INITIAL INVESTIGATION ON XYLOSE FERMENTATION FOR 
LIGNOCELLULOSIC BIOETHANOL PRODUCTION   
 
 
 
Except where reference is made to the work of others, the work described in this thesis is 
my own or was done in collaboration with my advisory committee.  This thesis does not 
include proprietary or classified information. 
 
 
 
 
Yanli Chen 
 
 
 
Certificate of Approval:  
  
  
  
  
Christopher B. Roberts                                                              Jin Wang, Chair  
Uthlaut Professor                                                                      Assistant Professor  
Chemical Engineering                                                              Chemical Engineering  
  
  
  
 
 
  
Mario R. Eden                                                                           George T. Flowers  
Associate Professor                                                                   Dean  
Chemical Engineering                                                               Graduate School  
 
 
 
  
  
 INITIAL INVESTIGATION ON XYLOSE FERMENTATION FOR 
LIGNOCELLULOSIC BIOETHANOL PRODUCTION  
  
 
 
 
 
Yanli Chen  
  
  
  
  
  
  
A Thesis  
Submitted to  
the Graduate Faculty of  
Auburn University  
in Partial Fulfillment of the  
Requirements for the  
Degree of  
Master in Science 
 
 
 
 
 
 
Auburn, Alabama 
May 9, 2009 
INITIAL INVESTIGATION ON XYLOSE FERMENTATION FOR 
LIGNOCELLULOSIC BIOETHANOL PRODUCTION  
 
 
 
 
Yanli Chen   
  
  
Permission is granted to Auburn University to make copies of this thesis at its discretion, 
upon request of individuals or institutions and at their expense.  The author reserves all 
publication rights.  
  
  
  
  
  
  
  
  
                                                                               Signature of Author  
  
  
  
                                                                               Date of Graduation 
  
 
 
 
 
 
 
 
 
 
 
 
 
iii
 
iv
VITA 
Yanli Chen, daughter of Peiwen Chen and Ailan Song, was born on March 20, 
1978 in Weihui, Henan Province of the People?s Republic of China. She received her 
high school diploma from the First High School of Weihui in July, 1996. Ms. Chen 
entered Northwest University at Xi?an of the People?s Republic of China in the fall of 
1996, and graduated with a Bachelor of Chemical Engineering (distinction) in July 2000. 
As a research assistant, she has worked at the Institute of Process Engineering, Chinese of 
Academy of Sciences for almost four years. In the fall of 2004, she entered the Graduate 
School of University of Houston at Houston, Texas, United States. With a Master of 
Science in Chemical Engineering, she graduated from University of Houston in 2006. 
Then, she entered the Graduate School at Auburn University in the Department of 
Chemical Engineering in August 2006 and began working in the field of systems biology.  
 
 
 
 
 
 
 
 
 
v
THESIS ABSTRACT 
INITIAL INVESTIGATION ON XYLOSE FERMENTATION FOR 
LIGNOCELLULOSIC BIOETHANOL PRODUCTION 
 
Yanli Chen  
Master of Science, May 9, 2009 
(M. S., University of Houston, 2006)  
(B.S., Northwest University at Xi?an, 2000) 
 
89 Typed Pages 
Directed by Jin Wang 
Bioethanol production from lignocellulosic biomass has drawn tremendous 
attention, not only because bioethanol has a number of attractive traits as an alternative 
fuel to fossil fuel, but also due to attractive characteristics of lignocellulosic biomass as a 
feedstock. However, there is lack of a cost-competitive bioethanol production process 
since significant challenges exist, such as pretreatment of biomass feedstock, 
optimization of hydrolysis process, xylose fermentation and so on. In this work, we 
concentrate on the issue of xylose fermentation. Xylose is the second most abundant 
fermentable sugar in biomass hydrolysate, so its fermentation is essential for economic 
conversion of lignocellulose to bioethanol. However, the native strains of ethanol-
producing microorganisms, such as S. cerevisiae, P. stipitis, E. coli and Z. mobilis, can 
not achieve effective cofermentation of glucose and xylose due to their intrinsic 
 
vi
limitations. This thesis presents two potential routes for solving the xylose fermentation 
issue, modified single strain and co-culture. The basic idea of modified single strain is to 
engineer native strains to selectively produce bioethanol from glucose and xylose by the 
use of recombinant DNA techniques. The basic idea of co-culture on ethanol production 
is to utilize two different microbial strains in a system and enable one to ferment glucose, 
the other to ferment xylose stably and effectively. Current research status, difficulties and 
possible strategies about the two routes are depicted in details.  
Furthermore, valuable future research directions of the two routes are recommended. 
Through a broad and complete review, some initial investigation results are provided.  
 
 
 
 
 
vii
ACKNOWLEDGMENTS  
First of all, I would like to thank Dr. Jin Wang for her guidance and support in 
this research and the preparation of this thesis. I also acknowledge Dr. Qinghua He for 
his valuable discussions. I would like to express my gratitude towards my committee 
members, Dr. Christopher B. Roberts and Dr. Mario R. Eden for their time and assistance 
in helping me complete this work. I would like to show my deep appreciation to Dr. 
Sang-Jin Suh and his group for their training on microbiological experimental skills.  
I would like to thank my group members, Mohammad Akanda, Hector Galicia 
and Meng Liang for useful discussions and enjoyable moments. Thanks are also due to Li 
Kang, Wei Yuan and other fellow graduate students (Aimee Poda, Melinda Hemingway, 
Sihe Zhang, Rui Xu, Qiang Gu, Ming Sheng and Ying Zhu) who have made my time 
spent at Auburn University both educational and enjoyable.   
Special thanks to my parents, Peiwen Chen and Ailan Song, and my younger 
sister, Yanzheng Chen, for their lasting love and support throughout my life. I would like 
to thank my husband Heping Liu, whose unconditional love and constant motivation have 
made this research possible. Especially, I want to thank my little lovely daughter, Lucy 
Jiayun Liu. Her birth during this work made me feel the great of being a mother.   
 
 
 
 
viii
The bibliography was prepared in the style: Biotechnology and Bioengineering  
Computer software used: Microsoft Word 2003, Microsoft Excel 2003, Visio 2000, 
Smart Draw 2008  
 
 
 
 
 
 
 
 
 
 
 
 
 
ix
TABLE OF CONTENTS 
LIST OF TABLES............................................................................................................. xi 
LIST OF FIGURES .......................................................................................................... xii 
CHAPTER 1: INTRODUCTION....................................................................................... 1 
CHAPTER 2: BACKGROUND......................................................................................... 4 
2.1 Biomass.................................................................................................................... 4 
2.2 Bioethanol ................................................................................................................ 6 
2.3 Conversion Process of Biomass to Bioethanol ........................................................ 7 
2.3.1 Dilute acid hydrolysis....................................................................................... 8 
2.3.2 Concentrated acid hydrolysis ........................................................................... 9 
2.3.3 Enzymatic hydrolysis ....................................................................................... 9 
2.4 Current Issues and Strategies ................................................................................. 12 
CHAPTER 3: MODIFIED SINGLE STRAIN  FOR GLUCOSE AND XYLOSE      
                       COFERMENTATION............................................................................... 16 
3.1 Successful Examples of Modified Single Strain.................................................... 16 
3.1.1 Recombinant S. cerevisiae ............................................................................. 17 
3.1.2 Recombinant E. coli ....................................................................................... 21 
3.1.3 Recombinant Z. mobilis.................................................................................. 22 
3.2 The Current Difficulties of Modified Single Strain Research ............................... 25 
3.3 Metabolic Network Modeling................................................................................ 26 
 
x
3.3.1 Introduction .................................................................................................... 26 
3.3.2 Reconstruction and Application of the S. cerevisiae Metabolic Network  
         Model ............................................................................................................. 29 
CHAPTER 4: CO-CULTURE FOR GLUCOSE AND XYLOSE COFERMENTATION 
                        ................................................................................................................... 36 
4.1 Description and Feature of Co-Culture Systems.................................................... 36 
4.2 Strains Used in Co-Culture Systems and Their Interactions.................................. 37 
4.3 Fermentation Mode and Condition of Co-Culture Systems................................... 40 
4.3.1 Fermentation Mode ........................................................................................ 40 
4.3.2 Fermentation Condition.................................................................................. 42 
4.4 Fermentation Performance..................................................................................... 43 
4.5 The Current Issues and Related Strategies in the Co-Culture Research ................ 44 
CHAPTER 5: CONCLUSIONS AND RECOMMENDATIONS.................................... 58 
5.1 Conclusions............................................................................................................ 58 
5.2 Recommendations for Future Work....................................................................... 60 
5.2.1. Stability of recombinant strains in continuous fermentation ........................ 60 
5.2.2. Genome-scale metabolic network model for P. stipitis................................. 60 
5.2.3. Different strain combinations for co-culture systems ................................... 61 
5.2.4. Effect of different fermentation parameters on fermentation     
           performance in different co-culture systems................................................ 62 
5.2.5. Kinetic model for co-culture systems............................................................ 62 
5.2.6. Metabolic network model for co-culture systems ......................................... 63 
BIBLIOGRAPHY............................................................................................................. 64 
 
xi
LIST OF TABLES 
 
Table 2.1     Cellulose, hemicellulose, and lignin content  
                    in various sources of biomass ?????????????????????????????????????????????????????????????????????? 14 
Table 3.1     Important traits for ethanol production ???????????????????????????????????????????????????????? 33 
Table 3.2     Comparison of recombinant strains for fermentation with  
                    pretreated corn fiber hydrolysates or glucose/xylose mixtures????????????????????? 33 
Table 4.1     The combinations of two strains used  
                    in different co-culture systems for ethanol production????????????????????????????????? 47 
Table 4.2     Statistics of the strain combinations  
                    in different co-culture systems????????????????????????????????????????????????????????????????????? 48 
Table 4.3     Summary of different co-culture systems ???????????????????????????????????????????????????? 49 
Table 4.4     Minimum requirements of performance  
                    parameters for commercial xylose conversion process ???????????????????????????????? 53 
Table 4.5     Comparison of fermentation performance of different single  
                    strain (wild type, recombinant strains) and co-culture systems???????????????????? 54 
 
 
 
 
 
xii
LIST OF FIGURES 
 
Figure 2.1     General composition of lignocellulosic biomass ???????????????????????????????????????? 15 
Figure 2.2     Schematic diagram of the conversion process  
                     of biomass to bioethanol????????????????????????????????????????????????????????????????????????????? 15 
Figure 3.1     Metabolic pathways for xylose utilization  
                      in fungi, yeast and bacteria ???????????????????????????????????????????????????????????????????????? 34 
Figure 3.2     Integrated process of microbial metabolic  
                      network model construction??????????????????????????????????????????????????????????????????????? 35 
Figure 4.1     Schematic of batch and continuous fermentors ?????????????????????????????????????????? 55 
Figure 4.2     Distribution of pH values in current co-culture systems????????????????????????????? 56 
Figure 4.3     The distribution of initial glucose/xylose concentration  
                      in current co-culture systems ????????????????????????????????????????????????????????????????????? 57 
 
 
 
 
1
CHAPTER 1 
INTRODUCTION 
Pressing economic and environmental factors, such as soaring crude oil prices, 
global warming and diminishing oil reserves, have been driving global interest in 
searching for renewable energy to replace fossil fuels (Stephanopoulos, 2007; Lin and 
Tanaka, 2006). As one of the renewable fuels, biofuels have been considered as potential 
alternatives to fossil fuel for the future. Biofuels are defined as liquid, solid, or gaseous 
fuels derived from renewable biological sources. Bioethanol has been regarded as a clean 
energy alternative due to its more complete combustion and lower carbon monoxide 
emissions. 
Currently, most commercial bioethanol is produced by the fermentation of starch 
or simple sugar from crops such as corn grain, sugar cane, and beets. Therefore, 
production costs are high due to higher source fees and operations fees. Moreover, the 
increasing and competing demand for bioethanol can result in serious supply scarcity and 
undesirable price increases of these agronomic crops. On the other hand, lignocellulosic 
biomass (such as agricultural residues, forestry wastes, waste paper, and energy crops) 
represents a non-food biomass resource and has been recognized as a potential 
sustainable source of mixed sugars to produce bioethanol due to vast quantities, low price 
and environmental benefits (DOE, 2006). In recent years, more and more attention has 
been focused on the conversion of lignocellulosic biomass to bioethanol production. 
 
2
Lignocellulosic biomass mainly consists of carbohydrate polymers (cellulose and 
hemicellulose) and phenolic polymers (lignin). Lower concentrations of various other 
compounds are also present. Cellulose and hemicellulose, which typically make up two-
thirds of lignocellulosic biomass by weight, are polysaccharides that can be hydrolyzed to 
sugars and then fermented to bioethanol. Lignin, accounting for 15-25% by weight due to 
different biomass sources, can not be used in fermentation processes.  
Generally, the conversion of lignocellulosic biomass to bioethanol involves three 
basic steps: 1) pretreatment to liberate cellulose and hemicellulose from lignin; 2) 
hydrolysis of complex polysaccharides to simple sugars with acids or special enzymes. 
Upon hydrolysis, cellulose and hemicellulose can be broken down to a mixture of 
monomeric hexose (glucose, mannose and galactose) and pentose (xylose and arabinose). 
Among these sugars, glucose is normally the most abundant followed by xylose; 3) 
Fermentation of mixed hexose and pentose sugars to produce ethanol by ethanologenic 
microbes (Stephanopoulos, 2007).  
Currently, converting lignocellulosic biomass to bioethanol is too expensive to be 
used on a commercial scale. Therefore, researchers are working to improve the efficiency 
and economics of the lignocellulosic biomass to bioethanol conversion process by 
focusing their efforts on the two most challenging steps: cellulose hydrolysis and sugar 
fermentation.  
Since glucose and xylose are the two principal components of lignocellulosic 
biomass hydrolysates, the issue of sugar fermentation mainly refers to glucose and xylose 
fermentation. Glucose fermentation has been well studied. However, xylose fermentation 
is still a problem. Currently, most research efforts of solving xylose fermentation in the 
 
3
process of bioethanol production from lignocellulosic biomass are mainly concentrated 
on two aspects: 1) genetic modification of glucose-fermenting microbial strains to 
ferment xylose; and 2) co-culture of glucose-fermenting and xylose-fermenting strains. 
This thesis presents the status of current research, difficulties and possible strategies, and 
future directions concerning these two methods. Moreover, some initial results regarding 
these two methods are given in the final chapter.  
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
4
CHAPTER 2 
BACKGROUND 
This chapter provides background information about biomass, bioethanol, and 
conversion process of biomass to bioethanol. Also, current issues and strategies of 
producing bioethanol are described.  
2.1 Biomass  
Biomass is defined as consisting of all plant and plant-derived materials including 
livestock manures (DOE, 2006). Through photosynthesis, plants use light energy from 
the sun to convert water and carbon dioxide to sugars that can be stored. Some plants, 
like sugar cane and sugar beets, store the energy as simple sugars. Other plants, like corn, 
potatoes and root crops, store the energy as more complex sugars, called starches. 
Currently, industrial ethanol production is carried out by using starchy materials such as 
corn, wheat starch and potatoes. However, bioethanol from starchy materials has put the 
effort into direct competition with the food industry. Lignocellulosic biomass is the non-
starch, fibrous part of plant materials. It is an attractive resource because it is renewable, 
abundant and low cost (Perlack et al., 2005). In recent years, more and more attention is 
being focused on the use of lignocellulosic biomass for the production of bioethanol via 
fermentation. Lignocellulosic biomass that can be used as feedstocks to produce 
bioethanol includes: 1) agricultural residues (leftover material from crops, such as corn 
 
5
stover and wheat straw); 2) forestry wastes (chips and sawdust from lumber mills, dead 
trees, and tree branches); 3) municipal solid wastes (household garbage and paper 
products); 4) food processing and other industrial wastes (black liquor, a paper 
manufacturing by-product); and 5) energy crops (fast-growing trees and grasses, such as 
switchgrass, poplar and willow) (DOE, 2006).  
For this discussion, the word ?biomass? will refer to ?lignocellulosic biomass.? 
The primary components of biomass are carbohydrate polymers (cellulose, hemicellulose) 
and phenolic polymers (lignin). Low concentration of various other compounds, such as 
proteins, acids, salts, and minerals, are also present. The general composition of biomass 
is shown in Figure 2.1.  
Cellulose is the most common form of carbon in biomass, accounting for 30%-
50% by weight. It is a glucose polymer linked by ?-1, 4 glycosidic bonds. The basic 
building block of this linear polymer is cellubiose, a glucose-glucose dimer. Hydrolysis 
of cellulose results in individual glucose monomers, which can be fermented to ethanol 
directly. Hemicellulose is a short, highly branched polymer containing five-carbon sugars 
(usually xylose and arabinose) and six-carbon sugars (glucose, galactose and mannose). It 
is at levels of between 20% and 40% by weight depending on the biomass types. 
Hemicellulose is more easily hydrolyzed than cellulose because of its branched, 
amorphous nature. When hydrolyzed, the hemicellulose from hardwoods releases 
products high in xylose (a five-carbon sugar). Lignin which provides structural integrity 
in plants is the largest non-carbohydrate fraction of lignocellulose. It makes up 15% to 
25% by weight of biomass. Unlike cellulose and hemicellulose, lignin can not be utilized 
in the fermentation process. However, it contains a lot of energy and can be burned to 
 
6
produce steam and electricity for the biomass-to-bioethanol process. The composition of 
cellulose, hemicellulose and lignin varies with the sources of biomass. Table 2.1 shows 
the composition of several selected agricultural residues, forestry wastes and energy 
crops. 
2.2 Bioethanol 
The principle fuel used as a gasoline substitute for road transport vehicles is 
bioethanol. Bioethanol has a number of advantages over fossil fuels. Firstly, it comes 
from a renewable resource. Secondly, it is biodegradable, low in toxicity and causes little 
environmental pollution. Bioethanol is a high octane fuel and can be added into gasoline 
as an octane enhancer. In the United States, ?ethanol is blended with gasoline at a 10:90 
ethanol-to-gasoline ratio to boost the fuel?s octane rating, which allows it to burn more  
cleanly, reducing urban smog? (Service, 2007). Thirdly, the use of bioethanol can reduce 
the greenhouse gas emissions. Relative to fossil fuels, greenhouse gas emissions are 
reduced about 18% by the use of corn-based ethanol, but it can be up to 88% if using 
cellulosic ethanol (Service, 2007). A closed carbon dioxide cycle can be formed by using 
bioethanol as fuels. After combustion of bioethanol, the released carbon dioxide is 
recycled back into crops because crops use carbon dioxide to synthesize cellulose during 
photosynthesis (Chandel et al., 2007). In addition, blending bioethanol with gasoline will 
help extend the life of the diminishing fossil oil supplies and ensure greater fuel security, 
avoiding heavy reliance on oil producing nations that have not always been very stable. 
Another advantage of encouraging bioethanol use is that the rural economy would receive 
a boost from growing the necessary crops and creating new employment opportunities 
 
7
(DOE, 2006). In addition, using agricultural and industrial residues to produce bioethanol 
can solve the waste disposal problem and provide environmental benefits. Currently in 
the United States, bioethanol is mainly produced from corn starch by the sugar 
fermentation process. However, the increased demand for bioethanol will result in serious 
problems, such as supply scarcity and dramatic increases in the cost of the food. 
Moreover, even converting all the starch to bioethanol, it can only reduce 10% of the 
gasoline demand (Service, 2007). Therefore, lignocellulosic bioethanol is thought to be 
the answer for solving these problems. We will introduce the conversion process of 
biomass to bioethanol in the following section.  
2.3 Conversion Process of Biomass to Bioethanol 
Basically, the overall process for converting lignocellulose to bioethanol is 
comprised of four major unit operations: pretreatment, hydrolysis, fermentation and 
product separation/distillation. Figure 2.2 shows in a block diagram the basic features of 
this process. 
Pretreatment is an important first step in the conversion process of biomass to 
bioethanol. This step reduces the biomass size and opens up the plant structure since 
native lignocellulosic biomass is extremely recalcitrant to hydrolysis. There are several 
pretreatment methods such as mechanical combination, steam explosion, ammonia fiber 
explosion, acid or alkaline pretreatment and biological treatment (Chandel et al., 2007).  
Each of these is suitable for different types of biomass. Currently, pretreatment is still one 
of the most expensive processing steps with the cost as high as 30 cents per gallon 
produced (Moiser et al., 2005).  Therefore, lowering the cost of the pretreatment process 
 
8
is necessary in order to achieve the production of bioethanol from lignocellulosic biomass 
on commercial scale.  
After pretreatment, the cellulose and hemicellulose portions need to be broken 
down further by enzymes or acids into monomeric sugars for the fermentation into 
ethanol. There are three principle methods of extracting sugars from biomass: dilute acid 
hydrolysis, concentrated acid hydrolysis and enzymatic hydrolysis.  
2.3.1 Dilute acid hydrolysis 
The dilute acid hydrolysis process is one of the oldest and simplest methods of 
extracting fermentable sugars from biomass. This process is carried out in two stages. In 
the first stage, the feedstock is mixed with 0.75% sulfuric acid and heated to 
approximately 50 ?C followed by transferring the material to the first stage acid 
impregnator where the temperature is increased to 190 ?C. Roughly, 80% of the 
hemicellulose and 29% of the cellulose are hydrolyzed in the first stage (Chandel et al., 
2007). The second stage is optimized to yield the more resistant cellulose fraction. 
Normally, this is achieved by using 0.4% sulfuric acid at 230 ?C. Then, the liquid 
hydrolysate is neutralized and recovered from the process. Dilute acid hydrolysis has 
some limitations. If higher temperatures or longer residence time are applied, the 
monomeric sugars derived from hemicellulose will degrade to form some fermentation 
inhibitors, such as furan compounds and weak carboxylic acids (Olsson and H?gerdal, 
1996). In order to remove these fermentation inhibitors, several chemical and biological 
methods could be used, such as ion exchange, charcoal adsorption and biological 
 
9
detoxification (Chandel et al., 2007). However, the detoxification method will increase 
operating cost. 
2.3.2 Concentrated acid hydrolysis 
The concentrated acid hydrolysis process can provide complete and rapid 
conversion of cellulose to glucose and hemicellulose to xylose with little degradation. In 
this process, 70% sulfuric acid is added to the biomass with 10% moisture content (in a 
ratio of 1.25 acid: 1 biomass) in a reactor. The temperature is controlled at 50 ?C for two 
to hours (Chandel et al., 2007). Water is then added to dilute the acid to 20-30% and the 
mixture is again heated to 100 ?C for one hour. The gel produced from this mixture is 
then pressed to release an acid sugar mixture and a chromatographic column is used to 
separate the acid and sugar mixture. The obtained acid can be reconcentrated through 
multiple effect evaporators. Approximately, 90% of both cellulose and hemicellulose can 
be depolymerized into their monomeric sugars with concentrated hydrolysis, so this 
process has the advantage of high sugar recovery efficiency (Chandel et al., 2007). In 
addition, monomeric sugar degradation can be avoided with low temperature and 
pressure in this process.  
2.3.3 Enzymatic hydrolysis 
Besides acid, cellulolytic enzymes which are highly specific can be used to 
hydrolyze the pretreated biomass into fermentable sugars. These enzymes can be 
produced from some bacterial and fungal sources or obtained from some commercial 
companies, such as Novozymes. However, the process is very expensive compared with 
 
10
acid hydrolysis due to the high enzyme cost.  Although the cost of cellulolytic enzyme 
has come down to 20 to 30 cents per gallon of ethanol produced, this conversion process 
cannot be competitive with the process of ethanol production from starch in corn kernels 
at a cost of 3 to 4 cents per gallon of ethanol (Stephanopoulos, 2007). 
After hydrolysis, the primary fermentable sugars in hydrolyzate are pentose and 
hexose, such as glucose and xylose. Different microorganisms are used to ferment these 
sugars to produce bioethanol, such as Saccharomyces cerevisiae, Pichia stipitis, 
Kluyveromyces marxianus, Candida shehatate, Zymomonas mobilis and Escherichia coli. 
Currently, the fermentation of a mixture of hexose and pentose is inefficient because no 
wild organism has been found that can convert all sugars into ethanol at a high yield 
(Ragauskas et al., 2006). For example, the most widely used glucose-fermenting yeast S. 
cerevisiae is able to effectively ferment glucose, but it can not take xylose as a carbon 
source. Although E. coli does metabolize xylose, it naturally produces a variety of acetic 
and lactic acids as fermentation products, so the bioethanol yield is comparatively low 
(Service, 2007). The lack of industrially suitable microbes for converting biomass into 
bioethanol has been a major technical problem.  
Fermentation can be preformed as a batch, fed batch and continuous process. 
Generally, batch fermentation is the most widely used mode by the fermentation industry. 
The kinetic properties of microorganisms, type of lignocellulosic hydrolysate and process 
economics aspects are the determinative factors for the choice of most suitable process 
mode. After fermentation, dilute bioethanol can be purified through distillation and/or 
filtration to produce the desired fuel-grade quality ethanol.  
 
11
During the biomass-to-bioethanol process, there are two significant steps: 
hydrolysis and fermentation. In order to improve the efficiency and economics of the 
ethanol production process and use it at commercial scale, intensive research has been 
conducted to improve biomass hydrolysis and fermentation technologies. 
Enzymatic hydrolysis performed separately from fermentation step is known as 
separate hydrolysis and fermentation (SHF). This process offers various processing 
advantages. It enables enzymes to operate at higher temperature (45-50 ?C) for increased 
performance and fermenting microbes to operate at moderate temperature (28-35 ?C) for 
optimizing of the sugars utilization (Lin and Tanaka, 2006). However, it has some 
shortcomings. For instance, the hydrolytic enzymes can be inhibited by hydrolysis 
products: glucose and short cellulose chains. In addition, the operating cost is higher 
since separated reactors for saccharification and fermentation are needed.  
In order to prevent inhibition of the hydrolytic enzymes by the reaction products, 
enzymatic hydrolysis and fermentation can be combined in one unit, termed as 
simultaneous saccharification and fermentation (SSF). This is a one-step process in which 
enzymatic hydrolysis and sugar fermentation are carried out in one reactor. Except 
keeping the glucose concentration low and prevent inhibition of the hydrolytic enzymes, 
SSF has other advantages, such as increasing the overall rate of biomass to bioethanol 
conversion, shortening fermentation time and reducing operating cost. However, it also 
has some drawbacks due to the low rate of cellulose hydrolysis and most ethanol 
producing microorganisms can not utilize all of the sugars derived after hydrolysis. By 
using promising ethanol bacteria recombinant E. coli K011, Klebsiella oxytoca and Z. 
mobilis, SSF process has been improved after including the co-fermentation of multiple 
 
12
sugar substrates present in the hydrolysate (Chandel et al., 2007). The new variant of SSF 
is known as simultaneous saccharification and co-fermentation (SSCF). SSCF also has 
the advantages of lower cost, higher ethanol yield and shorter processing time. The most 
promising conversion method is consolidated bioprocess (CBP), featuring cellulase 
production, cellulose hydrolysis and fermentation in one step. In this process, ethanol and 
all required cellulolytic enzymes are produced by a single microorganism. This can be 
achieved by expression of cellulases in fermenting organisms or transfer of the 
bioethanol-synthesizing pathway into a cellulase-producing organism. CBP will have the 
benefit of completely eliminating the cost of purifying cellulase (Stephanopoulos, 2007). 
Moreover, it has the potential to provide the lowest cost route for bioconversion from 
lignocellulosic biomass to bioethanol and other products (Lynd et al., 2005). However, 
this process can not become the leading process at present due to the relative lag in 
development of molecular biological methods to manipulate organisms.  
2.4 Current Issues and Strategies   
Lignocellulosic bioethanol is proposed as having such benefits as: reduction of 
greenhouse gas emissions, reduction of fossil fuel use, increased national energy security, 
increased rural development, a sustainable fuel supply for the future. Although significant 
advances have been made at bench scale toward the bioethanol generation from 
lignocellulose, there are still technical and economical barriers, which make the 
bioethanol program unsuccessful on a commercial scale. Currently, the challenges 
include 1) low bulk density feedstock; 2) high viscosity substrate; 3) optimization of 
 
13
hydrolysis and fermentation; 4) fermentability of substrate; 5) xylose fermentation; 6) 
cost challenges.  
?Xylose is a five-carbon sugar abundant in hardwoods and agriculture residues, so 
its fermentation is essential for economic conversion of lignocellulose to ethanol? 
(Jeffries et al., 2007).  The widely used microorganism in crop-based ethanol production, 
S. cerevisiae cannot ferment xylose, therefore, it has limited use for conversion of 
lignocellulose to ethanol. The ethanologenic bacterium Z. mobilis has the same limitation. 
Some xylose-fermenting microorganisms, such as P. stipitis, C. shehatate, and E. coli, 
have other drawbacks including low ethanol yield and low ethanol tolerance to hamper 
their large-scale utilization in biomass to bioethanol process (Zaldivar et al., 2001). In 
this thesis, we focused on the problem of xylose fermentation to do some initial 
investigation. To date, there are two viable strategies for solving xylose fermentation 
issue: modified single strain and co-culture. By introducing pathways for the xylose 
fermentation, some microorganisms have been engineered to coferment glucose and 
xylose effectively. The successful examples are recombinant E. coli, Z. mobilis and S. 
cerevisiae (Alterthum and Ingram, 1989; Amore et al., 1991; Takuma et al., 1991; Zhang 
et al., 1995; Eliasson et al., 2000; Sedlak and Ho, 2004). Besides the methods of 
engineered microbes, co-culture conversion strategy has been suggested as a 
breakthrough, high-payoff opportunity in bioethanol production (DOE, 2006). In co-
culture systems, there are two different microbial species which can ferment glucose and 
xylose respectively. We will explain the two routes for solving xylose fermentation 
problem in the following chapters in more details. 
 
 
14
 
 
 
 
Table 2.1 Cellulose, hemicellulose, and lignin content in various sources of 
biomass 
 
Feedstock Cellulose Hemicellulose Lignin Reference 
Corn stover 36.4 22.6 16.6 DOE, 2006 
Corn cob 42.0 39.0 14.0 Kuhad and Singh, 1993 
Rice straw 32.0 24.0 13.0 Kuhad and Singh, 1993 
Wheat straw 30.0 24.0 18.0 Kuhad and Singh, 1993 
Rice hulls 36.0 15.0 19.0 Kuhad and Singh, 1993 
Saw dust 55.0 14.0 21.0 Olsson and H?gerdal, 1996 
Willow 37.0 23.0 21.0 Olsson and H?gerdal, 1996 
Switchgrass 31.0 24.4 17.6 DOE, 2006 
Poplar 49.9 20.4 18.1 DOE, 2006 
 
 
 
 
  
 
 
 
 
 
Figure 2.1 General composition of lignocellulosic biomass (Lee et al., 2007)  
 
 
 
Figure 2.2 Schematic diagram of the conversion process of biomass to bioethanol 
 
 
 
15
 
16
CHAPTER 3 
MODIFIED SINGLE STRAIN  
FOR GLUCOSE AND XYLOSE COFERMENTATION  
The efficient fermentation of xylose is critical to attaining economically feasible 
processes for bioethanol production from lignocellulosic biomass since xylose is so 
abundant in biomass. However, there are no robust industrial microorganism strains that 
can ferment glucose and xylose simultaneously and efficiently. One detour to solve this 
problem is to genetically modify or adjust some potential strains by introducing pathways 
for xylose fermentation. The greatest successes have been made in the engineering of S. 
cerevisiae, E. coli and Z. mobilis. In this chapter, we will introduce several examples of 
successful recombinant strains, then address the current difficulties of modified single 
strain research. Finally, metabolic network modeling will be suggested to help solve these 
difficulties.  
3.1 Successful Examples of Modified Single Strain  
The characteristics required for an industrially suitable microorganism are 
summarized in Table 3.1 (Dien et al., 2003). However, no naturally occurring 
microorganisms can satisfy all of these features. The lack of industrially suitable 
microbes for converting biomass into ethanol has been a major technical barrier. Among 
all of the traits, ethanol yield is the most important one. In order to obtain a high ethanol 
 
17
yield, the stains should produce ethanol with few side productions, and metabolize all 
major sugars. The primary fermentable sugars in hydrolysate from cellulose and 
hemicellulose are pentose and hexose (e.g., glucose and xylose). However, the
 well-known fermentative yeast S. cerevisiae naturally converts glucose to ethanol, but 
does not ferment xylose. Although E. coli metabolize pentose sugars, it naturally 
produces a variety of acetic and lactic acids as fermentation products, but ethanol as a 
minor fermentation product (Service, 2007). By introducing pathways for the xylose 
fermentation, some microorganisms have been engineered to ferment both glucose and 
xylose and selectively produce ethanol. The successful examples are recombinant S. 
cerevisiae, recombinant E. coli, and recombinant Z. mobilis (Ingram et al., 1997; Bothast 
et al., 1999; Dien et al., 2003; Jeffries and Jin, 2004; Potera, 2006; Chu and Lee, 2007). It 
has been shown that metabolic engineering of microorganisms is a very efficient tool for 
increasing ethanol yield. Table 3.2 compared the fermentation performance of several 
genetically engineered microorganisms on pretreated corn fiber hydrolyzates or 
glucose/xylose mixtures. From this table, we can see that utilization of these recombinant 
strains in the fermentation with biomass hydrolyzates or glucose/xylose mixture can keep 
high ethanol yield (from 0.41~0.50 g g
-1
) which is very close to the theoretical ethanol 
yield from glucose or xylose (0.51g g
-1
) (Zhang et al., 1995) . In addition, high ethanol 
volumetric productivity can also be obtained by using these recombinant strains.  
 
3.1.1 Recombinant S. cerevisiae  
As a promising candidate for industrial bioethanol production, S. cerevisiae has 
attractive strengths, such as robustness, public acceptance, inhibitor tolerance and high 
 
18
ethanol productivity (Karhummaa et al., 2007). However, the native strain cannot ferment 
xylose. Xylose is a five-carbon monosaccharide that exists in the hydrolysate of biomass 
with large amount. Naturally, there are some bacteria, yeasts and fungi which can 
metabolize xylose. Bacteria generally accomplish conversion of xylose to xylulose in one 
step catalyzed by xylose isomerase (XI) encoded by the XYLA gene, while yeasts and 
fungi use different pathway to degrade xylose to xylulose (Bothast et al., 1999). Xylose 
reductase (XR) and xylitol dehydrogenase (XDH) are the two main enzymes used in this 
method of xylose degradation. XR, encoded by the XYL1 gene, is responsible for the 
reduction of xylose to xylitol and is aided by cofactors NADH or NADPH. Xylitol is then 
oxidized to xylulose by XDH, which is expressed through the XYL2 gene, and 
accomplished exclusively with the cofactor NAD
+
 (Chu and Lee, 2007). Therefore, there 
are two main pathways of xylose metabolism, ?Xylose Reductase-Xylitol 
Dehydrogenase? or XR-XDH pathway and ?Xylose Isomerase? (XI) pathway. After 
producing xylulose, both of XR-XDH and XI pathways proceed through enzyme 
xylulokinase (XK), encoded on gene XKS1, to further modify xylulose into xylulose-5-P 
where it then enters the pentose phosphate pathway for further catabolism. Figure 3.1 
clearly describes the two metabolic pathways for xylose utilization in bacteria, yeasts and 
fungi. The two different pathways have been applied to construct recombinant xylose-
fermenting S. cerevisiae.  
In 1991, researchers successfully obtained xylose-fermenting recombinant S. 
cerevisiae strains transformed with P. stipitis native genes XYL1 encoding xylose 
reductase (XR) (Amore et al., 1991; Takuma et al., 1991). The success of obtaining S. 
cerevisiae XYL1 transformants gave the first indication that the metabolic engineering 
 
19
approach could be used to improve xylose fermentation. Later, K?tter and Ciriacy (1993) 
isolated the XYL1 and XYL2 genes coding for XR and XDH from P. stipitis and expressed 
them in S. cerevisiae. Although this recombinant strain can metabolize xylose, the 
primary product is xylitol not ethanol. Moreover, the growth and fermentation rate is very 
slow. These limitations for this recombinant strain are probably caused by cofactor 
imbalance and an insufficient capacity of xylulose conversion by the pentose phosphate 
shunt (K?tter and Ciriacy, 1993). Since the native S. cerevisiae xylulokinase (XK) 
activity is insufficient for xylose or xylulose fermentation, over-expression is required to 
obtain high ethanol yields. In 1998, Ho et al. firstly constructed a recombinant S. 
cerevisiae strain 1400 (pLNH32) expressing the XR and XDH from P. stipitis and 
overexpressing XK from a high copy number yeast-E. coli shuttle plasmid. The ethanol 
yield can reach 0.46 g g
-1
 and the ethanol productivity is 1.15 g l
-1
 h
-1 
by using this strain 
to ferment a mixture of glucose (90 g l
-1
) and xylose (40 g l
-1
) (Ho et al. 1998). Another 
important recombinant strain, S. cerevisiae TMB3001, was metabolically engineered with 
a xylose-utilizing pathway by integrating the P. stipitis genes XYL1 and XYL2 and the 
endogenous XKS1 (Eliasson et al., 2000). This strain can convert xylose to ethanol under 
anaerobic condition as well as aerobic condition (Eliasson et al., 2000; Zaldivar et al., 
2002). Ethanol yields of 0.30 g g
-1
 was obtained from xylose-glucose mixtures in 
anaerobic chemostat cultures, with a dilution rate of 0.06 h
-1 
(Jeppsson et al., 2002). 
Although the recombinant S. cerevisiae TMB3001 can grow on xylose, the xylose 
utilization rate is low due to the imbalance of redox co-factors. Metabolic engineering of 
the ammonium assimilation in xylose fermenting S. cerevisiae has been used as a strategy 
to improve ethanol yield by modulating the redox metabolism (Roca et al., 2003; 
 
20
Grotkjaer et al., 2005). The stable recombinant strain TMB3001, carrying XR and XDH 
from P. stipitis and over-expressed endogenous XKS1 was modified by deleting the 
GDH1 gene encoding an NADPH-dependent glutamate dehydrogenase, and by over- 
expressing either the GDH2 gene encoding an NADH-dependent glutamate 
dehydrogenase or the glutamate synthase (GLT1) and glutamine synthase (GLN1) genes 
encoding the GS-GOGAT complex. These modifications in the ammonia assimilation 
pathway resulted in beneficial alterations of the redox metabolism and hereby of the 
xylose metabolism. The over expression of GDH2 strains increased ethanol yield from 
0.43 to 0.51 mol of carbon (Cmol) Cmol
-1
 mainly due to a 44% reduction of xylitol 
excretion, and the over expression of GS-GOGAT complex strains increased ethanol 
yield by 16% in carbon-limited continuous cultivation at a low dilution rate (Roca et al., 
2003). Grotkjaer et al. (2005) constructed a metabolic flux model for two recombinant S. 
cerevisiae trains TMB30001 and CPB. CR4. ?Metabolic flux analysis suggested that the 
ratio of the specific XR activity utilizing different co-factors (NADH/NADPH) was 
shifted from being primarily NADPH dependent for the TMB3001 strain to be partly 
NADH dependent for the CPB.CR4 strain. A higher NADH/NADPH ratio for co-factor 
usage by the XR is clearly beneficial for solving the redox imbalance for xylose 
fermenting S. cerevisiae (Grotkjaer et al., 2005)?.  
The redox co-factors imbalance problem associated with XR- and XDH- coupled 
reactions that are particularly harmful to efficient anaerobic xylose fermentation can also 
be circumvented by introducing XI pathway into S. cerevisiae. Heterologous expression 
of several xylose isomerases (XI) from different microorganisms in S. cerevisiae has 
been attempted (Sarthy et al., 1987; Amore et al., 1989; Moes et al., 1996; Walfridsson et 
 
21
al., 1996; G?rdonyi and Hahn-H?gerdal, 2003; Kuyper et al., 2003). However, the 
activity of XI when expressed in S. cerevisiae is generally reduced or lost due to some 
reasons, such as differences in internal pH between bacteria and yeasts, protein 
misfolding and unsuitable post-translational modifications (Bothast et al., 1999; Chu and 
Lee, 2007). The overall ethanol yields by using recombinant S. cerevisiae expressing XI 
have been low, and further development to improve XI activity and heterologous 
expression in S. cerevisiae through adaptation and metabolic manipulation will be 
required to achieve high ethanol yield.  
3.1.2 Recombinant E. coli  
Although E. coli has the natural ability to metabolize glucose, xylose, galactose, 
mannose and other sugars, the fermentation products are a mixture of acidic and neutral 
products. More important, the amount of ethanol is very small. Since 1986, the 
microbiologist Lonnie Ingram of the University of Florida and his research group have 
been working on the genetic engineering of different E. coli strains for ethanol production 
(Ingram et al., 1987; Ingram and Conway, 1988; Alterthum and Ingram, 1989; Ingram et 
al., 1991; Ohta et al., 1991; Ingram et al., 1997). They developed ethanologenic 
derivatives of E. coli B (ATCC11303) in which pyruvate metabolism was redirected to 
ethanol and carbon dioxide by the integration and functional expression of Z. mobilis 
genes encoding pyruvate decarboxylase (pdc) and alcohol dehydrogenase II (adhB). A 
deletion was also introduced into the fumarate reductase gene (frd) to minimize succinate 
production and increase ethanol yield (Ingram et al., 1997; Tao et al., 2001). The 
resulting strains, K011, can efficiently ferment high concentrations of glucose and xylose 
 
22
to ethanol at yields 103% to 106% of theoretical. ?The extra ethanol was thought to arise 
from fermentation of carbohydrates present in rich medium that were not account for in 
the sugar balance ?(Dien et al., 2003). Furthermore, E. coli K011 appears to grow more 
rapidly than the parent on plates and in broth (Tao et al., 2001). The major shortcomings 
of using recombinant E. coli in industrial fermentation are a narrow and neutral pH 
growth range (pH 6.0-8.0), less hardy cultures compared to yeast, and public perceptions 
regarding the danger of E. coli strains (Dien et al., 2003). Also, it has low ethanol 
tolerance (6.4% ethanol in the final fermenting solution) (Service, 2007). Regarding the 
stability of the recombinant E. coli K011, some researchers (Dumsday et al., 1999) 
compared it in batch and chemostat culture. They found that chemostat culture on glucose 
was remarkably stable, but on mannose, xylose and a xylose/glucose mixture, this 
recombinant strain progressively lost their hyperethanologenicity. Their results indicate 
the genetic instability of the recombinant E. coli K011 on xylose in the chemostat culture.  
3.1.3 Recombinant Z. mobilis  
As one of potential biocatalysts for bioethanol production, Z. mobilis has some 
attractive properties, such as high ethanol tolerance (up to 120 g/L ethanol), high ethanol 
yield, high specific ethanol productivity and public safety (Dien et al., 2003). However, 
wild-type Z. mobilis has an extremely limited substrate range, such as glucose, fructose 
and sucrose (Lawford and Rousseau, 2002), therefore, it can not ferment xylose. Much 
effort has been put into attempts to give Z. mobilis a xylose-fermenting capacity by the 
use of recombinant DNA techniques. Researchers at the National Renewable Resources 
Laboratory (NREL) were the first to construct a recombinant strain CP4 (pZB5) capable 
 
23
of fermenting xylose (Zhang et al., 1995). This recombinant strain was engineered by 
transforming four E. coli genes (xylose isomerase (xylA), xylulokinase (xylB), 
transketolase (tktA), and transaldolase (talB)) into Z. mobilis. The recombinant Z. mobilis 
CP4 (pZB5) was shown to grow on 25 g/L xylose as the sole carbon source at a specific 
growth rate of 0.057 h
-1
 and to produce ethanol as the principal fermentation product at 
86% of the theoretical yield. In the presence of a mixture of glucose and xylose (25 g/L 
each), the transformed strain CP4 (pZB5) can ferment both sugars to ethanol at 95% of 
theoretical yield within 30 hours. However, xylose was utilized much more slowly than 
glucose. The anaerobic fermentation of xylose and glucose to ethanol was
 
achieved 
through a combination of the pentose phosphate and
 
Entner-Doudoroff pathways. By 
using the same general strategy, they also constructed a recombinant strain Z. 
mobilis(ATCC 39676 (pZB186)) that can ferment arabinose (Deanda et al., 1996). Five 
genes, encoding L-arabinose isomerase (araA), L-ribulokinase (araB), L-ribulose-5-
phosphate-4-epimerase (araD), transaldolase (talB), and transketolase (tktA), were 
isolated from E. coli and introduced into Z. mobilis under the control of constitutive 
promoters. The engineered Z. mobilis strain grew on medium containing L-arabinose 25 
g/L as the sole carbon source, and produced ethanol at 98% of the maximum theoretical 
ethanol yield based on arabinose consumed. However, the yield based on arabinose 
supplied was lower (84%) due to residual unfermented arabinose. In the presence of 
glucose and arabinose (both 25 g/L), the overall ethanol yield (based on sugars supplied) 
was 84% of theoretical, due to incomplete utilization of arabinose. Arabinose was used at 
a much slower rate than glucose, and only after glucose was nearly depleted. The strain 
was stable in the presence of tetracycline or arabinose as the sole carbon source. However, 
 
24
only 40% of the cells retained their ability to ferment arabinose when grown on complex 
medium at 30 ?C and they completely lost their ability to ferment arabinose within seven 
generations at 37 ?C. Later, Chou et al. (1997) reported the construction of a single strain 
of Z. mobilis 206C (pZB301) capable of fermenting both xylose and arabinose. The strain 
contains seven plasmid-borne genes encoding xylose- and arabinose-metabolising and 
pentose phosphate pathway enzymes. This recombinant strains was capable of fermenting 
a mixture of 30 g/L glucose, 30 g/L xylose and 20 g/L arabinose to ethanol at an overall 
yield of ~82-84% of the theoretical (based on sugar supplied). Uptake of arabinose was 
slow compared to that of glucose and xylose, accounting for the long fermentation time 
(80-100 h at 30 ?C). Very recently, researchers in Rogers laboratory constructed a better 
recombinant strain ZM4 (pZB5) by  transforming pZB5 into their best Z. mobilis ethanol 
producing strain ZM4 (ATCC 31821) (Joachimsthal and Rogers, 2000). This strain 
demonstrated much higher ethanol tolerance than the CP4 derivatives on glucose/xylose 
mixtures. Furthermore, in order to enhance its genetic stability, researchers developed 
several genomic DNA?integrated strains of Z. mobilis 206C (pZB301) through the 
insertion of all seven genes necessary for xylose and arabinose fermentation into the 
Zymomonas genome (Mohagheghi et al., 2002). Four strains were selected from all the 
integrants developed and tested for stability. Based on the stability test, one of the 
integrants (AX101) was selected further to evaluate the cofermentation of glucose, xylose, 
and L-arabinose in batch and continuous modes (Mohagheghi et al., 2002). Lawford and 
Rousseau (2002) assessed the fermentation performance characteristics of strain AX101, 
in batch and continuous pH-controlled fermentations using different pure sugar synthetic 
biomass hydrolysate media (Lawford and Rousseau, 2002). They found that only 50% of 
 
25
the xylose (initial concentration 30 g/L) was consumed when both acetic acid (2.5 g/L, 
pH 5.5) and ethanol (30 g/L) were added to the medium (Lawford and Rousseau, 2002). 
The results indicated that acetic acid inhibition was exacerbated in the presence of 
ethanol. Although the recombinant Z. mobilis AX101 has several appealing properties as 
biocatalyst for ethanol production, such as higher ethanol process yield, low pH 
performance and minimal by-product formations, the main obstacle for its application in 
industrial process is the inhibition by acetic acid in continuous for ethanol productivity 
(Mohagheghi et al., 2002; Dien et al., 2003). Acetic acid is commonly found in 
hydrolysates and originates from acetyl side-chain groups of hemicellulose. To solve this 
problem, there are two possible routes: 1) adapting the strain AX101 to acetic acid or 2) 
removing acetic acid from the hydrolysate prior to the fermentation (Dien et al., 2003).  
3.2 The Current Difficulties of Modified Single Strain Research  
The successful construction of recombinant bacterial and yeast strains through 
genetic engineering exhibited an improved ability to ferment xylose to ethanol, which 
arguably represents the most significant development in xylose fermentation research in 
the past few decades. However, there are some limitations for the genetic manipulation 
strategy when attempting to develop recombinant strains for industrial bioethanol 
production process. These limitations include 1) gene insertion may cause pleiotropic 
effects in the host organism that are difficult to predict (McMillan, 1993); 2) recombinant 
constructs are often unstable in long-term culture; 3) fermentation performance on real 
hydrolyzates has seldom been evaluated; 4) formation of undesirable co-products that can 
inhibit the prolong fermentation.  
 
26
Actually, these limitations are result from one significant difficulty involving in 
lack of system-level understanding of cell metabolism. Mathematical models play an 
important role in the analysis of metabolism from a systems engineering perspective. The 
mathematical models widely used in metabolic engineering fall into two classes: kinetic 
models and stoichiometric models (Patil et al., 2004). Kinetic models describe the 
dynamic properties of metabolic network by combining kinetics information about 
specific cellular process with known stoichiometry. Due to the lack of kinetic data and 
the difference between in vivo and in vitro kinetic parameters for numerous biochemical 
reactions in metabolic network, application of kinetic models in metabolic engineering is 
in the preliminary stage. Different from kinetic models, stoichiometric models 
quantitatively describe steady-state flux distribution in metabolic network. Based on 
well-known stoichiometric coefficients, it does not require determination of parameters 
like kinetic constants.  Due to this advantage, stoichiometric models have been well 
utilized in metabolic engineering. We will introduce its application and give a successful 
example in the following part.  
3.3 Metabolic Network Modeling  
3.3.1 Introduction  
Researchers have long desired the ability to quantitatively describe metabolic 
fluxes through metabolic networks. So far this endeavor has been hampered by the need 
for extensive kinetic information describing enzyme catalysis within a living cell. 
Detailed information about all of the enzymes in a specific metabolic network has not 
always been available. Recently, this dilemma has been partially resolved by the 
development of mass balance-based stoichiometric models.  
Generally, the system of equations that described mass balances for the 
metabolites is: 
bvS
dt
dX
??=                                                                 (1) 
Where X is the vector of metabolite concentrations, where S is the stoichiometric 
matrix (m is the number of metabolites and n is the number of biochemical reactions in 
the metabolic network), where v is the vector of reaction fluxes, and where b is the vector 
of known substrate consumption rates, production formation rates and cell growth rates. 
The element  is the stoichiometric coefficient of the ith metabolite in the jth reaction. 
Since the rates of substrate consumption, product formation and cell growth are much 
slower than that of metabolic reactions, metabolites concentrations in the cell can be 
assumed to be at pseudo-steady state. This assumption reduces Eq. (1) to  
nm?
ij
S
bvS =?                                                                     (2) 
Typically, the number of metabolic fluxes is greater than the number of mass balance 
equations (i.e. ). Therefore, the system is underdetermined, which means more than 
one solution could exist.  Technically, the stoichiometry of the metabolic network does 
not uniquely specify the fluxes through the cell?s pathways, so that the number of 
possible flux distributions allowed by the stoichiometry is infinite. The cell?s choice of 
flux distribution is determined by a complex interplay of enzymatic and genetic 
regulatory events (Varma and Palsson, 1994). However, in the absence of detailed 
knowledge of these events, the metabolic flux distribution can be estimated by defining 
mn >
 
27
an ?objective? which underlies the cell?s behavior according to the evolutionary selection 
process. Linear programming provides an expedient method for finding a feasible steady-
state flux vector that maximizes an objective function. This technique is defined as Flux 
Balance Analysis (FBA).  
FBA has been successfully applied to describe metabolism in E. coli (Edwards 
and Palsson, 2000), Haemophilus influenzae (Edwards and Palsson, 1999) and S. 
cerevisiae (F?rster et al., 2003) and its recombinant strain (Jin and Jeffries, 2004). We 
will provide more details about the genome-scale reconstruction of the S. cerevisiae 
metabolic network model and its application for improved bioethanol production in the 
next section.  
The determination of a particular metabolic flux distribution was formulated as a 
linear programming problem, in which the solution that maximizes an objective function 
was identified. The linear programming problem can be formulated as the following 
system of equations:  
ii
vcZ
?
=Max          
bvS =?Subject to       
iii
v ?? <<    
 
28
Where Z is the objective function, and where c is the weight vector that determines the 
contribution of each flux to the objective function. The maximization of Z is subject to 
both equality and inequality constraints. The equality constraints are referenced as 
stoichiometric constraints (mass balance), and the inequality constraints are referenced as 
enzymatic capacity constraints (using an appropriate values). Each row of the 
stoichiometric constraints 
max
v
bvS =?  represents the steady state balance equation for a 
particular metabolite. If the i th row of the equation is associated with an internal 
metabolite, should be zero to reflect the pseudo-steady state assumption. On the other 
hand, if the i th row is associated with an external metabolite, which is transferred 
to/from the surrounding environment or is otherwise accumulated or depleted over time, 
the corresponding element of b should equal to the net exchange rate of accumulation rate 
of this metabolite as determined from experimental measurements. Inequality constraints 
are necessary to confine the feasible flux vectors between suitable lower and upper 
bounds, given by the vector 
i
b
? and ? . Generally, the flux vector is bounded in the 
interval if the i th reaction is irreversible or ,0[ ?) ),( ???  if it is reversible. 
Nevertheless, setting tighter bounds is beneficial in order to eliminate unrealistic or 
undesirable flux distributions from the feasible set. This should depend on the availability 
of experimental data for individual fluxes. As to objective functions, there are three types 
1) to represent exploration of the metabolic capabilities of a network; 2) to represent 
physiologically meaningful objectives (such as maximum cellular growth rate); and 3) to 
represent bioengineering design objectives. Different objective functions have been used 
to analyze metabolic networks, including minimization of ATP production, minimization 
of total nutrient uptake, maximization of metabolite production and maximization of 
growth rate (Savinell and Palsson, 1992).  
3.3.2 Reconstruction and Application of the S. cerevisiae Metabolic Network Model  
Along with its industrial importance, S. cerevisiae was the first eukaryotic 
organism whose genome was fully sequenced and annotated (Goffeau, 1997). With the 
availability of S. cerevisiae genome sequence database and other information, such as 
 
29
 
30
biochemical and physiological data, genome-scale metabolic network model for S. 
cerevisiae was first reconstructed by F?rster and his colleagues in 2003 (F?rster et al., 
2003). This genome-scale reconstructed metabolic network contains 584 metabolites and 
1175 metabolic reactions that were compartmentalized between the cytosol and the 
mitochondria, and included transport steps between the compartments and the 
environment. The reconstruction process consisted of collecting a set of biochemical 
reactions, constructing stoichiometric model of metabolism using mass balance, and 
gathering other necessary information, such as knowledge on the biomass composition, 
the growth-associated and nongrowth-associated ATP requirements.  
However, the process is iterative. The general process of reconstructing metabolic 
network model is depicted in Figure 3.2 (Covert et al., 2001). When the initial model is 
designed, it could be applied to some quantitative analytical methods to simulate cellular 
behavior under certain genetic and physiological conditions. Then computed results 
would be compared with experimental results. If no agreement was found, adjustments 
and corrections on the initial reaction list would need to be done until simulated results 
were in agreement with experimental results. Once an accurate metabolic network is 
reconstructed, then mathematical methods, such as convex analysis and linear 
programming, could be used to analyze structural properties, calculate metabolic 
capabilities, and examine the integrated functions of the reconstructed metabolic network. 
The resulting simulated results could be applied to provide valuable information for the 
development of metabolic engineering strategies for the construction of strains with 
desired and improved properties. 
 
31
Metabolic engineering is most effective when guided by systematic biochemical 
models to integrate the intrinsic variables and extrinsic changes (Jeffries, 2006). Bro et al. 
(2006) used the genome-scale metabolic network model of S. cerevisiae to simulate the 
effect of genome modification (gene insertion, gene overexpression and/or gene deletion) 
on cellular behavior (biomass, ethanol and glycerol yield). By testing different alternative 
strategies for engineering of the redox metabolism in S. cerevisiae, they identified the 
most optimal strategies in terms of reducing glycerol formation and increasing the 
ethanol yield on glucose or mixture of glucose and xylose and the best-scored strategies 
were predicted to completely eliminate formation of glycerol and increase ethanol yield 
with 10%. In addition, they also performed experiments to verify their prediction from 
their simulation results. By expressing the gene gapN which encodes non-
phosphorylating NADP
+
-dependent glyceraldehyde-3-phophate dehydrogenase (GAPN) 
in glucose-fermenting S. cerevisiae, they obtained a transformant that had a 40% lower 
glycerol yield while the ethanol yield increased with 3% without affecting the maximum 
specific growth rate by anaerobic batch cultivation on glucose. By expressing GAPN in a 
xylose-fermenting S. cerevisiae strain harbouring xylose reductase (XR) and xylitol 
dehydrogenase (XDH), they obtained a recombinant strain that could lead to an 
improvement in ethanol yield by up to 25% on xylose/glucose mixtures. Both the model 
predictions and the experimental results showed that heterologous expression of GAPN 
in S. cerevisiae seems to be an attractive strategy to redirect flux that is going to glycerol 
toward ethanol and that overexpression of GAPN may lead to a further reduction of 
glycerol yield and increase of ethanol yield either on glucose or on glucose/xylose 
mixture.  
 
32
In conclusion, their results showed that the genome-scale metabolic model 
represents a powerful tool for directing the construction of recombinant strains without 
doing numerous of biological experiments for trial. Therefore, it is necessary to build 
metabolic network models for other ethanol-producing strains, such as P. stipitis and Z. 
mobilis. . 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
33
 
Table 3.1 Important traits for ethanol production (Dien et al., 2003) 
Trait  Requirement  
Ethanol yield  >90% of theoretical  
Ethanol tolerance  >40 g l
-1
 
Ethanol productivity  >1 g l
-1
 h
-1
 
Robust grower and simple growth 
requirements 
Inexpensive medium formulation  
Able to grow in undiluted hydrolysates Resistance to inhibitors  
Culture growth conditions  
retard contaminants  
Acidic pH or higher temperature  
 
 
Table 3.2 Comparison of recombinant strains for fermentation with pretreated corn 
fiber hydrolyzates or glucose/xylose mixtures (adapted from Bothast et al., 1999) 
  
Strain Maximum
ethanol 
(g/L) 
Ethanol 
yield
a
 (g/g) 
Maximum 
ethanol 
volumetric 
productivity 
(g L
-1
h
-1
) 
Reference 
E. coli K011 34.7 0.41 1.16 Dien et al., 1997 
E. coli SL40 31.7 0.42 1.12 Dien et al., 1997 
Z. mobilis 
CP4(pZB5) 
22.6 0.45 1.04 Dien 
(unpublished 
data, 1998) 
Z.mobilis 
ZM4(pZB5)
c
 
62 0.46 1.29
b
 Bothast et al., 
1999 
Saccharomyces 
1400 
(pLNH32) 
21.0 0.50 1.60 Moniruzzaman et 
al., 1997 
a 
Grams of ethanol per gram of sugar consumed 
b 
Calculated from  Bothast et al.1999 (p 871) 
c 
Fermentation with a mixture of 65g/L each of glucose and xylose 
 
 
 
 
 
Figure 3.1 Metabolic pathways for xylose utilization in fungi, yeast and bacteria 
(edited from Bothast et al., 1999) 
 
 
 
 
34
 
 
Figure 3.2 Integrated process of microbial metabolic network model 
construction (Covert et al., 2001) 
 
 
 
 
35
 
36
CHAPTER 4 
CO-CULTURE FOR GLUCOSE AND XYLOSE 
COFERMENTATION 
Development in glucose and xylose cofermentation is progressing rapidly. 
Although recombinant strains can achieve glucose and xylose cofermentation, there is no 
robust modified strain which can be use in the bioethanol production from biomass at 
industrial scale due to their limitations. Some research on co-culture indicates that it has 
great potential to solve the problem of glucose and xylose cofermentation. In this chapter, 
we will give a broad overview of the current status of co-culture study on bioethanol 
production. Strains, fermentation mode, condition and performance for current co-
cultures systems will be compared and discussed. Finally, the issues about co-culture 
systems will be addressed.  
4.1 Description and Feature of Co-Culture Systems  
Currently most industrial bioconversions rely on pure microbial cultures, such as 
production of amino acids, nucleotides, antibiotics and enzymes; however, some other 
commercial processes, for instance, alcohol brewing, biological waste-water treatment 
and manufacture of dairy and conventional fermenting products (such as soy sauce, 
pickles) are accomplished by multiple microbial species (Taniguchi et al., 2004). 
Moreover, many environmental bioconversions are catalyzed by mixed microbial cultures 
 
37
with specialists ?working? together with an apparently stable fashion in natural systems. 
Based on some insights from commercial processes and natural systems, the main idea 
for the co-culture of ethanol production from cofermentation of glucose and xylose is to 
utilize two different microbial strains and enable each of them to carry out their own 
tasks stably and effectively. In particular, one microbe could ferment glucose and the 
other microbe could ferment xylose in the same environment, and there is no any 
negative effect between them, such as competition of substrate, inhibition of products and 
so on. In order to achieve this goal, the pivotal part is to find two appropriate microbial 
strains that can fully finish their special tasks in a co-culture system. In addition, the co-
culture system should be resistant and stable for long periods of time despite being 
contaminated by ?outside? microbes or other potentially toxic materials.  
4.2 Strains Used in Co-Culture Systems and Their Interactions 
The natural ethanologenic microorganisms are categorized as yeasts, fungi and 
bacteria. Examples of these yeasts are S. cerevisiae, Kluyveromyces marxianus, P. stipitis, 
C. shehatate and Pachysolen tannophilius. Examples of these bacteria are Z. mobilis, 
Zymobacter palmae, Clostridium cellulolyticum, E. coli and Bacillus stearothermophilus. 
Table 4.1 summarized all of the known strain combinations currently used in co-culture 
systems. In these co-culture systems, P. stipitis, P. tannophilu, C. shehatae and 
recombinant E. coli are usually responsible of for xylose fermentation; K. marxianus and 
S. cerevisiae or its respiratory deficient mutant are normally used as the glucose-
fermenting microbe.  
 
38
To gain greater insights, we performed a statistic analysis on 18 strain 
combinations and listed the result in Table 4.2. As shown in Table 4.2, the combination 
of P. stipitis and S. cerevisiae appeared eight times, and the combination of P. stipitis and 
respiratory deficient mutant S. cerevisiae appeared three times; the combination of C. 
shehatae and S. cerevisiae appeared two times in the total of 18 strain combinations. In 
addition, the combination of C. shehatae and respiratory deficient mutant S. cerevisiae 
appeared once. In respect to a bacteria and yeast combination, S. cerevisiae with E. coli. 
and P. tannophilis with Z. Mobilis appeared once each.  As to the combination of bacteria 
with bacteria, there appears to be no research in this area. Perhaps this is due to the lack 
of bacterial strains that can effectively ferment xylose. From Table 4.2, we can conclude 
that the most commonly used combination in current co-culture systems is the 
combination of P. stipitis and S. cerevisiae since both of them are yeasts, therefore they 
have much better compatibility as well as similar fermentation conditions, such as 
temperature and pH. Furthermore, S. cerevisiae has been regarded as one of the most 
efficient glucose-fermenting yeast with a number of attractive properties (safe, high 
ethanol tolerance etc.). And as one of the best xylose-fermenting yeasts, P. stipitis can 
convert xylose at a relatively high conversion rate and with high yields (Grootjen et al., 
1991b). Therefore, their combination should be considered as the natural choice. 
However, some researchers found that a limited oxygen supply was optimal for P. stipitis 
to ferment xylose (Grootjen et al., 1990b and 1991b). In order to solve the problem of 
limited oxygen condition for P. stipitis to ferment xylose lowering the S. cerevisiae 
fermentation yield, respiratory deficient (RD) mutants of S. cerevisiae have be used in 
some co-culture systems in stead of normal S. cerevisiae (Laplace et al., 1993b; 
 
39
Taniguchi et al., 1997a; Kordowska-Wiater and Targo?ski, 2001). Considering the 
function of xylose fermenting, P. stipitis probably could be combined with recombinant E. 
coli and Z. mobilis in co-culture systems. Also, C. shehatae could be associated with K. 
Marxianus, recombinant E. coli and Z. Mobilis for simultaneous glucose and xylose 
fermenting. Moreover, it is meaningful to test the combination of P. tannophilis with K. 
Marxianus, S. cerevisiae and/or its RD mutants, recombinant E. coli. These can be a 
potential direction for the future co-culture research. 
When selecting strains for a co-culture system, it is very important to understand 
the interactions and investigate the compatibility of these associated strains. Generally 
speaking, possible interactions between two microbial species can be categorized as 
neutralism, positive interactions (commensalisms, protocooperation and mutualism), and 
negative interactions (amensalism, competition, parasitism, and predation) (Taniguchi et 
al., 2004). Neutralism means that there is no ill or harmful effect between two different 
populations in the system. In commensalisms, one population benefits while the other 
remains unaffected. Protocooperation (synergism) benefits both populations, but the 
association is not obligatory. Mutualism is an obligatory relationship between two 
microbial populations that benefits both populations. Close mutualistic relationships are 
known as ?symbiosis.?  Different from positive interactions, amensalism and competition 
represent a negative relationship between two microbial populations. When one microbial 
population produces a substance that is inhibitory to the other population, the relationship 
is called amensalism. Competition occurs when two populations use the same resource, 
either space or a limiting nutrient. Parasitism exerts a negative influence on susceptible 
 
40
host populations and benefits the parasite. Predation typically occurs when one organism, 
the predator, engulfs and digests another, the prey.  
For an effective co-culture system, utilization of positive interactions and 
avoidance of negative interactions between two microorganisms should be the criterion 
for selecting microorganisms. Although Laplace and his coworker have presented the 
compatibility tests and used resistance to mitochondrial inhibitors assays to type and 
quantify both strains in the co-culture process (Laplace et al., 1992), there is no research 
on developing impactful experimental methods for studying the interactions or 
distinguishing different microbial species in co-culture systems. Obviously, this can be 
one of future directions for co-culture research as well.  
4.3 Fermentation Mode and Condition of Co-Culture Systems  
All of the co-culture systems that have been studied so far are summarized in 
Table 4.3. As shown in this table, most of these co-culture systems have good 
fermentation performances, such as higher overall ethanol yield, shorter fermentation 
time, and higher substrate utilization. It turned out that co-culture could be a very 
appealing strategy for increasing ethanol yield and reducing the whole processing cost of 
bioethanol production. In this part, we will discuss the fermentation modes and 
conditions used in these co-culture systems.  
4.3.1 Fermentation Mode 
Normally, there are two fermentation modes utilized in co-culture systems: batch 
and continuous. An obvious difference between these two modes is that there is no feed 
 
41
and outflow for a batch fermentor, but a continuous fermentor (chemostat) has its feed 
and outflow. A simple schematic of a batch fermentor and a typical chemostat is shown 
in Figure 4.1.  
From Table 4.3, we can see that most current co-culture systems are operated in a 
batch mode, because it is simple and the process is easy to control. However, the batch 
mode has some drawbacks. For instance, the existing glucose can suppress the xylose 
fermentation especially at the initial stage, because the xylose conversion is inhibited 
completely at glucose concentration of 2.3 g/L and higher (Grootjen et al., 1991b). On the 
other hand, when the glucose is close to depletion, the high ethanol concentration (around 
30 g/L) can inhibit xylose fermentation (Rouhollah et al., 2007). 
Different from batch mode, continuous fermentation can avoid the accumulation 
of ethanol and other inhibiting metabolites in the system due to the existence of outflow 
(Lebeau et al., 1997). Moreover, by adjusting the volumetric feed rate (dilution rate), the 
glucose concentration in the chemostat can be controlled below 2.3 g/L. Therefore, fast 
and simultaneous conversion of glucose and xylose with co-cultured systems can be 
easily achieved by continuous fermentation. If S. cerevisiae is used in a co-cultivated 
system, its high fermentative potential may also generate a glucose concentration low 
enough to allow fast xylose conversion (Laplace et al., 1993b). However, there are no 
good methods to get the two ethanologenic microorganisms in the outflow back into the 
fermentor or to separate the ethanologenic microorganisms within the solution in the 
outflow when using continuous mode. This is a big issue that limits the utilization of 
continuous fermentation in co-culture systems. We expect that this problem should be 
solved in the future co-culture research. 
 
42
4.3.2 Fermentation Condition 
The selection of fermentation conditions, such as temperature, pH, initial sugar 
concentration and composition, aeration or anaerobic environment and so on, should 
mainly depend on the two selected microorganism strains in a co-culture system. 
Normally, the fermentation temperature of a current co-culture system is at the range of 
28~30 ?C no matter which combination is used, because the optimal fermentation 
temperature for most monoculture, including S. cerevisiae, P. stipitis, E. coli, Z. mobilis, 
C. shehatae and P. tannophilis is around 30?C. The pH of fermentation medium is 
controlled at 4.5~7.0 by adding sodium hydrate or hydrogen chloride, since the optimal 
pH for the monoculture of S. cerevisiae, P. stipitis and Z. mobilis is 5.0,  but for E. coli is 
about 7.0 (Zaldivar et al., 2001). Figure 4.2 shows the distribution of pH selection in 
current co-culture systems. From this figure, a conclusion can be drawn that most current 
co-culture systems choose pH values at 5.0, since most current co-cultures are the 
combination of S. cerevisiae and P. stipitis, and pH of 5.0 is the optimal pH for these two 
yeast strains. For cofermentation of glucose and xylose, the initial total sugar 
concentration and composition play an important role in affecting the fermentation 
performance. However, very little research has been performed to determine the factor?s 
effect on the fermentation performance. As shown in Figure 4.3, different researchers 
have chosen different initial total sugar concentration and different glucose/xylose ratio 
in their co-culture systems according to their conventionality. Generally, the xylose 
composition changes from 20%~50% and most cases are at about 30% which is close to 
the xylose composition in the corn stove hydrolysate and other biomass hydrolysate. 
Some research indicates that S. cerevisiae regulates fermentation by sensing the presence 
 
43
of glucose, however, P. stipitis induces fermentative activity in response to oxygen 
limitation (Jeffries et al., 2007). Moreover, Skoog and Hahn-H?gerdal (1990) reported 
the maximum specific productivity and ethanol yield was achieved when the oxygen 
transfer rate below 1mmol/L/h on xylose fermentation by P. stipitis CBS6054. Normally, 
the oxygen transfer rate (OTR) or the average specific oxygen uptake rate was 
maintained at a low level by changing the airflow and/or agitation rate in some current 
co-culture systems (Laplace et al., 1993a,b; Taniguchi et al., 1997a, b). 
In addition to temperature, pH, initial total sugar concentration and composition, 
oxygen transfer rate, inoculum volume is another important factor which influences the 
fermentation performance. However, there is no related research that has been conducted 
so far. Generally, the inoculum volume in current co-culture systems is in the range of 
1%~3% (v/v).  As a result, a study on how the inoculum volume impacts the other 
fermentation conditions and how it impacts performance needs to be conducted in the 
future.  
4.4 Fermentation Performance 
Generally, there are several parameters to evaluate fermentation results, such as 
ethanol yield (Y
p/s
), volumetric ethanol productivity (Q
p
), specific ethanol production rate 
(q
p
), efficiency of substrate utilization (E) and so on. Of these parameters, the ethanol 
yield is the most vital parameter. It is calculated as grams of ethanol produced per gram 
of total sugars consumed (g ethanol/g substrate). The theoretical ethanol yield for glucose 
fermentation or xylose fermentation is 0.51 g ethanol/ g sugar (Krishnan et al., 1999). So 
far the highest ethanol yield for mono-culture has been achieved by using recombinant E. 
 
44
coli K011 and Z. mobilis ZM4. The recombinant strain E. coli K011 can achieve the 
yields of 0.48 and 0.52 g/g for glucose and xylose, respectively. They represent 94% of 
the theoretical yield for glucose, and more than 100% for xylose due to extra ethanol 
produced from the catabolism of other complex nutrient components in the medium 
(McMillan, 1993). When using the recombinant strain of Z. mobilis ZM4 (pZB5), sugar 
concentrations of about 50 g/L glucose and 50 g/L xylose can result in an ethanol 
productivity of approximately 5 gL?1h?1, and a yield of 0.50 g ethanol/g substrate, 
which is 98% of the theoretical yield (Joachimsthal and Rogers, 2000). From Table 4.3, 
we can see that the overall ethanol yield of different co-culture systems changes from 
0.29~0.50 g ethanol/ g substrate. The highest overall ethanol yield was obtained at 0.5 g 
ethanol/g substrate which is about 98% of the theoretical yield (Taniguchi et al., 1997a). 
By employing the strain combination of P. stipitis CBS5773 and respiratory deficient 
mutant S. cerevisiae no. 7 in batch fermentation, Taniguchi et al. obtained greater 
volumetric ethanol productivity with 0.94 gL
?1
h
?1
 and the maximum ethanol 
concentration with 37.5 g/L which is the highest value in all of the current co-cultures 
systems. Consequently, under some optimal conditions, excellent fermentation 
performance (high ethanol yield, high average volumetric productivity and final ethanol 
concentration) can be realized through co-culture.  
4.5 The Current Issues and Related Strategies in the Co-Culture Research  
Jeffries (1985) suggested the minimum values (Table 4.4) of some important 
performance parameters for a commercially attractive xylose conversion process. In order 
to compare the performance parameters of different single strains (wild type and 
 
45
recombinant strains) and co-culture systems, we extracted some experimental results for 
each case from literatures and listed them in Table 4.5. Comparing the results shown in 
Table 4.5 with the minimum requirements for commercial xylose conversion process, we 
can conclude that co-culture is a very appealing strategy to be used in the ethanol 
production at industrial scale. However, there are some shortcomings that exist for the 
current culture systems. We will address the issues of using the P. stipitis and S. 
cerevisiae co-culture as an example, since this co-culture is a well-studied case with the 
best fermentation performance so far.  
First of all, there is diauxic growth behavior for most microbes, such as P. stipitis, 
E. coli and other strains. This growth pattern was originally discovered by Monod (1942). 
When growing in multiple substrates, microbes with the diauxic behavior actively control 
substrate uptake by regulating their internal repertoire of enzymes. Normally, there are 
two growth phases. The first growth phase involves exclusive consumption of the 
preferred substrate. After the preferred substrate has been completely exhausted, the 
microorganisms undergo a lag period during which they switch to synthesizing the 
enzymes needed to metabolize the less preferred substrate. For example, when both of 
glucose and xylose simultaneously present in the environment, P. stipitis ferments 
glucose first, then ferments xylose after the glucose concentration is reduced to 2.3 g/L 
(Grootjen et al., 1991c). In this case, it is hard to achieve cofermentation of glucose and 
xylose simultaneously, especially at the beginning of batch fermentation. Continuous 
fermentation probably can solve this problem with proper dilute rate to keep the 
concentration of glucose in the fermentor is lower than 2.3 g/L. Unfortunately, there is 
little research about this strategy.  
 
46
Secondly, there is different fermentation mechanisms between two different 
strains. For instance, S. cerevisiae regulates fermentation by sensing the presence of 
fermentable sugars such as glucose, while P. stipitis induces fermentative activity in 
response to oxygen limitation. However, when oxygen exists in the medium, pyruvate is 
metabolized by cellular respiration but fermentation for S. cerevisiae and ethanol is not 
produced, therefore, the total ethanol yield will be reduced. In order to avoid the 
reduction of ethanol yield with limited oxygen condition for the co-culture system, it is 
better to use RD mutant S. cerevisiae cultivated with P. stipitis in the system. Another big 
issue is that there is lack of understanding on interspecies interactions (synergic effect, 
competitive effect and so on). In addition, the possible underlying metabolic changes 
caused by genetic or environmental perturbations (such as the metabolic products of 
another strain in the co-culture system) for each strain need to be studied. One possible 
way for solving this issue is the integration of experimental and computational methods. 
The approach includes developing a robust mathematical model that can describe the 
dynamic characteristics of the model system and predict the changes upon environmental 
conditions, devising effective measurement methods to measure some important 
intracellular metabolites concentrations in vivo by using nuclear magnetic resonance 
(NMR) spectroscopy and mass spectrometry combined with gas chromatography (GC-
MS).  
 
 
 
 
 
47
 
 
 
Table 4.1 The combinations of two strains used in different co-culture 
systems for ethanol production 
 
  Strains References 
P. stipitis + K. marxianus Rouhollah et al., 2007 
P. stipitis + S. cerevisiae Rouhollah et al., 2007 
Z. mobilis + P. tannophilus 
(Successive inoculation) 
Fu and Peiris, 2008 
S. cerevisiae + P. tannophilus Qian et al., 2006 
S. cerevisiae + recombinant E. coli Qian et al., 2006 
coimmobilized  P. stipitis + S. cerevisiae Bari et al., 2004 
Restricted catabolite repressed mutants 
of P. stipitis + S. cerevisiae 
Kordowska-Wiater and Targo?ski, 2002 
Restricted catabolite repressed mutants 
of P. stipitis + RD mutant of S. 
cerevisiae 
Kordowska-Wiater and Targo?ski, 2002 
P. stipitis + RD mutant S. cerevisiae Kordowska-Wiater and Targo?ski, 2001 
coimmobilized  S. cerevisiae + C. 
shehatae 
Lebeau et al., 1997 
P. stipitis + S. cerevisiae Taniguchi et al., 1997a 
P. stipitis + RD mutant S. cerevisiae Taniguchi et al., 1997a 
P. stipitis + RD mutant S. diastaticus Laplace et al., 1993a 
C. shehatae + S. cerevisiae Laplace et al., 1993b 
C. shehatae+ respiratory deficient 
mutant S. cerevisiae 
Laplace et al., 1993b 
coimmobilized P. stipitis + S. cerevisiae Grootjen et al., 1991a and 1990a 
immobilized P. stipitis 
+ suspended S. cerevisiae 
Grootjen et al., 1991a 
 
 
 
 
48
 
 
 
 
Table 4.2 Statistics of the strain combinations in different co-culture systems 
 
Strain Combination Appearance Times/Total Times 
P. stipitis + K. marxianus 1/18 
P. stipitis + S. cerevisiae 8/18 
P. stipitis + RD mutant S. cerevisiae 3/18 
P. stipitis + RD mutant S. diastaticus 1/18 
C. shehatae + S. cerevisiae 2/18 
C. shehatae + RD mutant S. cerevisiae 1/18 
E. coli + S. cerevisiae 1/18 
Z. mobilis + P. tannophilis 1/18 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
49
Table 4.3 Summary of different co-culture systems 
 
Co-culture 
Systems 
Fermentation 
Mode  
Fermentation  
Condition 
Performance  Reference 
P. stipitis 
CCUG18492+K. 
marxianus 
Batch Initial substrate concentration:30g/L 
glucose, 30g/L xylose, 12g/L 
mannose, 8g/L galactose  
pH: 4.5 
Working volume:100mL 
 
Y
p/s
: 0.36   
Y
x/s
: 0.08 
Q
pmax
: 1.08 
C
E,max
: 31.87 
T
F
: 72 h   
E: 99% 
Rouhollah et 
al., 2007 
P. stipitis 
CCUG18492 + 
S. cerevisiae  
Batch Initial substrate concentration: 30g/L 
glucose, 30g/L xylose, 12g/L 
mannose, 8g/L galactose  
pH: 4.5 
Working volume:100mL 
 
Y
p/s
: 0.41   
Y
x/s
: 0.08 
Q
pmax
: 0.77 
C
E,max
: 29.45 
T
F
: 60 h   
E: 94% 
Rouhollah et 
al., 2007 
Z. mobilis + P. 
tannophilus 
(Successive 
inoculation)  
Batch Initial substrate concentration:60g/L 
glucose and 40g/L xylose 
Temperature: 30?C 
Inoculations were carried out 
successively with firstly Z. mobilis, 
and after all the glucose had been 
converted to ethanol. P. tannophilus 
we then inoculated.  
Cofermentation with no aeration at 
glucose fermentation stage and full 
aeration at xylose fermentation stage 
 
Y
p/s
: 0.29   
X
max
: 2.41?10
7 
(Z. mobilis)  
X
max
: 9.4?10
7 
(P. 
tannophilus)  
Q
P
: 2.38 
T
F
: 12 h (Z. mobilis)  
T
F
: 120 h (P. tannophilus)  
 
Fu and 
Peiris, 2008 
Z. mobilis + P. 
tannophilus 
(Successive 
inoculation)  
Batch Initial substrate concentration:60g/L 
glucose and 40g/L xylose 
Temperature: 30?C 
Inoculations were carried out 
successively with firstly Z. mobilis, 
and after all the glucose had been 
converted to ethanol. P. tannophilus 
we then inoculated.  
Cofermentation with no aeration at 
glucose fermentation stage and a 
aeration level <1 mmol/L/h at xylose 
fermentation stage 
 
 
Y
p/s
: 0.33   
X
max
: 5.1?10
7 
(Z. mobilis)  
X
max
: 5.7?10
7 
(P. 
tannophilus)  
Q
P
: 2.32 
T
F
: 12 h (Z. mobilis)  
T
F
: 144 h (P. tannophilus)  
 
Fu and 
Peiris, 2008 
S. cerevisiae 
2.535 + P. 
tannophilis 
ATCC 2.1662 
Batch Treated or untreated Softwood 
hydrolysate was used as the substrate 
in the fermentation.  
Temperature: 30?C 
pH: 5.5 
150mL hydrolysate and 50mL 
inoculum of co-cultures  
 
For treated softwood 
hydrolysate fermentation by 
adapted co-culture  
Y
p/s
: 0.49 
Q
P
: 0.38 
Sugar consumed (%):>99 
Ethanol(g/L): 18.2 
 
Qian et al., 
2006 
S. cerevisiae + 
recombinant E. 
coli 
Batch Treated or untreated Softwood 
hydrolysate was used as the substrate 
in the fermentation.  
Temperature: 30?C 
pH: 7.0 
150mL hydrolysate and 50mL 
inoculum of co-cultures  
 
 
For treated softwood 
hydrolysate fermentation by 
adapted co-culture 
Y
p/s
: 0.45 
Q
P
: 0.71 
Sugar consumed (%):>99 
Ethanol(g/L): 17.1 
Qian et al., 
2006 
P. stipitis 
(NRRL Y-
11544) + S. 
cerevisiae 
coimmobilized 
in Ca-alginate 
gel beads 
Batch Mixed sugar syrups containing 45 
g/L glucose and 12 g/L xylose was 
used as substrate.   
Temperature: 30?C 
pH: 5.5   
 
 
 
Best conditions for the 
cofermentation of 40g/L 
glucose and 10 g/L xylose  
Y
p/s
: 0.396 
Conversion: 0.995
 g consumed/g 
initial
  
Bari et al., 
2004 
 
50
Co-culture 
Systems 
Fermentation 
Mode  
Fermentation  
Condition 
Performance  Reference 
Restricted 
catabolite 
repressed 
mutant P. 
stipitis 
CCY39501 + 
respiratory 
deficient mutant 
S. cerevisiae 
Ja(a)  
 
Batch  Initial substrate concentration:35g/L 
glucose and 15g/L xylose 
Temperature: 28?C 
pH: 5.5  
Fermentation time: 120 h 
1% v/v of respiratory deficient 
mutant S. cerevisiae V
30 
and 1% v.v 
of P. stipitis were inoculated in 150 
mL F
3
 medium.  
 
S
G
: 100.00% 
S
K
: 68.00% 
C
E,max
: 20.30 
Y
p/s
: 0.45   
Y
x/s
: 0.20 
Q
P
: 0.169 
 
Kordowska-
wiater and 
Targo?ski, 
2002 
P. stipitis 
CCY39501+ S. 
cerevisiae V
30
 
Batch Initial substrate concentration:35g/L 
glucose and 15g/L xylose 
Temperature: 28?C 
pH: 5.5   
Fermentation time: 96 h 
2% v/v of respiratory deficient 
mutant S. cerevisiae V
30 
and 1% v.v 
of P. stipitis were inoculated in 150 
mL medium.  
 
S
G
: 99.71% 
S
K
: 26.67% 
C
E,max
:15.00 
Y
p/s
: 0.39   
Y
x/s
: 0.14 
Q
P
: 0.318 
 
Kordowska-
Wiater and 
Targo?ski, 
2001 
P. stipitis 
CCY39501+ 
respiratory 
deficient mutant 
S. cerevisiae V
30
 
Batch Initial substrate concentration:35g/L 
glucose and 15g/L xylose 
Temperature: 28?C 
pH: 5.5  
Fermentation time: 96 h 
2% v/v of respiratory deficient 
mutant S. cerevisiae V
30 
and 1% v/v 
of P. stipitis were inoculated in 150 
mL medium.  
 
S
G
: 100.00 % 
S
K
: 99.67 % 
C
E,max
: 18.80 
Y
p/s
: 0.38   
Y
x/s
: 0.12 
Q
P
: 0.264 
 
Kordowska-
Wiater and 
Targo?ski, 
2001 
S. cerevisiae 
CBS 1200 + C. 
shehatae ATCC 
22984 
coimmobilized 
in a two-
chambered 
bioreactor 
 
Batch  Initial substrate concentration: 10g/L 
glucose and 4.5g/L xylose 
Initial cell loading:  
C. shehatae 0.65 mg dry wt/mL  
S. cerevisiae 5.00mg dry wt/mL 
 
 
Best conditions: 
T
m
: 230 h   
Total Ethanol 
produced:5.15g 
Y
p/s
: 0.47 
Y
x/s
: 0.088 
Q
P
: 7.5 
Lebeau et 
al., 1997 
P. stipitis 
CBS5773 + S. 
cerevisiae no. 7    
Batch Initial substrate concentration: 50g/L 
glucose and 25g/L xylose 
The initial concentration of P. stipitis 
and S. cerevisiae were tentatively 7.1 
g/L and 1.5g/L, respectively.   
pH: 5.0  
Working volume: 1L 
Fermentation time: 40h 
qO
2
 was controlled at 66.7 mg/g 
cell/h for glucose consumption and 
then controlled at  14.3 mg/g cell/h 
for xylose consumption .  
 
 
Y
p/s
: 0.39   
C
E,max
: 29.4 
Q
P
: 0.74 
 
Taniguchi et 
al., 1997a 
P. stipitis 
CBS5773 + 
respiratory 
deficient mutant 
S. cerevisiae no. 
7  
Batch Initial substrate concentration: 50g/L 
glucose and 25g/L xylose 
The initial concentration of P. stipitis 
and S. cerevisiae were tentatively 7.1 
g/L and 1.5g/L, respectively.   
pH: 5.0 (controlled by 2N NaOH and 
2N HCl) 
Working volume: 1L 
Fermentation time: 40h 
 
 
 
Y
p/s
: 0.50   
C
E,max
: 37.5 
Q
P
: 0.94 
 
Taniguchi et 
al., 1997a 
 
51
Co-culture 
Systems 
Fermentation 
Mode  
Fermentation  
Condition 
Performance  Reference 
P. stipitis 
CBS5773 + S. 
cerevisiae no. 7    
Batch Initial substrate concentration: 50g/L 
glucose and 25g/L xylose 
The initial concentration of P. stipitis 
and S. cerevisiae were tentatively 3.5 
g/L and 0.75 g/L, respectively.   
Fermentation time: 68h 
qO
2
 was controlled at 66.7 mg/g 
cell/h for glucose fermentation stage 
and 14.3 mg/g cell/h for xylose 
fermentation stage.  
 
Y
p/s
: 0.35   
C
E,max
: 26.2 
Q
P
: 0.39 
 
Taniguchi et 
al., 1997b 
P. stipitis 
CBS5773 
(fermentor A) + 
S. cerevisiae no. 
7 (fermentor B)   
Batch Initial substrate concentration: 50g/L 
glucose and 25g/L xylose 
The initial concentration of P. stipitis 
and S. cerevisiae were 7.0 g/L and 
0.75g/L, respectively.   
Filtered air was sparged at constant 
flow rate (0.2vvm) into fermentor A. 
Meanwhile, nitrogen gas was sparged 
at 0.2 vvm into fermentor B.   
Fermentation time: 56h 
Working volume in Fermentor A: 1L 
Working volume in Fermentor B: 1L 
Total working volume:2 L 
 
For P. stipitis in Fermentor 
A  
Y
p/s
: 0.44   
C
E,max
: 33.1 
Q
P
: 0.59 
For S. cerevisiae in 
Fermentor B 
Y
p/s
: 0.45   
C
E,max
: 33.7 
Q
P
: 0.60 
 
Taniguchi et 
al., 1997b 
Taniguchi 
and Tanaka, 
2004 
P. stipitis + 
respiratory 
deficient mutant 
S. diastaticus 
Continuous  3% (v/v) inoculum was used.  
Temperature: 30?C 
pH: 5.0 
Working volume: 1.5L 
O
2
 transfer rate:1.75 mmol/h 
Sugar mixture in fermentation 
medium: 70% glucose and 30% 
xylose  
 
 
 
When the initial substrate 
concentration did not exceed 
20 g/L, the substrate was 
entirely consumed.  
When the initial substrate 
concentration equaled to 
80g/L, the residual xylose 
concentration reached 20.5 
g/L.  
With the initial substrate 
concentration at 50 g/L, the 
xylose was entirely 
consumed when dilution rate 
did not exceed 0.006 h
-1
 
whereas the glucose was 
entirely consumed whatever 
the D is.  
 
Laplace et 
al., 1993a 
C. shehatae 
ATCC 22984 + 
S. cerevisiae 
CBS 1200 
Batch Initial substrate concentration: 14g/L 
glucose and 6g/L xylose (a mixture 
of 70% glucose and 30% xylose) 
Inoculums of each yeast strain were 
1.5%.  
Temperature: 30?C 
pH: 5.0 
Working volume: 1.5L 
Stirring speed:800rpm 
Aeration rate: 0.005vvw 
Oxygen transfer rate:1.75mmol/l/h 
 
T
F
: 14 h   
C
E,max
: 14.5 g/L 
Y
p/s
: 0.39 
S
G
: 100 % 
S
K
: 8 % 
 
 
Laplace et 
al., 1993b 
C. shehatae 
ATCC 22984 + 
respiratory 
deficient mutant 
S. cerevisiae 
CBS1200 
Batch Initial substrate concentration: 14g/L 
glucose and 6g/L xylose (a mixture 
of 70% glucose and 30% xylose) 
Inoculums of each yeast strain were 
1.5%.  
Temperature: 30?C 
pH: 5.0 
Working volume: 1.5L 
Stirring speed:800rpm 
Aeration rate: 0.005vvw 
Oxygen transfer rate:1.75mmol/l/h 
T
F
: 14 h   
C
E,max
: 14.7 g/L 
Y
p/s
: 0.40 
S
G
: 100 % 
S
K
: 6 % 
 
Laplace et 
al., 1993b 
 
52
Co-culture 
Systems 
Fermentation 
Mode  
Fermentation  
Condition 
Performance  Reference 
C. shehatae 
ATCC 22984 + 
respiratory 
deficient mutant 
S. cerevisiae 
CBS1200 
Continuous  Initial substrate concentration: 14g/L 
glucose and 6g/L xylose (a mixture 
of 70% glucose and 30% xylose) 
Inoculums of each yeast strain were 
1.5%.  
Temperature: 30?C 
pH: 5.0 
Working volume: 1.5L 
Stirring speed:800rpm 
Aeration rate: 0.005vvw 
Oxygen transfer rate:1.75mmol/l/h 
T
F
: 14 h   
C
E,max
: 14.5 g/L 
Y
p/s
: 0.39 
S
G
: 100 % 
S
K
: 8 % 
 
Laplace et 
al., 1993b 
coimmobilized 
P. stipitis CBS 
5773 + S. 
cerevisiae CBS 
8066 
Continuous  Temperature: 30?C 
pH: 5.0 
Working volume: 400mL 
 
 
Best condition: 
D: 0.07 
Influent concentration: 
Glucose: 40.9 g/L 
Xylose: 10.5 g/L 
S
G
: 100 % 
S
K
: 37 % 
 
Grootjen et 
al., 1991a 
immobilized P. 
stipitis CBS 
5773 + 
suspended S. 
cerevisiae CBS 
8066 
Continuous  Temperature: 30?C 
pH: 5.0 
Working volume: 400mL 
 
 
Best condition: 
D: 0.11 
Influent concentration: 
Glucose: 40.5 g/L 
Xylose: 11.2 g/L 
S
G
: 100 % 
S
K
: 11 % 
Grootjen et 
al., 1991a 
coimmobilized 
P. stipitis CBS 
5773 + S. 
cerevisiae CBS 
8066  
Continuous  The medium contained a mixture of 
glucose 40 g/L and xylose 10 g/L 
Temperature: 30?C 
pH: 5.0 
Working volume: 200mL 
Best condition: 
D: 0.375 
Y
p/s
: 0.4 
C
E,max
: 20 
 
   
Grootjen et 
al., 1990a 
Y
p/s
: Overall Ethanol yield (g ethanol/g substrate); X
max
: maximum biomass yield (CFU/ml); Q
P
: volumetric productivity of ethanol 
(g/L/h); Q
pmax
: maximum volumetric ethanol productivity (g/L/h); T
F
: time of maximum ethanol concentration to be reached (h);  
T
m
: time at which maximum ethanol produced reached (h); C
E,max
: maximum ethanol concentration (g/L); Y
x/s
: cell biomass yield (g 
cell biomass/g substrate); E: efficiency of substrate utilization (%); S
G
: glucose used (%); S
K
: xylose used (%); qO
2
: specific oxygen 
uptake rate, mg/g cell/h; D: dilution rate (h
-1
) 
 
 
 
 
 
 
 
 
 
 
53
 
 
Table 4.4 Minimum requirements of performance parameters for commercial 
xylose conversion process 
 
Performance parameter  Minimum level to achieve for 
commercialization  
Ethanol yield  0.40 g/g 
Final ethanol concentration 5% (w/v)  
Average volumetric productivity  1.4 g L
-1 
h
-1
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
54
Table 4.5 Comparison of fermentation performance of different single strain (wild 
type, recombinant strains) and co-culture systems 
 Strains  Carbon 
source 
(g/L) 
Ethanol  
Concent
ration 
(g/L) 
Ethanol 
yield  
(g g
-1
) 
Ethanol 
volumetric 
productivity 
(g L
-1 
h
-1
)  
Reference 
P. stipitis 
Wild type  
X50 - 0.43 0.9 Du Preez et 
al., 1986 
Recomb
inant S. 
cerevisi
ae 1400 
(XYL1, 
XYL2) 
integrate
d 
X50 2.7 0.05 0.02 Tantirungkij 
et al., 1993 
Recombinant 
S. cerevisiae 
(XYL1, XYL2, 
XKS1) 
integrated 
G90, 
X40 
60.0 0.46 1.15 Ho et al.,  
1998 
Recombinant 
S. cerevisiae 
TMB3001 
( XYL1, XYL2, 
XKS1) 
integrated 
X50 - 0.31 0.20 Jeppson et al., 
2002 
Recombinant 
Z.mobilis CP4 
(pZB5)  
G25, 
X25 
24.2 0.48 0.81 Zhang et al., 
1995 
Recombinant 
Z.mobilis 
CM4 (pZB5) 
G65, 
X65 
62.0 0.46 1.29 Joachimsthal 
et al.,  1999 
Single 
strain 
Recombinant 
E. coli 
K011(pdc, 
adhB, frd) 
integrated  
X80 41.6 0.52
a
 0.87 Ohta et al.,  
1991 
P. stipitis + S. 
cerevisiae 
G50, 
X25 
26.2 0.35 0.39 Taniguchi et 
al., 1997b 
Co-
cultur
e 
P. stipitis + 
RD mutant S. 
cerevisiae 
G50, 
X25 
37.5 0.50 0.94 Taniguchi et 
al., 1997a 
a 
Attributed to nutrient supplement; G stands for glucose; X stands for xylose  
 
Sterile
Medium
Reservoir
Outflow
Gas Sparger
Stirrer
Temperature
Control
pH Control
 
A. Continuous fermentor   
Motor
Paddle
Blade
Gas
Sparger
pH Control
Temperature
Control
 
B. Batch fermentor 
Figure 4.1 Schematic of batch and continuous fermentors 
 
 
55
 
 
The selection of pH in current co-culture systems
0
2
4
6
8
10
4.50 5.00 5.50 6.00 6.50 7.00
pH
Number of 
Co-culture 
Cas
e
s
 
Figure 4.2 Distribution of pH values in current co-culture systems 
 
 
 
 
 
 
 
 
 
 
 
56
 
10
14
30
35
40
45
50
60
4.5
6
10
12
15
25
30
40
40, 60
30, 30
25, 50
15, 35
12, 45
10, 40
6, 14
4.5, 10
0
1
2
3
4
N
u
m
b
er
 o
f
 C
o
-cu
l
t
u
r
e C
a
ses
G
l
u
c
o
s
e
 
Co
n
c
e
n
t
r
a
t
i
o
n
 
(
g
/
L
)
 
x
yl
o
s
e 
co
n
c
en
tr
a
ti
o
n
 (
g
/L
) 
 
Figure 4.3 The distribution of initial glucose/xylose concentration in current co-
culture systems 
 
 
 
 
57
 
58
CHAPTER 5 
CONCLUSIONS AND RECOMMENDATIONS  
Currently, research efforts of unraveling xylose fermentation in the process of 
bioethanol production from biomass are mainly focusing on two aspects: 1) genetic 
modification of glucose-fermenting microbial strains to ferment xylose; and 2) co-culture 
of glucose-fermenting and xylose-fermenting strains. This research presented some initial 
results regarding to these two routes. A comprehensive overview of each method has 
been given. In this chapter, we will draw some conclusions based on the investigated 
work. In addition, future research directions associated with these two routes will be 
presented. 
5.1 Conclusions  
The conclusions we can draw based on the initial investigation are listed below:  
1) Cofermentation of glucose and xylose can be achieved by genetic modification 
of glucose-fermenting microbial strains or co-culture of glucose-fermenting and xylose-
fermenting microbial strains.  
2) lthough successful construction of recombinant strains through genetic 
engineering makes it feasible for single strain to coferment glucose and xylose, 
there are some drawbacks that hinder application of recombinant strains on 
bioethanol production at industrial scale, such as pleiotropic effects caused by 
 
59
gene insertion, non-stability in long-term culture, inhibition of undesirable co-
products 
3) The above drawbacks of recombinant strains may result from a lacking 
of understanding of cell metabolism on a system level. It was shown that 
stoichiometric models of genome-scale metabolic network are powerful tools that 
can help study cells? metabolic functions from the perspective of systems 
engineering. Primary application of genome-scale metabolic network models 
includes prediction of outcomes of genetic manipulation, identification of novel 
metabolic functions and design of microbial strains for desirable products. 
4) Existing experimental results of current co-culture research show that it 
is a potential strategy to accomplish simultaneous cofermentation of glucose and 
xylose by selecting a proper combination of glucose-fermenting strains and 
xylose-fermenting strains. 
5) The most widely used strain combination during the current co-culture 
systems is the combination of P. stipitis and S. cerevisiae or its RD mutants. The 
best fermentation performance is obtained by P. stipitis CBS5773 and RD mutant 
S. cerevisiae no. 7 in batch fermentation. The overall ethanol yield is 0.50 g 
ethanol/ g substrate, ethanol volumetric productivity reaches 0.94 gL
?1
h
?1
 and the 
maximum ethanol concentration is 37.5 g/L. 
6) Although co-culture exhibits some appealing traits to improve the 
simultaneous cofermentation of glucose and xylose, there are some shortcomings 
of this strategy. For example, the diauxic effect exists for most xylose-fermenting 
strain candidates (P. stipitis and E. coli). Another issue is the difference of 
 
60
fermentation conditions for different strains in the co-culture system.  Solving 
these problems can hasten the process of co-culture utilization in bioethanol 
production from biomass on a commercial scale. 
5.2 Recommendations for Future Work  
We conclude with a summary of related research directions that deserve future 
investigation. 
5.2.1. Stability of recombinant strains in continuous fermentation  
When attempting to develop recombinant strains for industrial bioethanol 
production process, genetic stability and environmental acceptability should be 
considered. There is no research that has been conducted to test and improve the stability 
of recombinant strains in continuous fermentation so far. This needs further investigation.  
5.2.2. Genome-scale metabolic network model for P. stipitis 
Due to its importance on xylose fermentation, it is necessary to build a genome-
scale metabolic network model for P. stipitis. Since the complete genome of P. stipitis 
has been sequenced and annotated in 2007 (Jeffries, 2007), it is possible to set up a 
genome-scale metabolic model for it. The principles that have been developed and the 
experiences that have been gained from modeling E. coli (Varma and Palsson, 1993a, b; 
Varma and Palsson, 1994; Pramanik and Keasling, 1997) and S. cerevisiae (F?rster et al.,  
2003; Jin and Jeffries, 2004) could be directly applied to modeling P. stipitis. With the 
availability of genome sequences for more and more ethanologenic microorganisms, such 
 
61
as Z. mobilis, C. shehatae and P. tannophilis, building genome-scale metabolic network 
model for each of them will be possible in the future.  
5.2.3. Different strain combinations for co-culture systems 
Based on our investigation, we can draw a conclusion that the most used 
combination in current co-culture systems is P. stipitis and S. cerevisiae or its respiratory 
deficient mutant. Some research work has been done on this combination due to the 
better compatibility and good match for co-culture purpose. In respect to the combination 
of yeast and bacteria, S. cerevisiae with E. coli. and P. tannophilis with Z. Mobilis, has 
appeared only one time, respectively. As to the combination of bacteria and bacteria, 
there has been no research performed so far, it is probably due to unavailable bacterial 
strains which can effectively ferment xylose. Considering the function of xylose 
fermenting, P. stipitis could be cultured with recombinant E. coli or Z. mobilis 
simultaneously in the fermentation system. Since C. shehatae is also among the best 
yeasts for fermenting xylose, C. shehatae could be associated with K. Marxianus, 
recombinant E. coli or Z. mobilis for simultaneous glucose and xylose fermenting. 
Another promising area would be to test the combination of P. tannophilis with K. 
Marxianus, S. cerevisiae or its respiratory deficient mutants, recombinant E. coli 
respectively.  All of these would be good potential directions for the future co-culture 
research. 
 
 
62
5.2.4. Effect of different fermentation parameters on fermentation performance in 
different co-culture systems  
Fermentation parameters, such as pH, temperature, initial sugar concentrations 
and compositions, initial inoculum volume, and initial inoculum composition of the two 
strains, have significant influence on the fermentation performance. In addition, different 
co-culture systems require different optimal fermentation parameters. To date, there is 
very little research being done concerning these aspects. These aspects deserve future 
investigation for improving the potential application of co-culture in bioethanol 
production process.  
5.2.5. Kinetic model for co-culture systems  
In principle, kinetic models capture the dynamic properties of systems, and can be 
a powerful tool to help obtain optimum operating conditions for achieving sufficient 
profitability and reduce tests for eliminating extreme possibilities. Diverse kinetic models 
about mono-culture have been proposed in the literature (Thatipamala et al., 1992; Lee et 
al., 1995; Olsson and Hahn-H?gerdal, 1995; Birol et al., 1998; Lee and Huang, 2000; 
Leksawasdi et al., 2001; Hodge and Karim, 2002). However, none of these proposed 
models has put forth a kinetic model that can predict the cell, substrate, and ethanol 
concentrations for co-culture systems in either batch or continuous operation. Probably, a 
major problem associated with setting up this model is the lack of kinetic data and the 
difference between in vivo and in vitro kinetic parameters for co-culture systems. 
Therefore, building kinetic models for potential co-culture systems is also a good future 
research direction. 
 
63
5.2.6. Metabolic network model for co-culture systems  
By using the genome-scale reconstructed metabolic network models for each 
ethanol-producing microorganism, a combined metabolic network model for any 
potential co-culture system that captures the dominant metabolic interactions between 
two strains can be readily built. Elementary mode and extreme pathway analyses for the 
combined metabolic network model can give us valuable information, such as which 
genes are essential for producing ethanol, which enzymes would most likely be regulated 
for changing growth conditions, what kind of metabolic interactions would appear 
between two strains in the co-culture system. This research work could be one of the 
interesting directions for co-culture systems in the future as well. 
 
 
 
 
 
 
 
 
 
 
 
 
64
BIBLIOGRAPHY 
1. Alterthum F, Ingram LO. 1989. Efficient ethanol production from glucose, lactose, 
and xylose by recombinant Escherichia coli. Appl Environ Microbiol 55:1943-1948 
2. Amore R, Kotter P, Kuster C, Ciriacy M, Hollenberg CP. 1991. Cloning and 
expression in Saccharomyces cerevisiae of the NAD(P)H-dependent xylose 
reductase-encoding gene (XYL1) from the xylose-assimilating yeast Pichia stipitis. 
Gene 109:89-97 
3. Amore R, Wilhelm M, Hollenberg CP. 1989. The fermentation of xylose - an 
analysis of the expression of Bacillus and Actinoplanes xylose isomerase genes in 
yeast. Appl Microbiol Biotechnol 30:351-357. 
4. Bari IS, Cuna D, Nanna F, Braccio G. 2004. Ethanol production in immobilized cell 
bioreactors from mixed sugar syrups and enzymatic hydrolysates of steam-exploded 
biomass. Appl Biochem Biotechnol 114:539?557  
5. Birol G, Doruker P, Kirdar B, ?nsan Z? and ?lgen K. 1998. Mathematical 
description of ethanol fermentation by immobilized Saccharomyces cerevisiae. 
Process Biochemistry 33:763-771 
6. Bro C, Regenberg B, F?rster J, Nielsen J. 2006 In silico aided metabolic 
engineering of Saccharomyces cerevisiae for improved bioethanol production, 
Metab Eng 8:102-111  
 
65
7. Bothast RJ, Nichols, NN, and Dien BS. 1999. Fermentations with new recombinant 
organisms. Biotechnol Prog15:867-875  
8. Chandel AK, Chan ES, Rudravaram R, Narasu ML, Rao LV and Ravindra P. 2007. 
Economics and environmental impact of bioethanol production technologies: an 
appraisal. Biotechnol and Molecular Biology Review 2(1):14-32  
9. Chou YC, Zhang M, Mohagheghi A, Evans K & Finkelstein M. 1997. Construction 
and evaluation of a xylose/arabinose fermenting strain of Zymomonas mobilis. 
Abstracts 19th Symposium on Biotechnology for Fuels and Chemicals, Colorado 
Springs, May 4-8 
10. Chu BCH, Lee H. 2007. Genetic improvement of Saccharomyces cerevisiae for 
xylose fermentation. Biotechnol Advances 25: 425-441 
11. Covert MW, Schilling CH, Famili I, Edwards JS, Goryanin II, Selkov E and Palsson 
B?. 2001. Metabolic modeling of microbial strains in silico. Trends in Biochemical 
Sciences 26:179-186 
12. Deanda K, Zhang M, Eddy C, Picataggio S. 1996. Development of an arabinose-
fermenting Zymomonas mobilis strain by metabolic pathway engineering. Appl 
Environ Microbiol 62:4465? 4470 
13. Department of Energy, U. S. 2006. Breaking the biological barriers to cellulosic 
ethanol: a research roadmap resulting from the biomass to biofuels workshop  
14. Dien BS, Cotta MA, Jeffries TW. 2003. Bacteria engineered for fuel ethanol 
production: current status. Appl Microbiol Biotechnol 63:258-266 
 
66
15. Dien BS, Hespell RB, Wyckoff HA, Bothast RJ. 1998. Fermentation of hexose and 
pentose sugars using a novel ethanologenic Escherichia coli strains. Enzyme 
Microbial Technol 23:366-371 
16. Dien BS, Hespell RB, Ingram LO, Bothast RJ. 1997. Conversion of corn milling 
fibrous coproducts into ethanol by recombinant Escherichia coli strains K011 and 
SL40. World J Microbiol Biotechnol 13:619-625 
17. Dumsday GJ, Zhou B, Yaqin W, Stanley GA, Pamment NB. 1999. Comparative 
stability of ethanol production by Escherichia coli K011 in batch and chemostat 
culture. J Ind Microbiol Biotechnol 23:701-708 
18. Du Preez JC, Bosch M, Prior BA. 1986. Xylose fermentation by Candida shehatae 
and Pichia stipitis-effects of pH, temperature and substrate concentration. Enzyme 
Microb Technol 8:360-364 
19. Edwards JS & Palsson B?. 1999. Systems properties of the Haemophilus influenzae 
Rd metabolic genotype. J Biol Chem 274:17410-17416 
20. Edwards JS & Palsson B?. 2000. The Escherichia coli MG1655 in silico metabolic 
genotype: its definition, characteristics, and capabilities. Proc Natl Acad Sci USA 
97:5528-5533  
21. Eliasson A, Christensson C, Wahlbom CF, Hahn-H?gerdal B. 2000. Anaerobic 
xylose fermentation by recombinant Saccharomyces cerevisiae carrying XYL1, 
XYL2 and XKS1 in mineral medium chemostat cultures. Appl Environ Microbiol,  
66:3381-3386 
22. F?rster J, Famili I, Fu P, Palsson B?, Nielsen J. 2003. Genome-scale reconstruction 
of the Saccharomyces cerevisiae metabolic network, Genome Res 13:244-253 
 
67
23. Fu N and Peiris P. 2008. Co-fermentation of a mixture of glucose and xylose to 
ethanol by Zymomonas mobilis and Pachysolen tannophilus. World J Microbiol 
Biotechnol 24:1091-1097 
24. G?rdonyi M, Hahn-H?gerdal B. 2003. The Streptomyces rubiginosus xylose 
isomerase is misfolded when expressed in Saccharomyces cerevisiae. Enzyme 
Microb Technol 32:252-259 
25. Goffeau A. 1997. The yeast genome directory. Nature 387:5-6 
26. Grootjen DRJ, Jansen ML, Van der Lans RGJM, Luyben KchAM. 1991a. Reactors 
in series for the complete conversion of glucose/xylose mixtures by Pichia stipitis 
and Saccharomyces cerevisiae. Enzyme Microb Technol 13:828?833 
27. Grootjen DRJ, Meijlink LHHM, Van der lans RGJM, Luyben KChAM.1990.  
Cofermentation of glucose and xylose with immobilized Pichia stipitis and 
Saccharomyces cerevisiae. Enzyme Microb Technol 12: 860-864 
28. Grootjen DRJ, Meijlink LHHM, Vleesenbeek R, Van der Lans RGJM, Luyben 
KChAM. 1991b. Cofermentation of glucose and xylose with immobilized Pichia 
stipitis in combination with Saccharomyces cerevisiae. Enzyme Microb Technol 
13:530?536 
29. Grootjen DRJ, van der Lans RGJM and Luyben KChAM. 1991c. Conversion of 
glucose/xylose mixtures by Pichia stipitis under oxygen-limited conditions. Enzyme 
Microb Technol 13:648-654  
30. Grotkjaer T, Christakopoulos P, Nielsen J, Olsson L. 2005. Comparative metabolic 
network analysis of two xylose fermenting recombinant Saccharomyces cerevisiae 
strains. Metab Eng 7:437-444 
 
68
31. Ho NW, Chen Z, Brainard AP. 1998. Genetically engineered Saccharomyces yeast 
capable of effective co-fermentation of glucose and xylose. Appl Environ Microbiol  
64: 1852-1859 
32. Hodge DB and Karim MN. 2002. Modeling and advanced control of recombinant 
Zymomonas mobilis fed-batch fermentation. Biotechnol Prog 18:572-579 
33. Ingram LO, Alterthum F, Conway T, Ohta K. 1991. Ethanol production by 
Escherichia coli strains co-expressing Zymomonas pdc and adh genes. U. S. Patent 
No. 5, 000, 000 
34. Ingram LO, Conway T, Clark DP, Sewell GW, Preston JF. 1987. Genetic 
engineering of ethanol production in Escherichia coli. Appl Environ Microbiol  53: 
2420-2425 
35. Ingram LO, Conway T. 1988. Expression of different levels of ethanologenic       
enzymes from Zymomonas mobilis in recombinant strains of Escherichia coli. Appl 
Environ Microbiol 54:397?404 
36. Ingram LO, Gomez PF, Lai X, Moniruzzaman M, Wood BE, Yomano LP, York SW. 
1997. Metabolic engineering of bacteria for ethanol production.  Biotechnol Bioeng 
58:204-214 
37. Jeffries TW. 1985. Emerging technology for fermenting D-xylose. Trends in 
Biotechnology 3:208-212 
38. Jeffries TW. 2006. Engineering yeasts for xylose metabolism. Current Opinion in 
Biotechnology 17:320-326 
39. Jeffries TW, Grigoriev IV, Grimwood J, Laplaza JM, Aerts A, Salamov A, Schmutz 
J, Lindquist E, Dehal P, Shapiro H, Jin YS, Passoth V and Richardson PM. 2007.  
 
69
Genome sequence of the lignocellulose-bioconverting and xylose-fermenting yeast 
Pichia stipitis. Nature Biotechnology 25(3):319-326 
40. Jeffries TW, Jin YS. 2004. Metabolic engineering for improved fermentation of 
pentoses by yeasts. Appl Microbiol biotechnol 63:495-509  
41. Jeppsson M, Johansson B, Hahn-H?gerdal B, Gorwa-Grauslund MF. 2002. Reduced 
oxidative pentose phosphate pathway flux in recombinant xylose-utilizing 
Saccharomyces cerevisiae strains improves the ethanol yield from xylose. Appl 
Environ Microbiol 68:1604-1609 
42. Jin YS and Jeffries TW. 2004 Stoichiometric network constrains on xylose 
metabolism by recombinant Saccharomyces cerevisiae. Metab Eng 6:229-238 
43. Joachimsthal E, Hagget KD, Rogers PL. 1999.  Evaluation of recombinant strains of 
Zymomonas mobilis for ethanol production from glucose/xylose media. Appl 
Biochem Biotechnol 77?79:147?158 
44. Joachimsthal EL, Rogers PL. 2000. Characterization of a high productivity 
recombinant strain of Zymomonas mobilis for ethanol production from 
glucose/xylose mixtures. Appl Biochem Biotechnol 84(6):343?356 
45. Karhumaa K, Sanchez RG, Hahn-H?gerdal B, Gorwa-Grauslund MF. 2007. 
Comparison of the xylose reductase-xylitol dehydrogenase and the xylose isomerase 
pathways for xylose fermentation by recombinant Saccharomyces cerevisiae. 
Microbial Cell Factories 6:5-15 
46. Kordowska-Wiater M, Targo?ski Z. 2001. Ethanol production on the media 
containing glucose and xylose by coculture of Pichia stipitis CCY39501 and 
 
70
respiratory deficient mutant of Saccharomyces cerevisiae V30. Electronic Journal of 
Polish Agricultural Universities, Food Science and Technology 4(2):1-13  
47. Kordowska-Wiater M, Targo?ski Z. 2002. Ethanol fermentation on glucose/xylose 
mixture by co-cultivation of restricted glucose catabolite repressed mutants of 
Pichia stipitis with respiratory deficient mutants of Saccharomyces cerevisiae. Acta 
Microbiol Pol 51:345?352 
48. K?tter P and Ciriacy M. 1993. Xylose fermentation by Saccharomyces cerevisiae. 
Appl Microbiol Biotechnol 38:776-783  
49. Krishnan MS, Ho NWY, Tsao GT. 1999. Fermentation kinetics of ethanol 
production from glucose and xylose by recombinant Saccharomyces 1400 
(pLNH33). Appl Biochem Biotechnol 77-79:373-388 
50. Kuhad RC and Singh A. 1993. Lignocellulose biotechnology: current and future 
prospects. Crit Rev Biotechnol 13:151-172 
51. Kuyper M, Harhangi HR, Stave AK, Winkler AA, Jetten MS, de Laat WT, den 
Ridder JJ, Op den Camp HJ, van Dijken JP, Pronk JT. 2003.  High-level functional 
expression of a fungal xylose isomerase: the key to efficient ethanolic fermentation 
of xylose by Saccharomyces cerevisiae?  FEMS Yeast Res 4:69-78. 
52. Laplace JM, Delgenes JP, Moletta R. 1992. Alcoholic glucose and xylose 
fermentations by the coculture process: compatibility and typing of associated 
strains. Can J Microbiol 38: 654-658 
53. Laplace JM, Delgenes JP, Moletta R, Navarro JM. 1993a. Effects of culture 
conditions on the co-fermentation of a glucose and xylose mixture to ethanol by a 
 
71
mutant of Saccharomyces distaticus associated with Pichia stipitis. Appl Microbiol 
Biotechnol 39:760?763 
54. Laplace JM, Delgenes JP, Moletta R, Navarro JM. 1993b. Cofermentation of 
glucose and xylose to ethanol by a respiratory deficient mutant of Saccharomyces 
cerevisiae co-cultivated with a xylose-fermenting yeast. J Ferment Bioeng 75:207?
212 
55. Laplace JM, Delgenes JP, Moletta R, Navarro JM. 1993c. Ethanol production from 
glucose and xylose by separated and coculture process using high cell density 
systems. Process Biochem 28: 519-525 
56. Lawford HG and Rousseau JD. 2002. Performance testing of Zymomonas mobilis 
metabolically engineered for cofermentation of glucose, xylose, and arabinose. Appl 
Biochem Biotechnol 98-100:429-448 
57. Lebeau T, Jouenne T, Junter GA. 1997.  Simultaneous fermentation of glucose and 
xylose by pure and mixed cultures of Saccharomyces cerevisiae and Candida 
shehatae immobilized in a two-chambered bioreactor. Enzyme Microb Technol 
21:265?272 
58. Lebeau T, Jouenne T, Junter GA. 2007. Long-term incomplete xylose fermentation, 
after glucose exhaustion, with Candida shehatae co-immobilized with 
Saccharomyces cerevisiae. Microbiological Research 162:211-218  
59. Lee WC and Huang CT. 2000. Modeling of ethanol fermentation using Zymomonas 
mobilis ATCC 10988 grown on the media containing glucose and fructose. Biochem 
Eng J 4:217-227 
 
72
60. Lee YS, Lee WG, Chang YK and Chang HN. 1995. Modelling of ethanol 
production by Saccharomyces cerevisiae from a glucose and maltose mixture. 
Biotechnology Letters 17: 791-796  
61. Lee D, Owens VN, Boe A and Jeranyama P. 2007. Composition of herbaceous 
biomass feedstocks, South Dakota State University  Plant Science Department, June 
2007 
62. Leksawasdi N and Joachimsthal EL and Rogers PL. 2001. Mathematical modeling 
of ethanol production from glucose/xylose mixtures by recombinant Zymomonas 
mobilis, Biotechnology Letters 23:1087-1093  
63. Lin Y, Tanaka S. 2006. Ethanol fermentation from biomass resources: current state 
and prospects. Appl Environ Microbiol 69:627-642 
64. Lynd LR, van Zy WH, McBride JE and Laser M. 2005. Consolidated bioprocessing 
of cellulosic biomass: an update. Current Opinion in Biotech 16:577-583 
65. McMillan JD. 1993. Xylose fermentation to ethanol: a review. National Renewable 
Energy Laboratory 
66. Moes CJ, Pretorius IS, van Zyl C. 1996. Cloning and expression of the Clostridium 
thermosulfurogenes D-xylose isomerase gene (xylA) in Saccharomyces cerevisiae.   
Biotech Lett 18:269-274 
67. Mohagheghi A, Evans K, Chou YC, and Zhang M. 2002. Cofermentation of glucose, 
xylose, and arabinose by genomic DNA-integrated xylose/arabinose fermenting 
strain of Zymomonas mobilis AX101. Appl Biochem Biotechnol 98?100:885?898 
 
73
68. Moiser N, Wyman C, Dale B, Elander R, Lee YY, Holtzapple M, Ladisch M. 2005. 
Features of promising technologies for pretreatment of lignocellulosic biomass. 
Biores Technol 96:673-686 
69. Moniruzzaman M, Dien BS, Skory CD, Chen ZD, Hespell RB, Ho NWY, Dale BE, 
Bothast RJ. 1997. Fermentation of corn fiber sugars by an engineered xylose 
utilizing Saccharomyces yeast strain. World J Microbiol Biotechnol 13:341-346 
70. Monod J. 1942. Recherches sur la croissance des cultures bact?riennes (Studies on 
the growth of bacterial cultures). Hermann and Cie, Paris 
71. Ohta K, Beall DS, Mejia JP, Shanmugam KT and Ingram LO. 1991. Genetic 
improvement of Escherichia coli for ethanol production: chromosomal integration 
of Zymomonas mobilis genes encoding pyruvate decarboxylase and alcohol 
dehydrogenase II. Appl Environ Microbiol 57:893?900 
72. Olsson L and Hahn-H?gerdal B. 1995. Kinetics of ethanol production by 
recombinant Escherichia coli K011. Biotechnol Bioeng  45:356-365 
73. Olsson L and Hahn-H?gerdal B. 1996. Fermentation of lignocellulosic hydrolysates 
for ethanol production. Enzyme Microb Technol 18:312-331  
74. Patil KR, ?kesson M and Nielsen J. 2004. Use of genome-scale microbial models 
for metabolic engineering. Current Opinion in Biotechnology 15:64-69 
75. Perlack RD, Wright LL, Turhollow AF, Graham RL, Sotcks BJ, and Erbach DC. 
2005. Biomass as feedstock for a bioenergy and bioproducts industry: the technical 
feasibility of a billion-ton annual supply. DOE/GO-10 2005-2135, April. U.S. 
Department of Energy and U.S. Department of Agriculture 
 
74
76. Potera C. 2006. Process with biofuels will depend on, drive microbiology research. 
Microbe 1(7): 317-322 
77. Pramanik J and Keasling JD. 1997. Stoichiometric model of Escherichia coli 
metabolism: incorporation of growth-rate dependent biomass composition and 
mechanistic energy requirements. Biotechnol Bioeng 56:398-421 
78. Qian M, Tian S, Li X, Zhang J, Pan Y, Yang X. 2006.  Ethanol production from 
dilute-acid softwood hydrolysate by co-culture. Appl Biochem Biotechnol 134:273?
283 
79. Ragauskas AJ, Williams CK, Davison BH, Britovsek G, Cairney J, Eckert CA, 
Frederick WJ, Hallett JP, Leak DJ, Liotta CL, Mielenz JR, Murphy R, Templer R 
and Tschaplinski T. 2006. The path forward for biofuels and biomaterials. Science 
311: 484-489 
80. Roca C, Nielsen J, and Olsson L. 2003. Metabolic engineering of ammonium 
assimilation in xylose-fermenting Saccharomyces cerevisiae improves ethanol 
production. Appl Environ Microbiol 69(8):4732-4736 
81. Rouhollah H, Iraj N, Giti E and Sorah A.2007. Mixed sugar fermentation by Pichia 
stipitis, Sacharomyces cerevisiae, and an isolated xylose-fermenting Kluyveromyces 
marxianus and their cocultures. African J Biotechnology 6(9):1110-1114 
82. S?nchez S, BravoV, Castro E, Moya AJ, and Camacho F. 2002. The fermentation of 
mixtures of D-glucose and D-xylose by Candida shehatae, Pichia stipitis or 
Pachysolen tannophilus to produce ethanol. J Chem Technology and Biotechnol 
77:641-648 
 
75
83. Sarthy AV, McConaughy BL, Lobo Z, Sundstrom JA, Furlong CE, Hall BD. 1987. 
Expression of the Escherichia coli xylose isomerase gene in Saccharomyces 
cerevisiae.  Appl Environ Microbiol 53:1996-2000 
84. Savinell JM and Palsson B?. 1992. Network analysis of intermediary metabolism 
using linear optimization. I. development of mathematical formalism. J Theor Biol 
154: 421-454 
85. Sedlak M and Ho NW. 2004. Production of ethanol from cellulosic biomass 
hydrolysates using genetically engineered Saccharomyces yeast capable of 
cofermenting glucose and xylose. Appl Biochem Biotechnol 113-116:403-416 
86. Service RF. 2007. Biofuel researchers prepare to reap a new harvest. Science 315: 
1488-1491 
87. Skoog K and Hahn-H?gerdal B. 1990. Effect of oxygenation on xylose fermentation 
by Pichia stipitis. Appl Environ Microbiol 56:3389-3394 
88. Stephanopoulos G. 2007. Challenges in engineering microbes for biofuels 
production. Science 315:801-804 
89. Takuma S, Nakashima N, Tantirungkij M, Kinoshita S, Okada H, Seki T, et al. 1991.  
Isolation of xylose reductase gene of Pichia stipitis and its expression in 
Saccharomyces cerevisiae. Appl Biochem Biotechnol 28-29: 327-340 
90. Taniguchi M, Itaya T, Tohma T, Fujii M. 1997a. Ethanol production from a mixture 
of glucose and xylose by co-culture of Pichia stipitis and a respiratory-deficient 
mutant of Saccharomyces cerevisiae. J Ferment Bioeng 83:364?370 
 
76
91. Taniguchi M, Itaya T, Tohma T, and Fujii M. 1997b. Ethanol production from a 
mixture of glucose and xylose by a novel co-culture system with two fermentors and 
two microfiltration modules. J Ferment Bioeng 84(1):59-64   
92. Taniguchi M, Tanaka T. 2004. Clarification of interactions among microorganism 
and development of co-culture systems for production of useful substances. Adv 
Biochem Engin/Biotechnol 90:35-62 
93. Tantirungkij M, Nakashima N, Seki T, Yoshida, T. 1993.  Construction of xylose-
assimilating Saccharomyces cerevisiae. J Ferment Bioeng 75:83?88 
94. Tao H, Gonzalez R, Martinez A, Rodriguez M, Ingram LO, Preston JF, Shanmugam 
KT. 2001. Engineering a homo-ethanol pathway in Escherichia coli: increased 
glycolytic flux and levels of expression of glycolytic genes during xylose 
fermentation. J Bacteriol 183:2979?2988 
95. Thatipamala R, Rohani S and Hill GA. 1992. Effects of high product and substrate 
inhibitions on the kinetics and biomass and product yields during ethanol batch 
fermentation. Biotechnol Bioeng 40:289-297 
96. Toivari MH, Aristidou A, Ruohonen L, Penttila M. 2001. Conversion of xylose       
to ethanol by recombinant Saccharomyces cerevisiae: importance of xylulokinase       
(XKS1) and oxygen availability. Metab Eng 3:236-249 
97. Varma A and Palsson B?. 1993a. Metabolic capabilities of Escherichia coli. I. 
Synthesis of biosynthetic precursors and cofactors. J Theor Biol 165:477?502 
98. Varma A and Palsson B?. 1993b. Metabolic capabilities of Escherichia coli. II. 
Optimal growth patterns. J Theor Biol 165:503?522 
 
77
99. Varma A & Palsson B?. 1994. Metabolic flux balancing: basic concepts, scientific 
and practical use. Bio/Technology12:994-998 
100. Varma A and Palsson B?. 1994. Stoichiometric flux balance models quantitatively 
predict growth and metabolic by-product secretion in wild type Escherichia coli 
W3110. Appl Environ Microbiol 60:3724?3731 
101. Walfridsson M, Bao X, Anderlund M, Lilius G, Bulow L, Hahn-H?gerdal B. 1996. 
Ethanolic fermentation of xylose with Saccharomyces cerevisiae harboring the 
Thermus thermophilus xylA gene, which expresses an active xylose (glucose) 
isomerase. Appl Environ Microbiol 62:4648-4651 
102. Zaldivar J, Nielsen J and Olsson L. 2001. Fuel ethanol production from 
lignocellulose: a challenge for metabolic engineering and process integration. Appl 
Microbiol Biotechnol 56:17-34 
103. Zaldivar J et al. 2002. Fermentation performance and intracellular metabolite 
patterns in laboratory and industrial xylose-fermenting Saccharomyves cerevisiae. 
Appl Microbiol Biotechnol 59:436-442 
104. Zhang M, Eddy C, Deanda K, Finkelstein M, and Picataggio S. 1995. Metabolic 
engineering of a pentose metabolism pathway in ethanologenic Zymomonas mobilis. 
Science 267:240?243