.3. Cell volumes, cell counts per unit OD660, and internal xylose concentrations for the
xylose-grown wild-type and X strains .......................................................................................... 84
Table 3.1. Primers used in this study .......................................................................................... 110
Table 3.2. Differential expression of sugar transport genes in wild-type A. succinogenes grown
on glucose, fructose, mannose, and xylose. ................................................................................ 111
Table 3.3. Doubling times of the wild-type (WT) and evolved A. succinogenes strains on
lignocellulose sugars. .................................................................................................................. 112
Table 3.4. Fermentation balances of glucose-grown wild-type (WT) and evolved strains. ....... 113
Table 3.5. Mutations identified in the evolved strains. ............................................................... 114

ix

LIST OF FIGURES

Figure 1.1. Industrial products of succinate and maleic anhydride .............................................. 27
Figure 1.2. General succinate-producing mixed acid metabolism. ............................................... 28
Figure 1.3. Central metabolism of A. succinogenes 130Z. ........................................................... 29
Figure 2.1. Doubling times of evolving cultures during the evolution transfer series that led to
strain X1. ....................................................................................................................................... 77
Figure 2.2. Mutations in the A. succinogenes xylose utilization gene cluster. ............................. 78
Figure 2.3. Central metabolism of A. succinogenes 130Z ............................................................ 79
Figure 2.4. Principal component analysis of RNA sequencing results. ........................................ 80
Figure 2.5. Structure of E. coli K12 EF-G .................................................................................... 81
Figure S2.1. Quality of the purified RNAs. .................................................................................. 85
Figure S2.2: Determination of A. succinogenes xylose isomerase’s Km for xylose. .................... 86
Figure 3.1. Evolution scheme. .................................................................................................... 116
Figure 3.2. Doubling times of evolving cultures during the transfer series that generated strain
A1................................................................................................................................................ 117
Figure 3.3. Co-utilization of xylose and other sugars by A. succinogenes 130Z. ....................... 118
Figure 3.4. Growth and sugar consumption of strains 130Z and H1 on AFEX-pretreated corn
stover hydrolysate. ...................................................................................................................... 119
Figure 3.5. Proposed map of galactose use in A. succinogenes. ................................................. 120
Figure 3.6. A. succinogenes lactose operon and location of the mutations found in the series 1
evolved strains. ........................................................................................................................... 121
Figure 3.7. Effect of mutation 1548009 in strains C1 and H1 on the 5’ end of lacY. ................ 122

x

Chapter 1

Prospects for Actinobacillus succinogenes as a biocatalyst for conversion of lignocellulosic
sugars to succinate

1

Biobased industrial chemical production is becoming a necessity. Increasing fossil fuel
scarcity and a need to combat global warming have led to worldwide implementation of
emissions caps and increased popular demand for greener industry. In 2014, as part of the United
Nations Framework Convention on Climate Change, the United States announced plans to
reduce CO2 emissions by 26-28% compared to 2005 levels by 2020 (Secretary 2015). China
announced that it would reduce emissions by 40-45% of 2005 levels by the same 2020 goal (Wei
2015). The European Union has pledged to reduce overall greenhouse gas emissions by 40%
compared to 1990 by 2030 and plans to reduce emissions by 80-95% by 2050 (Intended
nationally determined contribution of the EU and its Member States 2015).
Finding alternate feedstocks and production methods for chemicals currently produced
from petroleum is important to help countries develop energy and chemical independence and
reduce reliance on oil, foreign and domestic. Biobased chemical production uses renewable
feedstocks as well, creating alternate and more environmentally friendly uses for biological
waste. In 2007, it was optimistically predicted that a switch from non-renewable to renewable
feedstocks for the production of bulk chemicals could eliminate non-renewable energy
consumption and the production of greenhouse gases and be used as a method to consume
existing greenhouse gases (Hermann et al. 2007). This number assumes the use of lignocellulosic
biomass and is completely theoretical, but a predicted overall consumption of greenhouse gases
shows the potential impact of implementing these technologies.

1.1 Succinate as an industrial chemical
In 2004, the United States Department of Energy established a list of top value-added
chemicals from biomass (Pacific Northwest Pacific Laboratory 2004). At the top of this list were

2

the 1,4-diacids, including succinic acid (or succinate). In 2006, a consortium of European
universities, research organizations, and industry members funded by the European Commission
Growth Programme (the BREW consortium) evaluated the opportunities and risks of a selection
of products of industrial biotechnology and confirmed succinate as a bulk chemical of
importance (Patel et al. 2006). Succinate is currently produced mostly petrochemically via
maleic anhydride, although the succinate produced for human consumption is typically made
fermentatively by bacteria. Succinate has four current primary markets. First, it is used as a
surfactant, foaming agent, and detergent extender. Second, succinate is used as an ion chelator to
prevent corrosion and metal pitting in electroplating. Third, it is used in the food market, where it
is used as a pH modulator, an anti-microbial agent, and for flavor. The fourth and smallest
market is in the production of pharmaceuticals (McKinlay et al. 2007b; Litsanov et al. 2014).
Research into the production of succinate has accelerated in the past ten years, and
several companies have set up production facilities across the world. BioAmber has set up a
3,000 metric ton/year facility in France; Reverdia has built a 10,000 metric ton per year facility
in Italy; and a 10,000 metric ton/year plant in Spain is a joint venture between BASF and Purac.
BioAmber also has a 50,000 metric ton/year facility as a joint venture with Mitsui in Ontario,
Canada, and Myriant has a 77,000 metric ton/year plant in Louisiana, USA. More
BioAmber/Mitsui plants have been set up in Thailand and Brazil, producing 65,000 metric
tons/year of succinate and 50,000 metric tons/year of 1-4 butanediol. Finally, PTT MCC
Biochem has a production facility in Thailand, producing 36,000 metric tons/year of succinate
(Cok et al. 2013).
Large-scale succinate production is not destined for use in just these existing four
markets, however. Succinate has strong potential as a commodity chemical, but the production

3

cost is still too high. In 2003 the theoretical cost of bio-based succinic acid was estimated as
$0.55-$1.10 per kg (Paster et al. 2003). In 2015, the cost of petrochemically produced succinic
acid was estimated at $2.84 per kg, while the maleic anhydride it is produced from was estimated
to sell for $1.33 per kg (Pinazo et al. 2015). However, the actual production cost of bio-based
succinate was still reported as the same 2003 figures. The actual costs are difficult to determine
and rely largely on proprietary information, but they could be approaching levels where biobased
succinate is competitive with petrochemical succinate, especially when the price of oil is high
(Pinazo et al. 2015; Patel et al. 2006).
In addition to its direct uses, succinate can be transformed into a variety of other
products. It can be esterified to dimethyl succinate or dehydrogenated and cyclized to maleic
anhydride, both of which are precursors to a wide range of known petrochemical products.
Maleic anhydride, specifically, is used to produce a wide variety of compounds (Figure 1.1).
1,4-butanediol is used as an industrial solvent as well as in the production of resins and in
2012 had a market of about 2 million tons per year (Business 2012). In the same year about 1
million tons of tetrahydrofuran was produced. Tetrahydrofuran has a wide range of uses,
including in the production of inks, adhesives, and magnetic tapes. γ-Butyrolactone is used in
textiles and as paint remover. In 2007, the worldwide production of γ-butyrolactone was
estimated at about 250,000 tons (McKinlay et al. 2007b). Adipic acid is used to produce foams,
lubricants, and food products. In addition, many of those chemicals are precursors in the
production of even more compounds (Zeikus et al. 1999). In combination with 1,4-butanediol,
succinate can be used to produce polybutylene succinate. Polybutylene succinate and its
derivatives have similar properties to polypropylene and polyvinylchloride, meaning existing
processing equipment can be used. Finally, even without conversion to other known platform

4

chemicals, succinate has increasing uses as a component in biobased nylons, polyesters, or other
polymers (Song and Lee 2006). In 2010, the market for succinate and its derivatives was
projected to be 245,000 metric tons per year, with a potential market for succinate-derived
polymers of 250 million metric tons per year (Bozell and Petersen 2010).
Succinate is produced as part of a mixed acid fermentation, often with acetate, lactate,
ethanol, and formate. The production of succinate from glucose without byproduct formation is
theoretically possible, but it has yet to be demonstrated. The theoretical yield from glucose
(Equation 1) is 1.71 mol/mol glucose, based on the number of electrons present in glucose. The
fermenting microbe also consumes the greenhouse gas carbon dioxide during production. CO2 is
a potential byproduct in biofuel production, as well as in many other industrial processes, and in
theory succinate production can be used to consume surplus CO2. Additionally, it will be
important to decrease byproduct formation as much as possible, both to increase succinate
production and to reduce the cost and complexity of downstream purification.
1 Glucose + 0.86 CO2  1.71 succinate2- + 1.74 H2O + 2.58 H+

(1)

1.2 Succinate-producing microbes of interest
Succinate is a common microbial fermentation product along with acetate, ethanol,
lactate, and formate (Figure 1.2). The general metabolic map in Figure 1.2 shows several paths to
succinate, as well as how the common byproducts are produced. In natural producers, succinate
is typically produced in the reductive branch of the tricarboxylic acid (TCA) cycle, but the
pathway requires reductant, as oxaloacetate (OAA) is reduced to malate by malate
dehydrogenase and fumarate is reduced to succinate by fumarate reductase. A homo-succinate
fermentation would simplify the downstream processing and lower the overall cost of

5

production, so it is often the targeted process. However, byproduct formation is the result of cells
meeting redox and/or energy requirements, so byproduct elimination is difficult. To these ends,
bacterial species with naturally high succinate production levels are under study and several
microbial workhorses of biotechnology have been engineered for succinate production.

1.2.1 Anaerobiospirillum succiniciproducens
A. succiniciproducens, isolated from a beagle dog (Davis et al. 1976), produces high
yields of succinate (1.38 mol/mol glucose) (Lee et al. 2003). Experiments testing the effect of pH
and CO2 availability correlated lower pH and higher CO2 availability with increased succinate
production and increased phosphoenolpyruvate (PEP) carboxykinase (PEPCK) activity
(Samuelov et al. 1991). That study was one of the first demonstrating that PEPCK can work in
an anaplerotic role, rather than a gluconeogenic one. However, A. succiniciproducens required a
growth medium supplemented with yeast extract, peptone, or corn steep liquor for complete
sugar consumption, even when grown on oak wood hydrolysate (Lee et al. 2003), increasing the
cost of potential industrial applications. Additionally, it will not co-consume lactose and glucose
(Lee et al. 2000), possibly indicating a wider sugar co-utilization problem that would have to be
solved by strain modification. Finally, A. succiniciproducens is a strict anaerobe and can cause
bacteremia in humans (Secchi et al. 2005), complicating industrial use even further.

1.2.2 Rumen bacteria
Several high succinate producing bacterial species have been isolated from cattle rumen.
Succinate produced in the rumen is quickly converted to propionate by other rumen
microorganisms (Samuelov et al. 1999) and that propionate is then absorbed as nutrient by the

6

cow. A large portion of this propionate is transported to the liver where it is used for
gluconeogenesis to maintain the cow’s blood glucose levels (Yost et al. 1977).
The succinate-producing rumen bacteria Actinobacillus succinogenes, Mannheimia
succiniciproducens, and Basfia succiniciproducens have been studied over the past 20 years for
potential use in industrial succinate production.

1.2.2.1 Actinobacillus succinogenes
A. succinogenes was first described in 1996 (Guettler et al. 1996a). It is a member of the
Pasteurellaceae family and is known for its naturally high levels of succinate production. It
grows on a variety of carbon sources, including sugars commonly found in lignocellulosic
biomass: glucose, xylose, and arabinose (Guettler et al. 1999; Lee et al. 2002; van der Werf et al.
1997). Its succinate yield is natively high and it can be further enhanced by increasing the
amount of available CO2 or adding H2 (van der Werf et al. 1997; McKinlay and Vieille 2008). In
addition to succinate, A. succinogenes produces formate, acetate, and ethanol as byproducts. It is
auxotrophic for cysteine, methionine, nicotinic acid, pantothenate, pyridoxine, and thiamine. A.
succinogenes is additionally auxotrophic for glutamate, due to an incomplete TCA cycle. The
absence of the genes encoding citrate synthase and isocitrate dehydrogenase makes A.
succinogenes unable to synthesize α-ketoglutarate, the precursor for glutamate (McKinlay et al.
2005). Wild-type A. succinogenes, strain 130Z, produces up to 80 g L-1 of succinate when grown
in optimized conditions (Guettler et al. 1996b). Several mutant strains isolated after chemical
mutagenesis with fluoroacetate no longer produce acetate in complex medium and produce 110 g
L-1 of succinate under optimized conditions (Guettler et al. 1996a).

7

A. succinogenes’s metabolism was characterized first by in vitro enzyme assays and
fermentation balances in different growth conditions in complex and defined media (McKinlay et
al. 2005; van der Werf et al. 1997). 13C metabolic flux analysis of glucose-grown cultures
(McKinlay et al. 2007a; McKinlay and Vieille 2008) and the completion of an annotated genome
(McKinlay et al. 2010) have led to a more thorough understanding of A. succinogenes.
Glucose is assimilated mainly through the Embden-Meyerhoff pathway (Figure 1.3). The
Entner-Doudoroff pathway does not seem to be present (van der Werf et al. 1997; McKinlay et
al. 2007a), and flux through the oxidative pentose phosphate pathway accounts for no more than
5% of the assimilated glucose. After glycolysis, flux from PEP branches toward OAA or
pyruvate.
In the pyruvate branch, PEP is converted to pyruvate by pyruvate kinase and pyruvate is
converted to acetyl-CoA and formate by pyruvate-formate lyase. Pyruvate dehydrogenase is also
active during anaerobic growth, although typically at low levels, converting pyruvate to acetylCoA and CO2 (McKinlay et al. 2007a). Acetyl-CoA is converted to either acetate or ethanol,
providing ATP (acetate) or fulfilling redox demands (ethanol) (van der Werf et al. 1997).
In the OAA branch, PEP is converted to OAA by PEPCK, also producing ATP.
However, the activity of PEPCK does not significantly change with changing bicarbonate or
hydrogen concentrations, even though the succinate yield does (McKinlay and Vieille 2008).
OAA is reduced to malate by malate dehydrogenase, malate is dehydrogenated to fumarate by
fumarase, and fumarate is finally reduced to succinate by fumarate reductase. No flux through a
potential glyoxylate shunt was detected, so that potential link between the two fermentative
branches is absent. The branches are not completely separate, though, as high fluxes through the
C4-decarboxylating enzymes malic enzyme and OAA decarboxylase were detected by 13C

8

metabolic flux analysis. (McKinlay et al. 2007a; van der Werf et al. 1997). This finding is
supported by the high malic enzyme and OAA decarboxylase activities measured in in vitro
enzyme assays (McKinlay et al. 2007a; van der Werf et al. 1997).
While much is known about the central metabolism of A. succinogenes, the uptake
mechanisms for most sugars are not known. Metabolic flux analysis confirmed that a
phosphotransferase system (PTS) is not used for glucose uptake (McKinlay et al. 2007a). It was
hypothesized then that the galactose permease (galP) is used for glucose transport in A.
succinogenes. The succinate exporters are still unknown as well, though several annotated genes
have moderate-to-high sequence similarity with known succinate transporters and putative
dicarboxylate transporters in other bacterial species (McKinlay et al. 2010)
Many well-known Pasteurellaceae are pathogenic, such as Haemophilus influenzae and
Pasteurella multocida, so it was important to determine that A. succinogenes was not.
Fortunately, genome sequencing enabled a search for known Pasteurellaceae virulence factors
(McKinlay et al. 2010). No known toxins were found, and no hemin or heme utilization proteins
were found either, meaning that iron uptake cannot involve the host heme. The genome does
encode possible homologs for type IV pili proteins, OapA and B (proteins implicated in cell
attachment during infection), as well as possible adhesins, although two involved genes have
frameshifts. A. succinogenes has been shown to be naturally competent (Joshi et al. 2014), so the
type IV pilus could be involved in that mechanism. It is currently unknown whether A.
succinogenes produces an adhesin, and if it does, what role it plays in the life cycle of the
bacterium. Most other typical Pasteurellaceae virulence factors were absent or incomplete. The
conclusions drawn from the genome analysis suggest that A. succinogenes is nonpathogenic,

9

although several potential pathogenic traits are present because of their roles in other, nondisease-causing, pathways (McKinlay et al. 2010).
As mentioned, natural competency genes were discovered in A. succinogenes. To
engineer the genome of A. succinogenes for either fundamental research or biotech applications a
method of transformation is required. The first indicator of potential natural competency was the
presence of 1,690 uptake signal sequences (USS) through the genome; only two of the fifteen
other Pasteurellaceae species had more. In addition to USSs, homologs for 23 of a regulon of 25
natural competency genes from H. influenzae were discovered in A. succinogenes. The two genes
absent in the A. succinogenes genome encode hypothetical proteins of unknown function. The
presence of these 23 genes, plus the master regulator for the competence regulon (cAMP
receptor protein) and the additional essential regulatory factor Sxy, suggested that natural
competence could be induced in A. succinogenes (McKinlay et al. 2010).
The ability to induce natural competence was confirmed and a natural transformationbased knockout method has been developed to aid strain engineering (Joshi et al. 2014). Linear
DNA fragments including USSs and an isocitrate dehydrogenase-based selection cassette flanked
by genomic DNA were used to knock out fumarate reductase (frd) and pyruvate formate lyase
(pflB). In addition, double mutants were made in the ΔpflB strain with citrate lyase, aconitase,
and β-galactosidase gene deletions. E. coli icd (encoding isocitrate dehydrogenase) was used as
the selection marker because it would complement A. succinogenes’ glutamate auxotrophy in the
presence of isocitrate. The icd gene was flanked with FRT sites to allow its removal by the yeast
flippase recombinase expressed on a plasmid. Markerless knockouts were confirmed by PCR,
enzyme assay (ΔpflB), blue/white color screening (ΔpflBΔlacZ), and by fermentation product

10

analysis (ΔpflB). Formate production was completely absent in the ΔpflB strain and pyruvate
accumulated above base levels, giving a clear confirmation of the knockout.
Wild-type A. succinogenes 130Z has been grown in continuous cultures for succinate
production as well. Growth in a bioreactor with glucose over 950 or 1,300 hours led to the
formation of biofilms that were stable for up to 7 days, but could not be purposefully maintained.
When glucose consumption was fastest, the succinate yield reached 1.39 mol/mol glucose with a
titer of 411 mM (Bradfield and Nicol 2014). Modification of the bioreactor agitator allowed for
biofilm retention, so growth on xylose-rich corn stover hydrolysate was tested. After 1,550 hours
the maximal succinate yield was 1.02 mol/mol xylose-rich hydrolysate, the titer was 335 mM,
and the volumetric productivity was 1.77 g/L/h (Bradfield et al. 2015).
Overall, several metabolic characteristics make A. succinogenes a potentially valuable
microbe for succinate production. A. succinogenes natively grows on a wide variety of
carbohydrates. Its use of a non-PTS transporter for glucose uptake means that half of PEP is not
pushed into the pyruvate fermentative branch, and that catabolite repression is less problematic.
Since PEPCK is used to convert PEP to OAA, ATP availability is increased. Succinate yields are
natively very high, and access to the sequenced genome and genetic tools should enable
engineering to increase this yield even further.
In addition to the potential use of A. succinogenes for industrial succinate production, we
can learn much about bacterial sugar consumption and succinate production. The glucose
transporter is currently unknown, even with a manually annotated genome and many other
sequenced bacteria to compare to. Many less common sugars are poorly utilized in most
industrial strains, as well. Succinate transporters are also not well defined in many bacteria, A.

11

succinogenes included. Studying the genome and metabolism of A. succinogenes can help fill in
many of these gaps.

1.2.2.2 Mannheimia succiniciproducens
M. succiniciproducens has been the subject of active research since its discovery. It is
also a member of the Pasteurellaceae and is closely related to A. succinogenes, sharing 78% of
open reading frames (McKinlay et al. 2010). It can also grow on a variety of carbohydrates,
including glucose, xylose, and fructose (Lee et al. 2002). It produces large amounts of succinate
and this production can be increased further with increased CO2 availability or H2
supplementation (Kim et al. 2009). M. succiniciproducens also produces ethanol, formate,
acetate, and lactate as byproducts during fermentation and its auxotrophies are similar to those of
A. succinogenes. While A. succinogenes converts PEP to OAA using only PEPCK, M.
succiniciproducens uses both PEPCK and PEP carboxylase. M. succiniciproducens’s genome
sequence (Hong et al. 2004) was used to create an in silico metabolic map (Kim et al. 2007),
predict ideal culture conditions (Kim et al. 2009), design microarray chips for transcriptomics,
and create a two-dimensional gel map for proteomics (Lee et al. 2008). These resources have led
to engineering of M. succiniciproducens to reduce byproduct formation (deletions of lactate
dehydrogenase, ldhA; pyruvate-formate lyase, pflB; and phosphotransacetylase/acetate kinase,
pta-ackA) increasing the succinate yield from 0.69 mol/mol glucose in the wild-type strain to
0.97 mol/mol glucose in mutant strain LPK7 (Lee et al. 2006). A ΔldhA, Δpta-ackA strain was
later grown on a combination of glucose and glycerol in a pH controlled medium and reached a
succinate yield of 1.15 mol/mol glucose equivalent, a titer of 768 mM, and a volumetric
productivity of 3.5 g/L/h (Choi et al. 2016). By adding ΔfruA to that strain and growing it on a

12

combination of sucrose and glycerol a yield, titer and productivity of 1.64 mol/mol glucose
equivalent, 664 mM, and 6.0 g/L/h were reached. Using a cell-recycling bioreactor system
pushed the productivity up to 38.6 g/L/h (Lee et al. 2016). Additionally, the final strain produced
succinate without byproducts.

1.2.3 Escherichia coli
E. coli is a perennial favorite in any endeavor that requires a significant amount of
engineering. Wild-type E. coli produces only a small amount of succinate, about 0.11 mol/mol
glucose (Sawers and Clark 2004). However, the genetic tools available for E. coli make it a
viable template for engineering.
Initial efforts to engineer E. coli emphasized increasing flux to the reductive branch of
the TCA cycle. This was done by overexpressing native PEPC, increasing succinate yields 3.75
times to 90.6 mM (Millard et al. 1996). Attempts were made to overexpress the native PEPCK as
well, but no effect was initially seen. However, when the more soluble sodium bicarbonate was
used instead of magnesium carbonate, overexpressing PEPCK did increase succinate
concentration (Kwon et al. 2006). Strain NZN111, with mutations in ldhA and pflB to block
lactate and formate production, was unable to grow fementatively on glucose, but a mutant
derivative, AFP111, with an additional mutation in ptsG, recovered that ability (Chatterjee et al.
2001). When grown aerobically and transitioned to an anaerobic production phase, AFP111
produced 840 mM succinic acid (Vemuri et al. 2002), one of the highest titers reported.
Dr. Lonnie Ingram’s group has built on previous knowledge and engineered E. coli
strains that produce large amounts of succinate from both glucose and xylose (Jantama et al.
2008a). They started by knocking out genes responsible for production of other major products,

13

lactate (ldhA), acetate (ackA), and ethanol (adhE). This strain, KJ012, grew poorly anaerobically
in glucose minimal medium and still produced mostly acetate. KJ012 was submitted to evolution
to improve growth rate and succinate yield. After 40 transfers, isolated, evolved strain KJ017
grew much faster than strain KJ012 and produced more succinate than other products, although
acetate, formate, and lactate were still major byproducts. Evolution had selected for a mutant
strain with increased energy efficiency, increased flux through glycolysis, and increased
succinate production for redox balance. To further reduce acetate and formate, pflB and focA
were knocked out. The resulting strain had an acetate auxotrophy, which was eliminated by
another round of selective evolution, giving rise to strain KJ060. Additional engineering rounds
culminated in strain KJ134 (Jantama et al. 2008b), which produces 1.61 mol succinate per mol
glucose, together with small amounts of byproducts, mostly acetate. This succinate yield is over
90% of the maximum theoretical yield for production from glucose and strain KJ134 was seen as
a success story.
Four important mutations were found in the KJ134 genome (Zhang et al. 2009). A
mutation in the pckA promoter made PEPCK expression constitutive, allowing PEPCK to replace
PEP carboxylase as the dominant PEP-carboxylating enzyme. This change in carboxylating
enzyme is important, because PEPC releases inorganic phosphate, whereas PEPCK produces
ATP instead. Second, a mutation in ptsI inactivated the glucose phosphotransferase system
(PTS), no longer forcing PEP into the pyruvate pathway. Finally, mutations in the D-galactose
transporter (galP) and glucokinase (glk) likely compensated for the loss of the PTS. This set of
four mutations resulted in a larger PEP pool for conversion to succinate. Interestingly, these
mutations result in metabolic pathways reminiscent of those found in the rumen bacteria
described above.

14

While strain KJ134 fermented glucose well and produced succinate at a high yield and
titer, it performed poorly on xylose. The Ingram group went back to strain KJ122 to attempt to
improve it (Sawisit et al. 2015). After sixteen serial transfers on xylose, strain AS1600a was
isolated. This strain grew much better on xylose than KJ122 and maintained a high succinate
production yield and titer. Strain AS1600a was sequenced, discovering an additional mutation in
galP. Expressing this mutation from a plasmid in a ΔgalP KJ122 strain resulted in a strain
(KJgalP*) that consumed xylose well but glucose less quickly than KJ122, similar to strain
AS1600a. Strains KJ122, AS1600a, and KJgalP* were also grown on sugar bagasse dilute acid
hydrolysate. This hydrolysate is a solution containing glucose, xylose, arabinose, and galactose
in addition to other products of the acid hydrolysis, including some growth inhibitors. In this
case, extra glucose was added to bring sugar proportions in line with a typical lignocellulose
hydrolysate (discussed more below). Strain KJ112 fermented the glucose completely and left
about 20% of the xylose unfermented after 144 h. Strains AS1600a and KJgalP fermented the
xylose nearly completely and left about 20% of the glucose unfermented after 144 h. All of the
strains fermented the small amounts of galactose and arabinose completely.
These E. coli strains all have high succinate production rates, titers, and yields, and
between strains KJ112 and AS1600a, as well as other strains created by other groups using the
information discovered, they could ferment most feedstocks to completion in a reasonable
amount of time. Indeed, at least two E. coli strains have been commercially licensed for
succinate production by Roquette (Roquette) and BioAmber (Nghiem et al. 2001; Bio-Amber)
for use in their production facilities.

15

1.2.4 Corynebacterium glutamicum
C. glutamicum is primarily known as an industrial amino-acid producer, has extensive
industrial applications, and is generally recognized as safe (GRAS). These characteristics and
available genetic tools make it a possible biocatalyst for succinate production as well. In C.
glutamicum, glucose is taken up by a combination of a glucose PTS and IolT, an inositol
permease (Tsuge et al. 2015). C. glutamicum is typically grown aerobically and transitioned to
an anaerobic production phase. Succinate production in C. glutamicum takes place primarily
through the reductive branch of the TCA cycle, although the oxidative branch is active as well
(Rados et al. 2014). Carbon enters the reductive TCA primarily through PEP carboxylase, but
also through pyruvate carboxylase (Tsuge et al. 2015). A ΔldhA strain that overexpresses the
pyruvate carboxylase (pyc) gene (strain ΔldhA-pCRA717) can produce 1.4 mol succinate per mol
glucose up to a final titer of 146 g/L after 46 hours, with a byproduct of 0.29 mol acetate per mol
glucose as well (Okino et al. 2008). Litsanov et al. constructed strain BOL-1 (Table 1.1) by
knocking out ldhA in addition to all known acetate production genes (acetyl-CoA:CoA
transferase, cat; phosphotransacetylase, pta; acetate kinase, ackA; and pyruvate:menaquinone
oxidoreductase, pqo) (Litsanov et al. 2012). BOL-1, when overexpressing pyc on a plasmid,
produced 1.16 mol succinate per mol glucose, but almost no acetate. It co-produced pyruvate and
α-ketoglutarate, however. Integrating pyc as well as the Mycobacterium vaccae formate
dehydrogenase gene (fdh) into the genome generated strain BOL-3, which could use formate as a
cosubstrate to increase succinate production. When overexpressing the glyceraldehyde 3phosphate dehydrogenase (gap) gene from a plasmid to overcome NADH-based feedback
inhibition in BOL-3, this strain produced 1.41 mol succinate per mol glucose, although it
consumed 89 mM formate along with 107 mM glucose during the process.

16

Jojima et al. have also showed that cells can be recycled from one anaerobic production
phase to the next, bypassing the aerobic growth phase and reducing the glucose lost to cell
growth. The authors recycled cells for 26 reaction cycles, losing only 0.7% productivity per
cycle. In recycling the cells, they reduced the overall glucose consumed by cell growth to only
5.1% of the total glucose fed to the cells, increasing the yield from 0.58 mol succinate/mol
glucose to 1.04 mol/mol (Jojima et al. 2016).

1.2.5 Saccharomyces cerevisiae
S. cerevisiae represents another established and robust microbe with GRAS status that is
currently used for industrial applications. It has a sequenced genome and variety of tools for
engineering (Goffeau 2000). It can grow on a wide range of carbon sources, defined media, and
in acidic conditions. This last characteristic is particularly interesting, since producing succinic
acid rather than succinate would lead to decreased costs in purification (Agren et al. 2013).
However, S. cerevisiae does not produce large amounts of succinate natively.
The first clue that S. cerevisiae could be engineered for succinate production came from
sake brewing, since succinate is one of the main taste components produced during sake
fermentation (Arikawa et al. 1999). It was observed that strains with disrupted succinate
dehydrogenase and fumarase produced 2.7-fold more succinate than the parent strain when
grown aerobically. The first successful engineering of S. cerevisiae involved knocking out two
components of succinate dehydrogenase (sdh1 and sdh2), one subunit of isocitrate
dehydrogenase (idh1), and a mitochondrial isozyme of isocitrate dehydrogenase (idp1) to force
flux through the glyoxylate shunt to succinate (Raab et al. 2010). Compared to the starting strain
AH22ura3, the quadruple mutant had increased succinate yield (0.111 vs 0.023 mol/mol

17

glucose), titer (30.7 vs. 6.4 mM), and production rate (0.02 vs. 0.005 g/L/h) as well as an altered
and decreased byproduct profile (Table 1.1). Another attempt to engineer S. cerevisiae started
with in silico prediction of metabolic engineering targets using a genome-scale metabolic map
(Otero et al. 2013). Three genes were targeted to link succinate production directly to growth.
First, succinate dehydrogenase was deleted to disrupt the TCA cycle after succinate formation.
Second, the genes for two isozymes encoding 3-phosphoglycerate dehydrogenase were deleted to
remove the ability to synthesize serine from glycolysis, forcing S. cerevisiae to synthesize serine
from glycine via the glyoxylate shunt, producing succinate as a necessary byproduct of the
process. Strain 8D was the result. Strain 8D was auxotroph for glycine and had grew slower than
the reference strain, even when supplemented with glycine (0.22/h vs. 0.33/h). An evolution
series was used to remove the glycine auxotrophy and restore the growth rate, resulting in strain
8D Evolved. Since flux was already driven through the glyoxylate shunt in this strain, isocitrate
lyase was then overexpressed from a plasmid to further drive flux through the pathway. This
final strain (8D Evolved + pICL1) had a lower growth rate (0.12/h), but increased succinate titer
(7.6 mM) and yield (0.076 mol/mol glucose). Another attempt at S. cerevisiae engineering based
on in silico predictions created strain Δdic1, knocking out an inner dicarboxylate mitochondrial
transporter. This strain produces 0.02 mol succinate/mol glucose with only a single mutation.

1.3 Succinate exporters
To produce succinate using a microbe on an industrial scale, it is important to know the
genes and proteins responsible for its transport. This enables transfer of enzymes between
organisms or enzyme engineering to increase succinate production.

18

The proteins responsible for bacterial succinate transport, and C4-dicarboxylate transport
in general, have largely been categorized into four families: dicarboxylate transport A (DctA),
tripartite ATP-independent periplasmic (TRAP) transporters, dicarboxylate uptake (Dcu)AB, and
DcuC (Janausch et al. 2002). Some of these families catalyze uptake, some export, and some
exchange, but transporters in all four families have been observed to act on the dicarboxylates
succinate, fumarate, and malate, as well as the dicarboxylic amino acid aspartate. One example
of a succinate exporter outside these families has been discovered in C. glutamicum (Huhn et al.
2011) as well.
DctA family transporters catalyze C4-dicarboxylate uptake during aerobic growth. They
have been identified in a variety of Gram-positive and negative bacteria, but, because they have
not been shown to export succinate or any other C4-dicarboxylates they will not be discussed
further.
TRAP transporters have two integral membrane-bound transport proteins and a
periplasmic solute-binding component protein. They are very structurally and functionally
different from the other C4-carboxylate transport proteins, but have only been shown to play a
role in C4-dicarboxylate uptake thus far.
DcuA and DcuB transporters function primarily as importers or exchangers and are
commonly used in fumarate-respiring anaerobes to exchange external fumarate for internal
succinate (Six et al. 1994; Janausch et al. 2002). Fumarate:succinate exchange activity has been
shown to be 2.6‒3 times higher than influx or efflux for DcuA and B transporters (Zientz et al.
1996), but both import and export of succinate and fumarate have been observed. DcuC has only
been observed in succinate-producing bacteria, often pathogens, and is structurally similar to
DcuA and DcuB, but primarily catalyzes export (Zientz et al. 1996).

19

Most of the work on the DcuAB and DcuC families of transporters has been done in E.
coli. DcuB and C are highly regulated by fumarate and the fumarate regulator in the Dcu gene
group, DcuSR, as well as by oxygen-dependent regulator FNR (Zientz et al. 1999; Golby et al.
1998). DcuA, though, was observed to be constitutively expressed (Golby et al. 1998). With nonfunctional DcuA and DcuB, DcuC was shown to be capable of assuming their role in import or
exchange as well (Zientz et al. 1999). E. coli DcuB has also been shown to be 2.3 times more
active than DcuA with respect to fumarate/succinate exchange (Six et al. 1994). Finally, an E.
coli strain carrying mutations inactivating DcuA, B, and C still exported succinate, albeit at a
much lower rate. The succinate export rate remained above the expected passive diffusion rate,
implying that E. coli has yet an additional succinate export system (Zientz et al. 1996).
C. glutamicum does not have any DcuAB or DcuC family transporters, but the SucE
succinate exporter has been identified. SucE was identified bioinformatically by comparing the
sets of proteins with predicted transmembrane domains from C. glutamicum, A. succinogenes,
and M. succiniciproducens and confirmed by functionally knocking out the gene (Huhn et al.
2011). A C. glutamicum ΔsucE strain exported 50% less succinate than the parental strain,
showing that SucE is involved in succinate export, but that redundancy in succinate export is
present in C. glutamicum as well. Additionally, SucE has been shown to not be sufficient for
succinate uptake or exchange.

1.4 Lignocellulosic feedstocks
Bio-based succinate production has the potential to replace current petroleum-based
succinate production, as well as maleic anhydride production. As of 2013, maleic anhydride
prices were between $5.9 and 9.0 per kg (Akhtar et al. 2014). The price of biobased succinate

20

depends on many factors, but the fermentation feedstock is an important one. Finding lowpriced, abundant, and widely available feedstocks is one of the main barriers to the production of
large amounts of inexpensive succinate. One promising source of sugar for fermentations is
lignocellulosic biomass.

1.4.1 Lignocellulose background
Terrestrial plant and algal cell walls are made of cellulose, hemicellulose, and lignin in
varying proportions. Cellulose, a β(14)-linked polymer of glucose, and hemicellulose are
polysaccharides. Hemicellulose has a wide variety of compositions across and within plant
species, but is composed primarily of xylose, arabinose, galactose, rhamnose, and mannose.
Lignin is a complex, crosslinked, phenolic polymer often simply burned for heat or energy.
Additionally, it can be modified and used in adhesives (Ghaffar and Fan 2014). Research into the
degradation of lignin is active, though, both to make lignocellulose hydrolysis easier and because
lignin represents a major source of aromatic and phenolic products for use in food additives and
flavoring (Bugg 2011; Laurichesse and Avérous 2014). White-rot fungus pretreatment of
lignocellulosic biomass has been the prevalent area of research because of its potential as a more
energy-efficient and less hazardous pretreatment method (Wan and Li 2012). The bacterium
Rhodococcus jostii RHA1 has been found to degrade lignin and has a sequenced genome. By
knocking out vanillin dehydrogenase in this strain, a strain was built that grew on minimal
medium with 2.5% wheat straw lignocellulose and 0.05% glucose (Sainsbury et al. 2013).

21

1.4.2 Pretreatment methods and byproducts
Because of the crystalline nature of cellulose, the complicated structure of hemicellulose,
and the protective aromatic mesh of lignin, the sugars in lignocellulose are generally
inaccessible, and lignocellulose must be broken down before the sugars can be used as feedstock.
Lignocellulose is mechanically ground down before being treated by one of several pretreatment
methods, including dilute acid, base, ammonia fiber expansion (AFEX), steam explosion, and
ionic liquid (Banerjee et al. 2009). These pretreatment steps decrease the cellulose crystallinity
and partially hydrolyze the polysaccharides, but they produce byproduct chemicals as well,
because of the nonspecific nature of the processes. Byproducts include carboxylic acids and
furan aldehydes like furfural and hydroxymethylfurfural, many of which are inhibitory to the
next step of pretreatment and to microbial growth (Banerjee et al. 2009). The second step uses
enzyme cocktails to break down the polysaccharides to monosaccharides or oligosaccharides that
are usable by fermenting bacteria. This enzyme treatment step is the most expensive pretreatment
step, and research to reduce enzyme loadings is ongoing.

1.4.3 Plant sources of lignocellulosic material
Many plants have come under consideration for use as fermentation feedstocks in recent
years. Existing crop plants, such as corn, rice, sugar cane, and wheat result in large amounts of
agricultural waste after harvest. Globally, between 600 and 900 million tons per year of rice
straw are produced. Most of the straw is burned or disposed of in other ways, and only a small
proportion is used, as animal feed or otherwise (Karimi et al. 2006). In the United States, more
than 90% of corn stover is left in the fields and tilled under. A portion of the stover is useful for
soil enrichment, but some of the waste could be used as fermentation feedstock (Glassner et al.

22

1999). A large proportion of sugar cane bagasse is burned for fuel or cogeneration of electricity
(Banerjee et al. 2009). While many of these agricultural waste products are currently used for
animal feed, soil development, or fuel, using them as feedstocks for bioindustries could prove to
be a more environmentally or economically sound decision.
Many other sources of biomass have been investigated for use as biotechnology
feedstocks as well. Some other major crops that have been included are: rye, oil palm (Sun et al.
2007), sorghum (Olson et al. 2012; Mullet et al. 2014), olive (Sánchez and Cardona 2008), and
cassava (Sawisit et al. 2014). Even urban food and paper waste (Lin et al. 2013; Sánchez and
Cardona 2008) have been tested. Grasses like Miscanthus species and switchgrass, are under
consideration as energy crops as well, because of their rapid growth rates and low maintenance
requirements (Sørensen et al. 2008; Sanderson et al. 2006). Because of its fast growth, relatively
low mineral content, and high biomass yield and polysaccharide content, Miscanthus could
represent an easy-to-grow crop to be used entirely for lignocellulose production (Brosse et al.
2012). Wood is also a possible source of lignocellulose, and poplar is being engineered to reduce
lignin production and improve sugar availability (Van Acker et al. 2013).

1.5 Succinate production from lignocellulose hydrolysates
Many of the microorganisms previously discussed for succinate production have been
grown on lignocellulose hydrolysates as well as pure sugars. The most recent or most successful
results are summarized below. A. succiniciproducens was grown on pretreated oak wood
hydrolysate and converted 88% of the 21 g/L glucose to succinate over approximately 25 hours.
Approximately 7 g/L xylose was also present, but was not consumed (Lee et al. 2003). M.
succiniciproducens was also grown on pretreated oak wood hydrolysate. In batch culture both

23

glucose and xylose were fully consumed in approximately 12 hours, producing succinate with a
yield of 56% and productivity of 1.17 g/L/h. In a continuous culture the yield was comparable,
but productivity increased to 3.19 g/L/h (Kim et al. 2004).
A. succinogenes has been grown on a wide variety of lignocellulosic feedstocks including
corn, wheat, and rice straw (Zheng et al. 2009), sugarcane bagasse (Borges and Pereira 2011),
and Jerusalem artichoke (Gunnarsson et al. 2014). Salvachúa and colleagues (Salvachúa et al.
2016) showed that A. succinogenes grown on dilute acid pretreated corn stover hydrolysate can
produce 0.74 g succinate/g sugars and to a titer of 42.8 g/L with productivity up to 1.27 g/L/h.
Additionally they monitored furfural levels and found that it was detoxified to furfuryl alcohol,
although an initial lag phase was present.
E. coli has been tested on many lignocellulosic feedstocks as well like cassava pulp
(Sawisit et al. 2014) and macroalgae (Bai et al. 2015; Olajuyin et al. 2016). The most successful
recent results with sugarcane bagasse have already been summarized in section 1.2.3.

1.6 Introduction to project
A general pattern seen in biobased production of fuels and chemicals is that the microbes
involved often have a narrow range of ideal sugar feedstocks. Glucose is typically the go-to
sugar, and many microbes grow slower on other sugars. In addition, catabolite repression by
glucose is a common problem that complicates the use of complex feedstocks, such as
lignocellulose hydrolysates. My project has centered on using A. succinogenes to convert
lignocellulosic hydrolysate sugars to succinate. A. succinogenes is an excellent model system
because of its ability to grow on lignocellulosic sugars and its natural ability to produce high
succinate yields. The wild-type A. succinogenes strain was selectively evolved to grow faster on

24

individual lignocellulosic sugars as well as on AFEX-pretreated corn stover hydrolysates, and
the evolved strains were sequenced. AFEX uses liquid ammonia at moderate temperature (60100 °C) and a rapid release of high pressure to solubilize lignin, partially hydrolyze
hemicellulose, decrystallize cellulose, and increase the surface area of the substrate.
Chapter 2 focuses on the characterization of X1, X2, and X3, the A. succinogenes strains
independently evolved for fast growth on xylose. These strains show increased growth rates on
xylose and increased succinate yields with respect to the wild-type strain, 130Z. In addition to
their interesting growth phenotypes, these strains share two sets of mutations in fusA (encoding
elongation factor G, EF-G) as well as upstream of and in xylE (encoding the xylose permease
XylE). These individual mutations were reintroduced to 130Z to test their effects on growth and
succinate production.
Chapter 3 expands the characterization of evolved strains, drawing special attention to the
G1 and H1 strains. Additionally, I investigated sugar transport in A. succinogenes. Strains G1
and H1 are the products of a single, sequential evolution series, in which A. succinogenes was
evolved for fast growth on xylose first, then on arabinose, galactose, cellobiose, and corn stover
hydrolysates. Strain G1 is able to grow solely on galactose, where wild-type 130Z is not. Strain
H1 is the culmination of the sequential evolution series and is able to grow quickly on corn
stover AFEX hydrolysates.
In Chapter 4, I discuss the suitability of A. succinogenes for conversion of lignocellulosic
sugars to succinate and summarize what was learned about sugar and succinate transport and
metabolism. I also use these insights to discuss future work and potential for strain engineering.

25

APPENDIX

26

APPENDIX
Chapter 1 Figures and Tables

Figure 1.1. Industrial products of succinate and maleic anhydride

Dimethylsuccinate

Succinonitrile
N-Methyl-pyrollidone

Polybutylenesuccinate

Succinate
Succindiamide

1,4-butanediol

Maleic anhydride

Adipic acid

γ-butyrolactone

Tetrahydrofuran

27

Figure 1.2. General succinate-producing mixed acid metabolism. Metabolites: AcCoA,
acetyl-CoA; GLX: glyoxylate; OAA, oxaloacetate; and PEP, phosphoenolpyruvate. Reactions:
ADH, alcohol dehydrogenase; AK, acetate kinase; CS, citrate synthase; EMP, EmbdenMeyerhof pathway; Fm, fumarase; Frd, fumarate reductase; GLX, glyoxylate shunt (isocitrate
lyase and malate synthase); MDH, malate dehydrogenase; MEnz, malic enzyme; PEPC, PEP
carboxylase; PEPCK, PEP carboxykinase; PFL, pyruvate-formate lyase; PyrC, pyruvate
carboxylase; PTA: phosphotransacetylase; and PyrK, pyruvate kinase.

Glucose
EMP
PEPC/PEPCK

PEP

PyrK

PyrC

Pyruvate

OAA
MDH

PFL

MEnz

Formate

CS

AcCoA

Malate

AcCoA
GLX

Isocitrate

Fm

ADH
Fumarate
Frd

Succinate

28

PTA

ADH

AK

Ethanol

Acetate

Figure 1.3. Central metabolism of A. succinogenes 130Z. Metabolites: AcCoA, acetyl-CoA;
OAA, oxaloacetate; and PEP, phosphoenolpyruvate. Reactions: ADH, alcohol dehydrogenase;
AK, acetate kinase; CS, citrate synthase; EMP, Embden-Meyerhof pathway; Fm, fumarase; Frd,
fumarate reductase; MDH, malate dehydrogenase; MEnz, malic enzyme; OAAdec, OAA
decarboxylase; OPPP, oxidative pentose phosphate pathway; PDH, pyruvate dehydrogenase;
PEPCK, PEP carboxykinase; PFL, pyruvate-formate lyase; PTA: phosphotransacetylase; and
PyrK, pyruvate kinase.

Glucose

OPPP
Xylose

EMP
CO2

PEP

PEPCK

PyrK

OAAdec

Pyruvate

OAA
MDH

PFL

MEnz

PDH

Malate

Formate

AcCoA

ADH

Fm
Fumarate

PTA

ADH

AK

Ethanol

Acetate

Frd
Succinate

29

Table 1.1. Succinate-producing bacterial and yeast strains

Organism

Actinobacillus succinogenes
130Z; wild type
130Z; wild type
Mannheimia succiniciproducens
MBEL55E; wild type
LPK7; ldhA::KmR, pflB::CmR, pta-ackA:SpR
PALFK; ΔldhA, Δpta-ackA, ΔfruA
Escherichia coli
C; wild type
AFP111 (pTrc9A-pyc; ΔpflAB::CmR,
ldhA::KmR, ptsG-, pyc+)
KJ012 (C, ΔldhA, ΔackA, ΔadhE)
KJ017 (KJ012, 40th transfer)
KJ060 (KJ017, ΔfocA, ΔpflB, 86th transfer)
KJ134 (KJ060, ΔmgsA, ΔpoxB, ΔtdcDE,
ΔcitF, ΔaspA, ΔsfcA, Δpta
AS1600a
AS1600a
Corynebacterium glutamicum
ΔldhA-pCRA717
BOL-3/pAN6-gap (Δcat, Δpqo, Δpta-ack,
Δldh, pycP4585, fdh)

a

b

Condition ; Medium

Molar yield
(succinate/
sugar)

Succinate
Titer
Vol. Prod.
(mM)
(g L-1 h-1)

Other
productsc

Reference

A
A, Fo

(Bradfield and Nicol 2014)
(Bradfield et al. 2015)

BR; Def., G, YE, CSL
BR; Def., XH, YE, CSL

1.39
1.02

411
335.3

NDe
1.77

B; P, YE
FB; YE
FB; YE, P, S, Gl

0.69
1.16
1.64

89
444
664

NDe
2.97
6.02 38.6

F, A, L
P, M, A, L
None

(Lee et al. 2006)
(Lee et al. 2006)
(Lee et al. 2016)

B; Def., G
FBf; YE, Tr

0.19
1.68

49
840

0.12
1.3

F, A, L, P
A, E

(Jantama et al. 2008a)
(Vemuri et al. 2002)

B; Def., G
B; Def., G
B; Def., G
B; Def., G

0.13
0.73
1.61
1.53

6
204
622
606

0.07
0.25
0.77
0.75

A, L
A, F, L
A, M, L, P
A, P, M

(Jantama et al. 2008a)
(Jantama et al. 2008a)
(Jantama et al. 2008a)
(Jantama et al. 2008b)

B; Def., G
B; Def., X

1.31
1.12

637
714

0.91
0.96

A,?
A,?

(Sawisit et al. 2015)
(Sawisit et al. 2015)

B; Def.d, G
B; Def.d, F, G

1.40
1.41

703
151

11.8
NDe

A, L, M, P
P, M, K, F, A

(Okino et al. 2008)
(Litsanov et al. 2012)

30

Table 1.1 (cont’d).
Organism

Saccharomyces cerevisiae
AH22ura3
8D
8D Evolved
8D Evolved + pICL1

Δdic1

Conditiona; Mediumb

B; Def., G
B; Def., G
B; Def., G
B; Def., G
B; Def. G

Molar yield
(succinate/
sugar)
0.023
0.031
0.046
0.076
0.02

a

Succinate
Titer
Vol. Prod.
(mM)
(g L-1 h-1)

6.4
3.4
5.1
7.6
1.9

0.005
ND
ND
ND
ND

Other
productsc

Reference

K, F, C, M
?
?
?
E, G

(Raab et al. 2010)
(Otero et al. 2013)
(Otero et al. 2013)
(Otero et al. 2013)
(Agren et al. 2013)

B, batch culture; FB, fed batch culture; BR, continuous bioreactor
Def., defined medium; CSL, corn steep liquor; F, formate; G, glucose; Gl, Glycerol; P, peptone; S, sucrose; X, xylose; XH, xyloserich lignocellulose hydrolysate; YE, yeast extract
c
Listed in order of abundance (mol/mol sugar consumed): A, acetate; C, citrate; E, ethanol; F, fumarate; Fo, formic acid; G, glycerol;
K, α-ketoglutarate, L, lactate; M, malate; P, pyruvate; ?, other products not mentioned
d
Growth is done in rich medium, cells are transferred to defined medium for succinate production
e
ND, not determined
f
After transitioning from aerobic growth to anaerobic production
b

31

REFERENCES

32

REFERENCES

Agren, R., J. M. Otero, and J. Nielsen. 2013. Genome-scale modeling enables metabolic
engineering of Saccharomyces cerevisiae for succinic acid production. J. Ind. Microbiol.
Biotechnol. 40:735-747.
Akhtar, J., A. Idris, and R. A. Aziz. 2014. Recent advances in production of succinic acid from
lignocellulosic biomass. Appl. Microbiol. Biotechnol. 98:987-1000.
Arikawa, Y., M. Kobayashi, R. Kodaira, M. Shimosaka, H. Muratsubaki, K. Enomoto, and M.
Okazaki. 1999. Isolation of sake yeast strains possessing various levels of succinateand/or malate-producing abilities by gene disruption or mutation. J. Biosci. Bioeng.
87:333-339.
Bai, B., J. Zhou, M. Yang, Y. Liu, X. Xu, and J. Xing. 2015. Efficient production of succinic
acid from macroalgae hydrolysate by metabolically engineered Escherichia coli. Biores.
Technol. 185:56-61.
Banerjee, S., S. Mudliar, R. Sen, B. Giri, D. Satpute, T. Chakrabarti, and R. A. Pandey. 2009.
Commercializing lignocellulosic bioethanol: technology bottlenecks and possible
remedies. Biofuels, Bioprod. Biorefin. 4:77-93.
Bio-Amber. [cited. Available from http://www.bio-amber.com/_media/en/sarnia/bioambersarnia-media-release-final.pdf.
Borges, E. R., and N. Pereira. 2011. Succinic acid production from sugarcane bagasse
hemicellulose hydrolysate by Actinobacillus succinogenes. J. Ind. Microbiol. Biotechnol.
38:1001-1011.
Bozell, J. J., and G. R. Petersen. 2010. Technology development for the production of biobased
products from biorefinery carbohydrates - the US Department of Energy's "Top 10"
revisited. Green Chem. 12:525-728.
Bradfield, M. F. A., A. Mohagheghi, D. Salvachúa, H. Smith, B. A. Black, N. Dowe, G. T.
Beckham, and W. Nicol. 2015. Continuous succinic acid production by Actinobacillus
succinogenes on xylose-enriched hydrolysate. Biotechnol. Biofuels 8:181.
Bradfield, M. F. A., and W. Nicol. 2014. Continuous succinic acid production by Actinobacillus
succinogenes in a biofilm reactor: steady-state metabolic flux variation. Biochem. Eng. J.
85:1-7.
Brosse, N., A. Dufour, X. Meng, Q. Sun, and A. Ragauskas. 2012. Miscanthus: a fast-growing
crop for biofuels and chemicals production. Biofuels, Bioprod. Biorefin. 6:580-598.

33

Bugg, T. D. H. 2011. The emerging role for bacteria in lignin degradation and bio-product
formation. Cur. Opin. Biotechnol. 22:394-400.
Business, I. C. Chemical industry awaits for bio-succinic acid potential 2012 [cited. Available
from http://www.icis.com/resources/news/2012/01/30/9527521/chemical-industryawaits-for-bio-succinic-acid-potential/.
Chatterjee, R., C. S. Millard, K. Champion, D. P. Clark, and M. I. Donnelly. 2001. Mutation of
the ptsG gene results in increased production of succinate in fermentation of glucose by
Escherichia coli. Appl. Environ. Microbiol. 67:148–154.
Choi, S., H. Song, S. W. Lim, T. Y. Kim, J. H. Ahn, J. W. Lee, M.-H. Lee, and S. Y. Lee. 2016.
Highly selective production of succinic acid by metabolically engineered Mannheimia
succiniciproducens and its efficient purification. Biotechnol. Bioeng. 113:2168-2177.
Cok, B., I. Tsiropoulos, A. L. Roes, and M. K. Patel. 2013. Succinic acid production derived
from carbohydrates: an energy and greenhouse gas assessment of a platform chemical
toward a bio-based economy. Biofuels, Bioprod. Biorefin. 8:16-29.
Davis, C. P., D. Cleven, J. Brown, and E. Balish. 1976. Anaerobiospirillum, a new genus of
spiral-shaped bacteria. Int. J. Syst. Bacteriol. 26:498-504.
Ghaffar, S. H., and M. Fan. 2014. Lignin in straw and its applications as an adhesive. Int. J.
Adhes. Adhes. 48:92-101.
Glassner, D., J. Hettenhaus, and T. Schechinger. 1999. Corn stover potential: recasting the corn
sweetner industry. In Perspectives on new crops and new uses, edited by J. Janick.
Alexandria: ASHS Press, 74-82.
Goffeau, A. 2000. Four years of post-genomic life with 6000 yeast genes. FEBS Lett. 480:37-41.
Golby, P., D. J. Kelly, J. R. Guest, and S. C. Andrews. 1998. Transcriptional regulation and
organization of the dcuA and dcuB genes, encoding homologous anaerobic C4dicarboxylate transporters in Escherichia coli. J. Bacteriol. 180:6586-6596.
Guettler, M. V., M. K. Jain, and D. Rumler. 1996a. Method for making succinic acid, bacterial
variants for use in the process, and methods for obtaining variants. U.S.: Michigan
Biotechnology Institute.
Guettler, M. V., M. K. Jain, and B. K. Soni. 1996b. Process for making succinic acid,
microorganisms for use in the process and methods of obtaining the microorganisms.
U.S.: Michigan Biotechnology Institute.
Guettler, M. V., D. Rumler, and M. K. Jain. 1999. Actinobacillus succinogenes sp. nov., a novel
succinic-acid-producing strain from the bovine rumen. Int. J. Syst. Bacteriol. 49:207–216.

34

Gunnarsson, I. B., D. Karakashev, and I. Angelidaki. 2014. Succinic acid production by
fermentation of Jerusalem artichoke tuber hydrolysate with Actinobacillus succinogenes
130Z. Ind. Crops Prod. 62:125-129.
Hermann, B. G., K. Blok, and M. K. Patel. 2007. Producing bio-based bulk chemicals using
industrial biotechnology saves energy and combats climate change. Environ. Sci.
Technol. 41:7915-7921.
Hong, S. H., J. S. Kim, S. Y. Lee, Y. H. In, S. S. Choi, J. K. Rih, C. H. Kim, H. Jeong, C. G.
Hur, and J. J. Kim. 2004. The genome sequence of the capnophilic rumen bacterium
Mannheimia succiniciproducens. Nature Biotechnol. 22:1275-1281.
Huhn, S., E. Jolkver, R. Krämer, and K. Marin. 2011. Identification of the membrane protein
SucE and its role in succinate transport in Corynebacterium glutamicum. Appl. Microbiol.
Biotechnol. 89:327-335.
Intended nationally determined contribution of the EU and its Member States. 2015. edited by L.
P. o. t. C. o. t. E. Union. Riga, Latvia.
Janausch, I. G., E. Zientz, Q. H. Tran, A. Kröger, and G. Unden. 2002. C4-dicarboxylate carriers
and sensors in bacteria. BBA-Bioenergetics 1553:39-56.
Jantama, K., M. J. Haupt, S. A. Svoronos, X. L. Zhang, J. C. Moore, K. T. Shanmugam, and L.
O. Ingram. 2008a. Combining metabolic engineering and metabolic evolution to develop
nonrecombinant strains of Escherichia coli C that produce succinate and malate.
Biotechnol Bioeng. 99:1140-1153.
Jantama, K., X. L. Zhang, J. C. Moore, K. T. Shanmugam, S. A. Svoronos, and L. O. Ingram.
2008b. Eliminating side products and increasing succinate yields in engineered strains of
Escherichia coli C. Biotechnol Bioeng. 101:881-893.
Jojima, T., R. Noburyu, M. Suda, S. Okino, H. Yukawa, and M. Inui. 2016. Improving process
yield in succinic acid production by cell recycling of recombinant Corynebacterium
glutamicum. Fermentation 2:5.
Joshi, R. J., B. D. Schindler, N. R. McPherson, K. Tiwari, and C. Vieille. 2014. Development of
a markerless knockout method for Actinobacillus succinogenes. Appl. Environ.
Microbiol. 80:3053-3061.
Karimi, K., S. Kheradmandinia, and M. J. Taherzadeh. 2006. Conversion of rice straw to sugars
by dilute acid hydrolysis. Biomass and Bioenerg. 30:247-253.
Kim, D. Y., S. C. Yim, P. C. Lee, W. G. Lee, S. Y. Lee, and H. N. Chang. 2004. Batch and
continuous fermentation of succinic acid from wood hydrolysate by Mannheimia
succiniciproducens MBEL55E. Enzyme Microb. Technol. 35:648-653.

35

Kim, T. Y., H. U. Kim, J. M. Park, H. Song, J. S. Kim, and S. Y. Lee. 2007. Genome-scale
analysis of Mannheimia succiniciproducens metabolism. Biotechnol Bioeng. 97:657-671.
Kim, T. Y., H. U. Kim, H. Song, and S. Y. Lee. 2009. In silico analysis of the effects of H2 and
CO2 on the metabolism of a capnophilic bacterium Mannheimia succiniciproducens. J.
Biotechnol. 144:184-189.
Kwon, Y. D., S. Y. Lee, and P. Kim. 2006. Influence of gluconeogenic phosphoenolpyruvate
carboxykinase (PCK) expression on succinic acid fermentation in Escherichia coli under
high bicarbonate condition. J. Microbiol. Biotechnol. 16:1448-1452.
Laurichesse, S., and L. Avérous. 2014. Chemical modification of lignins: towards biobased
polymers. Prog. Polym. Sci. 7:1266-1290.
Lee, J. W., J. Yi, T. Y. Kim, S. Choi, J. H. Ahn, H. Song, M.-H. Lee, and S. Y. Lee. 2016.
Homo-succinic acid production by metabolically engineered Mannheimia
succiniciproducens. Metab. Eng. 38:409-417.
Lee, P. C., S. Y. Lee, S. H. Hong, and H. N. Chang. 2002. Isolation and characterization of a
new succinic acid-producing bacterium, Mannheimia succiniciproducens MBEL55E,
from bovine rumen. Appl. Microbiol. Biotechnol. 58:663–668.
Lee, P. C., S. Y. Lee, S. H. Hong, H. N. Chang, and S. C. Park. 2003. Biological conversion of
wood hydrolysate to succinic acid by Anaerobiospirillum succiniciproducens. Biotechnol.
Lett. 25:111-114.
Lee, P. C., W. G. Lee, S. Kwon, S. Y. Lee, and H. N. Chang. 2000. Batch and continuous
cultivation of Anaerobiospirillum succiniciproducens for the production of succinic acid
from whey. Appl. Microbiol. Biotechnol. 54:23-27.
Lee, S. J., H. Song, and S. Y. Lee. 2006. Genome-based metabolic engineering of Mannheimia
succiniciproducens for succinic acid production. Appl. Environ. Microbiol. 72:1939–
1948.
Lee, S. Y., J. M. Kim, H. Song, J. W. Lee, T. Y. Kim, and Y. S. Jang. 2008. From genome
sequence to integrated bioprocess for succinic acid production by Mannheimia
succiniciproducens. Appl. Microbiol. Biotechnol. 79:11-22.
Lin, C. S. K., L. A. Pfaltzgraff, L. Herrero-Davila, E. B. Mubofu, S. Abderrahim, J. H. Clark, A.
A. Koutinas, N. Kopsahelis, K. Stamatelatou, F. Dickson, S. Thankappan, Z. Mohamed,
R. Brocklesby, and R. Luque. 2013. Food waste as a valuable resource for the production
of chemicals, materials and fuels. Current situation and global perspective. Energy
Environ. Sci. 6:426-464.

36

Litsanov, B., M. Brocker, and M. Bott. 2012. Toward homosuccinate fermentation: metabolic
engineering of Corynebacterium glutamicum for anaerobic production of succinate from
glucose and formate. Appl. Environ. Microbiol. 78:3325-3337.
Litsanov, B., M. Brocker, M. Oldiges, and M. Bott. 2014. Succinic acid. In Bioprocessing of
renewable resources to commodity bioproducts, edited by V. S. Bisaria and A. Kondo:
John Wiley & Sons, 435-472.
McKinlay, J. B., M. Laivenieks, B. D. Schindler, A. A. Mckinlay, S. Siddaramappa, J. F.
Challacombe, S. R. Lowry, A. Clum, A. L. Lapidus, K. B. Burkhart, V. Harkins, and C.
Vieille. 2010. A genomic perspective on the potential of Actinobacillus succinogenes for
industrial succinate production. BMC Genomics 11:680.
McKinlay, J. B., Y. Shachar-Hill, J. G. Zeikus, and C. Vieille. 2007a. Determining
Actinobacillus succinogenes metabolic pathways and fluxes by NMR and GC-MS
analyses of 13C-labeled metabolic product isotopomers. Metab. Engin. 9:177–192.
McKinlay, J. B., and C. Vieille. 2008. 13C-metabolic flux analysis of Actinobacillus
succinogenes fermentative metabolism at different NaHCO3 and H2 concentrations.
Metab. Engin. 10:55–68.
McKinlay, J. B., C. Vieille, and J. G. Zeikus. 2007b. Prospects for a bio-based succinate
industry. Appl. Microbiol. Biotechnol. 76:727–740.
McKinlay, J. B., J. G. Zeikus, and C. Vieille. 2005. Insights into Actinobacillus succinogenes
fermentative metabolism in a chemically defined growth medium. Appl. Environ.
Microbiol. 71:6651-6656.
Millard, C. S., Y. P. Chao, J. C. Liao, and M. I. Donnelly. 1996. Enhanced production of
succinic acid by overexpression of phosphoenolpyruvate carboxylase in Escherichia coli.
Appl. Environ. Microbiol. 62:1808–1810.
Mullet, J., D. Morishige, R. McCormick, S. Truong, J. Hilley, B. McKinley, R. Anderson, S. N.
Olson, and W. Rooney. 2014. Energy Sorghum–a genetic model for the design of C4
grass bioenergy crops. J. Exp. Bot. 65:3479-3489.
Nghiem, N., B. H. Davison, M. I. Donnelly, S.-P. Tsai, and J. G. Frye. 2001. ACS Symp. Ser.
784:160-173.
Okino, S., R. Noburyu, M. Suda, T. Jojima, M. Inui, and H. Yukawa. 2008. An efficient succinic
acid production process in a metabolically engineered Corynebacterium glutamicum
strain. Biotechnol. Prod. Proc. Eng. 81:459-464.
Olajuyin, A. M., M. Yang, Y. Liu, T. Mu, J. Tian, O. A. Adaramoye, and J. Xing. 2016. Efficient
production of succinic acid from Palmaria palmata hydrolysate by metabolically
engineered Escherichia coli. Biores. Technol. 214:653-659.

37

Olson, S. N., K. Ritter, W. Rooney, A. Kemanian, B. A. McCarl, Y. Zhang, S. Hall, D. Packer,
and J. Mullet. 2012. High biomass yield energy sorghum: developing a genetic model for
C4 grass bioenergy crops. Biofuels, Bioprod. Biorefin. 6:640-655.
Otero, J. M., C. D., K. R. Patil, S. G. Poulsen, L. Olsson, and J. Nielsen. 2013. Industrial systems
biology of Saccharomyces cerevisiae enables novel succinic acid cell factory. PLoS ONE
8:e54144.
Pacific Northwest Pacific Laboratory. 2004. Top value added chemicals from biomass volume I:
results from screening for potential candidates from sugars and synthesis gas, edited by
U. S. D. o. Energy.
Paster, M., J. L. Pellegrino, and T. M. Carole. 2003. Industrial bioproducts: today and tomorrow.
In Energetics, Incorporated. Columbia, MD.
Patel, M. K., M. Crank, V. Dornburg, B. Hermann, L. Roes, B. Hüsing, L. Overbeek, and F. R.
Terragni, E. 2006. The BREW project: medium and long-term opportunities and risks of
the biotechnological production of bulk chemicals from renewable resources–the
potential of white biotechnology. Utrecht, the Netherlands: Technology and Society
(STS)/Copernicus Institute.
Pinazo, J. M., M. E. Domine, V. Parvulescu, and F. Petru. 2015. Sustainability metrics for
succinic acid production: a comparison between biomass-based and petrochemical routes.
Catal. Today 239:17-24.
Raab, A. M., G. Gebhardt, N. Bolotina, D. Weuster-Botz, and C. Lang. 2010. Metabolic
engineering of Saccharomyces cerevisiae for the biotechnological production of succinic
acid. Metab. Engin. 12:518-525.
Rados, D., D. L. Turner, L. L. Fonseca, A. L. Carvalho, B. Blombach, B. J. Eikmanns, A. R.
Neves, and H. Santos. 2014. Carbon flux analysis by 13C nuclear magnetic resonance to
determine the effect of CO2 on anaerobic succinate production by Corynebacterium
glutamicum. Appl. Environ. Microbiol. 80:3015-3024.
Roquette, D. a. [cited. Available from
http://www.dsm.com/en_US/downloads/media/12e_09_dsm_and_roquette_commercializ
e_bio_based_succinic_acid.pdf.
Sainsbury, P. D., E. M. Hardiman, M. Ahmad, H. Otani, N. Seghezzi, L. D. Eltis, and T. D. H.
Bugg. 2013. Breaking down lignin to high-value chemicals: the conversion of
lignocellulose to vanillin in a gene deletion mutant of Rhodococcus jostii RHA1. ACS
Chem. Biol. 8:2151-2156.
Salvachúa, D., A. Mohagheghi, H. Smith, M. F. A. Bradfield, W. Nicol, B. A. Black, M. J.
Biddy, N. Dowe, and G. T. Beckham. 2016. Succinic acid production on xylose-enriched

38

biorefinery streams by Actinobacillus succinogenes in batch fermentation. Biotechnol.
Biofuels 9:28.
Samuelov, N. S., R. Datta, M. K. Jain, and J. G. Zeikus. 1999. Whey fermentation by
Anaerobiospirillum succiniciproducens for production of a succinate-based animal feed
additive. Appl. Environ. Microbiol. 65:2260-2263.
Samuelov, N. S., R. Lamed, S. E. Lowe, and J. G. Zeikus. 1991. Influence of CO2-HCO3- levels
and pH on growth, succinate production, and enzyme activities of Anaerobiospirillum
succiniciproducens. Appl. Environ. Microbiol. 57:3013–3019.
Sánchez, Ó. J., and C. A. Cardona. 2008. Trends in biotechnological production of fuel ethanol
from different feedstocks. Biores. Technol. 99:5270-5295.
Sanderson, M. A., P. R. Adler, A. A. Boateng, M. D. Casler, and G. Sarath. 2006. Switchgrass as
a biofuels feedstock in the USA. Can. J. Plant Sci. 86:1315-1325.
Sawers, R. G., and D. P. Clark. 2004. Fermentative pyruvate and acetyl-coenzyme A
metabolism. In Escherichia coli and Salmonella: Cellular and Molecular Biology, edited
by A. Bock: ASM Press.
Sawisit, A., K. Jantama, H. Zheng, L. P. Yomano, S. W. York, K. T. Shanmugam, and L. O.
Ingram. 2015. Mutation in galP improved fermentation of mixed sugars to succinate
using engineered Escherichia coli AS1600a and AM1 mineral salts medium. Biores.
Technol. 193:433-441.
Sawisit, A., S. S. Jantama, S. Kanchanatawee, and K. Jantama. 2014. Efficient utilization of
cassava pulp for succinate production by metabolically engineered Escherichia coli
KJ122. Bioproc. Biosyst. Eng. 38:175-187.
Secchi, C., V. V. Cantarelli, S. Pereira Fde, H. H. Wolf, T. C. Brodt, M. C. Amaro, and E.
Inamine. 2005. Fatal bacteremia due to Anaerobiospirillum succiniciproducens: first
description in Brazil. Braz. J. Infect. Dis. 9:169-172.
Secretary, O. o. t. P. 2015. FACT SHEET: U.S. Reports its 2025 Emissions Target to the
UNFCCC. United States of America.
Six, S., S. C. Andrews, G. Unden, and J. R. Guest. 1994. Eschericia coli possesses 2 homologous
anaerobic C-4-dicarboxylate membrane transporters (DcuA and DcuB) distinct from the
aerobic dicarboxylate transport system (Dct). J. Bacteriol. 176:6470-6478.
Song, H., and S. Y. Lee. 2006. Production of succinic acid by bacterial fermentation. Enzyme
Microb. Tech. 39:352-361.

39

Sørensen, A., P. J. Teller, T. Hilstrøm, and B. K. Ahring. 2008. Hydrolysis of Miscanthus for
bioethanol production using dilute acid presoaking combined with wet explosion pretreatment and enzymatic treatment. Biores. Technol. 99:6602-6607.
Sun, J. X., F. C. Mao, X. F. Sun, and R. C. Sun. 2007. Comparative study of hemicelluloses
isolated with alkaline peroxide from lignocellulosic materials. J. Wood Chem. Technol.
24:239-262.
Tsuge, Y., T. Hasunuma, and A. Kondo. 2015. Recent advances in the metabolic engineering of
Corynebacterium glutamicum for the production of lactate and succinate from renewable
resources. J. Ind. Microbiol. Biotechnol. 42:375-389.
Van Acker, R., J.-C. Leplé, D. Aertsd, V. Storme, G. Goeminne, B. Ivens, F. Légée, C. Lapierre,
K. Piens, M. C. E. Van Montagu, N. Santoro, C. E. Foster, J. Ralph, W. Soetaert, G.
Pilate, and W. Boerjan. 2013. Improved saccharification and ethanol yield from fieldgrown transgenic poplar deficient in cinnamoyl-CoA reductase. Proc. Natl. Acad. Sci.
U.S.A. 111:845-850.
van der Werf, M. J., M. V. Guettler, M. K. Jain, and J. G. Zeikus. 1997. Environmental and
physiological factors affecting the succinate product ratio during carbohydrate
fermentation by Actinobacillus sp. 130Z. Arch. Microbiol. 167:332–342.
Vemuri, G. N., M. A. Eiteman, and E. Altman. 2002. Succinate production in dual-phase
Escherichi coli fermentations depends on the time of transition from aerobic to anaerobic
conditions. J. Ind. Microbiol. Biotechnol. 28:325-332.
Wan, C., and Y. Li. 2012. Fungal pretreatment of lignocellulosic biomass. Biotechnol. Adv.
30:1447-1457.
Wei, S. U. 2015. Enhanced actions on climate change: China's intended nationally determined
contribution, edited by D. o. C. Change. China.
Yost, W. M., J. W. Young, S. P. Schmidt, and A. D. McGilliard. 1977. Gluconeogenesis in
ruminants: propionic acid production from a high-grain diet fed to cattle. J. Nutr.
107:2036-2043.
Zeikus, J. G., M. K. Jain, and P. Elankovan. 1999. Biotechnology of succinic acid production
and markets for derived industrial products. Appl. Biochem. Biotechnol. 51:545-552.
Zhang, X. L., K. Jantama, J. C. Moore, L. R. Jarboe, K. T. Shanmugam, and L. O. Ingram. 2009.
Metabolic evolution of energy-conserving pathways for succinate production in
Escherichia coli. Proc. Natl. Acad. Sci. USA 106:20180-20185.
Zheng, P., J. J. Dong, Z. H. Sun, Y. Ni, and L. Fang. 2009. Fermentative production of succinic
acid from straw hydrolysate by Actinobacillus succinogenes. Bioresour. Technol.
100:2425-2429.

40

Zientz, E., I. G. Janausch, S. Six, and G. Unden. 1999. Functioning of DcuC as the C-4dicarboxylate carrier during glucose fermentation by Escherichia coli. J. Bacteriol.
181:3716-3720.
Zientz, E., S. Six, and G. Unden. 1996. Identification of a third secondary carrier (DcuC) for
anaerobic C4-dicarboxylate transport in Escherichia coli: roles of the three Dcu carriers
in uptake and exchange. J. Bacteriol. 178:7241-7247.

41

Chapter 2

Regulation of D-xylose utilization in Actinobacillus succinogenes and evolution for fast
growth

42

2.1 Abstract
Petroleum processing currently produces a wide variety of chemicals, but alternative
renewable biological processes are under active development to reduce reliance on oil and
dampen environmental and health impacts of producing these chemicals. Succinate is a chemical
of special interest because if it were produced from a renewable feedstock in a cost-competitive
method, it could be used as chemical feedstock for a 25 million ton commodity chemical market.
Actinobacillus succinogenes is one of the best natural succinate producers and grows on more
than 20 carbohydrates, including those most commonly found in renewable feedstocks like
lignocellulose from agricultural waste or bioenergy crops. However, the two most common
sugars in lignocellulose from most sources are glucose and xylose, and A. succinogenes’ growth
on xylose is much slower than on glucose. Therefore, a strain of A. succinogenes that grows
quickly on xylose would be more useful industrially. Also, knowledge of how an evolved strain
grows quickly on xylose could be adapted for use in other bacterial species.
We have evolved three strains of A. succinogenes that grow faster on xylose. These X
strains also have higher succinate yields. Mutations in two genes, xylE and fusA, were identified
across all three X strains by whole genome sequencing. RNA sequencing showed that xylose
assimilation genes are more highly expressed in the X strains than in wild-type A. succinogenes,
but xylose isomerase activity and internal xylose pool size are unchanged. In addition, many
glycolytic genes as well as genes encoding proteins involved in succinate production from
phosphoenolpyruvate are upregulated, and genes encoding two proteins responsible for
distributing carbon flux between succinate and byproducts are downregulated. Multiplex natural
transformation was used for the first time in A. succinogenes to introduce the xylE mutation from
strain X2 to the wild-type strain, creating the X2-XylE strain. This strain has a 40% higher

43

succinate yield, but slower growth than wild-type A. succinogenes on xylose. Construction of a
second strain, X2-EFG, is in progress to test the fusA mutations.

2.2. Introduction
The consequences of oil-based industrial processes, as well as dwindling oil supplies and
fluctuating oil prices, have led to a surge in the development of biological processes to replace
petroleum-based fuel and chemicals (Wilke 1995, 1999). However, devoting a large part of the
United States’ agricultural production to producing fuels and chemicals is not a sustainable
strategy, due in large part to the competing need for food. An alternative to using or replacing
food crops such as corn would be using a lignocellulosic feedstock, like corn stover.
To further increase sustainability and adapt to changing markets, factories could balance
the production of low-value but high-volume biofuels, such as ethanol and biodiesel, with the
production of high-value lower-volume chemicals (Morris 2005). An example of a highly
desirable low-volume chemical is succinate. Currently succinate is produced from maleic
anhydride as a petroleum product, but if it could be produced renewably, in high yield, and for a
price competitive with that of maleic anhydride, it could grow from having a specialty chemical
market to having a large commodity market with demand for up to 25 million tons per year
(Bozell and Petersen 2010).
Biological production of industrially useful chemicals using a number of biocatalyst
organisms has been a major topic for research. Actinobacillus succinogenes, a capnophilic,
facultatively anaerobic member of the Pasteurellaceae, is a very good candidate as a biocatalyst
for succinate production. It naturally produces high amounts of succinate (up to 110 g/L in some
conditions) and grows on about twenty carbohydrates, including the most abundant sugars in

44

cellulose and hemicellose: glucose, mannose, xylose, and arabinose (Guettler et al. 1999). In
addition, an incomplete tricarboxylic acid (TCA) cycle and a relatively simple fermentation
product profile (primarily succinate, formate, and acetate) reduce the complexity of genetic
engineering projects. However, before A. succinogenes can be exploited to its full potential, its
growth rate on hemicellulosic sugars must be increased. As an example, after hydrolysis corn
stover is approximately 37% glucose, 21% xylose, 3% arabinose, 2% galactose, and 2%
mannose with the remainder accounted for as lignin, protein and other non-sugar products (Aden
et al. 2002). As mentioned, A. succinogenes grows on those sugars, except galactose, however its
doubling time on glucose (1.69  0.06 h) is much faster than on the other sugars. This is
especially a problem with xylose, the second most abundant sugar in lignocellulosic substrates,
on which A. succinogenes’ doubling time is at least 4.20  0.24 h. In this study, wild-type A.
succinogenes was evolved for faster growth on xylose in three parallel, independent experiments.
The genomes of the evolved strains were resequenced to identify the accumulated mutations.
RNA sequencing was used to measure variations in gene expression across the transcriptome.
Select individual mutations were also re-introduced in the wild-type strain and tested for their
effect on growth rate and succinate production.

2.3 Materials and Methods
2.3.1 Bacterial strains and media
A. succinogenes 130Z (ATCC 55618) was originally purchased from the American Type
Culture Collection and adapted to grow on medium AM3 (McKinlay et al. 2005). AM3 is a
phosphate-based chemically defined medium that contains ammonium chloride, vitamins,
minerals, cysteine, methionine, glutamate, NaHCO3, and kanamycin, as well a carbon source

45

(McKinlay et al. 2005). Cultures were grown anaerobically in 10 mL of AM3 at 37 °C with
shaking at 250 rpm unless stated otherwise. When a rich medium was required, cultures were
grown in medium B, which is a phosphate buffer-based medium containing yeast extract,
bactotryptone, and glucose (Joshi et al. 2014). For natural transformation experiments, cells were
grown in Brain Heart Infusion medium (BHI; Becton-Dickinson and Co., Franklin Lakes, NJ),
competence was induced in MIV medium (Poje and Redfield 2003) supplemented with 2 mM
cAMP, and BHI was used as the recovery medium. LB glucose agar plates used for colony
isolation after multiplex natural transformation contained, per liter: 10 g tryptone (BectonDickinson); 5 g yeast extract (Becton-Dickinson and Co); 5 g sodium chloride (Macron Fine
Chemicals, Avantor Performance Materials, Center Valley, PA); 15 g agar (Becton-Dickinson
and Co.); 9 g glucose (Mallinckrodt, Chesterfield, Derbyshire, UK); 600 mg polyvinylalcohol;
10 mg calcium chloride; and 160 mg 5-bromo-4-chloro-3-indolyl-β-D-galactopyranoside (X-gal)
(Denville Scientific, Holliston, MA). LB glucose plates used for numerations did not include Xgal, but were otherwise identical. All other chemicals, unless otherwise noted, were purchased
from Sigma-Aldrich (St. Louis, MO).
Escherichia coli Top10 (Invitrogen) was used for cloning. It was propagated in LB medium
in the presence of ampicillin (100 µg/mL) or kanamycin (50 µg/mL) when needed for plasmid
maintenance.

2.3.2. A. succinogenes evolution for fast growth on xylose
Three independent cultures of A. succinogenes were grown anaerobically in AM3 with 100
mM NaHCO3 and 60 mM xylose. Cultures (0.25 ml inoculum) were transferred to fresh medium
during the exponential growth phase (OD660 between 0.5 and 1.5). Evolution was considered

46

complete when the doubling time of the culture plateaued below 2.5 h. An example of the
reduction in doubling time over the course of evolution is shown in Figure 2.1.

2.3.3. Growth curves and fermentation balance analyses
Growth curves were generated by tracking absorbance at 660 nm using a Beckman DU650
spectrophotometer (Beckman Coulter, Inc., Indianapolis, IN). Glucose, xylose, and fermentation
products were quantified by high performance liquid chromatography (HPLC) as described
(McKinlay et al. 2005) using supernatant samples collected at two points during exponential
growth.

2.3.4. Plasmids and DNA manipulations
Plasmids used in this study are listed in Table 2.1. PCR products were cloned into pCR2.1
using the TOPO-TA cloning kit (Invitrogen, Carlsbad, CA). pLGZ920 was used to express
foreign genes in A. succinogenes under control of the A. succinogenes pckA promoter (ppckA)
(Kim et al. 2004). One Shot TOP10 chemically competent E. coli cells (Invitrogen) were
transformed as described by the manufacturer. Electrocompetent A. succinogenes cells used for
transformation of circular DNA were prepared as described (Joshi et al. 2014).
Linear DNA used for multiplex natural transformation was amplified by PCR using
Herculase II Fusion polymerase (Agilent Technologies, Inc., Santa Clara, CA) with the standard
manufacturer protocol. PCR products were purified from agarose gels using the Wizard SV Gel
and PCR Clean-Up kit (Promega, Madison, WI).

47

Primers were purchased from IDT (Coralville, Iowa), and the sequence of cloned DNA
fragments and genomic constructs was verified by Sanger sequencing (Genewiz, South
Plainfield, NJ).

2.3.5. Multiplex natural transformation
In multiplex natural transformations, wild-type A. succinogenes 130Z was naturally
transformed using two linear DNA fragments simultaneously. The first fragment (the selection
fragment) carried the E. coli icd gene in a lacZ deletion construct (ΔlacZ::icd) (Joshi et al. 2014),
and was used for transformant selection while the second fragment (the screening fragment)
carried an unmarked desired mutation as described (Dalia et al. 2014). Natural transformations
were performed as described (Joshi et al. 2014) with the following modification: transformation
was performed using 1.5 µg of selection fragment and 3-5 µg of screening fragment. Growth of
the transformants was selected in AM16-glucose-isocitrate liquid medium. That culture was used
to inoculate a second tube of AM16-glucose-isocitrate for a second round of selective growth,
before isolating colonies by plating 200 µL of OD660 1 culture diluted to 10-7 on LB-glucose-Km.
Transformants were tested for the double recombination of the ΔlacZ::icd construct by colony
PCR using primers CV1 and CV2 (Table 2.2) using standard PCR protocols (Ausubel et al.
1993). ΔlacZ::icd double recombinants were then tested for the presence of the unmarked
mutation by amplifying the targeted region using primers CV3 and CV4 (xylE) or CV 6 and CV7
(fusA), followed by sequencing of the PCR products using primer CV5 (xylE) or CV8 (fusA).
The icd marker was removed using the yeast flippase as described (Joshi et al. 2014).

48

2.3.6. Strain resequencing
Isolated colonies from the evolved cultures were obtained by two consecutive streakings of
the evolved cultures on LB-glucose plates. These evolved strains were resequenced by the
Department of Energy’s Joint Genome Institute (DOE JGI). Genomic DNA was prepared from
the A. succinogenes clonal cultures grown in AM3-glucose using the Promega Wizard Genomic
DNA extraction kit. DNA concentration was determined using a kit sent by the JGI. For each
strain sequenced, the JGI provided the list, genome location, and nature of all mutations.

2.3.7. Xylose isomerase assays
Actively growing A. succinogenes cultures (OD660 = 0.8) were centrifuged (4,500  g, 15
min, 4 °C) and concentrated 10-fold by resuspending in 50 mM 4-morpholinepropanesulfonic
acid (pH 7.0) containing 1 mM CoCl2 (buffer A). Bacterial suspensions were lysed using a
French pressure cell press (American Instrument Company, Urbana, Illinois). Xylose isomerase
assays were performed as described (Vieille et al. 1995) with the following modifications: the
200-µL reaction mixture contained 1 M glucose and 40 µL crude extract in buffer A. The
reaction was started by adding the crude extract. Reactions were incubated at 37 °C for 20 min
then stopped on ice. Fructose produced during the reaction was quantified using the cysteinesulfuric acid-carbazole method (Dische and Borenfreund 1951). After incubation at room
temperature, absorbance was measured at 537 nm on a Beckman DU650 spectrophotometer.
To determine the Km of A. succinogenes xylose isomerase for xylose, assays were run in
a coupled assay with sorbitol dehydrogenase. The assay was run at 30 °C in a 1-mL volume
containing 50 mM MOPS (pH 7.0), 1 mM CoCl2, 0.15 mM NADH, 1 U sheep liver sorbitol
dehydrogenase (Roche Diagnostics, Indianapolis, IN), 10 µL crude extract, and xylose (1 mM, 2

49

mM, 5 mM, 7 mM, 10 mM, 20 mM, 50 mM, 70 mM, and 200 mM). Reactions were started by
adding the substrate and monitored for NADH disappearance at 340 nm. The crude extracts were
from three independent xylose-grown cultures of the X1 strain. The Km was calculated using
non-linear least square fitting of the data to the Michaelis-Menten equation using GraphPad
Prism 6 (GraphPad Software, Inc., La Jolla, CA).

2.3.8. Internal xylose concentration measurement
A. succinogenes cultures grown in 10 mL of AM3-xylose to OD660 1 were filtered
through a nylon filter (47-mm diameter, 0.45-µm pore size, Sigma-Aldrich) that had been
prewashed twice with 10 mL water. The filter and cells were washed once with 10 mL of water,
then quickly transferred to a 50-mL round-bottomed centrifuge tube containing about 10 mL
liquid nitrogen to quench metabolism. The entire process, from filtering the culture to quenching
metabolism took approximately 40 seconds. The filter was thoroughly crushed to a powder in the
liquid nitrogen. The culture filtrate was passed through a second filter. This second filter was
immediately washed with water once, frozen in liquid nitrogen, and crushed as above to be used
as a negative control to measure how much free xylose bound to the filter. Dry crushed filters
were extracted with three mL of cold extraction solution (40 % methanol, 40 % acetonitrile, 20
% water, stored at -20 °C) at -20 °C for 15 min.
After the 3 mL extraction, the tube containing the crushed filter and extraction solution
was centrifuged (10 min, 16,000  g, 4 °C) and the supernatant was split between two 1.5 mL
microcentrifuge tubes. The supernatant was centrifuged again to pellet remaining filter
fragments, and this supernatant was transferred to two additional microcentrifuge tubes for
storage. The filter fragments were extracted a second time with 1.5 mL extraction solution. The

50

original crushed filter pellet was extracted two more times with 1.5 mL extraction solution, for a
total of one 3 mL extraction and three 1.5 mL extractions. The extraction supernatants were
evaporated under a stream of air and the combined residue was resuspended in 100 µL of 90 %
acetonitrile.
Xylose was quantified at the MSU Mass Spectrometry and Metabolomics Core using a
Waters (Milford, MA) Quattro Premier mass spectrometer with an Acquity UPLC BEH Amide
Column and a liquid phase gradient from 100% 10 mM ammonium formate in water to 100%
acetonitrile. Samples were diluted 100-fold in 90 % acetonitrile and 10 µL were injected. Xylose
peaks were quantified against xylose standards (1 nM, 5 nM, 10 nM, 50 nM, 100 nM, 500 nM, 1
µM, and 10 µM in 90 % acetonitrile).
To calculate the cell volume of xylose-grown A. succinogenes, the length and width of at
least 100 cells were measured for each strain on pictures obtained by scanning electron
microscopy (SEM, MSU Center for Advanced Microscopy) using the measurement feature of
the operating software for the JEOL JSM-7500F scanning electron microscope (JEOL, Peabody,
MA). The cells were assumed to be cylinders capped with hemispheres. Cell density was counted
by plating dilutions of OD660 1 cultures grown in AM3-xylose on LB-glucose and counting
CFUs. Xylose concentration, cell density, and cell volume were used to calculate internal xylose
concentration.

2.3.9. RNA purification and RT-qPCR
RNA purification was done using a modified QIAGEN RNeasy Midi Kit (QIAGEN,
Hilden, Germany) protocol. Sixty-mL A. succinogenes cultures were grown in AM3 to OD520
0.5-1. Twenty five-mL culture aliquots were mixed with 25 mL of -20 °C methanol and stored at

51

-20°C for at least 30 min to quench metabolism. After centrifugation (15 min, 4,500  g, 4 °C)
RNA was purified from the cell pellets with the QIAGEN RNeasy Midi Kit following the kit
instructions, including the optional on-column DNase digestion, and using the following buffer
volumes: 5 mL TE buffer (10 mM Tris-HCl, 1 mM EDTA, pH 8.0) containing 5 mg/mL
lysozyme (Roche, Indianapolis, IN), 4 mL RLT buffer, and 2.8 mL 200-proof ethanol.
RNA quality was validated visually on a 1% agarose gel prepared with DEPC-treated water
(Figure S2.1). RNA was mixed 50% v/v with formaldehyde before loading on the gel. PCR with
primers CV11 and CV12 (Table 2.2) was used to determine that the RNA preparations were free
of contaminating DNA (Figure S2.1). Complementary DNA was prepared with SuperScript II
reverse transcriptase as described in the manufacturer’s instructions (Life Technologies, Grand
Island, NY). RNA and cDNA concentrations were measured using a Qubit Fluorometer (Life
Technologies).
Real-time quantitative PCR (RT-qPCR) was performed using Power SYBR master mix
(Life Technologies) in a Bio-Rad iCycler iQ thermocycler (Bio-Rad, Hercules, CA), using
primers designed with IDT’s oligo analyzer tool (Table 2.2). Five candidate genes were tested as
ΔΔCt RT-qPCR quantification reference genes: 16S rRNA (Asuc_R0008), pckA
(phosphoenolpyruvate carboxykinase, Asuc_0221), rnp (RNase PH, Asuc_0420), rpoH (sigma
32 subunit, Asuc_0449), and rpoD (sigma 70 subunit, Asuc_0723). The 16S rRNA, rnp, and
rpoD genes are routinely used as reference genes for RT-qPCR in other bacterial species. The
pckA gene was chosen because it is constitutively and highly expressed in A. succinogenes (Kim
et al. 2004; van der Werf et al. 1997).
PCR efficiencies were measured for all primer pairs by performing RT-qPCR with
approximately 4 ng/µL cDNA and three 10-fold serial dilutions. cDNA prepared from the X1

52

strain grown on glucose was used in these experiments. Efficiency was then calculated as
described (Life Technologies).

2.3.10. RNA sequencing
The same RNA samples were submitted to the MSU RTSF for RNA sequencing.
Libraries were prepared using the Illumina TruSeq Stranded Total RNA Library Preparation kit,
and the Illumina Ribo-Zero Gram-Negative Bacteria kit was used for rRNA depletion. Libraries
were validated and quantified using the Qubit dsDNA assay, Caliper LabChipGX, and Kapa
Illumina Library Quantification qPCR kits. At the time, 48 total samples were submitted for
RNA sequencing, so libraries were divided into four pools of twelve. The sequenced libraries
were from wild-type A. succinogenes fermenting glucose, xylose, mannose, or fructose; strains
X1, X2, and X3 fermenting glucose or xylose; as well as wild-type A. succinogenes respiring
glucose with nitrate as the electron acceptor, glycerol with nitrate as the electron acceptor, and
glycerol in microaerobic conditions. Three biological replicate libraries were sequenced for each
strain/condition combination and the three libraries from the wild-type strain fermenting glucose
were included in each of the four pools. Each pool was loaded on one lane of an Illumina HiSeq
2500 High Output (v4) flow cell and sequenced in 1x50bp single end format using HiSeq SBS
reagents. Base calling was done by Illumina Real Time Analysis v1.18.64 and output was
demultiplexed and converted to FastQ format using Illumina Bcl2fastq v1.8.4. Only the libraries
from wild-type A. succinogenes and X strains grown on glucose or xylose are discussed in this
study.
Sequencing results were analyzed using the SPARTA pipeline (Johnson et al. 2016). The
pipeline uses Trimmomatic to trim reads and remove adapters, FastQC for read quality analysis,

53

Bowtie to map reads, HTSeq to count transcript/gene feature abundance, and edgeR for
differential gene expression analysis. As mentioned, 48 RNA samples were sequenced and
analyzed, but results from only 33 libraries are reported here.

2.3.11. Determination of fusidic acid minimum inhibitory concentration (MIC)
Three biological replicates of strains 130Z, X1, X2, and X3 were grown in AM3-glucose
and transferred to AM3-glucose containing 100 µg/mL, 400 µg/mL, 700 µg/mL, and 1,000
µg/mL fusidic acid. Cultures were incubated at 37 °C overnight to test for growth.

2.4. Results
2.4.1. Growth rates and fermentation balances of the xylose-evolved strains
Transferring A. succinogenes on AM3-xylose multiple times resulted in evolved strains that
showed a decreased doubling time on xylose, as illustrated for one evolution experiment in
Figure 2.1. When the generation time of the evolving cultures reached 2.5 h, one isolated colony
from each independent evolution experiment (named strains X1, X2, and X3) was characterized
in terms of generation time and fermentation balance. The generation times of strains X1, X2,
and X3 (the X strains) on xylose were 2.33  0.10 h, 3.22  0.24 h, and 3.48  0.12 h,
respectively, compared to 4.20  0.24 h for the wild-type strain (generation time ± standard
deviation, three biological replicates). On glucose, the generation times of strains X1, X2, and
X3 were 1.81  0.02 h, 2.54  0.06 h, and 2.03  0.02 h, respectively, compared to 1.69  0.06 h
for the wild-type strain. While no selection pressure was applied to maintain succinate
production, the evolution process did not negatively affect succinate production. On the contrary,
strains X1, X2, and X3 showed 30%, 21%, and 27% increased succinate yields on glucose, as

54

well as 28%, 26%, and 17% increased succinate yields on xylose, respectively, compared to the
wild-type strain (Table 2.3). Differences in succinate yields were statistically significant (p <
0.05) in all cases except for strain X3 on xylose. Succinate productivity (g/L/h) was much more
variable in the cultures grown on glucose. Strain X1 produced 14% more succinate, while strains
X2 and X3 produced 28% and 27% less, respectively, than the wild-type strain. On xylose,
however, all three strains produced significantly more succinate and strain X1’s productivity was
250% higher than that of the wild-type. Differences in succinate productivities were statistically
significant (p < 0.01) in all cases. Additionally, the ratio of succinate to acetate produced, a
measure of succinate production efficiency, increased for all three strains (Table 2.3). The
increase was significant (p < 0.05) for all X strains except X2 on glucose and X3 on xylose.

2.4.2. Mutations in the xylose-evolved strains
The mutations accumulated in the X strains are listed in Table 2.4. All mutations are single
nucleotide polymorphisms (SNPs, 13) or small insertions or deletions (indels, 7). Even though no
mutations are shared by any two strains, the three evolved strains accumulated mutations in two
shared regions: in or upstream of xylE, encoding the xylose-proton symporter (Figure 2.2), and in
fusA, encoding elongation factor G (EF-G). The probability that three independently evolved
strains have mutations in the same gene randomly is very low, suggesting that these shared
mutations are correlated with the increased growth rate on xylose and/or increased succinate
production.
The three strains accumulated other mutations, in addition to the mutations in these shared
regions. Strain X1 accumulated three SNPs in genes encoding NAD synthesis or scavenging
enzymes: one mutation in the nicotinamide-nucleotide adenylyltransferase gene and one

55

upstream of it, and the third in the nicotinate phosphoribosyltransferase gene. Strain X1 also had
a frame-shift deletion in the gene encoding acyltransferase 3 and a SNP in a ribonuclease gene.
Strain X2 contained only one additional mutation, a single base insertion in dnaK, which encodes
the chaperone protein DnaK. Strain X3 accumulated two mutations in the formate pathway: a
SNP in the pyruvate-formate lyase gene, pflB, and an eleven-base pair deletion in the gene
encoding formate dehydrogenase accessory protein FdhD. While potentially interesting for a
variety of reasons, these mutations were not shared by the three strains and will not be further
discussed in this study.

2.4.3. Transcription of the xylose utilization gene cluster
The A. succinogenes xylose utilization gene cluster comprises seven genes (Figure 2.2).
Encoded by these genes are xylose isomerase (XylA), xylulokinase (XylB), a xylose ATPbinding cassette (ABC) transporter (XylFGH), a xylose-proton symporter (XylE), and a
transcriptional regulator (XylR). Based on its similarity to E. coli XylR (WP_044374182, 60%
amino acid identity, 77% similarity), A. succinogenes XylR was expected to be an activator,
rather than a repressor.
Because the genome encodes two xylose uptake systems, mutations inactivating XylE are
unlikely to have a large effect on xylose uptake. However, xylE’s location immediately upstream
of xylR (Figure 2.2) suggests that mutations in or upstream of xylE can have a polar effect on
xylR expression and, in turn, on xylose gene cluster regulation. To address this hypothesis, the
transcript levels of xylR, xylA, xylE, and xylF were quantified by RT-qPCR in the evolved strains
as well as in strain 130Z.

56

Total RNA was extracted from glucose- and xylose-grown cultures. The RNA extracted
was of excellent quality and contained no detectable genomic DNA (Figure S2.1). Reference
genes for RT-qPCR had not previously been identified in A. succinogenes, and therefore five
candidate genes were tested: 16S rRNA, pckA, rpoD, rnp, and rpoH. PCR efficiency
measurements showed a linear relationship between the amount of DNA template added and
signal produced for rpoD, rnp, and rpoH, but not for 16S rRNA and pckA (data not shown),
excluding these two genes as potential reference genes. ΔΔCt values were calculated for xylR,
xylA, xylE, and xylF using each of rpoD, rnp, and rpoH as reference genes. These ΔΔCt values
were used to identify which of rpoD, rnp, and rpoH should be used as reference genes. ΔΔCt
values calculated with rpoD and rpoH were similar for the four xyl genes in glucose-grown
cultures, while rpoD and rnp gave similar ΔΔCt values in xylose-grown cultures. ΔΔCt and fold
change comparisons were therefore done using rpoD.
RT-qPCR results are shown in Table 2.5. All fold changes are relative to the wild-type
strain. The four genes (xylR, xylA, xylE, and xylF) showed lower transcript levels in the X strains
than in the wild-type strain in glucose-grown conditions. Growth on xylose resulted in a mixture
of expression changes, however, with transcript levels of some genes increased, some decreased,
and some changing across the three strains. The xylR and xylE genes showed increased transcript
levels in the three xylose-grown X strains, supporting our hypothesis that transcription changes
would be mirrored between the two genes. In contrast, xylF, encoding a subunit of the xylose
ABC transporter, was moderately downregulated in the X strains. Finally, xylA, encoding xylose
isomerase, appeared upregulated in strain X1, but downregulated in strains X2 and X3.
Because the RT-qPCR results were statistically noisy and difficult to interpret, the
transcript levels of the xyl genes in the evolved and the parental strains were compared by RNA

57

sequencing. In glucose-grown cultures, transcript levels of the xyl genes were not statistically
different between the parental and evolved strains (Table 2.6), as indicated by high false
discovery rates (FDR). In xylose-grown cultures, transcript levels of the xyl genes varied to
larger extents between the parental and evolved strains, and in a statistically significant manner.
Of note, xylR and xylE transcript levels were lower in the three X strains compared to the
parental strain, while transcript levels of all other xyl genes were higher in all three X strains.
Increases in transcript levels were not large, but might be sufficient to explain the increased
growth rates on xylose. The minimum and maximum fold changes were both seen in strain X3:
1.3-fold for xylA and 2.1-fold for xylF, respectively. Note though that, while the xylR and xylE
transcript levels were lower in the X strains than in the parental strain on xylose, they were still
approximately 25% higher (xylR) and 800% higher (xylE) than in the glucose-grown wild-type
strain. Essentially, in xylose-grown cultures, a smaller increase in xylR transcript levels seems to
have led to larger increases in xylFGH and xylAB transcript levels in the X strains.

2.4.4. Transcript levels and succinate production
To determine whether or not the mutations accumulated in the X strains affected general
transcription patterns in A. succinogenes, the transcriptomes of the wild-type and X strains
grown on glucose and xylose were compared by principal component analysis (Figure 2.4),
Principal component 1 explained 37.43% of the variance observed and principal component 2
explained 18.02%, for a total of 55.45% of the variance. The four strains grown on glucose
cluster together along the two principal components, suggesting that the mutations accumulated
in the X strains do not strongly affect gene expression during growth on glucose. In contrast, the
transcriptomes of the three X strains are clustered together, and away from that of strain 130Z

58

during growth on xylose, suggesting that independent mutations in the three X strains cause
similar transcriptional changes across the three mutant strains. As expected for growth on
different substrates, the transcriptomes of the four xylose-grown strains cluster away from the
glucose-grown strains, as well.
Succinate production in A. succinogenes is necessarily tied to production of reduced
cofactors, either NADH or NADPH. Glucose-grown cultures overexpressing glucose-6phosphate dehydrogenase were shown to produce more succinate (Guettler et al. 2014), and
predicted increased flux through the oxidative pentose phosphate pathway and increased
NADPH production have been linked to increased succinate production in stationary phase
fermentations (Bradfield and Nicol 2016; Guettler et al. 2014). Here, none of the oxidative
pentose phosphate pathway genes were upregulated in the X strains compared to the wild-type
strain in glucose- or xylose-grown cultures (Table S2.1). While not conclusive, these results
suggest that increased succinate production in the X strains is not tied to increased NADPH
production in the oxidative pentose phosphate pathway.
Compared to the parental strain, most of the genes encoding glycolytic enzymes showed
increased transcript levels in at least some of the glucose- and xylose-grown X strains, and the
phosphoglycerate mutase gene was upregulated in all three xylose-grown X strains. The
glyceraldehyde 3-phosphate dehydrogenase, phosphoglycerate kinase, and enolase genes were
upregulated in glucose- and xylose-grown X1 and X2 strains, while the phosphofructokinase,
fructose-bisphosphate aldolase, and triosephosphate isomerase genes were upregulated in at least
one of the three xylose-grown X strains. Most transcript level increases were around 1.5- to 2fold. If these transcript level increases correlate with increased protein levels and increased

59

glycolytic flux, they might explain the increased growth rates of the X strains on xylose (Table
S2.3).
Downstream of glycolysis, flux in A. succinogenes’s central metabolism is partitioned
largely through a node around phosphoenolpyruvate, pyruvate, and oxaloacetate, with an
additional reversible flux between malate and pyruvate (Figure 2.3) (McKinlay et al. 2007;
McKinlay and Vieille 2008). The pckA (encoding phosphoenolpyruvate carboxykinase, PEPCK)
and mdh (encoding malate dehydrogenase, MDH) genes showed 2-3 fold higher transcript levels
in strains X1 and X2 than in the parental strain in xylose-grown cultures. Higher PEPCK and
MDH levels may be enough to shift metabolic flux toward succinate production.
Oxaloacetate decarboxylase converts oxaloacetate to pyruvate (Figure 2.3) and is
responsible for some carbon being diverted from succinate production to byproduct formation.
The protein has three subunits encoded by oadA, oadB, and oadC. The three genes showed lower
transcript levels in the xylose-grown X strains, with these results being statistically significant,
except for oadA and oadC in strain X3 (Table S2.3). Malic enzyme (MaeB) is responsible for
the reversible cross flux between malate and pyruvate (Figure 2.3). Transcript levels for maeB
were somewhat lower (less than 2-fold) in the three xylose-grown X strains than the parental
strain, which suggests that the two fermentative pathways are more isolated from one another in
the X strains than in the wild-type strain when growing on xylose. Upregulation of pckA and mdh
combined with down-regulation of oadABC and maeB could mean that more
phosphoenolpyruvate is converted to oxaloacetate and ends up as succinate in the X strains than
in the wild-type in xylose-grown cultures.
Aside from genes specifically involved in sugar catabolism and succinate or byproduct
accumulation, overall transcript levels in the X strains and the wild-type strain were compared in

60

both glucose- and xylose-grown cultures. The top nine upregulated genes in glucose-grown X1
and X2 strains were part of the sulfur assimilation and cysteine synthesis pathway. While these
genes were between 4- and 32-fold upregulated compared to the parental strain, their transcript
levels remained low relative to average transcript levels. It is therefore unlikely that their large
upregulation would have a noticeable effect on growth. Indeed, the growth rates of strains X1,
X2, and 130Z did not change in the presence of varying concentrations of cysteine or sulfate
(both are already present in AM3) (data not shown).

2.4.5. Xylose isomerase activity, internal xylose concentration, and xylose uptake rate
To determine whether or not the higher transcript levels of the xylAB and xylFGH genes in
the xylose-grown X strains translate into higher protein activity levels, xylose isomerase activity
was quantified in crude extracts. No significant difference in xylose isomerase activity was
detected between the xylose-grown wild-type and X strains (Table S2.2), indicating that
increased xylA transcript levels did not lead to increased xylose isomerase activity.
Similar xylose isomerase levels do not preclude a faster flux through the xylose
assimilation pathway if a higher number of XylEFG ABC transporters in the membrane allows
faster xylose uptake and higher cytoplasmic xylose concentration. To address this possibility,
internal xylose concentrations were determined in the xylose-grown wild-type and X strains. No
differences in cell sizes, CFUs per mL (both values are needed to calculate internal
concentrations), and internal xylose concentrations (Table S2.3) were detected between the
xylose-grown wild-type and X strains. Note that the calculated internal xylose concentrations
were seven- to ten-fold higher than the xylose isomerase’s Km for xylose (10.5 ± 1.09 mM, R2 =
0.962) that had been determined independently (Figure S2.2). At these concentrations, a minor

61

concentration change would have little to no effect on flux through xylose isomerase. All these
results cast doubt on the co-located xyl mutations being responsible for the increased growth
rates of the X strains on xylose.

2.4.6. Mutations in fusA and translation elongation rates in the evolved strains
EF-G participates in protein synthesis by catalyzing the translocation of tRNA and mRNA
from the ribosome’s hybrid A/P site to P/P and from the hybrid P/E site to E/E during translation
(Spirin 1985). EF-G also participates in dissociating the ribosome into subunits after translation
(Hirashima and Kaji 1973; Zavialov et al. 2005). EF-G is a five-domain protein (Figure 2.5), in
which domains I and III perform the GTPase function (Caldon and March 2003); domains III
and V contain a common RNA recognition motif (Nagai et al. 1990); and domains III, IV, and V
help catalyze tRNA and mRNA translocation (Frank et al. 2007; Shoji et al. 2009). Knocking out
EF-G has a drastic negative effect on growth rate, and many mutations, especially those in
domains IV and V, impair the protein function (Shoji et al. 2009; Rodnina et al. 1997).
The antibiotic fusidic acid impairs EF-G activity by locking it on the ribosome in the GDPbound state (Koripella et al. 2012; Bodley et al. 1969), and resistance to fusidic acid has been
linked to a variety of mutations in EF-G. Fusidic acid-resistant Salmonella typhimurium had EFG mutations clustered in three regions: amino acids 66-161 (domain I), 413-471 (domain III),
and 628-681 (domain IV/V) (Johanson et al. 1996). Fusidic acid resistant Staphylococcus aureus
had EF-G mutations clustered in domain III as well (residues 404-464), and two mutations
(P406L and R464L) have been isolated in both bacterial species (Nagaev et al. 2001). The X1,
X2, and X3 strains each have a single amino acid substitution in EF-G and the three mutations
are located in the same area of domain V (Figure 2.5), which corresponds to the third mutation

62

cluster in S. typhimurium EF-G. We hypothesized that the mutations in domain V of strains X1,
X2, and X3 EF-G increase fusidic acid resistance as well as translation and growth rates of these
strains. The MIC of fusidic acid for wild-type A. succinogenes was between 100 µg/mL and 400
µg/mL, while it was between 400 µg/mL and 700 µg/mL for strain X1, and above 700 µg/mL for
strains X2 and X3. Because of the mode of action of fusidic acid, these results suggest that the
fusA mutations in the X strains increase their fusidic acid resistance as well.
Finally, with the discovery of an RNA polymerase-ribosome “expressome” complex
(Kohler et al. 2017), it is possible that accelerating translation could accelerate the entire
complex. This could explain the broad-scale increased transcription levels for the central carbon
metabolism genes seen in the RNA sequencing analysis.

2.4.7. Reintroduction of individual mutations into wild-type 130Z
Because no unmarked mutation had ever been recombined in the genome of A.
succinogenes, I attempted to do it with only two mutations, the xylE and fusA mutations found in
strain X2 (X2-XylE and X2-EFG, respectively). Multiplex natural transformation was recently
shown to work in Vibrio cholerae (Dalia et al. 2014), a naturally competent bacterial species. I
decided to choose this approach, because A. succinogenes is naturally competent as well, and
because we have already been using this property to build knockout strains (Joshi et al. 2014).
Strain ΔlacZ X2-XylE was built using multiplex natural transformation. This strain contains
mutation 544772 from strain X2, which introduces a stop codon 24 codons into xylE, as well as
ΔlacZ. ΔlacZ X2-XylE’s growth was compared to that of the wild-type strain on xylose. It grew
much slower than 130Z, with a generation time of 8.60 ± 0.46 h (105% higher than 130Z).
However, ΔlacZ X2-XylE’s succinate yield was 42% higher than that of 130Z and its formate

63

yield was 14% lower than 130Z (Table 2.3). The succinate:acetate ratio was 1.00, similar to the
X strains. X2-XylE reached a final OD660 of 2.56 ± 0.08, while strain 130Z grew to an OD660 of
5.07 ± 0.11 and X2-XylE had a biomass yield 15% lower than that of strain 130Z during
exponential growth, demonstrating that ΔlacZ X2-XylE channels more xylose toward
fermentation products than its parental strain. Finally, strain X2-XylE’s succinate productivity is
300% higher than the wild-type on xylose, even higher than strain X1. This shows that, while
increasing growth rate often increases succinate productivity in A. succinogenes, this is not
always true.
Note that the correct control for this experiment should have been the ΔlacZ strain instead
of the wild-type strain, but I was unable to get the ΔlacZ strain to grow on xylose at the time of
the experiment. Strain ΔlacZ X2-EFG has been constructed as well using multiplex
transformation, but the plasmid carrying the flippase recombinase has not been cured yet. The
growth rate and fermentation balance of this strain will be determined in xylose-grown cultures
as well.

2.5. Discussion
Succinate is currently produced from petroleum processing via maleic anhydride. A
sustainable biological production of this industrially important compound would have to rely on
fast and effective use of renewable feedstocks. Xylose is used by wild-type A. succinogenes, but
the doubling time is slower and succinate yields are lower than on glucose.
The X1, X2, and X3 strains were evolved for fast growth on xylose. Most of the growth
rate increase occurred in the first half of each of the evolution series, but the evolution process
was continued because the growth rate continued to improve. The X strains grow faster on

64

xylose than the wild-type strain and produce significantly more succinate when grown on either
glucose or xylose. Of all the mutations accumulated by the individual evolved strains, the three
X strains share mutations in or near two genes, xylE and fusA.
Our hypothesis for the effect of the xylE mutations on growth was that these mutations
upregulated xylR, which, in turn, upregulated xylAB and xylFGH. This hypothesis was first tested
with RT-qPCR. Reproducibility was poor and the results were not consistent between genes and
between strains. While xylR appeared upregulated in all X strains, this upregulation did not
translate into an upregulation of xylA and xylF: only xylA was upregulated, and only in strain X1.
The xylA and xylF genes appeared otherwise downregulated in the X strains.
The RNAseq results were more internally consistent, with variation between RNAseq
biological replicates lower than variations between technical RT-qPCR replicates, even though
the RNA samples used in both approaches were the same. The A. succinogenes xylose gene
cluster is organized similarly to that in E. coli, with the exception that xylE is located apart from
the other xyl genes in E. coli. In E. coli, xylFGH are cotranscribed, xylAB are cotranscribed, and
xylR is transcribed on its own. The xylFGH and xylAB operons are preceded by XylR binding
sites, while xylR is regulated otherwise (Song and Park 1997). Our RNAseq results did not
entirely align with a similar regulatory mechanism in A. succinogenes. Our results are consistent
with XylR being an activator, because expression of xylR in xylose-grown cultures coincides
with increased xylFGH and xylAB transcript levels, but they suggest the existence of a second
level of regulation as well. If XylR were the only regulator of the xylFGH and xylAB genes, these
genes should show lower transcript levels in the X strains than in the wild-type strain on xylose,
to match the trend in xylR transcript levels. Attempts to overexpress or delete xylR have been
unsuccessful.

65

Transcript levels do not necessarily correlate with protein levels and fluxes through a
pathway. Xylose isomerase activity levels and internal xylose concentrations were not
significantly different in the X strains compared to the wild-type strain during growth on xylose.
It is possible that the assay we developed to measure xylose concentration did not quench xylose
consumption fast enough. If this were the case, though, the reported cytoplasmic xylose
concentrations would be underestimated, and potential differences in xylose concentrations
between the X and the wild-type strains would have no effect on flux through xylose isomerase,
with the enzyme’s Km for xylose being only 10 mM. We were expecting xylose concentrations to
be in the same order of magnitude as the Km, as is commonly assumed for intracellular enzyme
substrates. Xylose is the initial substrate in the xylose-assimilation pathway, so this general
assumption might not apply in this case.
When I started this work, no method was available to introduce single, unmarked mutations
into A. succinogenes. This study is the first to demonstrate that multiplex transformation works
in A. succinogenes. Indeed, I was able to introduce the xylE mutation from X2 back into strain
130Z using the ΔlacZ::icd mutation as the selection marker. The inconvenience, of course, is that
I had to introduce a second mutation. When compared to strain 130Z, the double mutant showed
slower growth accompanied by an over 40% increased succinate yield during growth on xylose.
Because strain ΔlacZ has not been tested for growth on xylose, it is impossible yet to conclude
whether the X2-E mutation increases or decreases A. succinogenes’s growth rate on xylose. One
way to answer this question would be to reintroduce the wild-type lacZ gene in the double
mutant strain to generate a strain with the single X2-XylE mutation. It is unclear why a mutation
that potentially decreases the growth rate would be selected during the directed evolution
experiment, but if the xylE mutation did not happen first, mutations that took place earlier in the

66

evolution process could have produced a genetic background in which the xylE mutation had a
different effect on the growth rate.
While I have not been able yet to test the effect of the X2-EFG mutation on A.
succinogenes’s growth rate on xylose, I am only a few steps away from being able to do so. The
hypothesis I will be testing then is that the EF-G mutations in the X strains increase translation
rates and growth rates.
To our surprise, many genes encoding glycolytic enzymes, plus the genes encoding PEP
carboxykinase and malate dehydrogenase, were upregulated in the X strains compared to the
wild-type strain during growth on xylose. These transcript level increases were paired with
decreased transcript levels for the oxaloacetate dehydrogenase and malic enzyme genes.
Combined, these results could correlate with both the X strains’ increased growth rates on xylose
and their increased succinate yields. How these RNAseq results relate to the mutations found in
the X strains remains unknown at this point.
So far, I have been unable to identify a direct cause for the increased growth rates of the X
strains on xylose, but the fermentation balance of the ΔlacZ X2-XylE strain on xylose suggests
that the xylE mutation from strain X2 increases succinate production. The X strains are able to
grow quickly on xylose and have an increased succinate yield. While it is still not completely
clear why this is the case, the growth and metabolic phenotypes of these strains are interesting
and potentially useful for future applications.

2.6. Acknowledgements
This work was supported by MSU startup funds to CV, by USDA-NIFA award 201167009-30019 to CV, and by the DOE JGI Community Sequencing Program, Agreement No.

67

NPUSR006360. NM was supported in part by a recruiting fellowship from the MSU Plant
Research Laboratory. I thank Rajasi Joshi for building the pCR2.1-ΔlacZ::icd plasmid, Dr. Justin
Beauchamp for drawing Figure 2.5 and Vladislav Jdanov, Jacob Gibson, and Maeva Bottex for
their expert technical assistance. I thank Rhiannon LeVeque and the Mulks laboratory at
Michigan State University for sharing their RNA purification protocol. I thank Dr. Ben Johnson
for help with his SPARTA RNA sequencing analysis pipeline. I thank Dr. Robert Hausinger for
letting me use the GraphPad Prism 6 software in his laboratory. I am grateful to Jeff Landgraf of
the MSU Genomics Core for his help with RNA quality control methods and qPCR methods and
analysis.

68

APPENDIX

69

APPENDIX
Chapter 2 Figures and Tables

Table 2.1. Plasmids used in this study
Plasmids

Description
R

Source

R

Amp , Km , lacZα, cloning vector
pCR2.1 derivative, A. succinogenes ΔlacZ::FRTpCR2.1-ΔlacZ::icd
ppckA-icd-FRT
pLGZ920 derivative, S. cerevisiae flippase under
pCV933
control of A. succinogenes ppckA
pCR2.1

pCR2.1-xylE-X2
pCR2.1-fusA-X2

pCR2.1 derivative, includes mutation 544772 from
strain X2
pCR2.1 derivative, includes mutation 315072 from
strain X2

70

Invitrogen
(Joshi et al.
2014)
(Joshi et al.
2014)
This study
This study

Table 2.2. Primers for PCR and RT-qPCR
Primer Sequence

Use

CV1

CAAAAACGCAACGCACAATC

lacZ knockout confirmation (forward)

CV2

GATGGAAAATGGATTGAACGCAG

lacZ knockout confirmation (reverse)

CV3

GGTGCCTAATACGCACATGC

xylE mutation confirmation (forward PCR primer)

CV4

CGACTGCACTTTACGCTCTC

xylE mutation confirmation (reverse PCR primer)

CV5

GCAACACAAAACCCCAACAATG

xylE mutation confirmation (sequencing PCR primer)

CV6

CGATACGAAGTCTGTGTGCG

fusA mutation confirmation (forward PCR primer)

CV7

CTGAAGGCGAACGTCATGC

fusA mutation confirmation (reverse PCR primer)

CV8

CACTTTTCAGTTTCCTTTTGTTATTCCCC

fusA mutation confirmation (sequencing PCR primer)

CV9

CCGGATCAAGAGCTACCAAC

in pCR2.1, 1 kb upstream of MCS (forward)

CV10

CGAAACGATCCTCATCCTGT

in pCR2.1, 1 kb downstream of MCS (reverse)

CV11

TGCGTTACAACCCTGAAACA

frdB forward

CV12

TCTTTCGCACTTTCCAGCTT

frdB reverse

CV13

TGTTGGTCGAACGCATGAAA

qPCR pckA forward

CV14

CAACCGGTGTTCACCAAATAAG

qPCR pckA reverse

CV15

TGCGTGCCATGGTCGAT

qPCR rnp forward

CV16

TCCAGGGTGACGGAACGTT

qPCR rnp reverse

CV17

CGCGGCTAAAGCGTTGAT

qPCR rpoH forward

CV18

GCGCGCCACATGAATGA

qPCR rpoH reverse

CV19

CGGATCCGGTGCGTATGTAT

qPCR rpoD forward

CV20

CCCTTCCCGGTCAAGCA

qPCR rpoD reverse

CV21

GCTGGCGGCAGGCTTA

qPCR 16S forward

CV22

GCAAGCTTTCCACCCGTTAC

qPCR 16S reverse

71

Table 2.3. Fermentation balances of wild-type, evolved, and mutant A. succinogenes
strains. Results are the averages of three biological replicates ± standard deviations. Carbon
balancesa were all between 0.97 and 1.09.
Succinate
Succinate:acetate
Fermentation products (mM)

productivity

Strain

(g L-1 h-1)
Succinate

Formate

Acetate

Biomassb

ratio

Growth on glucose
130Z

58.5 ± 2.7

96.9 ± 4.3

67.8 ± 3.1

212.8 ± 4.0

0.520 ± 0.018

0.86 ± 0.06

X1

76.2 ± 2.9**

83.6 ± 3.4*

69.6 ± 2.6

187.7 ± 0.7**

0.593 ± 0.017**

1.09 ± 0.06**

X2

70.9 ± 7.1*

90.4 ± 9.1

73.9 ± 7.4

149.4 ± 0.9**

0.372 ± 0.003**

0.96 ± 0.14

X3

74.4 ± 4.4**

85.6 ± 5.4*

68.0 ± 4.0

156.2 ± 2.0**

0.380 ± 0.016**

1.09 ± 0.09*

Growth on xylose
130Z

49.9 ± 3.5

89.0 ± 6.7

66.1 ± 5.2

171.5 ± 1.2

0.162 ± 0.006

0.75 ± 0.08

X1

64.0 ± 2.6**

69.6 ± 4.0*

64.7 ± 3.3

134.0 ± 2.1**

0.566 ± 0.012**

0.99 ± 0.06*

X2

62.4 ± 3.3*

70.1 ± 4.1*

63.1 ± 3.6

109.3 ± 0.9**

0.336 ± 0.012**

0.99 ± 0.08*

X3

58.6 ± 4.5

76.8 ± 5.1

64.7 ± 4.5

103.2 ± 0.4**

0.261 ± 0.015**

0.91 ± 0.09

X2-E

71.0 ± 4.8**

72.4 ± 4.6*

71.9 ± 5.2

84.15 ± 6.27**

0.649 ± 0.044**

1.00 ± 0.01**

a

Carbon balance is the carbon in products/carbon in sugar consumed. It is assumed that one CO2
is fixed for each molecule of succinate produced.
b
Biomass was determined using assumed values of 567 mg dry cell weight/mL per OD660
(McKinlay et al. 2007) and a cell composition of CH2O0.5N0.2 (24.967 g/mol) (van der Werf et al.
1997)
* Significantly different (p < 0.05) from 130Z
** Significantly different (p < 0.01) from 130Z

72

Table 2.4. Mutations identified in strains X1, X2, and X3. (+): insertion; (Δ): deletion; (a  b): mutation changes a to b. NA: Does
not apply.
Positiona

DNA Mutationb

Effect on Protein

Elongation factor G, FusA

GT

A643E

Nicotinamide-nucleotide adenylyltransferase, NadR

CT

M288I

ASUC_RS02510
Nicotinamide-nucleotide adenylyltransferase, NadR
d
5’ of ASUC_RS02555 D-xylose proton-symporter, XylE

TA

E277V

Type

Locus

Protein function, name

315135

SNPc

ASUC_RS01455

532090

SNP

ASUC_RS02510

532124

SNP

544906

SNP

700429

SNP

ASUC_RS03310

Nicotinate phosphoribosyltransferase

885247

Indele

ASUC_RS04260

Acyltransferase 3

976810

SNP

ASUC_RS04730

SNP

544772
1209875
243883

Strain X1

CT

NA

CA

S441Y

ΔTGGACAA

Frameshift

Ribonuclease G

AG

Q55R

ASUC_RS01455

Elongation factor G, FusA

GT

T664K

SNP

ASUC_RS02555

D-xylose proton-symporter, XylE

GT

Y24-

Indel

ASUC_RS05745

Chaperone protein, DnaK

+A

Frameshift

SNP

ASUC_RS01090

Pyruvate-formate lyase, PflB

GA

S377F

315046

SNP

ASUC_RS01455

Elongation factor G, FusA

GT

R673S

543986

SNP

ASUC_RS02555

D-xylose proton-symporter, XylE

CA

Q286H

543990

Indel

ASUC_RS02555

D-xylose proton-symporter, XylE

ΔAAAACGGAAAGT

Deletion

544006

Indel

ASUC_RS02555

D-xylose proton-symporter, XylE

ΔT

Frameshift

544152

SNP

ASUC_RS02555

D-xylose proton-symporter, XylE

AC

L231R

1410818

Indel

ASUC_RS06635

Formate dehydrogenase accessory protein, FdhD

ΔTACGGAACAAT

Frameshift

1043606

Indel

ASUC_RS05040

Transcriptional antiterminator, BglG

+ TTTA

Frameshift

1046314

Indel

ASUC_RS05050

Cof-like hydrolase

ΔGGT

Deletion

1174181

SNP

ASUC_RS05580

HflC protein

TG

D125A

Strain X2
315072

Strain X3

a

Indicates the genomic location of the mutation or the first base of a multiple-base mutation.

73

Table 2.4 (cont’d).
b

A letter-number-letter notation indicates an amino acid substitution, while a number followed by a colon and nucleotide sequence
indicates the number of bases inserted or deleted and the sequence deleted or inserted.
c
SNP: Single nucleotide polymorphism.
d
: Non-protein-coding region.
e
Indel: Insertion or deletion.

74

Table 2.5. Fold changes of xylR, xylF, xylA, and xylE transcripts relative to wild type strain
130Z in glucose- and xylose-grown cultures of strains X1, X2, and X3 calculated from RTqPCR data. Changes are averages of three biological replicates, expressed as 2ΔΔCt values ±
standard deviations.
Glucose
xylR
xylF
xylA
xylE

Xylose

X1

X2

X3

X1

X2

X3

0.33 ± 0.86
0.26 ± 0.75
0.12 ± 0.49
0.30 ± 0.89

0.12 ± 0.44
0.09 ± 0.36
0.05 ± 0.21
0.12 ± 0.45

0.55 ± 2.07
0.62 ± 2.82
0.25 ± 1.31
0.58 ± 2.51

1.77 ± 2.67
0.76 ± 1.83
1.69 ± 4.55
2.44 ± 3.45

3.34 ± 2.40
0.44 ± 0.19
0.70 ± 0.39
4.45 ± 2.36

1.40 ± 1.48
0.59 ± 0.56
0.88 ± 1.13
1.54 ± 1.35

75

Table 2.6. Fold changes in transcripts of the xylose assimilation genes in the glucose- and
xylose-grown X strains compared to the parental strain, 130Z, calculated from RNAseq
data. Changes are expressed as log2 fold changes. False discovery rates (FDR) were calculated
using the Benjamini-Hochberg procedure.
Gene

X1

FDR X1

X2

FDR X2

X3

FDR X3

xylR

0.04

0.92

-0.08

0.90

-0.09

0.72

xylE

0.13

0.61

0.02

0.97

-0.41

0.003

xylF

0.12

0.78

0.04

0.97

0.13

0.65

xylG

0.46

0.06

0.27

0.60

0.39

0.08

xylH

0.47

0.09

0.36

0.47

0.35

0.20

xylA

0.24

0.64

0.32

0.64

0.21

0.60

xylB

0.28

0.20

0.23

0.53

0.12

0.63

xylR

-0.79

1.07 × 10-24

-0.92

1.84 × 10-32

-0.73

4.36 × 10-21

xylE

-2.79

1.47 × 10-120

-2.61

2.22 × 10-107

-0.48

4.89 × 10-5

xylF

0.95

5.75 × 10-8

0.90

3.22 × 10-7

1.08

1.07 × 10-24

xylG

0.95

2.82 × 10-8

0.82

2.20 × 10-6

0.77

8.68 × 10-6

xylH

0.86

1.06 × 10-7

0.77

2.29 × 10-6

0.59

3.27 × 10-4

xylA

0.65

3.62 × 10-16

0.62

5.04 × 10-15

0.35

1.81 × 10-5

xylB

0.90

6.04 × 10-11

0.85

9.86 × 10-10

0.81

4.87 × 10-9

Growth on glucose

Growth on xylose

76

Figure 2.1. Doubling times of evolving cultures during the evolution transfer series that led
to strain X1. Each point represents the average of three biological replicates. Error bars are
standard deviations (not always visible). Strain X1 was isolated from the culture obtained after
transfer 34.

77

Figure 2.2. Mutations in the A. succinogenes xylose utilization gene cluster. Gene lengths are
not shown to scale. Mutations are shown as triangles and labeled with the strain in which they
were found. Empty triangles indicate non-synonymous SNPs, filled triangles indicate stop
codons introduced by SNPs or indels.

78

Figure 2.3. Central metabolism of A. succinogenes 130Z. Metabolites: AcCoA, acetyl-CoA;
CO2, carbon dioxide; DHAP, dihydroxyacetone phosphate; E4P; erythrose 4-phosphate; G3P,
glyceraldehyde 3-phosphate; OAA, oxaloacetate; PEP, phosphoenolpyruvate; R5P, ribose 5phosphate; Ri5P, ribulose 5-phosphate; and X5P, xylulose 5-phosphate. Enzymes: ADH, alcohol
dehydrogenase; AK, acetate kinase; CS, citrate synthase; FBA, fructose-bisphosphate aldolase;
Fm, fumarase; Frd, fumarate reductase; G6PDH, glucose-6-phosphate dehydrogenase; G6PI,
glucose-6-phosphate isomerase; GAPDH, glyceraldehyde 3-phosphate dehydrogenase; MDH,
malate dehydrogenase; MaeB, malic enzyme; OAD, OAA decarboxylase; OPPP, oxidative
pentose phosphate pathway; PDH, pyruvate dehydrogenase; PEPCK, PEP carboxykinase; PFK,
6-phosphogluconolactonase; 6-phosphofructokinase; PFL, pyruvate-formate lyase; PGDH, 6phosphogluconate dehydrogenase; PGK, phosphoglycerate kinase; PGL, PGM,
phosphoglycerate mutase; PK, pyruvate kinase; PTA, phophotransacetylase; RPE, ribulosephosphate 3-epimerase; RPI, ribose 5-phosphate isomerase; TA, transaldolase; TK,
transketolase; TPI, triosephosphate isomerase; XylA, xylose isomerase; and XylB, xylulose
kinase.

79

Figure 2.4. Principal component analysis of RNA sequencing results. 130Z and the X strains
grown on glucose (open symbols) or xylose (filled symbols). Axis labels include the percent
variance explained by the principal component. 130Z (■), X1 (♦), X2 (▲), X3 (●).

80

Figure 2.5. Structure of E. coli K12 EF-G (PDB accession number 3j0eH). The structure is in
ribbon representation. The residues mutated in X1 (A643E), X2 (T664K), and X3 (R673S) are
shown in stick representation marked by the black box. The structure was visualized using The
PyMOL Molecular Graphics System, Version 1.3 (Schrödinger, LLC).

81

Table S2.1. Fold changes in transcripts levels of central metabolism genes. Glucose- and xylose-grown X strains compared to the
parental strain, 130Z, calculated from RNAseq data. Changes are expressed as log2 fold changes. False discovery rates (FDR) were
calculated using the Benjamini-Hochberg procedure. Cell shading indicates P < 0.05.
Glucose
Gene ID

Gene product

Xylose

X1

FDR

X2

FDR

X3

FDR

X1

FDR

X2

FDR

X3

FDR

ASUC_RS01190 Glucose-6-phosphate 1-dehydrogenase

-0.07

0.80

-0.07

0.91

0.00

0.99

0.05

0.76

0.02

0.89

0.00

0.98

ASUC_RS01195 6-Phosphogluconolactonase

-0.15

0.57

-0.13

0.77

0.01

0.98

-0.20

0.22

-0.14

0.43

-0.09

0.63

ASUC_RS01220 6-Phosphogluconate dehydrogenase

-0.08

0.87

-0.07

0.93

0.08

0.80

-0.25

0.26

-0.29

0.19

-0.51

1.6 × 10-2

ASUC_RS01145 Ribulose-phosphate 3-epimerase

0.08

0.83

-0.06

0.93

-0.16

0.48

-0.30

0.11

-0.28

0.14

Pentose Phosphate Pathway

ASUC_RS08195 Ribose 5-phosphate isomerase B

0.18

0.74

-0.07

0.95

0.29

0.39

0.42

0.15

ASUC_RS09695 Ribose 5-phosphate isomerase A

0.23

0.22

0.21

0.47

-0.23

0.15

0.35

2.2 × 10

ASUC_RS03040 Transaldolase

-0.08

0.85

-0.12

0.87

0.02

0.96

0.16

0.46

ASUC_RS00910 Transketolase

0.10

0.80

0.03

0.97

-0.08

0.80

0.46

-0.40

2.6 × 10-2

2.3 × 10

-2

0.15

0.65

0.34

2.8 × 10

-2

0.16

0.34

0.07

0.77

-0.25

0.23

0.33

0.09

0.64
-2

1.6 × 10

-2
-3

0.98

6.1 × 10-8

0.73

1.4 × 10

-2

1.60

1.6 × 10-8

ASUC_RS00915 Transketolase

0.26

0.59

0.11

0.93

-0.19

0.65

0.79

7.3 × 10

ASUC_RS01380 Transketolase

0.13

0.66

0.07

0.91

0.05

0.87

0.46

7.4 × 10-3

0.41

1.7 × 10-2

0.04

0.84

0.01

0.98

0.02

0.97

0.19

0.22

0.07

0.65

0.03

0.88

-0.18

0.23

Glycolysis and Gluconeogenesis
ASUC_RS07430 Glucose-6-phosphate isomerase
ASUC_RS03565 6-Phosphofructokinase

0.68

3.9 × 10

ASUC_RS02055 Fructose-1,6-bisphosphatase

-0.13

0.50

-6

0.40

0.05

-0.11

0.66

0.81

7.6 × 10

-0.06

0.91

-0.14

0.35

-0.19

0.16

-6

-2

-6

0.84

3.7 × 10

0.14

0.50

-0.16

0.25

-0.24

0.07

0.39

0.07

-0.16

0.49

-0.36

1.6 × 10-2

ASUC_RS02780 Fructose-bisphosphate aldolase

0.41

0.06

0.38

0.17

-0.12

0.68

0.52

1.3 × 10

ASUC_RS03500 Triosephosphate isomerase

0.02

0.97

0.01

0.99

-0.15

0.39

-0.15

0.34

-0.18

0.26

ASUC_RS05330 Glyceraldehyde 3-phosphate dehydrogenase

0.71

4.8 × 10-7

0.73

3.1 × 10-7

-0.10

0.70

1.02

5.4 × 10-9

1.06

1.7 × 10-9

0.29

0.13

ASUC_RS02775 Phosphoglycerate kinase

0.51

2.2 × 10-3

0.45

1.7 × 10-2

-0.21

0.33

0.58

1.5 × 10-3

0.46

1.4 × 10-2

-0.20

0.33

ASUC_RS02990 Phosphoglycerate mutase

0.34

2.1 × 10

-2

0.36

2.2 × 10

-2

-0.11

0.59

0.89

1.6 × 10

-9

0.82

2.9 × 10

-8

0.36

2.2 × 10-2

ASUC_RS10600 Enolase

0.37

4.4 × 10-2

0.39

4.5 × 10-2

-0.15

0.54

0.82

3.6 × 10-6

0.74

3.2 × 10-5

-0.09

0.66

82

Table S2.1 (cont’d).
Fermentative pathways
ASUC_RS01160 Phosphoenolpyruvate carboxykinase

0.87

1.3 × 10-8

0.72

1.3 × 10-5

-0.48

1.3 × 10-2

1.51

1.0 × 10-14

1.34

1.1 × 10-11

0.15

0.51

ASUC_RS06150 Pyruvate kinase

-0.33

0.06

-0.22

0.49

-0.12

0.59

0.15

0.39

0.29

0.08

0.24

0.16

-2

-3

ASUC_RS01590 Oxaloacetate decarboxylase (alpha)

-0.15

0.60

-0.15

0.74

0.01

0.98

-0.42

1.7 × 10

-0.45

9.4 × 10

-0.25

0.17

ASUC_RS01595 Oxaloacetate decarboxylase (beta)

-0.17

0.59

-0.20

0.68

0.02

0.96

-0.56

3.4 × 10-3

-0.64

7.4 × 10-4

-0.40

4.3 × 10-2

ASUC_RS01600 Oxaloacetate decarboxylase (gamma)

-0.04

0.92

-0.18

0.72

0.01

0.97

-0.74

8.2 × 10-5

-0.66

5.0 × 10-4

-0.19

0.37

1.63

-12

1.54

-11

0.46

0.06

-4

-0.66

2.0 × 10-5

ASUC_RS08350 Malate dehydrogenase

1.35

6.6 × 10

-16

1.22

1.4 × 10

-12

-0.25

0.40

4.5 × 10

ASUC_RS03460 Malic enzyme

-0.13

0.62

-0.18

0.61

0.09

0.68

-0.57

2.0 × 10

ASUC_RS04965 Fumarase

-0.06

0.91

0.05

0.95

-0.20

0.46

0.34

ASUC_RS09390 Fumarate reductase (flavoprotein)

-0.09

0.82

-0.08

0.92

-0.12

0.65

ASUC_RS09395 Fumarate reductase (iron-sulfur)

-0.12

0.81

-0.05

0.96

-0.09

ASUC_RS09400 Fumarate reductase (C)

-0.02

0.97

0.00

1.00

ASUC_RS09405 Fumarate reductase (D)

0.10

0.86

0.06

0.95

ASUC_RS01090 Pyruvate-formate lyase

-0.17

0.63

-0.09

0.91

5.1 × 10

0.14

0.29

0.21

0.26

0.27

-0.28

0.16

-0.30

0.13

-0.08

0.73

0.82

-0.22

0.42

-0.22

0.42

0.03

0.91

-0.04

0.94

-0.15

0.60

-0.03

0.93

0.31

0.25

0.09

0.82

0.02

0.95

0.29

0.31

0.43

0.12

-0.40

0.06

-0.19

0.43

0.31

0.12

0.40

4.1 × 10

-2
-2

4.2 × 10

-2

0.47

1.5 × 10

-2

-0.44

0.03

0.95

0.11

0.87

0.38

2.8 × 10

ASUC_RS06620 Formate dehydrogenase (beta subunit, iron-sulfur)

-0.13

0.79

-0.04

0.96

0.19

0.53

0.21

ASUC_RS06625 Formate dehydrogenase (alpha subunit)

-0.08

0.86

0.01

0.99

0.20

0.43

0.34

7.3 × 10

-14

5.0 × 10

-12

1.2 × 10

-13

ASUC_RS04895 Pyruvate dehydrogenase (E2)

-0.09
-0.10

0.81
0.79

-0.03
-0.04

0.97
0.96

5.8 × 10

-0.54

ASUC_RS06615 Formate dehydrogenase (gamma subunit)

ASUC_RS04890 Pyruvate dehydrogenase (E1)

-4

1.03
0.97

-2

0.40

3.9 × 10

0.43

0.09

0.75

0.08

0.80

0.12

0.21

0.37

0.18

0.46

-0.18

0.38

-0.21

0.30

0.24

0.25

-0.17

0.41

-0.23

0.27

0.26

0.20

ASUC_RS08605 Phosphate acetyltransferase

-0.10

0.79

-0.01

0.99

0.98

-0.01

0.97

-0.08

0.70

0.34

0.07

ASUC_RS08600 Acetate kinase

-0.44

1.2 × 10-2

-0.33

0.16

0.13

0.59

-0.93

3.5 × 10-8

-0.92

5.5 × 10-8

-0.73

2.0 × 10-5

ASUC_RS03080 Alcohol dehydrogenase

-0.38

0.07

-0.33

0.25

0.57

2.0 × 10-4

-0.72

1.1 × 10-4

-0.77

3.3 × 10-5

-0.33

0.10

83

Table S2.2. Xylose isomerase activity in the wild-type and the X strains. Results are the
averages ± standard deviations of three independent biological replicates.
Xylose isomerase activity
(µmol xylulose produced mg protein-1 min-1)
Strain

Growth on glucose

Growth on xylose

130Z

0.084 ± 0.004

0.125 ± 0.009

X1

0.112 ± 0.032

0.133 ± 0.046

X2

0.059 ± 0.014*

0.133 ± 0.016

X3

0.053 ± 0.008**

0.128 ± 0.004

* Significantly different from 130Z (p < 0.05)
** Significantly different from 130Z (p < 0.01)

Table S2.3. Cell volumes, cell counts per unit OD660, and internal xylose concentrations for
the xylose-grown wild-type and X strains.
Internal xylose
Strain Cell volume (µm3) Cells per mL of OD660 1 culture
concentration (mM)
130Z
X1

0.204 ± 0.071
0.173 ± 0.063

7.53 × 108 ± 2.35 × 108
6.56 × 108 ± 4.53 × 108

70.0 ± 11.1
99.7 ± 34.7 (p = 0.2308)

X2
X3

0.147 ± 0.048
0.185 ± 0.059

7.70 × 108 ± 9.71 × 108
9.61 × 108 ± 1.00 × 109

96.7 ± 17.6 (p = 0.0904)
72.0 ± 8.3 (p = 0.8149)

84

Figure S2.1. Quality of the purified RNAs. (A): Control for absence of contaminating DNA in
RNA samples. Lane 1: 1-kb ladder; lane 2: PCR using A. succinogenes genomic DNA; and lanes
3 to 5, PCRs using one biological replicate each of RNA samples from glucose-grown strains
X1, X2, and X3, respectively. PCRs used primers CV5 and CV6 and Taq polymerase. (B):
Representative RNA samples. Lane 1: 1-kb ladder; lane 2: 100-bp ladder; and lanes 3 to 5: one
biological replicate each of RNA samples from glucose-grown strains X1, X2, and X3,
respectively.
1
2
3
4
5
1
2
3
4
5

A

B

23S rRNA
16S rRNA

85

Figure S2.2: Determination of A. succinogenes xylose isomerase’s Km for xylose. Different
symbols represent three independent biological replicates, with each an average of three
technical replicates.

86

REFERENCES

87

REFERENCES

Aden, A., M. Ruth, K. Ibsen, J. Jechura, K. Neeves, J. Sheehan, B. Wallace, L. Montague, A.
Slayton, and J. Lukas. 2002. Lignocellulosic biomass to ethanol process design and
economics utilizing co-current dilute acid prehydrolysis and enzymatic hydrolysis for
corn stover. NREL Technical Report.
Ausubel, F. M., R. Brent, R. E. Kingston, D. D. Moore, J. G. Seidman, J. A. Smith, and K.
Struhl. 1993. Current Protocols in Molecular Biology. New York: Greene Publishing &
Wiley-Interscience.
Bodley, J. W., F. J. Zieve, L. Lin, and S. T. Zieve. 1969. Formation of the ribosome-G factorGDP complex in the presence of fusidic acid. Biochem. and Biophys. Res. Commun.
37:437-443.
Bozell, J. J., and G. R. Petersen. 2010. Technology development for the production of biobased
products from biorefinery carbohydrates - the US Department of Energy's "Top 10"
revisited. Green Chem. 12:525-728.
Bradfield, M. F. A., and W. Nicol. 2016. The pentose phosphate pathway leads to enhanced
succinic acid flux in biofilms of wild-type Actinobacillus succinogenes. Appl. Microbiol.
Biotechnol. 100:9641-9652
Caldon, C. E., and P. E. March. 2003. Function of the universally conserved bacterial GTPases.
Cur. Opin. Microbiol. 6:135-139.
Crutchfield, C. A., W. Lu, E. Melamud, and J. D. Rabinowitz. 2010. Mass spectrometry-based
metabolomics of yeast. In Yeast Metabolomics: Elsevier Inc., 393-426.
Dalia, A. B., E. K. McDonough, and A. Camilli. 2014. Multiplex genome editing by natural
transformation. Proc. Natl. Acad. Sci. U.S.A. 111:8937-8942.
Dische, Z., and E. Borenfreund. 1951. A new spectrophotometric method for the detection and
determination of keto sugars and trioses. J. Biol. Chem. 192:583-587.
Frank, J., H. Gao, J. Sengupta, N. Gao, and D. J. Taylor. 2007. The process of mRNA-tRNA
translocation. Proc. Natl. Acad. Sci. U.S.A. 104:19671-19678.
Guettler, M., R. Hanchar, S. Kleff, and S. Jadhav. 2014. Recombinant microorganisms for
producing organic acids. US patent 20140093925
Guettler, M. V., D. Rumler, and M. K. Jain. 1999. Actinobacillus succinogenes sp. nov., a novel
succinic-acid-producing strain from the bovine rumen. Int. J. Syst. Bacteriol. 49:207–216.

88

Hirashima, A., and A. Kaji. 1973. Role of elongation factor G and a protein factor on the release
of ribosomes from messenger ribonucleic acid J. Biol. Chem. 248:7580-7587.
Johanson, U., A. Ævarsson, A. Liljas, and D. Hughes. 1996. The dynamic structure of EF-G
studied by fusidic acid resistance and internal revertants. J. Mol. Biol. 258:420--432.
Johnson, B. K., M. B. Scholz, T. K. Teal, and R. B. Abramovitch. 2016. SPARTA: Simple
program for automated reference-based bacterial RNA-seq transcriptome analysis. BMC
Bioinformatics 17:1-4.
Joshi, R. J., B. D. Schindler, N. R. McPherson, K. Tiwari, and C. Vieille. 2014. Development of
a markerless knockout method for Actinobacillus succinogenes. Appl. Environ.
Microbiol. 80:3053-3061.
Kim, P., M. Laivenieks, J. McKinlay, C. Vieille, and J. G. Zeikus. 2004. Construction of a
shuttle vector for the overexpression of recombinant proteins in Actinobacillus
succinogenes. Plasmid 51:108–115.
Kim, T. Y., H. U. Kim, J. M. Park, H. Song, J. S. Kim, and S. Y. Lee. 2007. Genome-scale
analysis of Mannheimia succiniciproducens metabolism. Biotechnol Bioeng. 97:657-671.
Kohler, R., R. A. Mooney, D. J. Mills, R. Landick, and P. Cramer. 2017. Architecture of a
transcribing-translating expressome. Science 356:194-197.
Koripella, R. K., Y. Chen, K. Peisker, C. S. Koh, M. Selmer, and S. Sanyal. 2012. Mechanism of
elongation factor-G-mediated fusidic acid resistance and fitness compensation in
Staphylococcus aureus. J. Biol. Chem. 287:30257-30267.
Life Technologies. qPCR Education.
McKinlay, J. B., Y. Shachar-Hill, J. G. Zeikus, and C. Vieille. 2007. Determining Actinobacillus
succinogenes metabolic pathways and fluxes by NMR and GC-MS analyses of 13Clabeled metabolic product isotopomers. Metab. Engin. 9:177–192.
McKinlay, J. B., and C. Vieille. 2008. 13C-metabolic flux analysis of Actinobacillus
succinogenes fermentative metabolism at different NaHCO3 and H2 concentrations.
Metab. Engin. 10:55–68.
McKinlay, J. B., J. G. Zeikus, and C. Vieille. 2005. Insights into Actinobacillus succinogenes
fermentative metabolism in a chemically defined growth medium. Appl. Environ.
Microbiol. 71:6651-6656.
Morris, D. 2005. The carbohydrate economy, biofuels and the net energy debate. Washington,
D.C.: Institute for Local Self-Reliance, 24 p.

89

Nagaev, I., J. Björkman, D. I. Andersson, and D. Hughes. 2001. Biological cost and
compensatory evolution in fusidic acid-resistant Staphylococcus aureus. Mol. Microbiol.
40:433-439.
Nagai, K., C. Outbridge, T. H. Jessen, J. Li, and P. R. Evans. 1990. Crystal structure of the RNAbinding domain of the U1 small nuclear ribonucleoprotein A. Nature 348:515-520.
Poje, G., and R. J. Redfield. 2003. Transformation of Haemophilus influenzae. In Haemophilus
influenzae protocols, edited by M. A. Herbert: Springer Science & Business Media.
Rodnina, M. V., A. Savelsbergh, K. V. I., and W. Wintermeyer. 1997. Hydrolysis of GTP by
elongation factor G drives tRNA movement on the ribosome. Nature 385:37-41.
Shoji, S., S. E. Walker, and K. Fredrick. 2009. Ribosomal translocation: one step closer to the
molecular mechanism. ACS Chem. Biol. 4:93-107.
Song, S., and C. Park. 1997. Organization and regulation of the D-xylose operons in Escherichia
coli K-12: XylR acts as a transcriptional activator. J. Bacteriol. 179:7025-7032.
Spirin, A. S. 1985. Ribosomal translocation: facts and models. Prog. Nucl. Acid Res. Mol. Biol.
32:75-114.
van der Werf, M. J., M. V. Guettler, M. K. Jain, and J. G. Zeikus. 1997. Environmental and
physiological factors affecting the succinate product ratio during carbohydrate
fermentation by Actinobacillus sp. 130Z. Arch. Microbiol. 167:332–342.
Vieille, C., J. M. Hess, R. M. Kelly, and J. G. Zeikus. 1995. xylA cloning and sequencing and
biochemical characterization of xylose isomerase from Thermotoga neapolitana. Appl.
Environ. Microbiol. 61:1867-1875.
Wilke, D. 1995. What should and what can biotechnology contribute to chemical bulk
production? FEMS Microbiol. Rev. 16:89–100.
———. 1999. Chemicals from biotechnology: molecular plant genetics will challenge the
chemical and fermentation industry. Appl. Microbiol. Biotechnol. 52:135–145.
Zavialov, A., V. Hauryliuk, and M. Ehrenberg. 2005. Splitting of the posttermination ribosome
into subunits by the concerted action of RRF and EF-G. Mol. Cell 18:675-686.

90

Chapter 3

An Actinobacillus succinogenes strain evolved for fast growth on lignocellulosic sugars is
able to grow on galactose

91

3.1 Abstract
Actinobacillus succinogenes 130Z is among the best known natural succinate producers
and can grow on a variety of carbohydrates. Among those carbohydrates are the sugars
commonly found in lignocellulose hydrolysates, which means that there is an opportunity to use
A. succinogenes to convert hydrolyzed lignocellulose into succinate. Previous work with A.
succinogenes 130Z fermentation has focused on glucose, but growth on other sugars is much
slower. Not much is known about many of the sugar transporters in A. succinogenes and coconsumption of sugars has not yet been investigated. In a series of experiments growing strain
130Z on a number of sugar pairwise combinations, 130Z co-consumed glucose and xylose, but it
used xylose preferentially over arabinose, and fructose preferentially over xylose until fructose
was below about 7 mM. Strain 130Z was submitted to a series of directed evolution experiments
to produce strains that grow more quickly on xylose, arabinose, galactose, cellobiose, and
AFEX-pretreated corn stover hydrolysate. The final strain, H1, grew faster than 130Z on xylose,
arabinose, and the corn stover hydrolysate, and it was able to grow on galactose while 130Z
cannot. Additionally, many of the evolved strains produced a higher yield of succinate than wildtype A. succinogenes. Resequencing the evolved strains identified a number of mutations that
have been useful in further characterizing the strains. These mutant strains lay a foundation for
future investigation of sugar transporters and for the use of A. succinogenes for succinate
production from mixed-sugar feedstocks.

92

3.2 Introduction
The environmental effects of petroleum processing, a dwindling supply of oil, and
sometimes unpredictable oil prices have made biological processes an appealing alternative, and
development of these alternatives has accelerated (Wilke 1995, 1999). It is important that
feedstocks for these bioprocesses do not compete for food though, as competing with food makes
feedstocks less sustainable and more expensive. Rather than using starch and oils from food
crops, an alternative would be using a lignocellulosic feedstock, like corn stover. Lignocellulosic
feedstocks would improve sustainability, and would also require much less acreage to fill
demand.
An example of a highly desirable chemical with prospects of biological production is
succinate. Succinate is currently produced from maleic anhydride as a petroleum product, but the
conversion could be reversed and the market for succinate could expand into a commodity
market with demand for up to 25 million tons per year (Bozell and Petersen 2010). For biobased
succinate to be cost-competitive with maleic anhydride, though, succinate must not just be
produced renewably, but in high yield as well, and every production step must be optimized.
Actinobacillus succinogenes, a facultatively anaerobic, capnophilic member of the
Pasteurellaceae, is a very good candidate as a biocatalyst for succinate production. It produces up
to 110 g/L succinate in some conditions, and grows on about twenty different carbohydrates,
which is important for growing on mixed-sugar feedstocks. The sugars most commonly found in
cellulose and hemicellose, glucose, mannose, xylose, and arabinose, are among the sugars that A.
succinogenes can use (Guettler et al. 1999). The sugar composition of corn stover is primarily
glucose and xylose (37% and 21%, respectively) with arabinose, galactose, and mannose present
in lower amounts (3%, 2%, and 2%, respectively) (Aden et al. 2002). A. succinogenes can grow

93

on all these sugars, except galactose, but its doubling time on glucose (1.69  0.06 h) is much
faster than on the other sugars.
Little is known about sugar metabolism in A. succinogenes. In particular, the glucose
uptake mechanism has not been identified (McKinlay et al. 2010; McKinlay et al. 2007), and
how sugar co-utilization is regulated is unknown. In this chapter, I present RNA sequencing data
for glucose-, fructose-, mannose-, and xylose-grown cultures to better understand sugar uptake
pathways. I also report on sugar co-utilization experiments highlighting some sugar uptake
regulatory mechanisms. Finally, for A. succinogenes to be industrially useful, its growth rate on
hemicellulosic sugars must be increased. I serially evolved the wild-type strain for faster growth
on hemicellulosic sugars. The evolution process generated a strain that grew faster on xylose,
arabinose, and hydrolyzed lignocellulose and that was also able to grow on D-galactose. A.
succinogenes was also evolved to grow on xylose, arabinose, and cellobiose in independent,
parallel experiments, for comparison. The genomes of the serial and parallel evolved strains were
resequenced to identify the accumulated mutations.

3.3. Materials and Methods
3.3.1. Strains and media
A. succinogenes 130Z (ATCC 55618) was originally purchased from the American Type
Culture Collection and adapted to grow on medium AM3 (McKinlay et al. 2005). AM3 is a
phosphate-based chemically defined medium that contains ammonium chloride, vitamins,
minerals, cysteine, methionine, glutamate, NaHCO3, and kanamycin, as well a carbon source
(McKinlay et al. 2005). Cultures were grown anaerobically in 10 mL of AM3 at 37 °C with
shaking at 250 rpm unless stated otherwise. When a rich medium was required, cultures were
grown in liquid medium B, which is a phosphate buffer-based medium containing yeast extract,

94

bactotryptone, and glucose (Joshi et al. 2014) or on LB agar plates containing 25 mM glucose,
0.6 mg/mL polyvinyl alcohol, 10 µg/mL calcium chloride. Forty µL of 40 mg/mL 5-bromo-4chloro-3-indolyl-β-D-galactopyranoside (X-Gal) were spread on plates used to test LacZ
expression. Plates were incubated at 37 °C in stainless steel canisters under a CO2-enriched
atmosphere. All chemicals were purchased from Sigma-Aldrich (St. Louis, MO).

3.3.2. RNA purification and sequencing
RNA purification and sequencing were performed as in Chapter 2. Primers in Table 3.1
were used to test for genomic DNA contamination. Of the 48 sequenced RNA samples, results
from 21 samples are reported here: those from the wild-type strain grown on glucose (twelve
samples), fructose (three samples), mannose (three samples), and xylose (three samples).

3.3.3. Growth curves and fermentation balance analyses
Growth curves were generated by tracking absorbance at 660 nm using a Beckman
DU650 spectrophotometer (Beckman Coulter, Inc., Indianapolis, IN). Fermentation balance
analyses were performed by high performance liquid chromatography (HPLC) as described
(McKinlay et al. 2005) using supernatant samples collected at two points during exponential
growth. Sugar concentrations for co-utilization experiments were quantified using the above
HPLC method as well.

3.3.4. Strain evolution
A. succinogenes 130Z was serially evolved on D-xylose, L-arabinose, D-galactose,
cellobiose, and AFEX-pretreated corn stover hydrolysate (supplied by the Bruce Dale lab,

95

Michigan Biotechnology Institute, Lansing, MI) as illustrated in Figure 3.1, series 1. Two
additional independent evolution experiments were done each on xylose, arabinose, or cellobiose
for later comparison (Figure 3.1). During evolution, cultures of A. succinogenes were grown
anaerobically in AM3 with 100 mM NaHCO3 and sugar: 60 mM D-xylose or L-arabinose, 50
mM D-galactose, 25 mM cellobiose, or 10-40% v/v hydrolysate. These sugar concentrations,
including 10% v/v hydrolysate, were chosen so that all media had the same initial carbon
molarity. The hydrolysate concentration was increased to 20% after 14 transfers and 40% after
24 transfers to gradually increase the concentration of growth inhibitors. Cultures (0.25 ml
inoculum) were transferred to fresh medium during the exponential growth phase (OD660
between 0.5 and 1.5). Evolution was considered complete when the doubling time of the culture
plateaued. An example of the reduction in doubling time over the course of an evolution series is
shown in Figure 3.2.

3.3.5. Sequencing of the evolved strains
Isolated colonies from the evolved cultures, obtained by two consecutive streakings of the
evolved cultures on LB-glucose plates, were sequenced by the Department of Energy’s Joint
Genome Institute (JGI). Genomic DNA was prepared from the A. succinogenes clonal evolved
cultures grown in AM3-glucose using the Promega Wizard Genomic DNA extraction kit. DNA
concentration was determined using a JGI Genomic DNA Mass Standard Kit. For each strain
sequenced, the JGI provided the list, genome location, and nature of all mutations. NCBI
accession numbers for the evolved strains are X1: PRJNA337266, X2: PRJNA337281, X3:
PRJNA337282, A1: PRJNA337258, A2: PRJNA337259, A3: PRJNA337260, G1:

96

PRJNA337264, C1: PRJNA337261, C2: PRJNA337262, C3: PRJNA337263, and H1:
PRJNA337265. See Figure 3.1 for strain designations.
When necessary, mutations were confirmed by amplifying the mutated region using
primers purchased from Integrated DNA Technologies (Coralville, IA) and Sanger sequencing
performed by the Michigan State University Research Technology Support Facility or Genewiz
(South Plainfield, NJ).

3.3.6. Transmembrane helix prediction
Transmembrane helices in LacY were predicted using the Transmembrane Hidden Markov
v. 2.0 procedure (Sonnhammer et al. 1998; Krogh et al. 2001) available at
http://www.cbs.dtu.dk/services/TMHMM/.

3.4 Results
3.4.1 Glucose, mannose, fructose, and xylose assimilation pathways
The glucose transporter has not yet been identified in A. succinogenes. The genome
sequence (McKinlay et al. 2010) and metabolic flux analyses (McKinlay et al. 2007) indicate
that glucose is not transported through a phosphotransferase system (PTS) transporter. Transcript
data from A. succinogenes grown on glucose, fructose, xylose, and mannose were compared to
identify sugar transport genes that are specifically upregulated in glucose-grown cultures. The
genes encoding the maltose ABC transport system were highly upregulated in glucose-grown
cultures, compared to fructose-, mannose-, and xylose-grown cultures (Table 3.2). In E. coli,
MalFGK form the membrane-bound translocation complex of the ABC transporter, with MalFG
as the intrinsic membrane proteins and MalK the ATPase, while MalE is the periplasmic

97

maltose-binding protein. LamB is the maltoporin, and MalM is a periplasmic protein associated
with maltose, with yet no known function (Boos and Shuman 1998). One A. succinogenes gene
is annotated as malE, while another immediately downstream also has a best match to E. coli
malE (61% identity, 79% similarity in a protein BLAST). All of the other maltose uptake genes
are present and annotated in A. succinogenes. All of these genes have transcript levels 18- to 88fold higher during growth on glucose than during growth on fructose (Table 3.2), and are
similarly upregulated when compared to mannose or xylose cultures (data not shown). These
results suggest that the maltose transport system is used for glucose uptake in A. succinogenes.
The genes encoding mannose-specific PTS transport proteins did not show increased
transcript levels during growth on mannose compared to growth on glucose (Table 3.2).
Similarly, the mannose-6-phosphate isomerase-encoding gene was not upregulated on mannose.
During growth on fructose, the genes encoding the two annotated fructose-specific PTS
transport proteins as well as 1-phosphofructokinase were upregulated by approximately 50%
compared to glucose-grown cultures. Even though these genes were not highly upregulated,
differences in transcript levels were highly statistically significant.
During growth on xylose, the genes encoding the xylose ABC transporter system
components were upregulated 130- to 514-fold and the xylose-proton symporter-encoding gene
was upregulated 53-fold. The genes encoding xylose isomerase and xylulokinase were
upregulated 613- and 184-fold as well. All transcript level increases were highly significant, with
false discovery rates reported as zero by the SPARTA pipeline.
The HPr protein and enzyme E I are necessary for PTS transport, so transcript levels for
the genes encoding these proteins were compared across the four growth conditions. In glucosegrown cultures, the transcript levels of these two genes are approximately 29 times lower than

98

those of the most highly expressed gene in A. succinogenes and approximately 7 times lower
than that of PEPCK, which is the ninth-most transcribed gene (Supplemental Table 3.1).
Transcript levels were unchanged in fructose- and xylose-grown cultures (Table 3.2). Mannosegrown cultures showed 50% lower and 42% lower transcript levels for the HPr- and EI-encoding
genes, respectively, but still maintained transcript levels high enough that sufficient protein may
be produced for PTS transport. These results suggest that transcription for HPr protein and
enzyme E I is constitutive

3.4.2 Sugar co-utilization
A. succinogenes has been shown to grow on most of the sugars found in common
lignocellulosic feedstocks (Guettler et al. 1999), but very little is known regarding sugar coutilization and possible diauxic growth. Because glucose and xylose are the two major
lignocellulosic sugars, it is of particular interest to know whether A. succinogenes can use both
sugars simultaneously, or if glucose inhibits xylose utilization. The time course of sugar
consumption by A. succinogenes grown on 25 mM glucose and 30 mM xylose is shown in
Figure 3.3A. A linear fit (R2 = 0.992) to log-transformed data is shown to indicate the
exponential growth phase. Glucose and xylose were consumed simultaneously and no evidence
of a slope change for the linear fit was seen, indicating the absence of diauxic shift during growth
on these two sugars.
Because fructose seems to be the sugar that causes catabolite repression in Haemophilus
influenzae (Macfadyen et al. 1996b) and the A. succinogenes genome sequence suggests that
fructose is a PTS sugar in A. succinogenes, co-utilization of fructose and xylose was also tested
(Figure 3.3B). While fructose was consumed first, xylose consumption started before fructose

99

was completely exhausted. The xylose consumption rate increased once the fructose
concentration reached about 7 mM. At this point, though, the cultures rapidly reached stationary
phase and a clear diauxic shift could not be observed. A strict repression of xylose consumption
was not seen and co-utilization did occur, although the majority of the sugar consumed was
fructose until it was at a low concentration.
During E. coli growth on arabinose and xylose, arabinose is consumed first (Desai and Rao
2009). A. succinogenes did not co-utilize arabinose and xylose (Figure 3.3C). Xylose was
consumed first, and a small amount of arabinose was consumed at the beginning of the stationary
phase.
A. succinogenes did not consume the galactose in cultures grown on xylose plus galactose
(data not shown).

3.4.3 Growth rates of evolved strains
Compared to the parental, wild-type strain, evolved strains showed increased growth rates when
grown with the sugar on which they were evolved, and some of them showed slightly decreased
growth rates on glucose (Table 3.3). By the end of the evolution series, strain H1 grew at a rate
comparable to that of the wild-type strain on glucose and at rates comparable to those of the
other evolved strains on each other sugar, except on arabinose and cellobiose. Strain H1 still
grew faster than the wild-type strain on arabinose (5.25 h vs. 14.1 h). Of interest, H1 grew on
galactose, where the wild-type strain did not, even though the genes for galactose assimilation
are present in A. succinogenes (McKinlay et al. 2010). Growth on galactose was first observed
after evolution on arabinose, though strain A1 is not able to grow on galactose.

100

The wild-type and H1 strains were grown on AFEX-treated corn stover hydrolysate
(Figure 3.4). Strain H1 grew faster, with a doubling time of 1.42 ± 0.06 vs 1.82 ± 0.04 for the
wild-type strain (p = 0.0008). Wild-type 130Z consumed glucose and xylose simultaneously and
did not appear to consume the small quantities of arabinose or galactose over the course of the
culture. Strain H1, however, consumed glucose and did not consume the xylose over the course
of the culture. It may also consume the galactose, but it was difficult to quantify the exact
amount of galactose consumed, because the concentration was very low.

3.4.4 Fermentation balances of the evolved strains
Fermentation balances of the evolved strains were compared to those of the wild-type
strain during growth on glucose (Table 3.4). In strains X1, X2, and X3 grown on glucose, the
amount of succinate produced increased, the succinate to acetate ratio increased and biomass
production was lower. Increased succinate yield and productivity was observed in many of the
evolved strains, but not in strain H1, where it fell back to wild-type levels. The decreased
biomass production was observed in nearly every evolved strain, however.

3.4.5 Mutations in the evolved strains
Mutations accumulated in the evolved strains are listed in Table 3.5. Shortly after
resequencing was completed it was discovered that the sequencing results for strains A1 and A3
were identical. PCR and Sanger sequencing confirmed the results for strain A1 but not A3, so no
data are shown for strain A3.
A total of 54 mutations were discovered in the eleven evolved strains. Fifteen were
insertions or deletions and 38 were single nucleotide polymorphisms (SNPs). Forty-six mutations

101

were in protein-coding regions. Many mutations fall into one of two categories: sugar
transport/metabolism or general cell function (i.e., DNA/RNA synthesis, cell division). For
example, strain X1 has three mutations in non-coding regions (including upstream of xylE), two
in hypothetical proteins, and the remainder in genes unrelated to xylose metabolism (e.g.,
encoding elongation factor EF-G and a ribonuclease). Strain G1, on the other hand, has
mutations in general-function genes (e.g., in a DNA-directed RNA polymerase) as well as in
galactose-specific genes (i.e., encoding galactose mutarotase and a sugar transporter).
A few genes were mutated in different positions across multiple strains as well. For
example, fusA, encoding elongation factor EF-G was mutated in five independent locations in
strains X1, X2, X3, A1, and G1. Strain H1 was the product of a sequential evolution series, so 17
of the 23 mutations present in strain H1 were present in earlier strains in the evolution series as
well. However, since populations were not clonal when transferring from one evolution series to
the next, several mutations (e.g., in position 447420) disappear at one point in the evolution
process and reappear later, or are no longer present in some of the final strains.

3.4.6 Galactose assimilation
Galactose is often present in lignocellulosic materials, although most often in low
amounts. Still, an organism that consumes most lignocellulosic sugars would be beneficial for
lignocellulose hydrolysate-based fermentations. Acid production had been observed when A.
succinogenes was incubated on galactose (Guettler et al. 1999), but A. succinogenes typically
does not grow on galactose. Starting with the mixed population produced after evolution on Larabinose, A. succinogenes started being able to grow on galactose, though.

102

A. succinogenes has the full Leloir pathway for galactose assimilation (Figure 3.5)
(McKinlay et al. 2010), including galU, which encodes UTP-α-glucose-1-phosphate
uridylyltransferase, an enzyme essential for galactose utilization in E. coli. The A. succinogenes
pathway differs from the E. coli pathway by seemingly lacking the galactose-specific permease
GalP, the galactose repressor protein GalR, and the repressor (GalS) of the galactose/galactoside
ABC transporter operon. In the absence of GalP and GalS, D-galactose might not be an inducer
for A. succinogenes’s galactose/galactoside ABC transporter. Strains A1, G1, C1, and H1 contain
mutations in the lac operon (Figure 3.6). In E. coli, D-galactose can be taken up through the
lactose permease (Ganesan and Rotman 1966), so mutations in the lactose operon that deregulate
expression might allow galactose uptake. Two mutations in strain A1, a 2-bp deletion in the
transcriptional repressor gene, lacI, and a transition in the permease gene, lacY (mutations
1548338 and 1550511, Figure 3.6), could pave the way for growth on galactose, but these two
mutations are not sufficient for growth on galactose since strain A1 does not grow reliably on
galactose. Strain G1 contains a 4-bp insertion (mutation 1549341) in the 231-nt lacI–lacY
intergenic region, but a 1,461-bp deletion (mutation 1548009) in strains C1 and H1 deletes lacI,
as well as the entire intergenic region and the first 19 bp of lacY, plus it removes mutations
1548338 and 1548009. Strains C1 and H1 still grow on lactose (data not shown), so either lacY is
still expressed in a truncated, functional form, or LacY is not the only lactose transporter in A.
succinogenes. The 1,461-bp deletion places the remainder of lacY immediately downstream of
stop codons in all three reading frames (Figure 3.7) with no sequence resembling a ribosome
binding site in the upstream region. The ATG codon encoding Met52 is preceded by a sequence
(GGGTG) that might be able to function as a ribosome binding site. A truncated LacY missing
its 51 N-terminal residues would have lost its first transmembrane helix and first external loop

103

(Figure 3.7), but the permease’s N-terminal transmembrane helix does not seem to contain any
residues critical to the transport function (Kaback 2005). Strains C1 and H1 form blue colonies
on LB X-gal plates (not shown), indicating that LacZ is still expressed and that X-Gal can get
into the cell. Thus it is not completely clear at this point which mechanism strains C1 and H1 use
for lactose and galactose uptake, but it could be through a truncated LacY.
Two mutations potentially related to growth on galactose are present in strain G1.
Mutation 2090926, a SNP in the galactose mutarotase gene, could increase galactose mutarotase
activity, which may relieve a bottleneck very early in the galactose pathway. Mutation 1993108,
a SNP upstream of the UDP-glucose 4-epimerase gene, could decrease repression or activate
transcription of that gene, increasing the amount of UDP-glucose produced from UDP-galactose
late in the galactose pathway, and maybe pushing flux through hexose 1-phosphate
uridylyltransferase (Figure 3.6). Because this SNP is 185 nt upstream of galE, in a 224-nt noncoding region with a 71% AT content, it is impossible to identify promoters or regulatory
sequences based on sequence information alone. Note that strains C1 and H1 contain an
additional SNP in galU, encoding UTP-glucose-hexose-1-phosphate uridylyltransferase (Table
3.5). This mutation is clearly not inactivating GalU, but it might not affect galactose assimilation,
since strain H1 grows as well on galactose as strain G1 (Table 3.3).

3.5 Discussion
I used RNA sequencing to identify the main sugar uptake pathways in A. succinogenes.
The upregulation of all genes encoding the maltose ABC transport system exclusively in
glucose-grown cultures suggests that glucose uptake likely takes place through the maltose ABC
transport system. This result is not highly surprising, since no glucose-specific PTS system could

104

be found in the genome sequence (McKinlay et al. 2010), and metabolic flux analyses indicated
very little if any flux through a PTS system (McKinlay et al. 2007). This result is also aligned
with the absence of catabolite repression by glucose during the glucose-plus-xylose co-utilization
experiment, and with results in H. influenzae as well (Macfadyen et al. 1996). This result is
beneficial for growth on mixed-sugar substrates, such as lignocellulose hydrolysates, and should
be an advantage when using A. succinogenes for succinate production from mixed-sugar
feedstocks.
As expected from the genome sequence, the genes encoding the fructose PTS system and
the genes encoding the xylose ABC transporter and xylose-proton symporter were upregulated in
fructose- and xylose-grown cultures, respectively. That fructose uptake takes place through a
PTS system is aligned with fructose having an inhibitory effect on xylose uptake in the fructoseplus-xylose co-utilization experiment, and with observations of catabolite repression by fructose
in H. haemophilus as well (Macfadyen et al. 1996). In contrast, transcript levels of annotated
mannose transport genes are not higher in mannose-grown cultures. These results lead us to
hypothesize that mannose use may be constitutive in A. succinogenes, but further testing would
be required to confirm that hypothesis. Transcript levels of the genes encoding mannose
transporters are lower than the xylose, fructose, or maltose transporters when grown on xylose,
fructose, or glucose, respectively (Table S3.1), but may still be high enough to maintain
sufficient translation for growth on mannose.
In the xylose-arabinose co-utilization experiment, 130Z consumed xylose before
arabinose. In E. coli, AraC, the arabinose regulator protein, represses transcription of xyl genes
(Desai and Rao 2009), and XylR represses transcription of ara genes (Koirala et al. 2016). Desai
and Rao showed that xylose use was repressed in cultures induced with arabinose and grown

105

with 0.5-1.0 µM arabinose. Vice-versa, arabinose use was repressed in cultures induced with
xylose grown with 0.5-1.0 µM xylose. They also found that xylose-driven repression of
arabinose use was less strict. It is unknown whether A. succinogenes follows a similar pattern
and whether inducing a culture with arabinose would cause it to consume arabinose first in a
xylose-arabinose co-utilization experiment. For the purpose of growing A. succinogenes on
lignocellulosic hydrolysates, it would be important that A. succinogenes consume xylose first,
since xylose is present in much higher quantities than arabinose in many feedstocks.
While A. succinogenes 130Z has been tested for succinate production in a variety of
conditions, this study represents the first relatively large-scale evolution experiment to increase
growth rates on lignocellulosic sugars. In addition, succinate yield increased in many of the
evolved strains, even though the only selection during evolution was for increased growth rate.
When strains 130Z and H1 were grown on AFEX-pretreated corn stover hydrolysate,
strain H1 grew faster than 130Z. 130Z consumed glucose and xylose as predicted, but H1
consumed glucose and arabinose, but not xylose. Galactose concentrations were too low in the
sample to accurately measure a change during growth. This may have been a result of evolution
on arabinose following evolution on xylose in the sequential evolution series. However, no
mutations were observed in or around the arabinose gene cluster, so it is unclear what could have
caused a regulatory change between the xylose and arabinose regulons. Alternatively, if strain
H1 were grown with xylose before growth on lignocellulose hydrolysate, it may use xylose
before arabinose. Ideally, strain H1, or a strain based on mutations found in H1, would make a
versatile strain for consumption of sugar substrates, but catabolite repression like this adds
complications.

106

Fermentation balances of the evolved strains showed increased succinate production in
many strains and decreased biomass production in all but one. It is unclear what mutations cause
the altered fermentation balance and why succinate levels fell back to wild-type levels in the H1
strain. However, succinate levels did not fall below wild-type levels in any of the evolved strains,
and were higher than wild-type levels in many strains, even though no specific selection had
been applied to maintain succinate production. Additionally, with less of the carbon from sugar
being used for biomass production more carbon ends up in fermentation products, which
translates into increased yields.
Evolved strains from G1 on are able to grow on galactose. No obvious mutations in sugar
transporters could be found in G1, but two mutations are present in genes required for growth on
galactose. It is interesting to note that uptake may not be the step at which growth on galactose is
impeded in 130Z. The SNP in the galactose mutarotase gene could affect its activity, or the SNP
upstream of the gene encoding UDP-glucose 4-epimerase may release it from repression, which
may be a key regulatory point in the pathway.
Five independent fusA mutations occurred during the directed evolution process: in
strains X1, X2, X3, A1, and G1. Even though X1, A1, and G1 were part of the same sequential
evolution series, each strain has a different fusA mutation because of how the evolution was
performed. The fusA mutations in the X strains are near each other, while the mutations in strain
A1 and G1 are elsewhere in the protein. All five mutations appear to be at the interface between
domain I and domains IV/V, though, which extends our hypothesis put forth in Chapter 2: that
mutations at the interface between domain I and domains IV/V are responsible both for increased
fusidic acid resistance and increased translation rates. Fusidic acid resistance has yet to be tested
in strains A1 and G1, but mutations in the same regions of domain I and domains IV/V in

107

Salmonella typhimurium EF-G have been shown to increase fusidic acid resistance (Johanson et
al. 1996).
While it is not completely clear why the evolved strains of A. succinogenes discussed
here grow faster on lignocellulosic sugars, it is clear that they do. Additionally, some evolved
strains are able to grow on galactose, while the wild-type strain cannot. More research is required
to discover the regulatory interaction between arabinose and xylose use, as well as the
transporters used to take up galactose, but these strains and the mutations identified have shed
some light on these mechanisms. Additionally, the ABC transporter annotated as maltosespecific was identified as being used for glucose uptake. Combined with work being done to
identify succinate transporters in A. succinogenes, this work will be useful in the future to
construct a versatile strain of A. succinogenes for succinate production from mixed-sugar
feedstocks.

108

APPENDIX

109

APPENDIX
Chapter 3 Figures and Tables

Table 3.1. Primers used in this study
Primer name

Sequence

Use

CV1

TGCGTTACAACCCTGAAACA

frdB forward

CV2

TCTTTCGCACTTTCCAGCTT

frdB reverse

110

Table 3.2. Differential expression of sugar transport genes in wild-type A. succinogenes
grown on glucose, fructose, mannose, and xylose. Differential expression values for each set
of genes were calculated by comparing transcript levels of cultures grown on the two the
described sugars. Differential expression is conveyed as log2-fold differential expression. False
discovery rates (FDR) were calculated using the Benjamini-Hochberg procedure.
Locus tag
Mannose vs glucose

Gene product

Differential
expression

ASUC_RS01515

Mannose-specific PTS transporter (IIAB subunit)

-0.10

0.61

ASUC_RS01520

Mannose-specific PTS transporter (IIC subunit)

0.13

0.50

ASUC_RS01525

Mannose-specific PTS transporter (IID subunit)

0.30

0.10

ASUC_RS04850

Mannose-6-phosphate isomerase

0.29

0.05

ASUC_RS05155

HPr protein

-1.00

3.19 × 10-12

ASUC_RS05160

Enzyme E I

-0.79

1.76 × 10-8

ASUC_RS00470

Fructose-specific ABC transporter (IIA subunit)

0.61

3.20 × 10-8

ASUC_RS00460

Fructose-specific ABC transporter (IIBC subunit)

0.56

7.80 × 10-10

ASUC_RS00465

1-phosphofructokinase

0.60

2.00 × 10-8

ASUC_RS05155

HPr protein

0.06

0.68

ASUC_RS05160

Enzyme E I

0.15

0.27

ASUC_RS01680

Maltooligosaccharide binding protein (possible MalE)

5.82

5.65 × 10-57

ASUC_RS01675

MalE, maltose ABC transporter (periplasmic binding subunit)

5.20

2.46 × 10-52

ASUC_RS01670

MalF, maltose ABC transporter (intrinsic membrane subunit)

4.20

2.67 × 10-56

ASUC_RS01665

MalG, maltose ABC transporter (intrinsic membrane subunit)

4.17

3.49 × 10-69

ASUC_RS01685

MalK, maltose ABC transporter (ATPase subunit)

6.45

3.90 × 10-59

ASUC_RS01695

LamB, maltoporin

6.33

1.31 × 10-95

ASUC_RS01700

MalM, periplasmic, function unknown

5.34

1.04 × 10-79

ASUC_RS02555

XylE, xylose H+ symporter

5.73

0

ASUC_RS02565

XylF, xylose ABC transporter (periplasmic binding subunit)

7.05

0

ASUC_RS02570

XylG, xylose ABC transporter (intrinsic membrane subunit)

8.40

0

ASUC_RS02575

XylH, xylose ABC transporter (ATPase subunit)

9.01

0

ASUC_RS02580

XylA, xylose isomerase

9.26

0

ASUC_RS02585

XylB, xylulokinase

7.52

0

ASUC_RS05155

HPr protein

-0.30

0.03

ASUC_RS05160

Enzyme E I

0.00

0.99

FDR

Fructose vs glucose

Glucose vs fructose

Xylose vs glucose

111

Table 3.3. Doubling times of the wild-type (WT) and evolved A. succinogenes strains on
lignocellulose sugars. Strains ending with 1 are part of the serial evolution scheme. Doubling
times, in hours, are the averages of three biological replicates with standard deviations.
Strain

Glucose

Xylose

Arabinose

Galactose

Cellobiose

WT

1.69 ± 0.06

4.20 ± 0.24

14.1 ± 0.12

No growth

5.85 ± 0.80

X1

1.81 ± 0.02*

2.33 ± 0.10**

ND a

ND

ND

X2

2.54 ± 0.06**

3.22 ± 0.24**

ND

ND

ND

X3

2.03 ± 0.02**

3.48 ± 0.12**

ND

ND

ND

A1

1.57 ± 0.04*

ND

2.90 ± 0.04**

ND

ND

A2

1.86 ± 0.00**

ND

2.96 ± 0.01**

ND

ND

A3

2.14 ± 0.00**

ND

3.30 ± 0.13**

ND

ND

G1

2.01 ± 0.00**

ND

ND

5.10 ± 0.38

ND

C1

1.85 ± 0.09

ND

ND

ND

1.85 ± 0.04**

C2

1.62 ± 0.04

ND

ND

ND

2.07 ± 0.02**

C3

1.67 ± 0.05

ND

ND

ND

2.03 ± 0.05**

H1

1.41 ± 0.08**

2.70 ± 0.12**

4.46 ± 0.02**

5.25 ± 0.09

5.43 ± 0.06

a

ND: doubling time not determined
* Significantly different (p < 0.05) from wild-type
** Significantly different (p < 0.01) from wild-type

112

Table 3.4. Fermentation balances of glucose-grown wild-type (WT) and evolved strains.
Strains ending with 1 are part of the serial evolution scheme. All values are averages of three
biological replicates with standard deviations. Carbon recoveriesb were all between 0.96 and 1.2.
Products
(mmol/100 mmol glucose consumed)
Succinate

Formate

Acetate

Biomass a

Succinate
productivity
(g L-1 h-1)

WT

58.5 ± 2.8

96.9 ± 4.3

67.8 ± 3.1

212.8 ± 4.0

0.520 ± 0.018

0.86 ± 0.06

X1

76.2 ± 2.9**

83.6 ± 3.4*

69.6 ± 2.6

187.7 ± 0.7**

0.593 ± 0.017**

1.09 ± 0.06

X2

70.9 ± 7.1*

90.4 ± 9.1

73.9 ± 7.4

149.4 ± 0.9**

0.372 ± 0.003**

0.96 ± 0.14

X3

74.4 ± 4.4**

85.6 ± 5.4*

68.0 ± 4.0

156.2 ± 2.0**

0.380 ± 0.016**

1.09 ± 0.09

A1

94.2 ± 40.5

98.2 ± 42.9

77.7 ± 34.7

151.4 ± 0.7**

0.744 ± 0.148

1.21 ± 0.75

A2

89.3 ± 7.8**

95.2 ± 7.2

79.9 ± 8.1

144.1 ± 23.4**

0.707 ± 0.041**

1.12 ± 0.15

A3

76.0 ± 8.7*

82.4 ± 9.9

68.1 ± 8.9

217.6 ± 0.6

0.325 ± 0.008**

1.12 ± 0.19

G1

88.5 ± 12.3*

96.6 ± 14.7

78.4 ± 12.2

145.9 ± 1.0**

0.460 ± 0.022*

1.13 ± 0.24

C1

79.5 ± 11.8*

95.2 ± 15.0

73.0 ± 11.5

97.0 ± 1.2**

0.646 ± 0.054*

1.09 ± 0.23

C2

69.5 ± 16.4

81.6 ± 19.5

74.0 ± 17.6

169.9 ± 0.9**

0.544 ± 0.031

0.94 ± 0.31

C3

69.1 ± 6.2

73.7 ± 6.2**

69.9 ± 6.4

181.7 ± 0.6**

0.452 ± 0.031*

0.99 ± 0.13

H1

56.8 ± 6.2

70.9 ± 13.0*

61.0 ± 12.6

137.5 ± 2.6**

0.511 ± 0.043

0.93 ± 0.22

Strain

Succinate:
acetate ratio

a

Biomass production is calculated using the difference in OD660 during the growth period
measured using the equation: (567 x ΔOD660) / 24.967 as described (van der Werf et al. 1997).
b

The carbon balance is calculated using the equation:
(mM succinate × 3) + (mM formate) + (mM acetate × 2) + biomass
mM glucose consumed

A carbon balance of 1 indicates that all carbon atoms from the substrate are represented in the
products.
* Significantly different (p < 0.05) from wild-type
** Significantly different (p < 0.01) from wild-type

113

Table 3.5. Mutations identified in the evolved strains. (●): mutation present; (-): mutation absent; (+): insertion; (Δ): deletion;
(a  b): mutation changes a to b. NA: Does not apply.
Genome
position X1

Strains
X2

X3

A1

A2 G1 C1

C2

C3 H1

Locus

Protein function

DNA Mutation

Effect on
protein

315135

●

-

-

-

-

-

-

-

-

-

ASUC_RS01455

Elongation factor EF-G

GT

A643E

885247

●

-

-

-

-

-

-

-

-

-

ASUC_RS04260

Acyltransferase 3

ΔTGGACAA

Frameshift

532090

●

-

-

●

-

●

●

-

-

●

ASUC_RS02510

Nicotinamide-nucleotide adenylyltransferase

CT

M288I

532124

●

-

-

●

-

●

●

-

-

●

ASUC_RS02510

Nicotinamide-nucleotide adenylyltransferase

TA

E277V

544906

●

-

-

●

-

●

●

-

-

●

5’ of ASUC_RS02555

D-xylose proton-symporter XylE

CT

NA

700429

●

-

-

●

-

●

●

-

-

●

ASUC_RS03310

Nicotinate phosphoribosyltransferase

CA

S441Y

976810

●

-

-

●

-

●

●

-

-

●

ASUC_RS04730

Ribonuclease, Rne/Rng family

AG

Q55R

315072

-

●

-

-

-

-

-

-

-

-

ASUC_RS01455

Elongation factor EF-G

GT

T664K

544772

-

●

-

-

-

-

-

-

-

-

ASUC_RS02555

D-xylose proton-symporter XylE

GT

Y24  stop

1209875

-

●

-

-

-

-

-

-

-

-

ASUC_RS05745

Chaperone protein DnaK

+A

Frameshift

1043606

-

●

●

-

-

-

-

-

-

-

ASUC_RS05040

Transcriptional antiterminator BglG

+ TTTA

Frameshift

1046314

-

●

●

-

●

-

-

●

-

-

ASUC_RS05042

Cof-like, haloacid dehalogenase domain protein hydrolase

+ GGT

+ Thr

1174181

-

●

●

-

●

-

-

●

-

-

ASUC_RS05145

HflC protein

TG

D125A

243883

-

-

●

-

-

-

-

-

-

-

ASUC_RS01090

Pyruvate-formate lyase

GA

S377F

315046

-

-

●

-

-

-

-

-

-

-

ASUC_RS01455

Elongation factor EF-G

GT

R673S

543986

-

-

●

-

-

-

-

-

-

-

ASUC_RS02555

D-xylose proton-symporter XylE

CA

NA

543990

-

-

●

-

-

-

-

-

-

-

ASUC_RS02555

D-xylose proton-symporter XylE

ΔAAAACGGAAAGT

ΔLSVF

544006

-

-

●

-

-

-

-

-

-

-

ASUC_RS02555

D-xylose proton-symporter XylE

ΔT

Frameshift

544152

-

-

●

-

-

-

-

-

-

-

ASUC_RS02555

D-xylose proton-symporter XylE

AC

L231R

1410818

-

-

●

-

-

-

-

-

-

-

ASUC_RS05350

Formate dehydrogenase subunit FdhD

ΔTACGGAACAAT

Frameshift

316716

-

-

-

●

-

-

-

-

-

-

ASUC_RS01455

Elongation factor EF-G

AG

V116A

1175149

-

-

-

●

-

-

-

-

-

-

ASUC_RS05146

HflK protein

CT

R202H

2144477

-

-

-

●

-

-

-

●

-

-

ASUC_RS06032

Glutamate-cysteine ligase

ΔCAT

ΔM

1548338

-

-

-

●

-

●

-

-

-

-

ASUC_RS07280

LacI transcriptional regulator

ΔTA

Frameshift

447420

-

-

-

●

●

-

●

-

-

●

ASUC_RS02045

Transcriptional regulator, TetR family; AcrR protein

CA

T12N

273011

-

-

-

●

-

●

●

-

-

●

5’ of ASUC_RS01250

Putative ascorbate-specific PTS transporter

CT

NA

1530880

-

-

-

●

-

●

●

-

-

●

ASUC_RS07190

Molybdopterin synthase sulfurylase MoeB

CA

E27  stop

114

Table 3.5 (cont’d).
Genome
position X1

Strains
X2

X3

A1

A2 G1 C1

C2

C3 H1

Locus

Protein function

DNA Mutation

Effect on
protein

1550511

-

-

-

●

-

●

●

-

-

●

ASUC_RS07285

LacY lactose permease

CT

S354F

537773

-

-

-

-

●

-

-

-

-

-

ASUC_RS02530

L-arabinose transport system permease

+ TCA

+I

1043636

-

-

-

-

●

-

-

-

-

-

ASUC_RS05040

Transcriptional antiterminator, BglG

+A

Frameshift

316816

-

-

-

-

-

●

-

-

-

-

ASUC_RS01455

Elongation factor EF-G

GA

R83C

1549341

-

-

-

-

-

●

-

-

-

-

+ TTAA

NA

1607942

-

-

-

-

-

●

-

-

-

-

ASUC_RS07535

Putative cold-shock DNA-binding domain protein

ΔCG

Frameshift

530359

-

-

-

-

-

●

●

-

-

●

ASUC_RS02495

DNA-directed RNA polymerase, alpha subunit

TG

L290W

1175359

-

-

-

-

-

●

●

-

-

●

ASUC_RS05585

HflK protein

CT

R132H

1807995

-

-

-

-

-

●

●

-

-

●

ASUC_RS08515

Glutathione S-transferase domain

AG

D119D

1993108

-

-

-

-

-

●

●

-

-

●

5’ of ASUC_RS09410

UDP-glucose 4-epimerase

GA

NA

2090926

-

-

-

-

-

●

●

-

-

●

ASUC_RS09850

Galactose mutarotase

GA

D59N

62742

-

-

-

-

-

-

●

-

-

●

ASUC_RS00280

UTP-glucose-1-phosphate uridylyltransferase

AT

L15M

1004969

-

-

-

-

-

-

●

-

-

●

5’ of ASUC_RS04870

Putative mannose-specific PTS transporter

AG

Neutral

1548009

-

-

-

-

-

-

●

-

-

●

ASUC_RS07280 and
ASUC_RS07285

LacI transcriptional regulator and LacY lactose permease

Δ1461 bp

1047582

-

-

-

-

-

-

-

●

-

-

ASUC_RS05055

PTS system, beta-glucoside-specific IIABC subunit

AT

F357Y

1337504

-

-

-

-

-

-

-

●

-

-

ASUC_RS06205

Exodeoxyribonuclease V, beta subunit

GT

R949L

7008

-

-

-

-

-

-

-

-

●

-

ASUC_RS00030

Na+/K+ transporter

GT

F414L

1046142

-

-

-

-

-

-

-

-

●

-

ASUC_RS05050

Cof-like, haloacid dehalogenase domain protein hydrolase

CA

G184C

1047580

-

-

-

-

-

-

-

-

●

-

ASUC_RS05055

PTS system, beta-glucoside-specific IIABC subunit

TC

S358G

1174212

-

-

-

-

-

-

-

-

●

-

ASUC_RS05580

HflC protein

CT

A115T

412528

-

-

-

-

-

-

-

-

-

●

ASUC_RS01895

LuxR family nitrate/nitrite response regulator protein

GT

E31  stop

517380

-

-

-

-

-

-

-

-

-

●

ASUC_RS02370

30S ribosomal protein S10

GC

V57L

532007

-

-

-

-

-

-

-

-

-

●

ASUC_RS02510

Nicotinamide-nucleotide adenylyltransferase

GA

T316M

581764

-

-

-

-

-

-

-

-

-

●

ASUC_RS02745

Transcriptional regulator, TetR family; AcrR protein

ΔA

Frameshift

2088039

-

-

-

-

-

-

-

-

-

●

5’ of
ASUC_RS09840

Hexose 1-phosphate uridylyltransferase

GA

NA

5’ of ASUC_RS07280 LacI transcriptional regulator and LacY lactose permease
and ASUC_RS07285

115

Figure 3.1. Evolution scheme. Each arrow represents one set of transfers. Each strain
designation represents a population from which the strain was isolated.

116

Figure 3.2. Doubling times of evolving cultures during the transfer series that generated
strain A1. Each point is an average of three biological replicates. Error bars are standard
deviations, often too small to be seen. Strain A1 was isolated from the culture obtained after
transfer 51.

117

Figure 3.3. Co-utilization of xylose and other sugars by A. succinogenes 130Z. A: growth on
xylose and glucose; B: growth on xylose and fructose; and C: growth on xylose and arabinose.
(▲): Xylose; (■): glucose; (●): fructose; (×): arabinose; (◊): Log10 of absorbance at 660 nm. The
results represent the average of three biological replicates. R2 = 0.992, 0.999, and 0.997 for linear
regressions in panels A, B, and C, respectively. Error bars are standard deviations.

118

Figure 3.4. Growth and sugar consumption of strains 130Z and H1 on AFEX-pretreated
corn stover hydrolysate. A: Strain 130Z; B: Strain H1. (▲): Xylose; (■): glucose; (●):
galactose; (×): arabinose; (◊): Log10 of absorbance at 660 nm. The results represent the average
of three biological replicates. R2 = 0.996 and 0.991 for linear regressions in panels A and B
respectively. Error bars are standard deviations.

119

Figure 3.5. Proposed map of galactose use in A. succinogenes. 1: galactose/galactoside ABC
transporter, MglABC; 2: galactose mutarotase, GalM (ASUC_RS09850, EC:5.1.3.3); 3:
galactokinase, GalK (ASUC_RS09845, EC:2.7.1.6); 4: hexose 1-phosphate uridylyltransferase,
GalT (ASUC_RS09840, EC:2.7.7.12); 5: UDP-galactose 4-epimerase, Gal E (ASUC_RS09410,
EC:5.1.3.2); 6: phosphoglucomutase (ASUC_RS07415, EC 5.4.2.2); and 7: UTP-α-glucose-1phosphate uridylyltransferase, GalU.

120

Figure 3.6. A. succinogenes lactose operon and location of the mutations found in the series
1 evolved strains. Filled triangles: deletion 1548388 and insertion 1549341; open triangle:
transition 1550511; bracket: large deletion 1548009.

121

Figure 3.7. Effect of mutation 1548009 in strains C1 and H1 on the 5’ end of lacY. 5’ end of
lacY in the wild-type strain (A) and in strains C1 and H1 (B). The 5’ end of the lacY open
reading frame is highlighted in yellow, the ribosome binding site is highlighted in green, and the
start codon is green. The underline represents the 3’ end of the 1,461-nt deletion found in strains
C1 and H1. In B, a stop codon immediately upstream of the truncated lacY is in red. A possible
ribosome binding site (highlighted in green) and start codon (in green letters) are indicated. (C)
Transmembrane helices in LacY, predicted by TMHMM.
(A)
TCTCAGTAAAAGGAGTTAGAAATGGAAAATAATTCTGCCACTCGAGCCTATTATTTGACTAACCGTAATTACTGGTT
ATTCAGCGGCTATTTCTTTGTATTTTTCTTTATTATGGCAACCTGCTATCCGTTTCTCGGAATTTGGCTCGGTGATA
TCAACGGTTTATCGGGTGAAAATGCAGGCATGGTATTTGCCATGATGTCGTTTTTCGCACTCTGCTTTCAACCTGTG

(B)
CTAATTTCGCATAATGTATATTATGTTAAATAAATTAAATCTCGAGCCTATTATTTGACTAACCGTAATTACTGGTT
ATTCAGCGGCTATTTCTTTGTATTTTTCTTTATTATGGCAACCTGCTATCCGTTTCTCGGAATTTGGCTCGGTGATA
TCAACGGTTTATCGGGTGAAAATGCAGGCATGGTATTTGCCATGATGTCGTTTTTCGCACTCTGCTTTCAACCTGTG

(C)

122

Table S3.1. Transcript levels of transporters in A. succinogenes 130Z grown on glucose, mannose, fructose, or xylose. Levels
reported are averages ± standard deviations of normalized counts produced by the SPARTA differential gene expression analysis
pipeline.
Locus tag

Gene product

Glucose

Mannose

Fructose

Xylose

ASUC_RS05155

HPr protein

1289 ± 170

643 ± 60.2

1347.5 ± 38.1

1048 ± 63.2

ASUC_RS05160

Enzyme E I

1277.5 ± 173.5

740.1 ± 91.3

1422.1 ± 275.7

1279.5 ± 59.5

ASUC_RS01515

Mannose-specific PTS transporter (IIAB subunit)

361.2 ± 80

337.5 ± 13.2

ASUC_RS01520

Mannose-specific PTS transporter (IIC subunit)

226.2 ± 44.7

248.2 ± 31.2

ASUC_RS01525

Mannose-specific PTS transporter (IID subunit)

444.6 ± 92.6

545.6 ± 84.3

ASUC_RS04860

Mannose-specific PTS transporter (IID subunit)

235.6 ± 16.5

268.6 ± 48.8

ASUC_RS04865

Mannose-specific PTS transporter (IIC subunit)

118.9 ± 5.8

120.1 ± 9.1

ASUC_RS04870

Mannose-specific PTS transporter (IIAB subunit)

194.9 ± 10.5

229.4 ± 28.1

ASUC_RS04850

Mannose-6-phosphate isomerase

93.9 ± 3.3

114.6 ± 7.4

ASUC_RS00470

Fructose-specific ABC transporter (IIA subunit)

662.3 ± 56.2

1011.8 ± 279.4

ASUC_RS00460

Fructose-specific ABC transporter (IIBC subunit)

714.1 ± 45.6

1051.6 ± 143.1

ASUC_RS00465

1-phosphofructokinase

314.6 ± 26.3

476.5 ± 74.9

ASUC_RS01680

Maltooligosaccharide binding protein (possible MalE)

1733.8 ± 622.8

30.6 ± 2.5

ASUC_RS01675

MalE, maltose ABC transporter (periplasmic binding subunit)

2062.8 ± 590.8

56.1 ± 5.3

ASUC_RS01670

MalF, maltose ABC transporter (intrinsic membrane subunit)

974.1 ± 247.9

53.1 ± 6

ASUC_RS01665

MalG, maltose ABC transporter (intrinsic membrane subunit)

860.8 ± 170.3

47.7 ± 4.4

ASUC_RS01685

MalK, maltose ABC transporter (ATPase subunit)

2210.8 ± 837.6

25.3 ± 2.9

ASUC_RS01695

LamB, maltoporin

3230.9 ± 836.4

40.2 ± 2.2

ASUC_RS01700

MalM, periplasmic, function unknown

1517.8 ± 391.9

37.5 ± 2.5

ASUC_RS05055

PTS beta-glucoside transporter subunit IIABC

123

Table S3.1 (cont’d).
Locus tag

Gene product

Glucose

ASUC_RS02555

XylE, xylose H+ symporter

19.9 ± 1.9

1061.9 ± 7.5

ASUC_RS02565

XylF, xylose ABC transporter (periplasmic binding subunit)

7.3 ± 1.3

969 ± 53.1

ASUC_RS02570

XylG, xylose ABC transporter (intrinsic membrane subunit)

3.9 ± 0.6

1323.8 ± 82.4

ASUC_RS02575

XylH, xylose ABC transporter (ATPase subunit)

3.8 ± 0.8

1945.1 ± 77.8

ASUC_RS02580

XylA, xylose isomerase

17.1 ± 4.8

10490.3 ± 263.5

ASUC_RS02585

XylB, xylulokinase

17.2 ± 1.9

3174.6 ± 277.1

124

Mannose

Fructose

Xylose

REFERENCES

125

REFERENCES

Aden, A., M. Ruth, K. Ibsen, J. Jechura, K. Neeves, J. Sheehan, B. Wallace, L. Montague, A.
Slayton, and J. Lukas. 2002. Lignocellulosic biomass to ethanol process design and
economics utilizing co-current dilute acid prehydrolysis and enzymatic hydrolysis for
corn stover. NREL Technical Report.
Boos, W., and H. Shuman. 1998. Maltose/maltodextrin system of Escherichia coli: transport,
metabolism, and regulation. Microbiol Mol Biol Rev 62:204-229.
Bozell, J. J., and G. R. Petersen. 2010. Technology development for the production of biobased
products from biorefinery carbohydrates - the US Department of Energy's "Top 10"
revisited. Green Chem. 12:525-728.
Desai, T. A., and C. V. Rao. 2009. Regulation of arabinose and xylose metabolism in
Escherichia coli. Appl. Environ. Microbiol. 76:1524-1532.
Domalski, E. S., T. L. Jobe Jr., and T. A. Milne. 1986. Thermodynamic data for biomass
conversion and waste incineration, edited by S. T. I. Program.
Ganesan, A. K., and B. Rotman. 1966. Transport systems for galactose and galactosides in
Escherichia coli. J. Mol. Biol. 16:42-50.
Guettler, M. V., D. Rumler, and M. K. Jain. 1999. Actinobacillus succinogenes sp. nov., a novel
succinic-acid-producing strain from the bovine rumen. Int. J. Syst. Bacteriol. 49:207–216.
Johanson, U., A. Ævarsson, A. Liljas, and D. Hughes. 1996. The dynamic structure of EF-G
studied by fusidic acid resistance and internal revertants. J. Mol. Biol. 258:420-432.
Joshi, R. J., B. D. Schindler, N. R. McPherson, K. Tiwari, and C. Vieille. 2014. Development of
a markerless knockout method for Actinobacillus succinogenes. Appl. Environ.
Microbiol. 80:3053-3061.
Kaback, H. R. 2005. Structure and mechanism of the lactose permease. C. R. Biol. 328:557-567.
Koirala, S., X. Wang, and C. V. Rao. 2016. Reciprocal regulation of L-arabinose and D-xylose
metabolism in Escherichia coli. J. Bacteriol. 198:386-393.
Krogh, A., B. Larsson, G. Von Heijne, and E. L. L. Sonnhammer. 2001. Predicting
transmembrane protein topology with a hidden Markov model: application to complete
genomes. J. Mol. Biol. 305:567-580.
Macfadyen, L. P., I. R. Dorocicz, J. Reizer, M. H. Saier Jr, and R. J. Redfield. 1996. Regulation
of competence development and sugar utilization in Haemophilus influenzae Rd by a
phosphoenolpyruvate: fructose phosphotransferase system. Mol. Microbiol. 21:941-952.

126

McKinlay, J. B., M. Laivenieks, B. D. Schindler, A. A. Mckinlay, S. Siddaramappa, J. F.
Challacombe, S. R. Lowry, A. Clum, A. L. Lapidus, K. B. Burkhart, V. Harkins, and C.
Vieille. 2010. A genomic perspective on the potential of Actinobacillus succinogenes for
industrial succinate production. BMC Genomics 11:680.
McKinlay, J. B., Y. Shachar-Hill, J. G. Zeikus, and C. Vieille. 2007. Determining Actinobacillus
succinogenes metabolic pathways and fluxes by NMR and GC-MS analyses of 13Clabeled metabolic product isotopomers. Metab. Engin. 9:177–192.
McKinlay, J. B., J. G. Zeikus, and C. Vieille. 2005. Insights into Actinobacillus succinogenes
fermentative metabolism in a chemically defined growth medium. Appl. Environ.
Microbiol. 71:6651-6656.
Sonnhammer, E. L. L., G. Von Heijne, and A. Krogh. 1998. A hidden Markov model for
predicting transmembrane helices in protein sequences. In Proc. ISMB-98. 21-28.
Wilke, D. 1995. What should and what can biotechnology contribute to chemical bulk
production? FEMS Microbiol. Rev. 16:89–100.
———. 1999. Chemicals from biotechnology: molecular plant genetics will challenge the
chemical and fermentation industry. Appl. Microbiol. Biotechnol. 52:135–145.

127

Chapter 4

Development of a markerless knockout method for Actinobacillus
succinogenes

128

The work described in this chapter is part of a previously published study by Rajasi Joshi,
the primary author, Dr. Claire Vieille, and previous members of the Vieille lab.

My contribution to this report included:



Development of a solid selective medium for use with A. succinogenes



Development of a technique for introducing knockout mutations to A. succinogenes using
electroporation



Drafting parts of the manuscript related to the bullet points above



Reading and editing the entire manuscript

The article can be viewed at the following URL:
http://aem.asm.org/content/80/10/3053

129

Chapter 5

Conclusions and Future Directions

130

5.1 Introduction
Agricultural wastes and biofuel crops have the potential to be converted to a wide range
of industrially important chemicals using succinic acid as a chemical intermediate. As a
biocatalyst, A. succinogenes has been shown to be among the best natural succinate producers,
and much that was learned from its metabolism has been used to engineer other bacterial species
for succinate production. The results presented in this dissertation demonstrate the succinatecentered metabolism of A. succinogenes and A. succinogenes’s ability to adapt and evolve to
grow on a variety of sugars. Growth rates of evolved strains are faster than those of the wild-type
strain, and evolved strains were obtained that can grow on galactose. Succinate yield also
increased in many of the evolved strains. The information learned about sugar transporters and
metabolism, as well as cell metabolism in strains that make more succinate, can be useful for
future work with A. succinogenes or with other species.

5.2 Evolution of A. succinogenes for growth on lignocellulose hydrolysate sugars
5.2.1 Xylose
Wild-type A. succinogenes is able to grow on xylose, but the X strains, evolved for fast
growth on xylose, grow faster and make 20-25% more succinate. Two genes were mutated in all
three strains, xylE and fusA. When the xylE mutation from strain X2 was introduced back into
wild-type A. succinogenes, it increased the succinate yield by more than 40%, but also more than
doubled the generation time of the strain. While it is unclear why this mutation would be
maintained during evolution selecting for growth rate, it is possible that other mutations could
influence the effect that this mutation has on growth rate. The effect of the fusA mutation on
growth on xylose has not yet been directly tested, but a strain carrying the fusA mutation from

131

strain X2 has been constructed. The A1 and G1 strains have mutations in fusA as well, which
argues that the fusA mutations are at least partially responsible for increased growth rates.
RNA sequencing revealed that most glycolysis genes, as well as the phosphoenolpyruvate
carboxykinase and malate dehydrogenase encoding genes were significantly upregulated. In
contrast, the oxaloacetate dehydrogenase and malic enzyme encoding genes were downregulated
in the xylose-grown X strains. These changes in transcript levels are likely correlated with
increased growth rates on xylose and maybe related to increased succinate production as well.
The specific changes in transcript levels could not be attributed to specific mutations, but the
fusA mutation could have widespread effects on transcription rates and could be at least partially
responsible.

5.2.2 Galactose
A. succinogenes grows on a variety of sugars and sugar alcohols, but galactose is not one
of them. After evolution for fast growth on arabinose, though, a strain was evolved that could
grow on galactose. Two mutations in the first evolved strain that grew on galactose, the G1
strain, targeted loci involved in galactose assimilation: the galactose mutarotase gene and the
region upstream of the UDP-glucose 4-epimerase gene. The G1 strain had a doubling time of
about 5.1 h on galactose and this growth rate was maintained in the H1 strain, which also
contains these mutations. Galactose is typically present in lignocellulose hydrolysates in low
amounts, although it is more common in some hydrolysates than others, so having access to a
strain that can grow on galactose could be useful. Understanding how galactose is taken up by A.
succinogenes would be useful to know for applications across other species as well, so strain G1
requires some additional investigation.

132

5.2.3 Other sugars and AFEX corn stover hydrolysate
The directed evolution produced strains that grew quickly on arabinose, cellobiose, and
AFEX-pretreated corn stover hydrolysate. Strains A2 and A3 were evolved in parallel to grow
faster on arabinose and strains C2 and C3 were evolved in parallel for fast growth on cellobiose.
I have dedicated less time to analyzing the A and C strains and strain H1 than the X strains, but
there should be interesting mutations and gene expression phenotypes to uncover with additional
research, in particular in the arabinose-evolved strains, whose growth rates on arabinose
increased the most compared to the growth improvements observed on other sugars.
Strain A2 and A3 were evolved in parallel with the series 1 evolution experiment, but an
error during resequencing resulted in the genome of A1 being sequenced twice instead of A3. As
a result, nothing is currently known about mutations in strain A3. Resequencing of strain A3 is
planned soon and comparisons between the mutations in the A strains will lead to some
interesting information about arabinose assimilation in A. succinogenes.
The C strains grow much faster than the wild-type strain on cellobiose, but strain H1 lost
this ability after evolution on lignocellulose hydrolysate. H1 grows as quickly as many of the
evolved strains on individual sugars, but co-consumption has possibly become an issue. Wildtype A. succinogenes consumes xylose alongside glucose, although at a slower rate, and
consumes xylose before arabinose when grown with just those two sugars. When grown on
lignocellulose hydrolysate, wild-type A. succinogenes follows the expected order of
consumption: glucose and xylose are co-consumed and the small amounts of arabinose and
galactose are not consumed during the course of growth. Strain H1, however, consumes glucose,
and possibly arabinose and galactose, without consuming xylose. It seems possible that the
mutation upstream of xylE forces strain H1 to use the xylose ABC transporter, which is likely

133

downregulated by the presence of arabinose. What is unclear, however, is what caused the
catabolite repression priority to shift from xylose to arabinose consumption during evolution and
if it is possible to maintain fast grown on arabinose without inhibiting growth on xylose in the
presence of arabinose.
Finally, an annotated maltose-specific transport system was identified as likely
responsible for glucose uptake. It was known that A. succinogenes does not use a PTS transporter
for glucose uptake and this is the first time that non-PTS glucose transporters were identified. It
is still unclear which transporters are used for galactose and cellobiose uptake, so deeper analysis
of the mutations discovered in these strains could be useful for identifying them.

5.2.4 Discussion of the evolution experiment protocol
The directed evolution experiments resulted in a set of very interesting strains, but
improvement could definitely be made to the methods used. After each step of the sequential
evolution, the culture should have been spread on plates and colonies isolated so that each new
strain was started from a clonal population. The clonal populations would need to be grown to
verify that their growth rate matches that mixed population, as well. Additionally, some frozen
stocks were made over the course of evolution, but a stock should have been made during each
transfer. This would have allowed me to sequence cultures from frozen stocks and determine the
order in which the different mutations arose. If the reversal of the priority of arabinose and
xylose use is due to the order of those steps (xylose first, arabinose second), it would have been
better to reverse the order of those steps as well.

134

5.3 Future work
5.3.1 X strain testing and combination with other engineering
The X strains are a good baseline for future work, especially strain X2 which contains
three mutations only. Introducing the xylE mutation from strain X2 to the wild-type A.
succinogenes had a large positive effect on succinate yield and other research in the Vieille lab
could be combined with this mutation to further increase the succinate yield. After the fusA
mutation from X2 has been introduced to wild-type A. succinogenes, the two mutations could be
combined to see how much of the X strain phenotype can be replicated. The ΔlacZ mutation
should be removed if possible as well, to be left with only the desired xylE and fusA mutations.
Additionally, since strain A1 has a mutation in fusA, the X2-EFG strain should be tested on
arabinose to see if the X2 fusA mutation increases the growth rate.
Adding the ΔpflB mutation to this synthetic X2-XylE-EFG strain would likely eliminate
formate production and decrease acetate production. Work is underway to identify the primary
succinate transporters in A. succinogenes, and overexpressing a succinate transporter would
possibly further increase succinate production.

5.3.2 Further work with A, G, C, and H strains
While a majority of research time has been spent on the X strains, the A, G, C, and H
strains have their own sets of mutations accumulated during adaptive evolution that need to be
characterized. The A strains had the largest increase in growth rate on their respective sugar
compared to the wild-type and, while arabinose is not often present in large quantities in
lignocellulose, it is almost always present. The G1 strain grows on galactose, where the wildtype strain does not, even though genes for galactose import and catabolism are annotated in the

135

genome. It would be interesting if a strain that grows on galactose could be evolved directly from
wild-type A. succinogenes, although limited attempts to do so have failed so far. The wild-type
strain does grow on lactose, though, so most of the adaptation was expected to happen in
transport and possibly initial phosphorylation. However, no sugar transporter mutations were
identified in strains able to grow on galactose and the two mutations in galactose-related genes
identified in strain G1 encode proteins that are part of the known Leloir galactose pathway.
Confirming how strain G1 takes up galactose would continue to add to the knowledge of sugar
transporters in A. succinogenes. The H1 strain has an altered priority for xylose and arabinose
consumption, so it would be useful to investigate this further. Growing the strain on xylose
before lignocellulose hydrolysate may be all that is required for it to prioritize xylose
consumption. If that is not the case, it may be necessary to look more closely at accumulated
mutations to discover what caused the altered priority. In addition, RT-qPCR or RNA
sequencing of strains A1, G1, and H1 could shed more light on transcriptional changes and draw
attention to transporters or regulators whose altered expression levels are related to arabinose and
galactose consumption.

5.3.3. Continuing development of genetic tools and other methods for A. succinogenes
The multiplex natural transformation method used to introduce mutations X2-XylE and
X2-EFG into wild-type A. succinogenes has been used in other bacteria, specifically Vibrio
cholerae, but the method for use in A. succinogenes could be improved. Method optimization
could make this a very powerful engineering tool for many experiments where the FRT scar
from the selection cassette makes complex or small mutations otherwise difficult or impossible.
Some optimization has been attempted based on the amount of DNA added during

136

transformation, but more could be done. For example changing the induction and recovery times
and enriching the AM16-isocitrate selection medium could increase transformation rates.
Development of genetic tools has been an iterative and slow process in A. succinogenes, but
much of the work to improve the natural transformation protocol has been ongoing, so the
groundwork for optimization of multiplex transformation is well established.
Currently, the S. cerevisiae flippase is used to remove the isocitrate selection cassette,
and in the X2-XylE and X2-EFG strains has left an FRT site in the ΔlacZ mutation. Curing the
plasmid carrying the flippase requires growing the strain with acridine orange, a mutagen. Once
the plasmid is cured, the ΔlacZ mutation remains, which could affect growth. Instead of this
procedure, which has the possibility of introducing additional unknown mutations and leaving
the ΔlacZ mutation, the strain could be retransformed with the wild-type lacZ gene, selecting for
transformation on AM16-lactose. This method would leave only the desired mutation in the
strain and would not leave FRT sites, meaning several mutations could more easily be introduced
in sequence without risk of large scale recombination facilitated by FRT sites.
RNA extraction and sequencing were also successfully performed for the first time from
A. succinogenes in the work presented in this dissertation, with assistance from the Mulks
laboratory. This opens a door for future transcription analyses and expands the number and types
of –omics experiments that are possible. RNA sequencing of strain H1 growing on glucose,
xylose, arabinose, galactose, cellobiose, and a lignocellulose hydrolysate may make it easier to
draw conclusions about the effects of the accumulated mutations, and would allow for more
exploration of the metabolism of A. succinogenes. The transcriptome of strain H1 growing on
hydrolysate would be especially interesting, because A. succinogenes displays a strong tolerance

137

to the byproducts of lignocellulose hydrolysis and knowing which genes are upregulated in
presence of these byproducts could help in identifying genes that encode detoxifying proteins.

138