Development of xyloseutilizing and inhibitortolerant yeast strains for bioethanol production

Vrije Universiteit Brussel
Katholieke Universiteit Leuven
Universiteit Antwerpen
Interuniversity Program Molecular Biology (IPMB)

Development of xylose-utilizing and inhibitor-tolerant
yeast strains for bioethanol production

Thesis submitted in partial fulfillment of the requirements for the Degree of
Master of Science in Molecular Biology

Tung Thanh Dinh

Promoter: Prof. Johan Thevelein
Co-promoter: Dr. Françoise Dumortier

Supervisor: Mekonnen M. Demeke

Laboratory of Molecular Cell Biology
Department of Molecular Microbiology
Faculty of Science
K.U.Leuven

Academic year 2010-2011


CONFIDENTIAL DOCUMENT
This thesis is a piece of examination that has not been corrected after the defense.
The results of this thesis might be used for a patent application. Therefore, all the
data of this document should be considered as confidential. They may by no
means made public; and there should not be any reference to them. To know the

date of public release, the promoter of this thesis can be contacted.


TABLE OF CONTENTS

TABLE OF CONTENTS............................................................................................................ i
LIST OF FIGURES .................................................................................................................. iv
LIST OF TABLES .................................................................................................................... vi
ACKNOWLEDGEMENT ....................................................................................................... vii
ABSTRACT............................................................................................................................ viii
1. INTRODUCTION ................................................................................................................. 1
2. LITERATURE REVIEW ...................................................................................................... 2
2.1 BIOETHANOL PRODUCTION – AN OVERVIEW ..................................................... 2
2.1.1 First generation bioethanol ........................................................................................ 2
2.1.2 Second generation bioethanol.................................................................................... 3
2.2. INDUSTRIAL REQUIREMENTS, CURRENT STATUS AND CHALLENGES ....... 4
2.2.1 Industrial requirements .............................................................................................. 4
2.2.2 Current status ............................................................................................................. 5
2.2.3 Challenges ................................................................................................................. 9
2.3. INHIBITORS IN HYDROLYSIS PRODUCTS OF LIGNOCELLULOSES .............. 10
2.3.1. Origin ...................................................................................................................... 10
2.3.2. Effects and Mechanisms ......................................................................................... 11
2.3.2 Solutions for inhibitor.............................................................................................. 12
2.4. XYLOSE FERMENTATION ....................................................................................... 15
2.4.1. Xylose fermentation in yeast .................................................................................. 15
2.4.2. Recombinant xylose-fermenting strain development ............................................. 16
3. OBJECTIVES ...................................................................................................................... 20
4. MATERIAL AND METHODS ........................................................................................... 21
i

4.1. CULTURING MEDIA.................................................................................................. 21
4.1.1. Inoculation media ................................................................................................... 21
4.1.1. Screening media ..................................................................................................... 21
4.2. YEAST STRAINS ........................................................................................................ 22
4.3. YEAST MANIPULATION .......................................................................................... 23
4.3.1. Sporulation.............................................................................................................. 23
4.3.2. Tetrad analysis ........................................................................................................ 23
4.3.3. Mating type determination...................................................................................... 23
4.3.4. Spot test .................................................................................................................. 23
4.3.5. OD 600 measurement ................................................................................................ 24
4.3.6. Dry weight measurement ........................................................................................ 24
4.3.7. Fermentation ........................................................................................................... 24
4.4. DNA MANIPULATION............................................................................................... 25
4.4.1. DNA isolation ......................................................................................................... 25
4.4.2. Polymerase chain reaction (PCR) ........................................................................... 25
4.4.2. Agarose gel electrophoresis .................................................................................... 25
4.5. SUGAR AND METABOLITE ANALYSIS ................................................................ 26
4.5.1. Rate of sugar consumption ..................................................................................... 26
4.5.2. Ethanol, Sugar and metabolite analysis .................................................................. 26
4.6. GENOME SHUFFLING ............................................................................................... 26
5. RESULTS AND DISCUSSION .......................................................................................... 28
5.1. EXPERIMENTS ON ETHANOL RED ........................................................................ 28
5.1.1. Segregant preparation ............................................................................................. 28
5.1.2. High-throughput inoculating segregants with non-specified amounts of cells ...... 28
5.1.3. High-throughput inoculating segregants with specified amounts of cells .............. 29
5.1.4. Fermentation of Ethanol Red segregants ................................................................ 32
ii

5.2. FERMENTATION OF SACCHAROMYCES CEREVISIAE STRAINS ....................... 33
5.2.1. Effect of furan derivative on fermentation of ER, BY4742, and TMB3400 strains
.......................................................................................................................................... 33
5.2.2. Fermentation of ISO12 ISOB57 ER TMB3400 strains in 50%, 60%, and 70%
Hydrolysate and containing YP and 13% glucose............................................................ 35
5.3. GENETIC MAPPING BASED ON AMTEM .............................................................. 38
5.3.1. Genetic mapping strategy ....................................................................................... 38
5.3.2. AMTEM for inhibitor tolerance phenotype............................................................ 38
5.4. GENOME SHUFFLING ............................................................................................... 40
6. CONCLUSION AND RECOMMENDATION ................................................................... 42
7. BIBLIOGRAPHY ................................................................................................................ 44

iii


LIST OF FIGURES

Figure 2.1: General principle of bioethanol production
Figure 2.2: Structures of cellulose and hexose components of hemicelluloses
Figure 2.3: The required stages in case cellulose is the main substrate
Figure 2.4: Inhibitors in the lignocellulosic biomass
Figure 2.5: Schematic view of inhibition mechanisms of inhibitors
Figure 2.6: Three main fermentation modes used in fermentation technology
Figure 2.7: Xylose (a,b) and arabinose (c,d) utilization pathways in bacteria (a,c) and in fungi
(b,d)
Figure 2.8: Metabolic pathways used by yeasts: the nonoxidative pentose phosphate pathways
Figure 5.1: OD 600 values of ER segregants on 65% liquid hydrolysate in 7 days
Figure 5.2: ER segregants from 1A to 44D on 60% solid hydrolysate in YPD pH=6 after 2
days
Figure 5.3: ER segregants from 1A to 44D on 70% solid hydrolysate in YPD pH=6 after 2

days
Figure 5.4: ER segregants from 1A to 44D on 80% solid hydrolysate in YPD pH=6 after 3
days
Figure 5.5: Spot test of ER segregants on 50% solid hydrolysate after 4 days
Figure 5.6: Spot test of ER segregants on 60% solid hydrolysate after 4 days
Figure 5.7: OD 600 values of ER and its 12 selected segregants and in 60% and 70% liquid
hydrolysate media supplemented with YP and 2% glucose after 48hrs
Figure 5.8: Sugar consumption of ISO12, ER and its 6 segregants in 60% spruce hydrolysate
+ YP + 13% glucose
Figure 5.9: Effect of furan derivatives on fermentation of BY4742
iv


Figure 5.10: Effect of furan derivatives on fermentation of ER
Figure 5.11: Effect of furan derivatives on fermentation of TMB3400
Figure 5.12: Sugar consumption of ISO12, ISOB57, ER, and TMB3400 in 50% hydrolysate
supplemented with YP and 13% glucose
Figure 5.13: Sugar consumption of ISO12, ISOB57, ER, and TMB3400 in 60% hydrolysate
supplemented with YP and 13% glucose
Figure 5.14: Sugar consumption of ISO12, ISOB57, ER, and TMB3400 in 70% hydrolysate
supplemented with YP and 13% glucose
Figure 5.15: Ethanol yields measured from the residual fermentation medium of ISO12,
ISOB57, ER, and TMB3400 in 50%, 60%, and 70% hydrolysate supplemented with YP and
13% glucose
Figure 5.16: AMTEM strategy employed to identify the genes or mutation responsible for
inhibitor-tolerant and xylose-fermenting phenotype
Figure 5.17: OD 600 values of ER and ISOB57 after 72hrs growth in hydrolysate
supplemented with YP and 2% glucose
Figure 5.18: Spot test result of F1 hybrid [ISOB57 x AMS928α] segregants on 40%
hydrolysate supplemented with YP and 2% glucose solid medium after 2 days

Figure 5.19: Spot test result of F1 hybrid [ISOB57 x AMS928α] segregants on 50%
hydrolysate supplemented with YP and 2% glucose solid medium after 4 days
Figure 5.20: OD 600 values of initial mating products along with MDX1051, MDX1052,
MDX1053, MDX1054, ER, and TMB3400 in YP + 2% xylose after 24hrs and 48hrs (with
same initial OD 600 of 1)
Figure 5.21: OD 600 values of mating mix in 50% and 60% hydrolysate supplemented with YP
and 2% xylose after 24hrs and 48hrs
Figure 5.22: Sugar consumption of the final mating mix in YP supplemented with 2% xylose

v


LIST OF TABLES
Table 2.1: Annual world ethanol production by country
Table 2.2: The most commonly used strategies applied to S. cerevisiae for pentose
fermentation
Table 4.1: Yeast strains used in experiments
Table 4.2: Composition of buffers and medium used in genome shuffling experiment
Table 5.1: Composition of the spruce hydrolysate supplied by SEKAB Company

vi


ACKNOWLEDGEMENT
First, I would like to thank Belgian government namely BTC for financing my Master
study. I have really enjoyed the time living in this beautiful country. I also want to show my
most sincere gratitude to my promoters Professor Johan Thevelein and Dr. Francoise
Dumortier for giving me the opportunity to do my thesis at Molecular Cell Biology
Laboratory (MCB), KULeuven. I really appreciate my supervisor Mekonnen M. Demeke for
his devoted support during my time at the lab.

I want to devote the thesis to my parents who brought me up with love and care. My
family including my younger brother has been always behind me when I was in trouble and
shared happiness with me in the time of success. Without my family, I would not be what I
am today.
I would like to thank all IPMB teachers because of precious knowledge I have learnt
from them, as well as administrative staffs for their support during my study period. My
deepest gratitude to IPMB coordinator Professor Eddy Van Driessche and Mr Rudi Willems
for their enthusiastic guidance and assistance. I really enjoyed the time with my IPMB
classmates and the memory of our class will always be in my heart.

vii


ABSTRACT
Second generation bioethanol production has been considered as a solution for energy
crisis in the future due to fossil fuel depletion. Despite being promising, this technology
confronts several challenges to reach its full potential. One of them is the presence of
inhibitors generated during the hydrolysis of lignocellulosic biomass. Another which is not
less important is the incapacity of the most commonly used microorganism in the industry,
Saccharomyces cerevisiae, to ferment pentose sugar such as arabinose and xylose.
The first objective of this study was to isolate a segregant of an inhibitor tolerant
industrial yeast strain, Ethanol Red, and subsequently use this segregant for mapping
genomic regions responsible for inhibitor tolerance. By tetrad analysis, 188 segregants were
isolated and applied to microbial techniques in order to find segregants which are at least as
tolerant as ER. They were then spotted on solid spruce hydrolysate media (60%, 70%, and
80%) without pre-determined amount of cells. Segregants which grew nearly as good as ER
were subsequently subjected to Spot test for further screening and 12 segregants with best
growth were kept. Afterwards, these 12 segregants were inoculated in liquid hydrolysate
media and only 6 segregants which were able to grow on highly inhibiting 70% hydrolysate
were chosen for fermentation evaluation. As a result, we manage to get 1 segregant (8D) that

ferments better than ER in 60% spruce hydrolysate containing YPD.
With the aim of combining the inhibitor-tolerance and xylose-fermenting phenotypes
capacity from strains possessing one of these two phenotypes, genome shuffling of Ethanol
Red (inhibitor tolerant strain) along with four xylose utilizing yeast strains was performed.
After the spores of these 5 strains were isolated and crossed all together, the resulting
diploids were able to grow in both xylose and inhibitor containing medium. The mating
products were proven to grow significantly faster in YPXylose than their xylose-fermenting
parents but fermented xylose with very slow rate. This indicates that good growth on xylose
does not mean good fermentation to ethanol and further genome shuffling cycle and
evolutionary adaptation in anaerobic condition should be done to obtain strains of better
xylose fermenting capacity.

viii


1. INTRODUCTION

The fossil fuel all over the world has been on the way of exhaustion and to meet the
ever increasing demand of energy, especially for transport purpose, human being has to
search for alternative renewable sources. One of the solutions is to convert biomass of the
environment into biofuel. For instance, biomass like starch or lignocellulosic material can be
hydrolyzed into monomeric sugars which are then fermented by microorganism, usually
yeast, into ethanol. This kind of biofuel called bioethanol has become one of the most
developed energy sources nowadays. It not only alleviates our dependency on fossil fuel but
also reduces green house gas emission (Farrell et al., 2006). Plenty of companies and
laboratories have invested enormous amount of money and resources in research of
bioethanol production.
Lignocellulosic material is the most abundant renewable resource. Approximately,
plants on earth can supply 1.3x1010 tons of wood per year which is equivalent to 7x109 tons
of coal and sufficient to meet human’s energy requirement. Every year, about 180 million

tons of cellulosic feedstocks are produced (Demain et al., 2005). Moreover, if people can take
advantage of inexpensive lignocelluloses from agriculture and forestry residues, it can help to
reduce pollution and greenhouse effect problems.
Although being promising, the second-generation bioethanol production has confronted
several challenges. One of them is the presence of inhibitor generated during the hydrolysis
of lignocellulosic biomass. Another which is not less importance is the incapacity of the most
commonly used microorganism in the industry, Saccharomyces cerevisiae, in fermenting
pentose sugar such as arabinose and xylose (Träff, 2001). The aim of this study was to
unravel the genes or mutations responsible for inhibitor tolerance and xylose-fermenting
capacity of Saccharomyces cerevisia strains. Thereby, a strain which is inhibitor-tolerant and
at the same time xylose-fermenting could be developed.

1


2. LITERATURE REVIEW
2.1 BIOETHANOL PRODUCTION – AN OVERVIEW
2.1.1 First generation bioethanol
Bio-ethanol can be produce from a wide variety of carbohydrate sources: mono-, di-, or
polysaccharides. The biomass sources from which ethanol is produced in the industry can be
sweet juice (e.g., sugar cane or molasses) and starch (e.g., corn, wheat, barley).
2.1.1.1 Sucrose-to-Ethanol
The most common disaccharide used for bioethanol production is sucrose which is
fermented by industrial yeast like Saccharomyces cerevisiae. First, sucrose which is
composed of glucose and fructose is hydrolyzed by invertase (an enzyme produced by yeast)
into these two monosaccharides. Afterwards, zymase (an enzyme complex produced by
yeast) ferments glucose and fructose into ethanol and carbon dioxide. Theoretically, one ton
of sucrose can be converted into 511 kg ethanol, yet the practical efficiency is only about
92% of this theoretical value (Pandey, 2009).
In the industry, the main source of sucrose is sugarcane and sugar beet or in some cases

sweet sorghum. Brazil has been the main producer of fuel produced from sugarcane juice,
giving 16 billion liter of ethanol in 2005, 2 billion of which were exported. European Union
(EU) mainly produces ethanol from sugar beet juice which plays a less important role than
wheat previously. However, usage of wheat has increased significantly thanks to incentive of
EU policy for energy crops. In 2005, about 950 million liters of bio-ethanol were produced in
the EU (F.O. Licht 2006).
2.1.1.2 Starch-to-Ethanol
To convert starch to ethanol, the polymer is first broken down by the action of
glucoamylase enzyme into dextrose or D-glucose. This is followed by the fermentation and
complementary processes to produce ethanol. In the industry the starch feedstock are mainly
grains like corn, wheat, or barley among which corn is the dominant source worldwide
containing 60-70% starch (Pandey, 2009).
There are two methods of converting starch to ethanol, namely the dry and wet mills. In
the dry mill method, the grain which is ground to powder is hydrolyzed and the sugar present
in the hydrolysate is fermented to ethanol while the remaining is converted into distiller
2


grains. This co-product of ethanol can be used as animal feed (figure 2.1). Carbon dioxide, a
by-product, may also be taken for other usage such as in beverage industry. This method is
the most commonly used in industrial field. The wet mill method is also applied in a large
number of factories. The grains are wet milled to separate various components like starch,
protein, fiber and germ before being converted into different products (Pandey, 2009).
The United States is the leading producer of bioethanol from corn, producing 8 billion
liters in 2002 and up to 15 billion liters in 2005. Production of fuel ethanol from corn is
responsible for 93% of the whole 18.5 billion liters of U.S. bioethanol yield in 2006. It has
been shown that corn-to-ethanol production helps to increase energy gain but only
insignificantly reduces total CO 2 emission (Demirbas, 2009).
Figure 2.1: General principle of bioethanol production (Demirbas, 2009)

Ethanol production from plant substrates has been proven to increase the available
source for gasoline blending and help to reduce greenhouse gas emission and other gaseous
pollutants (Bergeron, 1996). However, it on the other hand is not cost-effective and
influences the food supply and biodiversity (Mousdale, 2008).
2.1.2 Second generation bioethanol
To meet the ever increasing energy demand especially for transportation and to reduce
green house gas emission, ethanol has been produced from other feedstock than saccharides
and starch. Lignocellulosic material appears to be a good candidate and in some countries
inulin also has attracted attention. Second-generation in which bioethanol is produced from
lignocellulosic materials like agricultural and forestry by-products can help to decrease the
use of fossil fuel. However, a number of challenges have to be circumvented, for instance
developing enzymes for lignocellulose hydrolysis, microorganism which can use a variety of
substrates, and so on. Second generation bioethanol production has been developed since
1920s by Scholler and was commercially marketed in former Soviet Union until the late
1980s (Keller, 1996).
Figure 2.2: Structures of cellulose and hexose components of hemicelluloses

As mentioned above, second-generation bioethanol is produced from lignocellulosic
biomass containing cellulose, hemicelluloses, and lignin which is a cementing substance,
making cross-linking and hindering the hydrolysis process. Cellulose, a structural polymer in
plants, is highly insoluble, mixed with other polysaccharides like hemicelluloses and
3


protected from enzymatic attack in woods by the presence of lignins. Cellulose contents in
plant vary from 38 to 57% (Mousdale, 2008). While cellulose is a linear polymer of glucose,
hemicelluloses is a branched heteropolymer of a mixed group of polysaccharides with
different structures of two or three types of sugars (sometimes O-methylated or O-acetylated)
and a sugar acid (figure 2.2). The major components of hemicelluloses are xylose, arabinose
(pentoses) and glucose, galactose and mannose (hexoses). To convert biomass into ethanol,

first cellulose and hemicellulose have to be chemically or enzymatically hydrolyzed into
monomeric sugars (figure 2.3) which are subsequently fermented into ethanol (Mousdale,
2008). If hemicellulosic sugars are also considered as substrates, either hemicellulases or
other microorganisms are needed for simultaneous or subsequent fermentations.

Figure 2.3: The required stages of conversion of biomass into ethanol using cellulose as the main
substrate

The promising future of second-generation bioethanol production has been statistically
persuasive. It is estimated that the total biomass energy crops potential for bioethanol
production is around 1.3-2.3 billion tons, being capable of supplying 30-50% U.S. gasoline
consumption. In contrast, even if all the cultivated areas were devoted for energy, the
production of bioethanol from corn would only meet 12% of energy demand. According to
the figures from Canada, in 2004, 2025 million liters of ethanol was produced from wheat,
barley, corn, and potatoes while the amount from nonfood crops like straw, wood residues,
and forest residues already reached about 11500 liters (Mousdale, 2008). Recently, there have
been some policies and initiatives to increase the ethanol production from cellulosic
materials. This has promoted the interest in bioengineering solutions to the challenges in
lignocellulosic bioethanol, in the hope of creating a sophisticated biofuel program
(Economic, Financial, Social Analysis and Public Policies for Fuel Ethanol Phase 1, Natural
Resources Canada, Ottawa, November 2004).
2.2. INDUSTRIAL REQUIREMENTS, CURRENT STATUS AND CHALLENGES
2.2.1 Industrial requirements
Up to now, most bioethanol production is first generation and it is a tradition to use
yeast as fermenting microorganism to produce ethanol. Among yeast, S. cerevisiae is the
most commonly used one in the industry nowadays thanks to its capacity of fermenting fast
and efficiently. Compared to other yeast and filamentous fungi, S. cerevisiae possesses plenty
4

of traits which are favorable in the industrial conditions, especially its high tolerance toward
ethanol and as a result high concentration of ethanol can be reached (Hahn-Hägerdal et al.,
2007). It can grow under aerobic and anaerobic condition and ferment well in industrial
context. It can also tolerate a wide range of pH, working optimally at acidic pH, which
enables fermentation by S. cerevisiae less susceptible to infection than bacterial fermentation.
S. cerevisiae can survive up to about 40oC and optimally between 30-35oC which is easily
achievable in the industry (Skoog and Hahn-Hägerdal, 1988).
Because S. cerevisiae has been used from a long time ago in human history for the
production of food, it is “generally regarded as safe” (GRAS. Since, it has been also utilized
perpetually in the industry, S. cerevisiae can tolerate high concentration of sugar and ethanol
(above 10%), which allows non-sterile process operation and relatively high osmotic
pressure. Furthermore, S. cerevisiae isolated from industry shows robustness against
inhibitory compounds in fermentation process (Hahn-Hägerdal et al., 2005; Lindén et al.,
1992).
2.2.2 Current status
Nowadays, the United States and Brazil are the two world’s largest ethanol producers.
For Brazil, the biomass which has been used is sugar cane while for the United States, corn
has been utilized. Sugar cane consists of a large portion of sugar in which the content of
sucrose can be nearly 20% (www.suedzucker.de, www.nedalco.com). Sucrose can be
hydrolyzed into glucose and fructose and then fermented by S. cerevisiae to bioethanol. For
the production of ethanol from sugar cane, the viscous residue left after sugar crystallization
has high osmolarity, which makes it possible to be stored for a long period of time without
microbial infection. In Brazil, fed-batch model along with yeast recycling has been applied
commonly. Since the fermentation is performed with high cell density, the rate is relatively
fast in which fermentation is almost completed in 6-11hrs and S. cerevisiae has been almost
exclusively used (Amorim et al., 2004).
Table 2.1: Annual world ethanol production by country (adapted from
www.ethanolrfa/industry/statistics)

USA

2008
(million gallons)
9000.0

2009
(million gallons)
10,600.00

Brazil

6472.2

6577.89

Country

5


European Union

733.6

1039.52

China

501.9

541.55

Thailand

89.8

435.20

Canada

237.7

290.59

Other

128.4

247.27

Colombia

79.29

83.21

India

66.0

91.67

26.4A

56.80

Australia

2.2.2.1 Hydrolysis of lignocellulosic biomass
There are three widely used methods of hydrolysis in industry, namely, the two-step
dilute acid hydrolysis, concentrated acid hydrolysis and enzymatic hydrolysis. After these
processes, a variety of inhibitors are generated and the composition of which depends on type
of lignocellulosic origin. In the three methods, the biomass materials are usually
mechanically pretreated to enhance the accessibility of the substrates. Besides, each method
has its own advantages and disadvantages. Acid hydrolysis is fast and easy to operate but its
drawbacks are the non-selectivity of the hydrolysis and inhibitors formation. In contrast,
enzymatic hydrolysis has higher substrate specificity and under milder condition of
temperature and pH, problem of corrosion by acid can be avoided. However, the use of
expensive biocatalyst lowers the cost-effectiveness of the method (Demirbas, 2009).
2.2.2.1.1 Dilute Acid Hydrolysis
Hydrolysis by diluted acid is performed under high temperature and pressure with
reaction times within seconds or minutes, which is favorable for continuous processing. Most
of the processes can only reach sugar efficiency of about 50%. The reason is sugar
degradation and the fact that furfural and other degradation products are poisonous to the
fermenting microorganisms. The most striking advantage of dilute acid hydrolysis is its fast
rate reaction, which facilitates continuous processing. However, combination of acid and high
temperature and pressure necessitates special reactor material, which can be very costly
(Demirbas, 2009).
Dilute acid hydrolysis is the first technology which has been developed. The hydrolysis
is performed in two stages due to the differences between the hemicelluloses and the

cellulose (Harris et al. 1985), maximizing the conversion yield. Since 5-carbon sugars
6


degrade more rapidly than 6-carbon sugars, one way to decrease sugar degradation is to have
a two-stage process. In the first step, which is called prehydrolysis step, a mild condition is
applied and hemicellulose is broken down to obtain 5-carbon sugars, while in the second step
a more harsh condition is performed to hydrolyze the more resistant cellulose component
producing 6-carbon sugars. The two hydrolyzed streams can be fermented into ethanol either
together or separately and in the latter case they are mixed and ethanol is then distilled. The
prehydrolysis can also be carried out physically (steam pretreatment, milling, freeze
explosion), biologically (white rot fungi), by other acids (phosphoric acid, sulfuric acid,
sulfur dioxide, by alkaline (sodium hydroxide, ammonia), or by organic solvents (ethylene
glycol) (Vallander and Eriksson, 1990; Saddler et al., 1993). In this step, hemicelluose is
liquefied and converted to mono- and oligosaccharides (Olsson and Hahn-Hägerdal, 1996). A
process devised by National Renewable Energy Laboratory (NREL) gives the following
steps: for stage 1: 0.7% sulfuric acid, 190oC and for stage 2: 0.4% sulfuric acid, 215oC
condition. The liquid hydrolysates are retrieved and converted into alcohol while the
remaining cellulose and lignin in the solids are used as boiler fuel for electricity or steam
production. Yields of 89% for mannose, 82% for galactose, and 50% for glucose can be
achieved. In another process, by applying very low acid and temperature condition of
autohydrolysis of sawdust, a yield of 70% glucose was obtained (Ojumu and Ogunkunle,
2005).
2.2.2.1.2 Concentrated Acid Hydrolysis
Concentrated acid processes use mild temperatures and pressures are only involved in
pumping materials from vessel to vessel. Reaction times are usually longer than that of
diluted acid hydrolysis. This method generally applies concentrated sulfuric acid before
diluting with water to dissolve and hydrolyze the substrate into sugars. The process gives a
complete and quick conversion to sugars with little degradation. The decisive factors required
for the economic feasibility are the optimization of the recovery of sugars and the cost

effectiveness of the acid recycling. The biggest advantage of the concentrated acid method is
its high sugar yield. The acid and sugar can be separated by ion exchange followed by the reconcentration of acid via multiple evaporators. The application of low temperature and
pressure allows the use of low-cost materials such as fiberglass tanks and piping. However,
the process is slow and cost effective acid recovery systems have been difficult to develop.
Without acid recovery, large quantities of lime must be used to neutralize the acid present in

7


the sugar mixture. Hence, a significant amount of calcium sulfate needs to be removed, which
adds more cost to the process (Demirbas, 2009).
In short, concentrated acid is used to decrystalize cellulose and the dilute acid is used to
hydrolyze it into sugar. Acids have to be retrieved to ensure the economic feasibility. In a
process in Japan which was commercialized in 1948, concentrated sulfuric acid was used and
a striking feature is the usage of a membrane to separate acid from sugar in the product
stream. The efficiency of acid recovery was very high 80% (Wenzl, 1970). In another method
developed by the Soviet Union, concentrated hydrochloric acid was utilized and in this case
the prehydrolysis and hydrolysis were operated in one step, resulting in a variety of
compounds, some of which interfere with the subsequent fermentation step. In the time of
World War II, scientists at the U.S. Department of Agriculture’s Northern Regional Research
Laboratory in Peoria, Illinois improved the process, working with concentrated sulfuric acid
on corn cobs and achieving a yield of 15-20% xylose and 10-20% glucose which was readily
fermented to ethanol at 85-90% theoretical yield (Pandey, 2009). Another modification in the
United States in 1992 is the addition of recycling of dilute acid from the hydrolysis step and
the improvement of sulfuric acid recycle, which helps to reduce the use of it. Moreover,
recycling of acid increases the cost-effectiveness of the process. According to U.S.b Patent
5,188,673, the use of 30-70% weight of sulfuric acid at temperature below 100oC acid results
in high conversion of biomass (80-90% cellulose and hemicelluloses are hydrolyzed) but low
product yield caused by degradation and the recovery of acid. The drawback of this model is
clearly demonstrated by the fact that concentrated acids are toxic and corrosive, which

necessitates reactors resistant to corrosion (Von Sivers and Zacchi, 1995).
2.2.2.1.3 Enzymatic Hydrolysis
The third basic method is enzymatic hydrolysis including two technological
approaches: enzymatic and direct microbial conversion methods. Prior to enzymatic
hydrolysis, the chemical pretreatment of cellulosic biomass is necessary. The hydrolysis is
conducted by using cellulolytic enzymes. Different type of cellulases can be used to break
down cellulose and hemicelluloses, for example a mixture of endoglucanases, exoglucanases,
β-glucosidases, and cellobiohydrolases (Ingram and Doran, 1995; Laymon et al., 1996). The
endoglucanases randomly break the cellulose chains to create shorter polysaccharides, while
exoglucanases attack the non-reducing ends of these shorter chains and remove cellobiose
moieties. β-glucosidases hydrolyze cellobiose and other oligosaccharides to glucose
8


(Philippidis and Smith, 1995). To work efficiently, enzymes need to be in good contact with
their corresponding substrates. This necessitates a pretreatment process to separate
hemicelluloses and lignin as well as break down the crystal structure of cellulose and as a
result cellulose and hemicellulose molecules can be exposed for access. A new generation of
enzymes is required for cost-effective hydrolysis of cellulose to glucose since the technical
bottleneck of the enzymatic approach is the low specificity of currently used enzymes and
enzyme production cost. The lignin component is also unconvertible during hydrolysis and
fermentation processes (Ching, 2008). Due to the fact that lignin impedes the hydrolysis by
obstructing the access of cellulase to cellulose and by irreversible binding to the enzymes, it
has to be removed out of the biomass. Lignin can be separated from hemicellulose and
cellulose in the pretreatment step. This improves the hydrolysis efficiency significantly
(McMillan, 1994).
Because of the low hydrolysis yield of the previously described dilute-acid hydrolysis
(50-60%) (Jones and Semrau, 1984), enzymatic hydrolysis has been developed in which the
hydrolysis yield of 80-90% can be achieved theoretically (Söderström et al., 2003; Bura et al.,
2003). Nevertheless, no large-scale ethanol production process based on enzymatic

hydrolysis has been operated. To achieve efficient hydrolysis of cellulose, the lignocelluloses
have to be first mechanically treated to interrupt the matrix of cellulose-hemicellulose-lignin
(Galbe and Zacchi, 2002). Material is processed into small size particles which in addition
enhances mass transfer. The mild condition of enzymatic hydrolysis reduces saccharide
degradation and inhibitor formation (Jones and Semrau, 1984).
2.2.3 Challenges
The first challenge is in hydrolysate production. In the beginning, hydrolysates were
generated by acid hydrolysis and concentrated acid was used at first which gives high sugar
yield. Nevertheless, due to its corrosive nature, the recovery of acid was very costly. To
reduce acid consumption, single-state dilute-acid (0.4% H 2 SO 4 ) hydrolysis process at high
temperature was then applied, however, saccharide degradation was observed (Jones and
Semrau, 1984). This problem can be circumvented by using a two-stage procedure in which
hemicelluloses is hydrolyzed in the first step (150-190oC) and cellulose is then hydrolyzed in
the second step (190-230oC) (Nguyen et al., 1999; Wayman et al., 1984).
Another challenge is that on one hand S.cerevisiae can utilize sucrose, glucose,
fructose, galactose and mannose (Lindén et al., 1992; Nilsson et al., 2002) but on the other
9


hand it cannot ferment a number of monosaccharides, disaccharides (cellobiose and
xylobiose), trisaccharides originated from starch cellulose and hemicelluose, neither can it
utilize pentose sugar like xylose and arabinose (Lynd et al., 2002; Zhang and Lynd, 2005).
Major disadvantages of current industrial ethanol fermentation are related to the
metabolism of S.cerevisiae. Under anaerobic conditions a significant amount of glycerol is
generated resulting from the formation of the reduced co-factor NADH. Furthermore, the
hexose sugar galactose is only utilized upon the depletion of glucose, which makes the
fermentation of galactose-rich materials slower (Oura, 1973).
2.3. INHIBITORS IN HYDROLYSIS PRODUCTS OF LIGNOCELLULOSES
2.3.1. Origin
Hydrolysis of pretreated lignocellulosic biomass results in D-glucose as main

components as well as D-galactose, D-mannose and D-rhamnose (hexoses) and that of
hemicelluloses gives rise to D-xylose and L-arabinose (pentoses derived from
hemicelluloses). Also, uronic acids, for instance -glucuronic and 4-O-methylglucuronic acids
are generated from the hydrolysis of hemicelluloses. Further degradation of monomeric
sugars and lignin could bring about 3 main groups of inhibitors: (1) furan derivatives; (2) low
molecular weight fatty acids (mainly acetic acid, formic acid and levulinic acid); and (3)
phenolic compounds (Almeida et al., 2007)
Furan derivatives: The furan compounds including 5-hydroxymethyl-2-furaldehyde
(HMF) and 2-furaldehyde (furfural) are generated from the dehydration of hexoses and
pentoses, respectively (Dunlop, 1948). The amount of furans depends on the type of raw
material and pretreatment methods. For example, HMF content can vary from 2.0g/L to
5.9g/L depending on whether one-step or two-step dilute acid hydrolysis is used (Larsson et
al., 1999; Nilvebrant et al., 2003). In contrast, HMF is not present in wheat straw treated by
wet-oxidation method (Klinke et al., 2003). The concentration of furfural is usually lower
than that of HMF, though it is still high enough (about 1g/L) to inhibit fermentation.
Figure 2.4: Inhibitors in the lignocellulosic biomass (Almeida et al., 2007)

Low molecular weight fatty acids: Formic and levulinic acid are the major low
molecular weight fatty acids present in lignocellulosic hydrolysates. They are formed by the
breakdown of HMF, while acetic acid is generated by the deacetylation of hemicelluloses.

10


Moreover, additional quantity of formic acid can be produced from furfural under acidic
condition at high temperature (Dunlop, 1948).
Phenolic compounds: consists of a variety of phenolics that are formed by lignin and
carbohydrate degradation during acid hydrolysis (Popoff and Theander, 1976). The amount
and kind of phenolic compounds depends on the biomass nature.
2.3.2. Effects and Mechanisms

The inhibitors generated during pretreatment and hydrolysis have been shown to inhibit
either the microorganism growth or ethanol production (figure 2.5).
Furan derivatives: HMF and furfural inhibit growth, cause longer lag phase, and
therefore lower ethanol yield. Several mechanisms have been proposed to explain the
inhibition effects of furan on fermentation. Through in-vitro measurement, furfural and HMF
have been shown to inhibit alcohol dehydrogenase (ADH), pyruvate dehydrogenase (PDH)
and aldehyde dehydrogenase (ALDH) (Modig et al., 2002). Crude-cell extracts from cultures
which contained furfural demonstrated the decrease in activity of ADH, hexokinase, and
glyceraldehyde-3-phosphate dehydrogenase in the glycolysis pathway (Banerjee et al., 1981).
The reduction of furans by yeast cells could lead to NAD(P)H depletion. This was proven by
the fact that after furfural was added to the medium, levels of excreted acetaldehyde
increased (Palmqvist et al., 1999). Moreover, furfural in S.cerevisiae results in the
accumulation of reactive oxygen species which damages several organs such as vacuole and
mitochondrial membranes, chromatin and actin (Almeida et al., 2007). In short, furan
derivatives make yeast cells to devote their energy for fixing the damage and at the same time
decrease the intracellular ATP and NAD(P)H levels by inhibiting enzymes or consuming
cofactors.
Figure 2.5: Schematic view of inhibition mechanisms of inhibitors (Almeida et al., 2007)

Low molecular weight fatty acids: The inhibition effect of these acids has been linked
with intracellular anion uncoupling and accumulation as well as a decrease in biomass
formation (Russel, 1992). The undissociated form of these fatty acids can diffuse from the
fermentation media through the plasma membrane, dissociate under high intracellular pH,
and thus lower the pH in the cytosol (Verduyn et al., 1992). To counteract this phenomenon,
the ATPase in the plasma membrane has to pump out protons at the expense of ATP
hydrolysis. As a result, there is less ATP available for biomass formation. However, low
11


levels of these acids enhance ethanol yield (Larsson et al., 1999) because it stimulates ATP

production. In contrast, when the concentrations are high the ATP demand becomes so high
that the cell has to acidify the cytosol (Larsson et al., 1999). The anionic form of the acid
inside the cell is captured and the undissociated acid will diffuse into the cell until
equilibrium is reached. Additionally, these weak acids have been known to lower the uptake
of aromatic amino acids in the medium due to strong inhibition of Tat2p amino acid
permease (Bauer et al., 2003) and consequently inhibit yeast growth.
Phenolic compounds: The inhibition mechanism of phenolics in S. cerevisiae has not
been clarified still it is assumed that these compounds can partition into cell membrane and
therefore disrupt membrane barrier integrity (Hage et al., 2001; Heipieper et al., 1994).
Phenolic compounds with weak acidity may interfere with the electrochemical gradient due
to the fact that protons are transported back through the mitochondrial membranes (Terada,
1990).
S. cerevisiae innate tolerance to furan and phenolics: Some strains of S. cerevisiae
have shown furan tolerance owning to their ability to convert HMF and furfural to less
inhibitory compounds. Under either aerobic or anaerobic conditions, HMF is reduced to 2,5bis-hydroxymethylfuran (HMF alcohol) while furfural is converted to furfuryl alcohol which
can be oxidized to formic acid under aerobic conditions (Palmqvist et al., 2000). S. cerevisiae
also possesses the innate ability to metabolize phenolics present on the lignocellulosic
hydrolysate (Klinke et al., 2003). This could be caused by the presence of phenylacrylic acid
decarboxylase (PAD) which is a enzyme capable of metabolizing aromatic acid such as
cinnamic, p-coumaric and ferulic acids since these acids inhibit the ethanol production of S.
cerevisiae (Goodey et al., 1982).
2.3.2 Solutions for inhibitor
To deal with the intrinsic inhibitors in the lignocellulosic hydrolysate, three strategies
have been studied and implemented: Fermentation technology, Detoxification, and Strain
development.
2.3.2.1 Fermentation technology
In fermentation technology there are 3 main fermentation mode used, namely the batch,
fed-batch and continuous modes. For the second and the third modes, the rate of substrate
addition can be controlled. Since yeast is able to convert inhibitors in the hydrolysate into less
12

toxic compound, it can be exploited to optimize the detoxification and get higher ethanol
yield (Satyanarayana, 2009).
Figure 2.6: The three main fermentation modes and typical substrate ( — ) and product (---) profiles
(Satyanarayana, 2009)

2.3.2.2 Detoxification
There are three methods which can be used for detoxification namely biological,
physical and chemical methods (Olsson and Hahn-Hägerdal, 1996). The choice of method is
based on the source of lignocellulosic hydrolysate and the used microorganisms.
Lignicellulosic hydrolysates are various in their inhibitor content and microorganisms are
also inhibitor-tolerant to different extent (Pandey, 2009).
2.3.2.2.1 Biological Detoxification Methods
Biological detoxification exploits specific enzymes or microorganism which act on
inhibitors present in hydrolysates and alter their composition. Usage of enzymes like
peroxidases and laccase obtained from lignolytic fungus Trametes versicolor results in two to
three fold increases in maximal ethanol productivity from hemicelluloses hydrolysate of
willow owning to their effects on acid and phenolic compounds (Jönsson et al. 1998). The
fungus Trichoderma reesei has been used in inhibitor degradation of hemicelluloses
hydrolysate obtained after steam pretreatment of willow, which help to increase the maximal
ethanol productivity by about three times and ethanol yields by four times (Palmqvist et al.
1997). This results from the fact that T. reesei removes acetic acid, furfural, and benzoic acid
derivatives. The use of microorganism has also been proposed to selectively remove
inhibitors from lignocellulosic hydrolysates.

2.3.2.2.2 Physical Detoxification Methods
One of the methods of physical detoxification is by vacuum evaporation to reduce the
concentration of volatile compounds like low molecular weight fatty acids, for example
acetic acid, furfural, and vanillin present in the hydrolysate. However, this approach increases

fairly the concentration of nonvolatile inhibitors and consequently the level of fermentation
inhibition.

13


2.3.2.2.3 Chemical Detoxification Methods
For chemical detoxification, inhibitors can be precipitated or ionized under particular
pH values in which the toxicity degree of inhibitors is altered (Mussatto, 2002). Toxic
compounds can be alternatively adsorbed on activated charcoal (Dominguez, Gong, and
Tsao, 1996; Mussatto and Roberto 2001), on diatomaceous earth (Ribeiro et al. 2001) and on
ion exchange resins (Larsson et al. 1999; Nilvebrant et al. 2001). Inhibitors in hydrolysate
can also be eliminated through treatment of alkali, ozone, ion-exchange resins and enzymes
(Alriksson et al., 2006; Jönsson et al., 1998; Nilvebrant et al., 2003; Nilvebrant et al., 2001;
Santos et al., 2003). Nevertheless, this will raise the production cost and most detoxification
procedures potentially reduce the saccharide component in the hydrolysate (Larsson et al.,
1999b; Nilvebrant et al., 2003).
2.3.2.3 Strain development
2.3.2.3.1. Recombinant S. cerevisiae
Improvement of yeast tolerance to inhibitors present in lignocellulosic hydrolysates can
be achieved by overexpressing homologous or heterologous genes encoding for enzymes
which provide resistance to certain inhibitor(s). An example is the homologous alcohol
dehydrogenase (ADH6p) gene encoding for a NADPH-dependent enzyme which can reduce
HMF and furfural in yeast. Overexpressing of this gene gives rise to a strain that can take in 4
times more of HMF in defined medium under aerobic or anaerobic conditions (Petersson et
al., 2006). For tolerance to phenolics compounds, the PAD1 gene encoding phenylacrylic
acid decarboxylase was overexpressed in S. cerevisiae cultured in ferulic and cinnamic acid
containing media. The transformants were capable of converting ferulic acid and cinnamic
acid 1.5 and 4 times faster, respectively than the control strain under aerobic conditions,
which lead to 50 to 100% improvement in ethanol productivity (Larsson et al., 2001).

Another candidate is the heterologous Trametes versicolor laccase gene, the enzyme of which
can reduce oxygen molecule into water and radicals that in turn can metabolize and hence
eliminate phenolics (Larsson et al., 1999). Laccase-expressing strain cultivated in hydrolysate
in the presence of coniferyl aldehyde at the concentration of 1.25mmol/L showed growth
while the control strain did not. Moreover, this recombinant S. cerevisiae strain converted
aldehyde faster and gave higher ethanol productivity in the hydrolysate supplemented with
coniferyl aldehyde (Larsson et al., 2001). It can be concluded that recombinant S. cerevisiae
strains acquiring an increased tolerance for inhibitors have performed better in the
14


hydrolysates. Determining of gene(s) or enzyme(s) involved in inhibitor conversion has been
a promising strategy to obtain more robust S. cerevisiae strains.
2.3.2.3.2. Improved S. cerevisiae strains via evolutionary engineering
Inhibitor-tolerant S. cervisiase strains have been developed by strain adaptation
exploiting its ability of adapting to inhibitory hydrolysates. Significant improvement in
fermentation efficiency can be achieved by short pre-cultivation on hydrolysate (Alkasrawi et
al., 2006). Another approach is by continuous transferring of yeast to increased concentration
of hydrolysate. An example is the case of sequential adding of increasing concentration of
HMF and furfural to synthetic media. After more than 100 times of adding, the two S.
cerevisiae strains namely 307-12H60 and 307-12H120 demonstrated better reduction
capacity of HMF at high concentration of as 30 and 60 mmol/L, respectively. Furthermore,
they both grew and fermented glucose faster than the control strain Y-12632 (Liu et al.,
2005).
2.4. XYLOSE FERMENTATION
2.4.1. Xylose fermentation in yeast
S. cerevisiae which is a facultatively fermentative microorganism can convert sugars
completely to CO 2 and H 2 O in either aerobic or microaerobic condition, producing large
amounts of ethanol. In anaerobic condition, S. cerevisiae does not grow because essential
substances such as unsaturated fatty acids and sterols cannot be synthesized due to the

absence of O 2 (Lagunas, 1986). However, S. cerevisiae uses a narrow range of fermentable
substrates. Glucose, fructose, sucrose, galactose, mannose, and maltose are easily
metabolized. In contrast, cellobiose, lactose, xylose, rhamnose, sorbose, and maltotetraose
cannot be used.
Figure 2.7: Xylose and arabinose utilization pathways in bacteria (a,c) and in fungi (b,d)
(Satyanarayana, 2009)

S. cerevisiae which is used in bioethanol industry traditionally cannot ferment Dxylose. However, it can slowly metabolize D-xylulose which is an isomer of xylose. The
wild-type S. cerevisiae genome possesses genes for both xylose reductase and xylitol
dehydrogenase. Therefore, it can convert xylose into xylulose (figure 2.7) and the resulting
xylulose after being phosphorylated by the action of xylulokinase can enter the pentose
phosphate pathway (PPP) (Toivari et al., 2004). This is a biochemical pathway for xylose
15