<%BANNER%>

Metabolic Characterization and Engineering of Enterobacter Asburiae Strain Jdr-1 to Develop Biocatalysts for Efficient H...

Permanent Link: http://ufdc.ufl.edu/UFE0024266/00001

Material Information

Title: Metabolic Characterization and Engineering of Enterobacter Asburiae Strain Jdr-1 to Develop Biocatalysts for Efficient Hemicellulose Utilization
Physical Description: 1 online resource (118 p.)
Language: english
Creator: Bi, Changhao
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: 4, biocatalyst, bioenergy, enterobacter, fermentation, hemicellulose, metabolic, methylglucuronoxylan
Microbiology and Cell Science -- Dissertations, Academic -- UF
Genre: Microbiology and Cell Science thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: METABOLIC CHARACTERIZATION AND ENGINEERING OF ENTEROBACTER ASBURIAE STRAIN JDR-1 TO DEVELOP MICROBIAL BIOCATALYSTS FOR EFFICIENT HEMICELLULOSE UTILIZATION Acid pretreatment is commonly used to release pentoses from the hemicellulose fraction of cellulosic biomass for bioconversion. The predominant pentose in the hemicellulose fraction of hardwoods and crop residues is xylose in the polysaccharide methylglucuronoxylan, in which as many as 1 in 6 of the ?-1,4 linked xylopyranose residues are substituted with ?-1,2-linked 4-O-methylglucuronopyranose. Resistance of the ?-1,2-methylglucuronosyl linkages to acid hydrolysis results in release of the aldobiuronate 4-O-methylglucuronoxylose (MeGAX) which is not fermented by bacterial biocatalysts currently used for bioconversion of hemicellulose. Enterobacter asburiae JDR-1, isolated from colonized hardwood (sweetgum), was shown to efficiently ferment both MeGAX and xylose in acid hydrolysates of sweetgum xylan, producing predominantly ethanol and acetate. 13C-NMR studies defined the Embden-Meyerhof pathway for metabolism of glucose and the pentose phosphate pathway for xylose metabolism. MeGAX-defective mutants isolated after Tn5 transposon mutagenesis led to identification of oxi, ptsI, xylA and xylB genes required for efficient MeGAX metabolism. Based on both genetic and physiological studies, including rates of substrate utilization, product formation and molar growth yields, MeGAX utilization by E. asburiae JDR-1 requires transportation into the cell and hydrolysis by a novel pathway to release methanol, glucuronic acid and xylose. E. asburiae JDR-1 with deletion of the pflB gene encoding pyruvate formate lyase (E. asburiae E1) was transformed with a plasmid (pLOI555) carrying pdc and adhB genes derived from Zymomonas mobilis to generate a recombinant strain capable of completely fermenting both xylose and MeGAX in a hydrolysate of sweetgum xylan to ethanol with a yield at 99% of the theoretical maximum. A second strain, E. asburiae L1 was engineered by deletion of pflB and als genes to produce D(-)lactate at a yield approximately 100% of the theoretical maximum in hydrolysate of sweetgum xylan. The engineered strains of E. asburiae JDR-1 provide novel microbial biocatalysts for maximal conversion of carbohydrate compounds in acid hydrolysates of hemicellulose into ethanol and other fermentation products.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Changhao Bi.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: Preston, James F.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2010-08-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024266:00001

Permanent Link: http://ufdc.ufl.edu/UFE0024266/00001

Material Information

Title: Metabolic Characterization and Engineering of Enterobacter Asburiae Strain Jdr-1 to Develop Biocatalysts for Efficient Hemicellulose Utilization
Physical Description: 1 online resource (118 p.)
Language: english
Creator: Bi, Changhao
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: 4, biocatalyst, bioenergy, enterobacter, fermentation, hemicellulose, metabolic, methylglucuronoxylan
Microbiology and Cell Science -- Dissertations, Academic -- UF
Genre: Microbiology and Cell Science thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: METABOLIC CHARACTERIZATION AND ENGINEERING OF ENTEROBACTER ASBURIAE STRAIN JDR-1 TO DEVELOP MICROBIAL BIOCATALYSTS FOR EFFICIENT HEMICELLULOSE UTILIZATION Acid pretreatment is commonly used to release pentoses from the hemicellulose fraction of cellulosic biomass for bioconversion. The predominant pentose in the hemicellulose fraction of hardwoods and crop residues is xylose in the polysaccharide methylglucuronoxylan, in which as many as 1 in 6 of the ?-1,4 linked xylopyranose residues are substituted with ?-1,2-linked 4-O-methylglucuronopyranose. Resistance of the ?-1,2-methylglucuronosyl linkages to acid hydrolysis results in release of the aldobiuronate 4-O-methylglucuronoxylose (MeGAX) which is not fermented by bacterial biocatalysts currently used for bioconversion of hemicellulose. Enterobacter asburiae JDR-1, isolated from colonized hardwood (sweetgum), was shown to efficiently ferment both MeGAX and xylose in acid hydrolysates of sweetgum xylan, producing predominantly ethanol and acetate. 13C-NMR studies defined the Embden-Meyerhof pathway for metabolism of glucose and the pentose phosphate pathway for xylose metabolism. MeGAX-defective mutants isolated after Tn5 transposon mutagenesis led to identification of oxi, ptsI, xylA and xylB genes required for efficient MeGAX metabolism. Based on both genetic and physiological studies, including rates of substrate utilization, product formation and molar growth yields, MeGAX utilization by E. asburiae JDR-1 requires transportation into the cell and hydrolysis by a novel pathway to release methanol, glucuronic acid and xylose. E. asburiae JDR-1 with deletion of the pflB gene encoding pyruvate formate lyase (E. asburiae E1) was transformed with a plasmid (pLOI555) carrying pdc and adhB genes derived from Zymomonas mobilis to generate a recombinant strain capable of completely fermenting both xylose and MeGAX in a hydrolysate of sweetgum xylan to ethanol with a yield at 99% of the theoretical maximum. A second strain, E. asburiae L1 was engineered by deletion of pflB and als genes to produce D(-)lactate at a yield approximately 100% of the theoretical maximum in hydrolysate of sweetgum xylan. The engineered strains of E. asburiae JDR-1 provide novel microbial biocatalysts for maximal conversion of carbohydrate compounds in acid hydrolysates of hemicellulose into ethanol and other fermentation products.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Changhao Bi.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: Preston, James F.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2010-08-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024266:00001


This item has the following downloads:


Full Text





METABOLIC CHARACTERIZATION AND ENGINEERING OF ENTEROBACTER
ASBURIAE STRAIN JDR-1 TO DEVELOP MICROBIAL BIOCATALYSTS FOR
EFFICIENT HEMICELLULOSE UTILIZATION




















By

Changhao Bi


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

2009




























2009 Changhao Bi





























To my parents and my wife









ACKNOWLEDGMENTS

I thank my family, for without their support, I would not have been able to complete

this work. And I would like to thank my supervisor, Dr. Preston and the rest of my committee.

In the past 5 years, I have gained much knowledge and developed many skills in the field of

microbiology and molecular biology with the relentless help of my supervisor Dr. Preston

and my committee members. However, what has affected me most is the serious attitude for

scientific research my supervisor and his fellow professors demonstrated. Also, I thank my

lab colleagues for the kind help and support to my research.










TABLE OF CONTENTS

page

A C K N O W L E D G M E N T S ..............................................................................................................4

LIST OF TABLES.......... ......... ......... ........ ................... ........8

LIST OF FIGURES ................ ....... .................................... 9

AB STRACT ........... .................... .............................................. 10

CHAPTER

1 LITERATURE REVIEW .....................................................................12

Advantages of Utilizing Lignocellulosic Biomass to Produce Ethanol and Chemicals.........12
N national E energy Security .............................................................12
Economic Advantages ...... ................. ... .... ...................... 13
Environmental Advantages ......................... ................ 15
Constituents of Lignocellulose and Hemicellulose .......... ............................................16
Lignocellulose .......................................................16
Hemicellulose ............... ............... ........................... ........ 17
Pretreatment Processes of Lignocellulose ....................... ..................18
Pretreatment Methods ..................... ........ ............................. 18
Dilute Acid Pretreatment...... ......... ........ ..................19
Cost of Pretreatment Methods .........................................21
Development of Ethanol Producing Biocatalysts toward Hemicellulose Utilization.........21
Escherichia coli......................................................... ................22
Zymom onas mobilis .................. ..........................................................23
Klebsiella oxytoca ................................................ ..... ...24
Yeasts............................................... ........ .................... .........25
O objectives of this R research .......................................................................................... ... ....... 26

2 COMPLETE FERMENTATION OF XYLOSE AND
METHYLGLUCURONOXYLOSE DERIVED FROM
METHYLGLUCURONOXYLAN BY Enterobacter asburiae STRAIN JDR-1 ..................29

Introduction ...................................... ........................................... 29
M materials and M methods ........................ ................... ...............30
Preparation of Substrates and Culture M edia.......................... ...............30
Isolation and Identification of E. asburiae JDR-1......................................................32
Substrate Utilization and Fermentation Product Analysis.............................................32
Determination of Metabolic Pathways by 13C-NMR .................................................33
Determination of Molar Cell Dry Weight Yield ......................................................34
Results ............... .. .................. .... .. ... ........ ............... 35
Identification and Characterization of E. asburiae JDR-1 ............................................35
Utilization of Acid Hydrolysates of Methylgluronoxylan by E. asburiae JDR-1 ..........36
Substrate Preference of E. asburiae JDR-1 ......... ........................36
Ferm entation Characteristics .............. ..............................................7
Central M etabolic Pathways Determined by 13C-NM R ...............................................38










Growth and Projected ATP Yields with Different Substrates.......................................40
Discussion..................................................... .........41
Fermentation of M eGAX by E. asburiae JDR-1 ...........................................................41
Central Metabolic Pathways Used by E. asburiae JDR-1.............................................41
Possible Pathway of M eGAX M etabolism .................................. ............................. 43
Bioenergetics of E. asburiae JDR-1 ....................................... ............... 45
Role of E. asburiae JDR-1 in Soil Ecology and Bioprocessing.............. ....46
B iotechnological A applications ............................................... ............... 46

3 GENETIC ENGINEERING OF Enterobacter asburiae STRAIN JDR-1 FOR
EFFICIENT D(-)LACTIC ACID PRODUCTION FROM HEMICELLULOSE
HYDROLY SATE............... ....................... ................... .......... .. ...............58

Introduction ...................................... ......... .............................. 58
M materials and M methods ................................... ...............................................59
Bacterial Strains, Media, and Growth Conditions...................................................59
Genetic M ethods............... . ...............60
Deletion ofpflB and als Gene in E. asburiae JDR-1 ............ ..... ............ 60
Fermentation .......................... .. ... .. ... .. ........ 62
Determination of Lactate Isomers Produced by E. asburiae L1 ............................. 62
R esults............... .... ........... ......... ........ .. . .........................63
Fermentation Characteristics of the Wild type Strain E. asburiae JDR-1 ....................63
Fermentation Characteristics of the Engineered Strains E. asburiae El and L1 ............64
D-Lactate Was Produced by E. asburiae LI ...........................................65
Discussion............... ............ .................. .......... 66

4 GENETIC ENGINEERING Enterobacter asburiae STRAIN JDR-1 FOR
EFFICIENT ETHANOL PRODUCTION FROM HEMICELLULOSE
HYDROLY SA TE ..................72.................................................

Introduction ...................................... ......... .............................. 72
M materials and M methods ................................... ...............................................73
Bacterial Strains, Media, and Growth Conditions...................................................73
Genetic methods ....................................... .. ............ ...............73
Transformation of Plasmid Carrying PET Operon into E. asburiae JDR-1....................73
Integration of PET Operon into E. asburiae JDR-1 Chromosome ...............................74
Plasmid Stability in E. asburiae JDR-1 .................................................. .......... 75
Assay of PDC Activity .............. .................. .......... ..................... 75
R esults............... .... ........... .. .............. ...... .... .. .. ......... ...............76
Fermentation Characteristics of the Wild Type Strain E. asburiae JDR-1 .....................76
Fermentation of E. asburiae 4666 and E. asburiae 4672 in Glucose...........................77
Fermentation of E. asburiae JDR-1 (pLOI297) and E. asburiae JDR-1
(pLO I555) .......... ................................77
Fermentation of Strain E. asburiae El (pLOI555) Compared with E. coli KO 11
and Other E. asburiae JDR-1 D erivatives ........................................ ............... 78
PD C A activity in E. asburiae Strains............................................................ ......79
Plasmid Stability in E. asburiae JDR 1 ................................... ..... ............... 80
Discussion............... ............ .................. .......... 80

5 GENETIC DEFINITION OF THE MEGAX UTILIZATION PATHWAY IN
E nterobacter asburiae JD R -1 ............................................................................................ 9 1









Introduction............... ................... ......................................... 91
M materials and M methods ............................ ... .... .... ..... ..............91
Degenerate Primer Method to Determine Presence of GH 67 genes ..............................91
Determination of Genes Involved in MeGAX Utilization Process with Mutant
Library M ethod ...................................... .......... .. ............................... 92
Determination of MeGAX Utilization Genes by Transforming E. coli with
Cosmid Library and Topo Plasmid Library .............................................................93
Obtaining Candidate Genes by Sequencing Cosmid Containing P-xylosidase
G ene and G enom e C om prison ............................................ .........................93
Deletion of Candidate Genes to Confirm Their Involvement in MeGAX
Utilization Process ............................................... .... .... 94
R results .................. ................... ................. .. ........ .............. ........ 94
MeGAX Defective Mutants of E. asburiae JDR-1 ....................................................94
Candidate Genes Determined From Cosmid and Plasmid Library
Transform action M ethod ...................... ............. ...........................97
Obtaining Candidate Genes for Function Confirmation by Gene Deletion .................97
Discussion............... ............ .................. .......... 99

6 CONCLUSION..................... .............. ... ........ 104

REFERENCE LIST .................. ........................ ...................................107

BIO GRAPH ICAL SK ETCH .............................................................................................. ....... 118









LIST OF TABLES


Table page

2-1. Fermentation products formed by E. asburiae JDR-1 from monosaccharides derived
from hemicellulose. .......................................... ........ ........ 48

2-2. Fermentation products of E. asburiae JDR-1 derived from MeGAXn. Acetate,
ethanol and format concentrations were determined in duplicate cultures with
HPLC and m ethanol w ith GC. ................................................ ............... 48

2-3. Distribution of 13C in fermentation products formed in anaerobic cultures of E.
asburiae JDR-1 and E. coli B grown with differentially 13C labeled xylose and
glucose. .............................................49

2-4. Anaerobic molar cell dry weight and ATP yield from different substrates calculated
based on estimated YATP, 8, for all substrates in E. asburiae JDR-1............................49

3-1. B bacterial strains and plasm ids. ........................................ ...................67

3-2. Comparing fermentation products of wild type and genetically engineered E.
asburiae JDR-1 strains. ................................................. ........68

3-3. Specific consumption rates and specific production rates ofE. asburiae LI in
5g/liter acid hydrolysate of sweetgum xylan.a................ ...............69

4-1. Bacterium strains and plasmids for engineering ethanologenic E. asburiae........................85

4-2. Comparison of sugar fermentation products of wild type and genetically engineered
E. asburiae JDR-1. Fermentations were carried out at 300C in ZH minimal media
for 48 hours as described in the Materials and Methods section. ...................................86

4-3. Fermentation products from acid hydrolysates of sweetgum xylan. Fermentations
were carried out at 300C in ZH minimal media for 48 hours as described in the
Materials and Methods section. Results were averages of 3 experiments.........................86

4-4. Specific consumption rates and specific production rates in acid hydrolysates of
sweetgum xylan (5g /liter)a. Results were averages of 3 experiments.............................87

4-5. Specific activity of PDC in cell crude extract from E. asburiae JDR-1 derived
strains. Results were averages of 3 experiments................ ..............87

4-6. Plasmid stability of pLOI297 and pLOI555 in E. asburiae JDR-1. Results were
averages of 3 experim ents ................................................................................... ............87

5-1. Growth rate of ptsT in different substrate with and without cAMP...................................101









LIST OF FIGURES


Figure page

1-1. M odel for corn stover cell w alls. ...................................................................................... 28

2-1. Scheme for the release of xylose and MeGAX by dilute acid hydrolysis of sweetgum
xylan..................................................................................................................50

2-2. Microscopic images of E. asburiae JDR1. ....................................................... 51

2-3. Aerobic growth, substrate utilization, and formation of products from acid
hydrolysates of M eG A X n........................................................................................... ...........52

2-4. Aerobic growth of E. asburiae JDR-1 on different combinations of sugar substrates...........53

2-5. Pathway determination for the metabolism of xylose and glucose by E. asburiae
JD R -1 ... ..... .. ... ...... ....... ......................... .................................... 55

3-1. Fermentation time course for different strains in media containing 0.5% sweetgum
xylan hydrolysate. ........................... ............................70

3-2. Diagram to illustrate deletion of als andpflB genes modifying mixed-acid
fermentation of E. asburiae JDR-1 into a homolactate production pathway in E.
asburiae L .. ................................................. .........71

4-1. HPLC profile of fermentation products of E. asburiae JDR- 1, E. coli KO 11 and E.
asburiae El (pLOI555) in 0.5% sweetgum xylan hydrolysate with MOPS......................88

4-2. Fermentation time course of E. asburiae JDR-1 A), E.coli KO 11 B), E. asburiae
JDR-1 (pLOI555) C) and E. asburiae El (pLOI555) D) in media of buffered
sw eetgum xylan hydrolysate................................................. 89

5-1. Function and regulation of PTS ............................................................................102

5-2. HPLC profile of aerobic culturing media of E. asburiae JDR-1 xylA...... ...... .......... 102

5-3. Enzyme activity of GH88 enzyme unsaturated -glucuronyl hydrolase (EC 3.2.1.-)
(Hashimoto et al., 1999)...................... ............... 103









Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

METABOLIC CHARACTERIZATION AND ENGINEERING OF ENTEROBACTER
ASBURIAE STRAIN JDR-1 TO DEVELOP MICROBIAL BIOCATALYSTS FOR
EFFICIENT HEMICELLULOSE UTILIZATION

By

Changhao Bi

August 2009

Chair: James F. Preston
Major: Microbiology and Cell Science

Acid pretreatment is commonly used to release pentoses from the hemicellulose

fraction of cellulosic biomass for bioconversion. The predominant pentose in the

hemicellulose fraction of hardwoods and crop residues is xylose in the polysaccharide

methylglucuronoxylan, in which as many as 1 in 6 of the P-1,4 linked xylopyranose residues

are substituted with a-1,2-linked 4-O-methylglucuronopyranose. Resistance of the

a-1,2-methylglucuronosyl linkages to acid hydrolysis results in release of the aldobiuronate

4-O-methylglucuronoxylose (MeGAX) which is not fermented by bacterial biocatalysts

currently used for bioconversion of hemicellulose. Enterobacter asburiae JDR-1, isolated

from colonized hardwood (sweetgum), was shown to efficiently ferment both MeGAX and

xylose in acid hydrolysates of sweetgum xylan, producing predominantly ethanol and acetate.

13C-NMR studies defined the Embden-Meyerhof pathway for metabolism of glucose and the

pentose phosphate pathway for xylose metabolism. MeGAX-defective mutants isolated after

Tn5 transposon mutagenesis led to identification of oxi, ptsl, xylA and xylB genes required for

efficient MeGAX metabolism. Based on both genetic and physiological studies, including

rates of substrate utilization, product formation and molar growth yields, MeGAX utilization

by E. asburiae JDR-1 requires transportation into the cell and hydrolysis by a novel pathway

to release methanol, glucuronic acid and xylose.









E. asburiae JDR-1 with deletion of the pflB gene encoding pyruvate format lyase (E.

asburiae El) was transformed with a plasmid (pLOI555) carryingpdc and adhB genes

derived from Zymomonas mobilis to generate a recombinant strain capable of completely

fermenting both xylose and MeGAX in a hydrolysate of sweetgum xylan to ethanol with a

yield at 99% of the theoretical maximum. A second strain, E. asburiae LI was engineered by

deletion ofpflB and als genes to produce D(-)lactate at a yield approximately 100% of the

theoretical maximum in hydrolysate of sweetgum xylan. The engineered strains of E.

asburiae JDR-1 provide novel microbial biocatalysts for maximal conversion of carbohydrate

compounds in acid hydrolysates of hemicellulose into ethanol and other fermentation

products.









CHAPTER 1
LITERATURE REVIEW

Advantages of Utilizing Lignocellulosic Biomass to Produce Ethanol and Chemicals

The main driving forces to develop renewable supplies of energy are derived from 1)

National energy security: to decrease the dependence of imported fuels from certain areas with

political instability and prepare for depletion of fossil fuels; 2) Economic advantages: to develop

less expensive renewable fuels in large amounts without increasing food prices; and 3)

Environmental advantages: to alleviate the greenhouse effect and make use of industrial,

agricultural and municipal wastes.

National Energy Security

Developing biobased sustainable fuel production is crucial for national energy security.

Societies are mainly sustained by petroleum-based energy currently, which is a non-renewable

resource with limited reserves. As the human population increased 4-fold in the twentieth

century, global energy consumption increased 16-fold to reach 13 trillion watts. When the

population reaches 8 toll billion by 2050, the energy consumption is expected to be 27 to 42

trillion watts (Whitesides and Crabtree 2007; Hoffert et al. 2002). The estimated amount of oil in

determined reservoirs is approximately 1 trillion barrels while the total possible reserves might

be as high as 2 trillion barrels with new discoveries, expansion of existing fields and new

extracting technologies. Since current oil production and consumption is about 30 billion barrels

per year, the global oil reserve might become extremely limited within 40 to 50 years and

exhausted in 30 to 70 years (Wesseler 2007). Although some reports present a more optimistic

estimation of the oil reserve, the growth and increase in global economy is expected to accelerate

the rate of petroleum depletion. The United States consumes approximately 30 billion barrels oil

per year and about 30 % of total oil produced worldwide. As much as 60% of the oil U.S.









consumed is imported (Gray et al. 2006). The dependence of oil from foreign sources negatively

affects the U.S. national energy security. Development of a renewable transportation fuel is

urgently needed to reduce the dependence on imported oil.

Biomass is the only known renewable bioresource that can be used for production of liquid

transportation fuel (Wyman 2003). Domistic production of ethanol from biomass will reduce the

dependence on imported fuels (Lin and Tanaka 2006). The U.S. Department of Energy has

proposed an objective of producing 30% of the U.S. transportation fuels from biomass by 2030,

based on the availability of lignocellulose in the U.S. (Himmel et al. 2007; Hahn-Hagerdal et al.

2006). The DOE and USDA reported that the United States can produce approximately 1.3

billion dry tons of lignocellulose per year, including 933 million tons from agricultural and 368

million tons from forest resources (Zhang 2008). One billion dry tons of lignocellulosic biomass

contain sufficient carbohydrate to produce 80-130 billion gallons of bioethanol. Therefore the 1.3

billion dry tons of lignocellulose produced per year in the United States is capable of producing

approximately 60 billion gallons of bioethanol to meet the goal of replacing 30% transportation

fuel in the U.S. In addition, the ocean also has a great potential for biofuel production with its

lage surface area for marine biomass production. Microalgae and macroalgae, resources for

biofuel production, are reported to produce 81-150 dry ton/ha/year, if saturating levels of CO2

are provided (Gao and Mckinley 1994).

Economic Advantages

Lignocellulosic biomass is abundant, renewable and relatively inexpensive. Development

of biomass as an energy resource has the potential for production of cheaper alternative fuels.

Production of biomass-based fuel does not compete with food sources and does not increase food

prices. Furthermore, land not fit for food crops can be used to culture energy crops, albeit will









lower annual yields. Production of large amounts of bioethanol may help stablelize the price of

transportation fuels.

Biomass is produced by photosynthesis, during which light energy is converted to

chemical energy stored as fixed CO2 in biomass. Lignocellulosic biomass is the most abundant

renewable biological resource, with a production of approximately 2 x 1011 dry tons per year

(Reddy and Yang 2005). Most lignocellulosic biomass is available at lower prices than crude oil.

For example, the crude oil ($40 to 80$ per barrel) costs about $7.1 to $14.2/GJ, which is much

higher than the price of lignocellulose (up to $3/GJ), based on equivalent energy content.

Although the values of lignocelluloses are variable according to feedstock composition, the price

of lignocellulose-based products is expected to be competitive with fossil resource-derived

products if processing costs of lignocelluloses are reduced. When municipal waste is used, the

reduction of waste removal fees brings additional economic advantages (Wyman 2003).

In 2006, fermentation of corn starch accounted for more than 93 % of the total ethanol

production in the U.S. (Lin and Tanaka 2006). Increasing ethanol production in the near future

will require more corn to be diverted away from food and feed sources, which will inevitably

increase the price of food and other downstream products of corn. For example, during the years

2004 to 2006, corn kernel price rose by more than 70% from $2.25 per bushel to $4 per bushel.

The Chinese government even banned building new ethanol production facilities based on grains

in 2006 (Zhang 2008). However, using lignocellulose as feedstock to produce ethanol does not

compete with food and feed sources, or increase the food prices which negatively impacts the

society welfare (Gray et al. 2006; Sticklen 2006). Furthermore, bioenergy crops do not compete

with agricultural crops for land committed to cultivation, since land of lower quality can be used

for energy crops production, albeit the lower annual yield. In addition, cultivation of crops for









energy production generally requires less fertilizer and energy inputs compared to food crops

(Wyman 1999).

Environmental Advantages

Using biomass-based fuel has the potential to alleviate the greenhouse effect and reduce

the pollution of industrial, agricultural and municipal wastes. Fossil fuels, the only sources of net

CO2 production, are the primary sources of the greenhouse gases. Fossil fuel combustion is

responsible for approximately 82% of net greenhouse gas emission, as much as about 7.0 billion

tons of carbon per year (Lal 2004; Eissen et al. 2002). Since the industrial revolution in

nineteenth century, nearly 270 billon tons of carbon from the combustion of fossil fuels have

accumulated in the atmosphere and the atmospheric CO2 level has increased from 275 to 380

ppm. Consequently, the global average near-surface temperature has risen by 0.60.20C in the

twentieth century (Lal 2004). The CO2 level is expected to pass 550 ppm by the middle of this

century and cause further increase of temperature according to current momentum (Kheshgi et al.

2000; Galbe and Zacchi 2002). Because CO2 generated from ethanol is recycled into plants and

algae by photosynthesis, no net CO2 is produced with utilization of ligcellulosic ethanol.

Therefore development of ligcellulosic ethanol as an alternative fuel has the potential to reduce

accumulation of CO2 gas and reduce the greenhouse effect. Ethanol is also a potent fuel

oxygenate. When blended with gasoline, ethanol is able to improve engine combustion and

reduce CO2 emissions (Jeuland et al. 2004). Additionally, utilization of ethanol also reduces the

amount of ozone forming compounds, such as SOx, NOx, and CO (von Blottnitz and Curran

2007).

Large volumes of waste material are produced every year. Although portions of the waste

can be sold as animal feed or burned to generate electricity, most waste material that can not be

utilized is usually buried in land fills. There are environmental concerns about the large amount









of hazardous matter produced by buried wastes. Household and office lignocellulosic waste

(paper, yard waste, etc.) also contribute to the material currently being placed in land fills

(Prasad et al. 2007). With development of biomass-based ethanol, most of these lignocellulosic

wastes are able to be processed as feedstock to produce ethanol, thus largely reducing the amount

of land fills, removing disposal problem of wastes and reducing pollution from wastes.

Constituents of Lignocellulose and Hemicellulose

Lignocellulose

In general, lignocellulosic biomass is the mixture of structural components of plant matter

as opposed to the sugar storage compounds, starch and sucrose. In other words, lignocellulosic

biomass is the inedible portion of plant material including wood, plants, grasses, crop residues,

and ligocellulosic wastes.

The primary components of plant cell walls are cellulose, hemicellulose and lignin.

Cellulose is the most abundant carbohydrate polymer in the world and is the major structural

component of plants, ranging from 36% to 50% of total plant mass. Cellulose in plant walls is

constituted by associated crystalline fibers consisting of many individual cellulose strands

assembled by hydrogen bonds and Van der Waal's forces. These single cellulose strands are

0-1,4-linked glucose polymers (Lynd et al. 1999). Quantitative data of cellulose accessibility

clearly suggests that only a small fraction of P-glucosidic bonds of crystalline cellulose are

accessible to cellulases (Hong et al. 2007). The second most abundant constituent in biomass is

the hemicellulose. Polymer methylglucuronoarabinoxylan is the major component of

hemicellulose in graminaceous plants representing up to 35% of total plant mass (Lynd et al.

1999). In hardwoods hemicelullose is composed primarily of methylglucuronoxylan ranging

from 20% to 35% of total lignocellulose (Puls 1997). Lignin is composed of phenyl-propane









derivatives, randomly linked through carbon-carbon bonds by an enzymatic dehydrogenation

process, and associated to hemicellulose to form a network (Monties 2005).

Hemicellulose

The major hemicellulosic component of hard wood, methylglucuronoxylan, is a 0-1,4

linked xylose chain polymer, in which the xylose groups are randomly substituted with a-1,2

linked 4-O-methylglucuronate residues at certain frequency. Hemicelluloses are located among

the lignin and the cellulose fibers and covalently linked at many points of their main chain and

substitution group with the lignin network. For example, two types of covalent cross-links are

identified between hemicellulose and lignin: diferulic acid bridges and ester linkages between

lignin and glucuronic acid attached to xylans (de Vries and Visser 2001). Hemicelluloses also

associate with cellulose strands by hydrogen bonds to form layers of network around them. So

that the xylan layer has both covalent linkages to lignin and non-covalent interaction with

cellulose to maintain the plant cell wall and helps the associated cellulose fibers against

degradation by cellulases (Beg et al. 2001).

Xylans from different sources, such as grasses, cereals, softwood, and hardwood, differ in

composition. In various sources, the frequency of methylglucuronate substitution is different,

ranging from 1 of every 6 xylose residues to 1 of every 10 xylose residues (Jacobs et al. 2001,

Sunna and Antranikian 1997). The 13C-NMR studies of methylglucuronoxylan from the

hardwood sweetgum illustrate the frequency of methylglucuronate substitution is 1 in every 6

xylose residues (Preston et al. 2003; Puls 1997). Methylglucuronoarabinoxylan in softwood has a

higher 4-O-methylglucuronate content than hardwood xylans, but can contain arabinose at a

frequency of about one for every six xylose residues. The a-L-arobinofuranose units are linked

by a-1,3-glycosidic bonds to the C-3 position of the xylose groups in xylan main chain.

Softwood xylans are shorter than the hardwood xylans, with a degree of polymerization from 70









to 130 xylose groups and less branched (Xiao et al. 2001; Sunna and Antranikian 1997). A model

for the corn stover cell wall is shown in Fig 1-1 (Saha 2003). The xylans and lignins, are highly

cross-linked by diferulic linkages, and constitute a network where the cellulose fibers fit in

underneath.

Pretreatment Processes of Lignocellulose

Pretreatment Methods

In lignocellulose biorefineries, biological conversion oflignocellulose generally has three

main steps: 1) lignocellulose pretreatment, which converts the original lignocellulose structure to

enzyme-reactive cellulose and pentoses; 2) enzymatic cellulose hydrolysis, by which cellulases

hydrolyze reactive cellulose to fermentable sugars; and 3) fermentation, which produces ethanol

or other bio-based chemicals from released sugars (e.g., lactic acid, succinic acid) (Ragauskas et

al. 2006). Conversion of lignocellulosic biomass to fermentable sugars is the most challenging

step in the process of utilization, due to the complex nature of lignocelluloses. Utilization of the

carbohydrates of lignocellulosic biomass requires a physico-chemical pretreatment. Following

the first pretreatment step, most methods require an enzymatic cellulose hydrolysis step, in

which celluases from fungi are supplemented to degrade the cellulose compartment into glucose.

Various pretreatment options are available to fractionate, separate, solubilize and hydrolyze

cellulose, hemicellulose, and lignin components. These methods include dilute sulfuric acid,

flow through, pH controlled water, ammonia fiber explosion (AFEX), ammonia recycle

percolation (ARP) and lime pretreatment. Except for AFEX, all the other pretreatment methods

function primarily by removing hemicellulose and lignin. Since the hemicellulose and lignin

fractions are associated within the cellulose matrix, the methods which remove either

hemicellulose or lignin increase cellulose accessibility and make it susceptible to enzymatic

hydrolysis (Saha 2003; Teymouri et al. 2004).









AFEX is a physico-chemical pretreatment in which lignocellulosic materials are exposed

to liquid ammonia at high temperature and pressure for a period of time, and then the pressure is

quickly reduced to induce the ammonia to explode. The explosion significantly improves the

enzymatic saccharification rates of various herbaceous crops and grasses. However, the AFEX

pretreatment does not significantly solubilize hemicellulose compared to acid pretreatment and

acid catalyzed steam explosion (Meshartree et al. 1988; Vlasenko et al. 1997).

The effects of lime pretreatment depend on the lignin content of the materials (Hu and

Wen 2008). The mechanism of this pretreatment method is saponification of intermolecular ester

bonds which crosslink xylan hemicellulose and other components. By removal of these

crosslinks, the enzymatic accessibility of the treated lignocellulosic materials increases. Dilute

NaOH pretreatment of lignocellulosic materials causes swelling of materials, which leads to

internal surface area increase, crystallinity decrease, linkage separation between lignin and

carbohydrates and lignin structure disruption (Tarkow and Feist 1969).

Concentrated sulfuric and hydrochloric acids have been used to treat lignocellulosic

materials. Although they are powerful agents for cellulose hydrolysis, concentrated acids are

toxic, corrosive and hazardous. Pretreatments with concentrated acids also require expensive

corrosion-resistant reactors. In addition, this method is not very economical, due to the higher

cost of concentrated acids (vonSivers and Zacchi 1995).

Dilute Acid Pretreatment

Dilute acid pretreatment is currently developed as a preferred pretreatment method (Saha

et al. 2005a; Saha et al. 2005b). This method employs mild acid conditions (0.5-3.0% H2SO4)

with temperatures from 1300C to 2000C and pressures from 3 atm tol5 atm. There are two major

types of dilute acid pretreatment processes: high temperature (>160 'C), continuous flow process

for low solid loading (5-10%) (Converse et al., 1989); and low temperature (<160 'C), batch









process for high solid loading (10-40%) (Esteghlalian et al. 1997). In dilute acid hydrolysis, the

acid, temperature and pressure function together to release single sugars and small oligomers

from the hemicellulose portion almost completely. In addition, dilute acid treatment disrupts and

exposes the cellulose portion for effective enzymatic hydrolysis. The dilute acid treatment was

reported to convert approximately 83% of the total xylose content to free xylose and slightly

more than 2% to xylooligomers. The release of hemicellulose proportion may be the major factor

for the nearly complete hydrolysis of cellulose (-92%) with the subsequent cellulase treatment

(Eggeman and Elander 2005). In this process, the high xylan to xylose conversion yields is also

economically favorable because xylan accounts for up to a third of the total carbohydrate in most

lignocellulosic materials (Hinman et al. 1992).

There are some drawbacks and limitations of the dilute acid pretreatment method. The

most detrimental feature is the formation of acid hydrolysis side products such as furfural and

a-1,2-methylglucuronoxylose (Jones et al. 1961; Zaldivar et al. 1999). Furfural forms from the

acid and heat catalyzed dehydration of xylose. The formation of this side product inhibits

microbial growth and fermentation and reduces the net convertible xylose concentration

(Zaldivar et al. 1999). The aldouronate, a-1,2-methylglucuronoxylose (MeGAX), is produced

from acid hydrolysis of methylglucuronoxylan and methylglucuronoarabinoxylan due to the

stability of the a-1,2 glycosyl linkage between methylglucuronate and xylose. This linkage is

thought to form an internal lactone between the carboxylate residue on the glucuronic acid and a

hydroxyl on the xylose in main chain under acidic conditions (Jones et al. 1961). While the

arabinose and acetyl substitutions are released, the substituted aldouronate, MeGAX, is not

hydrolyzed in the pretreatment. It is also unable to be utilized by current ethanologenic

biocatalysts (Rodriquez et al 2001). Considering the frequency of substitution of









methylglucuronate is from 1 for every 10 to 1 for every 6 xylose residues, current bioconversion

process can only recover 83% to 90% of the total xylose fraction. This suggests a biocatalyst

capable of utilizing the currently unusable MeGAX portion of the dilute acid hydrolysate will

increase the efficiency of bioconversion started with dilute acid pretreated biomass.

Cost of Pretreatment Methods

Research demonstrated none of these pretreatment methods had a clear advantage

economically (Eggeman and Elander 2005). The direct fixed capital for the dilute acid, AFEX,

ARP, and lime methods are approximately the same. As for the dilute acid method, the primary

cost is associated with equipment required to handle corrosive conditions. Thus the pretreatment

reactors account for most of the cost, and a minor cost is associated with chemical supply

requirements. AFEX pretreatment, on the other hand, requires costly pure ammonia, but less

expensive reactors and equipment. Dilute acid, AFEX and lime pretreatments result in the

lowest total fixed capital per gallon of annual ethanol production capacity making these the most

preferred methods for large scale pretreatment currently. The cost for both dilute acid and AFEX

pretreatment is calculated to be $3.72 per gallon ethanol (Eggeman and Elander 2005).

Development of Ethanol Producing Biocatalysts toward Hemicellulose Utilization

The development of industrially suitable microorganisms for converting biomass into fuel

ethanol is a major technical roadblock for a mature bioethanol industry. Over the last several

decades, microorganisms are developed as biocatalysts to convert lignocellulosic sugars into

ethanol (Ingram et al. 1999b; Jeffries 2005). Microorganisms capable of fermenting sugars not

fermentable by brewers' yeast are needed to utilize complex single carbohydrates or sugar

oligomers derived from lignocellulosic biomass. The greatest successes have been achieved in

the engineering of Gram-negative bacteria: Escherichia coli, Klebsiella oxytoca and Zymomonas

mobilis (Dien et al. 2003).









Escherichia coli

E. coli has several advantages as a biocatalyst for ethanol production. These advantages

include the ability to ferment a wide range of sugars, no requirements for complex growth factors,

well established genetic manipulation methods and prior industrial usage. The construction of E.

coli strains to selectively produce ethanol is one of the first successful applications of microbial

metabolic engineering. Development of ethanologenic E. coli includes a combination of genetic

engineering and metabolic adaption evolution. The adhB andpdc genes were co-expressed under

the control of the native lac promoter, and the construct was named the PET (production of

ethanol) operon (Ingram et al. 1987). The PET operon was inserted into the pyruvate format

lyase gene (pfl) to eliminate this enzyme competing for pyruvate. The integration recombinants

were further screened for increased chloramphenicol resistance in order to enhance the

expression level of PET operon gene products. In addition, the fumarate reductase gene (frd) in

the succinate fermentation pathway was disrupted to eliminate succinate production. The final

strain, E. coli KO 11, is able to convert glucose and xylose to ethanol at yields 103 to 106% of

theoretical maximum in rich media (Ingram et al. 1999).

Long-term adaptation evolution method was used by Yomano et al. in 1998 on medium

supplemented with ethanol to increase the ethanol tolerance of KO 11 by approximately 10%. E.

coli LYO1A was an isolate from an adapted culture. The adaptation reduced time to ferment 140

g/l xylose from 120 h by E. coli KO11 to 96 h by E. coli LYO1A (Yomano et al. 1998). E. coli

LYO 1 also has a higher tolerance to hydrolysate-associated inhibitors, including aldehydes,

alcohols and organic acids (Zaldivar and Ingram 1999; Zaldivar et al. 1999; Zaldivar et al., 2000).

While both KO 11 and LYO 1 achieve high ethanol yields and titers in rich media, both strains

perform poorly in minimal media (Jarboe et al. 2007; York and Ingram 1996). To enable

ethanologenic E.coli to ferment in inexpensive minimal media, a new strain was constructed









from strain SZ110, which was developed from KO11 for lactate production in mineral salts

medium. Strain LY160, with the restored pflB gene and PET operon inserted down stream of the

rrlE gene promoter of SZ 110, has higher yield of ethanol in minimal medium than previously

constructed ethanologenic E.coli (Yomano et al. 2007).

Non-recombinant ethanologic E.coli was also developed by isolating mutant strains

capable of anaerobic growth derived from AldhA ApflB double deletion strains. The selected

strain SE2378 ferments glucose and xylose to ethanol with an 82% yield (Kim et al. 2007). It

was found this strain had an essential mutation within the pyruvate dehydrogenase operon to

enable this enzyme to be active in anaerobic condition. With this enzyme, an additional NADH

is produced compared to pyruvate formatelyase from one pyruvate. With activation of pyruvate

dehydrogenase, a homo-ethanologenic pathway in E. coli is oxi-reductively balanced. E. coli

based biocatalysts are still under development to enhance their traits for better industrial

application.

Zymomonas mobilis

Zymomonas mobilis (Z. mobilis) is a Gram-negative microorganism with appealing

properties as a biocatalyst for ethanol production. It has a homoethanologenic fermentation

pathway and high ethanol tolerance. Even compared with Saccharomyces sp. it has a higher

ethanol yield and specific ethanol productivity (Rogers et al. 2007). Zymomonas is also the only

microorganism that metabolizes glucose anaerobically using the Entner-Doudoroff (ED)

pathway but not the Embden-Meyerhoff (EM) or pentosephosphate pathway. Because less ATP

is produced by the ED pathway, Zymomonas produces relatively less biomass and more carbon is

converted to fermentation products than bacteria using EM pathway. However, Z. mobilis has a

major disadvantage for biomass conversion. This organism has a limited substrate range, glucose,

fructose and sucrose (Sprenger 1996).









Strains of Z. mobilis capable of fermenting xylose and arabinose were developed to extend

their substrate range. The genes of xylose isomerase and xylulose kinase from E.coli were

expressed on a plasmid using strong constitutive promoter from Z. mobilis (Lawford and

Rousseau 1991). The transformed strain CP4 (pZB5) is able to ferment on xylose with an ethanol

yield of 86%. A newer strain, Z. mobilis AX101, ferments both arabinose and xylose with

necessary recombinant genes integrated in chromosomal DNA (Lawford and Rousseau 2002).

However, AX101 ferments arabinose more slowly than xylose, and the fermentations are often

incomplete. This strain also has low tolerance to acetic acid, especially in the presence of ethanol.

Z. mobilis continues to be developed in many labs and companies as a potential ethanologenic

biocatalyst.

Klebsiella oxytoca

Klebsiella oxytoca is an enteric bacterium isolated from paper and pulp streams and some

other sources of wood. This microorganism is capable of growing at a pH as low as 5.0 and at

350C. K. oxytoca has the potential to be developed as a biocatalyst for cellulose bioconversion,

since it grows on a wide variety of sugars including hexoses and pentoses, as well as cellobiose

and cellotriose (Freer and Detroy 1983; Wood and Ingram 1992). Similar to the construction of E.

coli KO 11, the PET operon was integrated in the pfl gene of K. oxytoca chromosome along with

a cat marker. The best isolate from recombinants, K. oxytoca P2, ferments glucose (100 g/1) or

cellobiose (100 g/1) to ethanol with yields of 44-45 g/l within 48 h (Wood and Ingram 1992).

K. oxytoca P2 was further engineered to express heterogeneous cellulases for reducing

enzyme cost. Two extracellular endoglucanase genes (CelZ and CelY) from Erwinia

c hi y%\,llini were integrated into the chromosomal DNA of K. oxytoca P2 with the required

auxiliary transporter gene expressed on a plasmid (pCPP2006). However, the engineered









cellulolytic K. oxytoca SZ21 strain still utilized cellulose (Sigmacell 50) poorly without

supplemention of additional cellulases (Zhou et al. 2001; Zhou and Ingram 2000).

Yeasts

The traditional organism used for ethanol production, yeast, is also an important

engineering target for developing biocatalysts for lignocellulose conversion. Since most yeasts

do not ferment xylose, the prevailing attempts are to engineer yeast to utilize xylose. With the

introduction of genes encoding xylose reductase (XR) and xylitol dehydrogenase (XDH) from

Pichia stipitis into Saccharomyces, the recombinant strain still produces ethanol at low yield

from xylose with low growth rate (Amore et al. 1991; Slininger et al. 1985). There are more

problems in genetically engineering yeast than bacteria due to the biological gap between

prokaryotes and eukaryotes, such as the difference in internal pH between bacteria and yeasts,

incorrect protein folding, incorrect post-translational modification and cofactor imbalance. These

problems were reported to cause low ethanol yield, and low specific growth rate for yeast as

biocatalysts for fuel ethanol production from lignocellulosic biomass (Dombek and Ingram 1987;

Jeffries and Jin 2004; Ostergaard et al. 2000). However, improvements in ethanol productivity

and yield were achieved recently by several research groups. Matsushika and Sawayama

engineered Saccharomyces strains by optimizing the XR/XDH/ xylulokinase (XK) ratio

(Matsushika and Sawayama, 2008). They also introduced an XDH with altered coenzyme affinity

into and XR and XK expressing Saccharomyces strain to enhance the xylose fermentation

(Matsushika et al., 2008a,b). Xylose transporters in Trichoderma reesei (Xltl) were

heterologously expressed in Saccharomyces strains expressing XR, XDH and XK (Hector et al.,

2008) to increase xylose utilization. Future improvements in ethanol productivity are possible

with better understanding of the xylose utilization mechanism in yeast and new metabolic

engineering methods.









Objectives of this Research

The goal of this study is to develop biocatalysts to utilize the currently unusable MeGAX

in the dilute acid hydrolysate of hemicellulose to increase the efficiency of hemicellulose

bioconversion. A new Enterobacter asburiae strain JDR-1(E. asburiae JDR-1) capable of

MeGAX utilization was isolated by our lab for the purpose of MeGAX utilization research and

metabolic engineering.

Most E. asburiae species isolated are short motile rods with peritrichous flagella

(Hoffmann et al. 2005). Some of the early isolated Enterobacter asburiae species were

characterized as human pathogens (Brenner et al. 1986). However, some other strains of E.

asburiae were isolated from soil environments with implicated activity to mobilize phosphate

from calcium phosphate for plant nutrition (Sharma et al. 2005). Enterobacter is also one of the

most commonly isolated bacterial genuses as plant endophytes. E. asburiae was found as an

endophyte in cotton, cucumber, common bean, rice and sweetpotato (McInroy and Kloepper

1995; Elbeltagy et al. 2001; Asis and Adachi 2004). Enterobacter asburiae colonizes different

plant species and establishes endophytic populations in various tissues. For example a systemic

colonization of E. asburiae JM22 in cotton plants was reported to specifically locate in the root

surfaces, within epidermal cells, and inside intercellular spaces of the root cortex close to the

conducting elements (Quadt-Hallmann and Kloepper 1996). Besides gaining entrance to plants

through natural openings or wounds, endophytic bacteria appear to actively penetrate plant

tissues using hydrolytic enzymes like cellulases and pectinases. However, once inside the plants,

endophytes seem to utilize exudates that enter the intercellular spaces as substrates rather than

break down the more complex cell wall materials as the major food source (Hallmann et al.

1997).









Scope of this work is : 1) Evaluate the potential of E. asburiae JDR-1 for efficient

conversion dilute acid hydrolysates and characterize its metabolic properties to understand the

MeGAX utilization process. 2) Genetically engineer E. asburiae JDR-1 to homogeneously

produce ethanol or other fermentation products. 3) Define the MeGAX utilization process

through identification of genes involved, which is also for engineering current Enterobacteraciae

biocatalysts to utilize MeGAX in hemicellulose hydrolysate.














Proteln-Polysaccharide linkage?


Figure 1-1. Model for corn stover cell walls (Saha 2003) With permission from Springer

Science + Business Media. Copyright @ Springer Science + Business Media









CHAPTER 2
COMPLETE FERMENTATION OF XYLOSE AND METHYLGLUCURONOXYLOSE
DERIVED FROM METHYLGLUCURONOXYLAN BY ENTEROBACTER ASBURIAE
STRAIN JDR-1

Introduction

The quest for alternatives to petroleum resources for production of fuels and chemicals has

been motivated in part by economic incentives associated with limited and diminishing supply,

and from the recognition of the contribution that increasing levels of carbon dioxide derived

from these resources have contributed to global warming (Kheshgi et al. 2000). Microbial

fermentation products derived from resources renewable through photosynthesis are particularly

attractive alternatives that may be generated with existing technologies (McMillan 1997). The

development of yeast and bacterial biocatalysts has been applied to the commercial production of

ethanol as an alternative fuel from starch and sucrose derived from commodity crops, e.g., corn

and sugarcane (Dien et al. 2003; Jarboe et al. 2007). To expand production of ethanol and

chemical feedstocks from renewable resources that do not economically impact these

commodities, lignocellulosic resources, including forest and agricultural residues, have become

targets for bioconversion of cellulose and hemicellulose to fermentable sugars (Jarboe et al. 2007;

Shanmugam and Ingram 2008; vonSivers and Zacchi 1996).

Current industrial methods for pretreatment of lignocellulose for bioconversion to ethanol

solubilize the hemicellulose fraction by dilute acid hydrolysis, releasing the pentoses for

fermentation (Preston et al. 2003). The predominant structural polymer in the hemicellulose

fraction of hardwoods and crop residues is methylglucuronoxylan (MeGAXn), a 0-1,4 linked

xylan in which xylose residues are periodically substituted with a-1,2-linked

4-0-methyl-glucuronic acid (MeGA) (Preston et al. 2003). Resistance of the a-1,2 glucuronosyl

linkages to dilute acid hydrolysis results in the release of the aldobiuronate,









methylglucuronoxylose (MeGAX), which is not fermented by bacterial biocatalysts currently

used to convert hemicellulose-derived xylose to ethanol, e.g., E. coli KO11 (Preston et al. 2003;

Rodriguez 2001; Ingram et al. 1999). The frequency of MeGA substitutions on the xylose

residues of methylglucuronoxylan ranges from less than one in ten in crop residues to one in six

to seven in hardwoods, sweetgum and yellow poplar, and as much as 18% to 27% of the

carbohydrate may reside in this unfermentable fraction (primarily as MeGAX) following dilute

acid pretreatment (Rodriguez 2001). A scheme for the release of xylose and MeGAX by dilute

acid hydrolysis of sweetgum xylan is depicted in Fig. 2-1 below.

In this chapter, I describe the isolation and characterization of a bacterium, E. asburiae

JDR-1, with the ability to ferment MeGAX as the sole carbon source. E. asburiae JDR-1 was

defined with respect to its metabolic potential for the conversion of MeGAX, xylose and glucose

to specific fermentation products. As a Gram-negative member of the Enterobacteriaceae, the

genetic basis for this metabolic potential may be applied to the further development of bacterial

biocatalysts for conversion of hemicellulose-derived pentoses to biobased products.

Materials and Methods

Preparation of Substrates and Culture Media

Sweetgum methylglucuronoxylan (MeGAXn) was prepared from sweetgum stem wood

(Liquidambar styraciflua) as previously described and characterized by 13C-NMR (Hurlbert and

Preston 2001; Kardosova et al. 1998). Dilute acid hydrolysates of methylglucuronoxylan were

prepared by hydrolysis with 0.1 N H2S04 (4 g methylglucuronoxylan in 400 ml 0.1 N H2S04) at

121 'C for 60 min, followed by neutralization with BaCO3. Anion exchange resin (Bio-Rad

AG2-X8) in the acetate form was used to adsorb the charged aldouronates; the uncharged xylose

and xylooligosaccharides, mainly small amounts of xylobiose, were eluted with water. The

aldouronates were then eluted with 20 % (v/v) acetic acid. After concentration under vacuum at









50 'C, aldouronates were separated on a 2.5 cm x 160 cm BioGel P-2 column (BioRad, Hercules,

CA) with 50 mM formic acid as the eluent. The formic acid was removed from the purified sugar

sample fractions by lyophilization. MeGAX and MeGAX2 were identified by thin layer

chromatography (TLC) analysis using MeGAX and MeGAX2 standards structurally defined by

13C and H-NMR spectrometry (Zuobi-Hasona 2001). Xylobiose and xylotriose were obtained

and purified from MeGAXn digested with Paenibacillus sp. strain JDR-2 XynAi catalytic

domain (CD), a recombinant GH10 endoxylanase XynAi CD overexpressed in E. coli. The

substrate containing 30 mg/ml MeGAXn was prepared with 10 mM sodium phosphate buffer, pH

6.5. Digestions were initiated by the addition of 3.5 U of XynAi CD into 50 ml substrate and

incubated with rocking at 300C for 24 h. An additional 1 U was added after 24 h and incubation

was continued for 40 h. Aldouronates, xylobiose, and xylotriose were separated with the P2

column and identified by TLC (St John et al. 2006). Total carbohydrate concentrations related to

substrate preparations were determined by the phenol-sulfuric acid assay (Dubois et al. 1956),

with xylose as the reference. The conditions of acid hydrolysis generated mostly MeGAX and a

small amount of MeGAX2 from MeGAXn, with no aldouronates larger than MeGAX2 detected.

MeGAX3 was prepared from GH10 endoxylanase-catalyzed depolymerization of sweetgum

MeGAXn and then purified by gel filtration on BioGel P4 (St John et al. 2006).

Minimal medium containing the substrates described above was prepared upon mixing

sterile substrate solutions (2x concentration) with the same volume of a 2x solution of Zucker

and Hankin mineral salts (ZH salts) at pH 7.4 (Zucker and Hankin 1970). Neutralized MeGAXn

acid hydrolysate (0.5% w/v) was also added to ZH salts directly as a growth substrate. Where

indicated, some media preparations were supplemented with 0.1% yeast extract (YE medium).









Isolation and Identification of E. asburiae JDR-1

E. asburiae JDR-1 was isolated from discs of sweetgum stem wood (Liquidambar

styraciflua) buried, soon after cutting, about one inch below the soil surface in a sweetgum stand

for approximately three weeks. Discs were suspended in 50 ml sterile deionized water and

sonicated in a 125 Watt Branson Ultrasonic Cleaner water bath for 10 min. The sonicate was

inoculated into 0.2% (w/v) MeGAX YE medium and incubated at 370C. Cultures were streaked

on MeGAX minimal medium agar plates. Isolated colonies were passed several times between

MeGAX broths and agars until pure. Exponential phase cultures growing on 0.2 % MeGAX

minimal media were cryostored in 25% sterile glycerol at -700C.

The purified isolate was submitted to MIDI Labs (http://www.midilabs.com) for partial 16s

rRNA sequencing and FAME analysis. BBL EnterotubeTM II (Becton, Dickinson and Company,

USA) inoculation was also used to identify the isolate based upon metabolic capability using the

standard protocol. The Differential Interference Contrast (DIC) microscopic photos of E.

asburiae JDR-1 growing in MeGAX minimal medium at exponential phase were obtained with a

Zeiss DIC microscope at 40x15-fold magnification. The negative stain electron microscopic

photo was taken with a Zeiss EM10OA electron microscope.

Substrate Utilization and Fermentation Product Analysis

Growth and substrate utilization analysis was performed in cultures actively aerated by

shaking. For preparing inocula, cultures of E. coli B (ATCC 11303) and E. asburiae JDR-1 from

cryostored samples were directly streaked on Luria-Bertani (LB) agar plates. After overnight

incubation at 370C, isolated colonies were picked to inoculate liquid media specified for a

particular experiment. Growth studies were performed at 370C in 16 mm x 100 mm test tubes

containing 6 ml medium. Optical densities of cultures were measured at 600 nm (OD600) with a

Beckman DU500 series spectrophotometer. The correlation of cell density and OD600 is CDW/L









(g cell dry weight/L)=0.49*OD600oo+0.02 was experimentally determined. Sample dilutions were

made to obtain OD600 readings between 0.2 and 0.8 absorbance units which, corrected for

dilution factors, provided turbidity values for growth studies. Individual 6 ml cultures for study

were inoculated with 12 .il (0.2% volume) of overnight cultures and maintained at 370C with

constant shaking (Eberbach shaker set at "low").

Batch fermentations under anaerobic conditions at 370C were conducted in 13 mm x 100

mm screw cap tubes containing 3.0 ml medium. Inocula (0.5% [v/v]) were from overnight

aerobic cultures grown in the same medium. After inoculation, nitrogen gas was used to flush

and saturate the sealed batch culture. The tubes were set in a Glas-Col minirotator at 60 rpm.

For analysis of substrates and fermentation products, the culture supernatants were passed

through 0.22 um filters and subjected to HPLC analysis. Products were resolved on a Bio-Rad

HPX-87H column with 0.01 N H2S04 as the eluent at 650C. Samples were delivered with a 710B

WISP automated injector and chromatography controlled with a Waters 610 solvent delivery

system at flow rate of 0.5 ml/min. Products were detected by differential refractometry with a

Waters 2410 RI detector. Data analysis was performed with Waters Millennium Software. To

determine and quantify methanol, unfiltered supernatants from fermentation cultures were also

analyzed by gas chromatography (6890N Network GC system, Agilent Technologies), using

isopropanol as an internal standard. This detection method was used since diffusion during

HPLC precluded quantitative detection of methanol by differential refractometry.

Determination of Metabolic Pathways by 13C-NMR

The central metabolic pathways utilized by E. asburiae JDR-1 during glucose and xylose

fermentation were evaluated with 13C-NMR (Scott and Baxter 1981). Cultures were grown in LB

medium to mid-exponential phase at 370C. Cultures (0.5 ml) were centrifuged and the cells

washed with 2x ZH salts solution. The cell pellets were suspended in 1.0 ml 0.5% [2-13C]xylose









(99% enrichment; Omicron Biochemicals Inc, IN) in ZH minimal medium. Similar

fermentations were also prepared with 1.0 ml 0.5% [1-13C]glucose, or 1.0 ml 0.5%

[6-13C]glucose ZH minimal medium using D-[l-13C]glucose or D-[6-13C]glucose (99%

enrichment; Cambridge Isotope Laboratories, Andover, MA). Fermentations were carried out

under anaerobic conditions at 370C for 8 hours. Cells were removed by centrifugation, and the

supernatants analyzed by HPLC (after filtration) and 13C-NMR spectrometry. NMR spectra were

obtained using a VXR300 NMR spectrometer (NMR facility of the Department of Chemistry,

University of Florida) operating in the Fourier transform mode as follows: 75.46 MHz; excitation

pulse width, 7.0 s; pulse repetition delay, 0 s; spectral width, 16502; 256 acquisitions. Acetone

(30 .il) containing 13C at natural abundance in 700 [il sample was used as an internal reference of

31.07 ppm for the 13C methyl carbon (Kardosova et al. 1998). Individual carbon atoms for

fermentation products were identified by shift assignments and quantified by comparison with

standards (13C at natural abundance) of known concentrations.

Determination of Molar Cell Dry Weight Yield

For molar growth yield experiments (Bauchop and Elsden 1960; Smalley et al. 1968;

Gunsalus 1961), anaerobic growth was performed in 50 ml minimal medium containing either

0.26% glucose, 0.36% xylose, 0.35% glucuronate or 0.2% MeGAX as sole carbon source with

the fermentation conditions described above. After 24 hours of growth and complete utilization

of the carbon source, cells were harvested by centrifugation and the resulting pellets were

washed twice with deionized water. The pellets were dried to constant weight in a Sargent

vacuum dryer at 600C for up to 36 hours. The culture supernants were analyzed by HPLC to

determine substrate consumption. The molar cell dry weight yield was calculated as cell dry

weight (gram) divided by consumed substrate (mole).









Results

Identification and Characterization of E. asburiae JDR-1

A novel bacterial strain able to grow on MeGAX minimal medium was obtained by John D

Rice of our lab (Bi et al 2009) and subsequently identified with three tests. The partial 16S rRNA

sequence (accession number EU117142, Gene Bank, NCBI), amplified using primers

corresponding to E. coli 16S rRNA positions 005 and 531 (526bp), provided an alignment with

99.5% identity within the sequence of Enterobacter asburiae (MIDI Aerobic Bacteria Database

version 4.0, January 1999). Results of FAME (fatty acid methyl ester) analysis indicated this

strain had the greatest similarity index with Enterobacter asburiae species (0.766) compared

with any other entry in the MIDI database. A biocode of 32061, obtained from the Enterotube II

(BBL) test, also corresponded to Enterobacter asburiae species. Based upon these three criteria,

the isolate was identified within Enterobacter asburiae species and designated as Enterobacter

asburiae strain JDR-1. The strain has been deposited with the Agriculture Research Service

Patent Culture Collection of the USDA, Peoria, IL., under NRRL number NRRL B-S0074.

When exponential phase cultures were observed by optical DIC microscopy, E. asburiae

JDR-1 appeared as short motile rods (Fig 2-2 A). Negative stain electron microscopy revealed 3

[tm x 1 [tm cells with peritrichous flagella (Fig 2-2 B). The microscopy study was done the

Laboratory for Electron Microscopy and Imaging, Department of Microbiology and Cell Science

and kindly assisted by Ms. Donna Williams. These morphological characteristics were similar to

those of other isolates of Enterobacter asburiae (Hoffmann et al. 2005). When grown on LB

agar plates, colonies of E. asburiae JDR-1 were morphologically indistinguishable from E. coli

colonies.









Utilization of Acid Hydrolysates of Methylgluronoxylan by E. asburiae JDR-1

The unique ability of E. asburiae JDR-1 to grow on the aldobiuronate MeGAX as the sole

carbon source suggested a potential for the complete metabolism of the carbohydrates generated

by the dilute acid pretreatment currently applied for the release and fermentation of xylose in

hemicellulose fractions. To evaluate this potential, E. asburiae JDR-1 was grown aerobically in

minimal medium comprised of neutralized MeGAXn acid hydrolysate and Zucker and Hankin

mineral salts. Based upon HPLC analysis of media samples taken during different stages of

growth, E. asburiae JDR-1 utilized MeGAX completely in minimal media with MeGAXn

hydrolysate after it depleted xylose (Fig. 2-3 A). Biphasic growth occurred as E. asburiae JDR-1

switched from utilizing xylose to MeGAX. In contrast to E. asburiae JDR-1, E. coli B consumed

only the free xylose with the MeGAX concentration in the medium remaining constant.

Concentrations of xylose and MeGAX in hydrolysate medium of sweetgum MeGAXn, as

determined by HPLC, were 0.206% w/v and 0.036% w/v, respectively. Therefore, E. asburiae

JDR-1 utilized 17.5% more substrate (mass amount) than E. coli B, which was unable to utilize

MeGAX (Fig. 2-3 B). Under aerobic conditions, both E. asburiae JDR-1 and E. coli B formed

acetic acid during exponential growth phase that was metabolized upon complete utilization of

the carbon sources in the MeGAXn hydrolysates. E. asburiae JDR-1 was also able to grow in

xylobiose and xylotriose minimal medium, which E. coli B could not utilize. However, E.

asburiae JDR-1 was unable to utilize MeGAX2 and MeGAX3 (data not shown).

Substrate Preference of E. asburiae JDR-1

E. asburiae JDR-1 was found to grow aerobically in minimal medium containing different

sole carbon sources, such as glucose, xylose, mannitol, maltose, rhamnose, mannose, glucuronate

and glycerol. As noted above, it was able to quantitatively metabolize MeGAX, but was unable

to utilize MeGAX2 generated by acid hydrolysis, or MeGAX3 generated by a GH10









endoxylanase. When growing in a minimal medium containing an eqimolar mixture of glucose

and xylose, E. asburiae JDR-1 displayed a diauxic growth pattern typical of species of

Enterobacteraceae (Fig. 2-4 A). Glucose (8 mM) was consumed within approximately 8 hours,

while xylose utilization began when glucose was almost entirely consumed and was depleted in

14 hours.

To study the process by which MeGAX was utilized, E. asburiae JDR-1 was grown in

minimal medium containing both xylose and glucuronate, products that might be generated from

MeGAX. A single phase growth curve was observed in which both substrates were consumed by

15 hours (Fig. 2-4 B). This is similar to its single phase growth curve on MeGAX, in which the

6.5 mM substrate was depleted in about 11 hours (Fig. 2-4 C). The similarity in growth pattern

with MeGAX and the combination of xylose and glucuronate as carbon sources supports the

possibility that free glucuronate and free xylose may be released during the metabolism of

MeGAX.

Fermentation Characteristics

Fermentation experiments were performed to evaluate the potential of E. asburiae JDR-1

as a biocatalyst for the production of biobased products, and define the processes involved in the

metabolism of MeGAX. Using limiting amounts (0.25% w/v) of substrates and cultivation under

anaerobic standing conditions, E. asburiae JDR-1 was able to ferment all major sugars

constituting hemicellulose, including D-glucose, D-xylose, D-mannose, L-arabinose and

D-galactose. The major products from xylose and galactose fermentation were acetic acid and

ethanol present in similar molar quantities. Acetic acid, ethanol and small amounts of lactic acid

were produced from glucose, mannose and arabinose (Table 2-1). Small amounts of formic acid

and very small amounts of fumaric and succinic acids were detected in most fermentations. The









HPLC profiles indicate that E. asburiae JDR-1 performs mixed acid fermentation as does E. coli,

but with preferential formation of acetate and ethanol over lactate.

With sweetgum MeGAXn hydrolysate as substrate, E. asburiae JDR-1 consumed 99% of

the substrate when the pH was maintained above 5, giving the major products acetic acid and

ethanol (Table 2-2). With glucuronic acid as carbon source, acetic acid was the major

fermentation product. To study the process of MeGAX metabolism, the presumed degradation

products of MeGAX, xylose and glucuronate, both at 11 mM, were used as substrates. The

predominant products were 20.4 mM acetate and 5.25 mM ethanol. Major fermentation products

from 4.0 mM MeGAX were 8.1 mM acetic acid, 1.2 mM ethanol, and 4.3 mM methanol.

Central Metabolic Pathways Determined by 13C-NMR

The total quantities of ethanol, acetate and lactate were determined by HPLC and the

quantities of 13C labeled products were quantified from integration of differentially labeled

compounds detected in the 13C-NMR spectra. This allowed determination of the fraction of each

fermentation product that was differentially labeled with 13C, which helped to illustrate the

central metabolic pathways E. asburiae JDR-1 uses. The quantities of each product and the

fractions labeled with 13C are presented in Table 2-3.

To determine the primary pathway of xylose metabolism by E. asburiae JDR-1,

comparisons were made for the fermentation of [2-13C]xylose with cultures of E. coli B. For E.

coli B, employing only the pentose-phosphate pathway to metabolize xylose, the prominent shift

signals in the 13C-NMR spectrum of the fermentation products were assigned to [1-13C]ethanol at

57.6 ppm, [2-13C]lactate at 68.8 ppm, and [1-13C]acetate at 181.0 ppm. Shift signals at 71.0 ppm

and 74.5 ppm were assigned to the a- and 0- anomers of unused [2-13C]xylose, and the signal at

30.6 ppm to the methyl carbons of the acetone standard (Fig. 2-5 B). Fractions of labeled versus

total acetate, ethanol, and lactate with E. coli B were 0.26, 0.27, and 0.31, respectively, which









was slightly less than the theoretical fraction 0.4 expected for metabolism through the

pentose-phosphate pathway (Table 2-3). The lower quantities of labeled products as fractions of

the total found for E. coli may reflect accuracy limitations for integration against the 13C-acetone

standard, as these products all showed similar fractions (0.26-0.31) were labeled.

When E. asburiae JDR-1 fermented [2-13C]xylose, a 13C-NMR spectrum for fermentation

products was obtained with prominent signals for [1-13C]ethanol, [2-13C]lactate, and

[1-13C]acetate at expected shift positions (Fig. 2-5 A). The fractions of labeled ethanol, labeled

acetate and labeled lactate to their total amounts were 0.43, 0.4 and 0.45, respectively (Table 2-3),

and nearly identical to the theoretical fraction of 0.4. Moreover, the fractions of labeled acetate

and ethanol were not higher than the fraction of labeled lactate. These results establish that the

pentose phosphate pathway is the main metabolic pathway for xylose utilization in E. asburiae

JDR-1.

To determine the primary pathway E. asburiae JDR-1 utilizes to metabolize glucose,

[1-13C]glucose and [6-13C]glucose were used as fermentation substrates. Similar 13C-NMR

spectra of fermentation products were obtained from [6-13C]glucose and [1-13C]glucose (Fig. 2-5

C, 2-5 D). Shift signals at 92.4 and 96.2 ppm were assigned to the a- and 0- anomers of unused

[1-13C]glucose (Fig 2-5 C); signals at 60.9 and 60.1 ppm were assigned to the a- and 0- anomers

of unused [6-13C]glucose (Fig. 2-5 D). The signal at 30.6 ppm was assigned to the methyl

carbons of the acetone standard. Excepting the shift signals for reference and unused substrates,

the prominent signals in both spectra were for [2-13C]ethanol at 17.1 ppm, [2-13C]acetate at 22.2

ppm and [3-13C]lactate at 20.3 ppm with similar distributions for both substrates. The absence of

[1-13C]lactate indicates that no [1-13C]glucose was metabolized through the Entner-Douderoff

(ED) pathway. Moreover, the fractions of all labeled products of their total amounts were similar









for fermentation of [6-13C]glucose and [1-13C]glucose; and these fractions for [6-13C]glucose

were not higher than those found for [1-13C]glucose (Table 2-3), indicating little or no

[1-13C]glucose went through the pentose-phosphate pathway. Collectively, these results establish

that the Embden-Meyerhof (EM) pathway is the main metabolic pathway for glucose utilization

in E. asburiae JDR-1.

Growth and Projected ATP Yields with Different Substrates

To understand the bioenergetics in the process of MeGAX fermentation by E. asburiae

JDR-1, molar cell dry weight yields were determined after 24 hours of growth with glucose,

xylose, glucuronate and MeGAX as sole carbon sources in Zucker-Hankin minimal medium. The

experiment was performed three times and the average approximate YM values were about 10 g

per mole of substrate for growth on xylose and glucuronate, 20 g for growth on glucose, and 30 g

for growth on MeGAX (Table 2-4). The experimental YATP in anaerobic growth has been

reported in the range of 8 to 12 gram cell dry weight per mole of ATP for bacteria (Russell and

Cook 1995). An estimated YATP value at the lower end of this range, 8, was used here since this

is for anaerobic growth in batch cultures in minimal medium with a relatively low concentration

of carbon source (BAUCHOP and Elsden 1960; Gunsalus 1961). The apparent ATP yields per

mole of substrate were calculated based on the estimated YATP of 8 as 1.3 mole of ATP produced

from either xylose or glucuronate, 2.6 from glucose and 4.0 from MeGAX (Table 2-4). These

apparent ATP yields allow an estimate of the relative ATP yields obtained for the different

substrates without considerations of maintenance energy or overflow metabolism (Russell and

Cook 1995), providing insight into the metabolism of MeGAX. The ratios of the molar growth

yields obtained with xylose, glucuronate, and MeGAX as carbon sources are 1:1:3, indicating

that the requirement for MeGAX transport is less than that for separate transport of xylose and

glucuronate.









Discussion

Fermentation of MeGAX by E. asburiae JDR-1

E. asburiae JDR-1 is the first bacterium that has been described to ferment MeGAX. The

complete utilization of MeGAX by E. asburiae JDR-1 clearly indicates a metabolic potential not

shared with E. coli B, as well as the ethanologenic strains E. coli KO11 (Rodriguez 2001) and

Klebsiella oxytoca P-2 (Qian et al. 2003).

Based upon the studies presented here, E. asburiae JDR-1 uses the EM glycolytic pathway

to metabolize glucose and the pentose phosphate pathway to metabolize xylose. This is the first

definitive study to determine the role of these pathways in any Enterobacter spp., although the

findings are not unexpected as the role of these pathways have been well defined in other

Enterobacteriaceae (Moat 2003). The roles of these pathways in central metabolism are

supported by genome annotations in Enterobacter sp. 638 (Complete sequence of chromosome

of Enterobacter sp. 638, NC_009436, NCBI genome database).

The formation of equal molar amounts of ethanol and acetic acid from xylose under

anaerobic conditions suggests that pyruvate serving as a precursor for acetic acid and ethanol

might be converted to acetyl CoA by pyruvate format lyase (Moat 2003). The ability of E.

asburiae JDR-1 to completely convert MeGAX to acetic acid, ethanol and methanol suggests

that demethylation may occur as an early event in the metabolic processing. The metabolic

potential of E. asburiae JDR-1 to ferment mannose, galactose, and glucose, as well as xylose and

arabinose, defines its potential to convert all of the pentoses and hexoses that comprise

hemicellulosic biomass to fermentation products.

Central Metabolic Pathways Used by E. asburiae JDR-1

In E. coli, the xylulose-5-phosphate enters the pentose-phosphate pathway, yielding

fructose-6-phosphate and glyceraldehyde-3-phosphate (Moat 2003). Fructose-6-phosphate can be









further metabolized to glyceraldehyde-3 -phosphate. [2-13C]xylose metabolism by the

phosphoketolase pathway is expected to yield [1-13C]acetate or [1-13C]ethanol and unlabeled

pyruvate; thus no label will be found in lactate (Gunsalus et al. 1955). However, if organisms use

the pentose-phosphate pathway, the [1,2-13C] pyruvate produced is expected to account for

one-fifth of the total pyruvate, with [2-13C] pyruvate accounting for another one-fifth of the total

pyruvate. The remaining three fifths of the pyruvate do not carry the 13C label (Patel et al. 2006).

[1,2- 13C] pyruvate can be reduced to [1,2-13C]lactate or converted to [1-13C] ethanol or

[1-13C]acetate by pyruvate format lyase and subsequent enzymes; [2-13C] pyruvate can be

reduced to [2-13C]lactate or converted to [1-13C]ethanol and [1-13C]acetate. Therefore, if an

organism uses only the pentose-phosphate pathway to metabolize [2-13C]xylose, both

[l-13C]ethanol and [l-13C]acetate would account for two-fifths of total ethanol and acetate

produced, whereas [1,2-13C]lactate and [2-13C]lactate together would account for two-fifths of

total lactate. The fraction of labeled carbon two in lactate would be two-fifths of the total carbon

two in lactate, the same fraction as labeled acetate and ethanol to total acetate and ethanol. If

both the phosphoketolase and pentose phosphate pathways were used, label would be found in

the lactate, but the ratio of labeled ethanol or acetate to total ethanol or acetate would be higher

than ratio of labeled lactate to total lactate. The labeling patterns obtained in this study indicate

that the pentose-phosphate pathway is the main if not exclusive metabolic pathway for xylose

utilization in E. asburiae JDR-1.

Three major pathways, Embden-Meyerhof (EM), pentose-phosphate (PP), and

Entner-Doudoroff (ED) pathways, are used by bacteria to catabolize glucose or other sugars into

pyruvate (Moat 2003). To determine the primary pathway E. asburiae JDR-1 utilizes to

metabolize glucose, [1-13C]glucose and were used as fermentation substrates. For either the EM









or ED pathway, labeled carbon derived from [6-13C]glucose would be found in carbon three of

the intermediate pyruvate. Following subsequent metabolic processing, [2-13C]ethanol,

[2-13C]acetate and [3-13C]lactate would be produced. No labeled carbons would be found at the

other carbon positions of these compounds. However, different pathways give different labeled

fermentation products for [1-13C]glucose. If [1-13C]glucose were utilized by the EM pathway, all

labeled carbon would also be found at carbon three position of pyruvate, and the amounts of

labeled products and their ratios to total products would be the same as those derived from that

of [6-13C]glucose. If the pentose-phosphate pathway were utilized, carbon one of [1-13C]glucose

would be oxidized to the carboxyl group in 6-phosphogluconate (6PG) which would then be

oxidized to CO2 by glucose-6-phosphate dehydrogenase (Moat 2003). After the two oxidation

steps, no other labeled fermentation products would be obtained except for labeled CO2. If the

ED pathway is used, carbon one of [1-13C]glucose would become carbon one in one of the two

pyruvate molecules generated from 2-keto-3-deoxy-6-phosphogluconate. In the subsequent

fermentation process, [1-13 C]pyruvate would be cleaved by pyruvate format lyase, yielding

acetate or ethanol without the 13C label, and [l-13C]pyruvate would be reduced to [1-13C]lactate.

The HPLC and NMR data (Fig. 2-5 C and D, and Table 2-3) showed that both [6-13C]glucose

and [1-13C]glucose gave similar products fermented by E. asburiae JDR-1 with similar

distributions of labeled carbons. These results establish that the EM pathway is the main

metabolic pathway for glucose utilization in E. asburiae JDR-1.

Possible Pathway of MeGAX Metabolism

The processing of aldouronates formed by the action of GH10O endoxylanases, e.g.

MeGAX3, has been shown to involve the removal of the a-1,2-linked 4-O-methylglucuronosyl

moiety from the reducing terminus of P-1,4-linked xylotriose by the action of a GH67









a-glucuronidase (Shulami et al. 1999). a-Glucuronidases (GH67) from some bacteria have been

shown to act on MeGAX as well as MeGAX3 (Nong at al. 2005; Shulami at al. 1999). Consensus

primers based on GH67 family genes (http://www.cazy.org/fam/GH67.html) may be used to

amplify aguA genes from bacterial genomic DNA prior to partial sequencing and cloning. The

application of this strategy was not successful in identifying a gene encoding a GH67

a-glucuronidase in E. asburiae JDR-1 (G. Nong and C. Bi, unpublished). Either this strain

synthesizes an a-glucuronidase completely different from the GH67 that has been well defined,

or a novel metabolic process may be used in which the cleavage of the glycosidic bond follows

other reactions.

The fermentation products derived from xylose and glucuronate were similar to those

obtained from MeGAX, although the ethanol yield from MeGAX was lower. This result

suggested that xylose and glucuronate might be released in the process of MeGAX utilization.

This possibility was also supported by the similarity in the growth patterns of E. asburiae JDR-1

growing in MeGAX minimal medium and the glucuronate plus xylose minimal medium (Fig.

2-4).

Due to the small amount of methanol observed relative to system noise of the GC system,

accurate quantitative data for methanol as a fermentation product was not obtained in media

containing acid hydrolysates of MeGAXn. Methanol was quantitatively detected in stoichiometric

equivalence with MeGAX during the metabolism of MeGAX, indicting methanol is derived

from the methyl group released from methylglucuronate as an early event in the metabolism of

MeGAX (Table 2-2). Based on the results obtained from this study, the process of MeGAX

catabolism by E. asburiae JDR-1 may involve the release of methanol, glycosidic bond cleavage

to release glucuronate and xylose, and catabolism of these carbohydrates to generate ethanol and









acetic acid as predominant fermentation products. An elimination reaction for the release of

methanol prior to glycosidic bond cleavage is a possibility, in which case the product of

glycosidic bond cleavage would be delta-4,5-hexuronate.

Bioenergetics of E. asburiae JDR-1

Glucose was shown to be metabolized in E. asburiae JDR-1 by the EM pathway. If the

PTS system is used to import glucose as in E. coli (Simoni et al. 1976), the net formation of 3

moles of ATP result from 1 mole of glucose when major fermentation products are ethanol and

acetate. For xylose utilization via the pentose-phosphate pathway, presumably requiring one

mole ATP to import xylose with an ABC transporter as in E. coli (Linton and Higgins 1998), 1.5

moles of ATP would be produced per mole of xylose when fermentation products were ethanol

and acetate (Hasona et al. 2004). Glucuronate has been shown to be transported into cells with

the UxuT transporter in E. coli. This transporter bears significant homology to ExuT in Erwinia

ci h/l)/%/wiheii, which has been defined as responsible for the transport of glucuronate as well as

galacturonate through an active process in which energy is consumed. Considering energy

consumed in active transport, the maximal ATP yield in the process of glucuronate fermentation

may be 2 moles or less per mole of glucuronate when the major fermentation product is acetic

acid (Hugouvieuxcottepattat and Robertbaudouy 1987; Robertbaudouy et al. 1981; SanFrancisco

and Keenan 1993). The theoretical ATP yields from glucose, xylose and glucuronate were close

to the calculated ATP yields based on the estimated YATP, 8, suggesting that the presumed

transportation systems and metabolism of these substrates are correctly interpreted. The ratios of

the molar growth yields obtained with xylose, glucuronate, and MeGAX as carbon sources are

1.0:1.0:3.2 (Table 2-4), clearly indicating that the bioenergetic requirement for MeGAX

transport is less than that for separate transport of xylose and glucuronate.









Role of E. asburiae JDR-1 in Soil Ecology and Bioprocessing

Most Enterobacter asburiae species have been identified and characterized as human

pathogens (Brenner et al. 1986). However, strains of E. asburiae have been isolated from soil

environments and implicated in the mobilization of phosphate for plant nutrition from calcium

phosphate (Sharma et al. 2005). The ability of E. asburiae JDR-1 to utilize MeGAX, as well as

xylobiose and xylotriose, may reflect its evolution in a soil environment with hemicellulose

providing a carbon resource. With the exception of Erwinia species, the Enterobacteriaceae are

not known for their ability to secrete endoxylanases and endoglucanases. For Erwinia spp., the

secreted endoxylanases are members of glycohydrolase family 5 that generate oligosaccharides

all of which are substituted with MeGA residues, and are not further metabolized by the cell

(Hurlbert and Preston 2001; Preston et al. 1992). The role of these enzymes in Erwinia species

may be related to their colonization and maceration of plant tissues, and thus may serve a

virulence function related to pathogenesis. Enterobacter spp. have been found associated with

plant roots and may provide commensal, if not symbiotic, relationships that contribute to positive

development of the plant (Sharma et al. 2005). However, none have been identified which

secrete endoxylanases. E. asburiae JDR-1 is unable to utilize the aldotetrauronate product of

GH10 endoxylanase, MeGAX3, which serves as a substrate for several gram-positive bacteria

that secrete GH10 endoxylanases. At present, the generation of the aldobiuronate MeGAX is

only known to result from the acid hydrolysis of methylglucuronoxylans. Since MeGAX is the

only aldouronate this bacterium will use, this metabolic potential has yet to be linked to the

biological generation of aldouronates from hemicellulosic biomass.

Biotechnological Applications

The predominant structural polymer in the hemicellulose fraction of hardwoods and crop

residues is methylglucuronoxylan (Preston et al. 2003), and depending on the source, as much as









27% of the carbohydrate may occur in the form of MeGAX following dilute acid pretreatment

(Rodriguez 2001). Capturing the xylose, including that found in MeGAX, in acid hydrolysates of

hemicellulose will increase the efficiency of converting lignocellulosic biomass to targeted

products. Our current research on the metabolism of E. asburiae JDR-1 has established its

potential for the engineering of gram-negative bacteria for bioconversion of aldouronates as well

xylose and other pentoses and hexoses that may be derived from the hemicellulose fraction of

any lignocellulosic source. The definition of the genetic basis for MeGAX metabolism may

allow its application to the engineering of established bacterial biocatalysts, e.g. E. coli KO 11,

for the efficient conversion of cellulosic biomass to biofuels and chemical feedstocks.












Table 2-1. Fermentation products formed by E. asburiae JDR-1 from monosaccharides derived
from hemicellulosea.

Fermentation products (mM)
Substrate (0.25% w/v) Fr
Acetic Acid Ethanol Lactic Acid .a
acid
D-Xylose 10.10.1 10.20.7 0 1.60.3

D-Glucose 7.20.3 9.70.5 1.80.1 1.60.4

D-Mannose 7.50.1 9.20.2 0.90.2 3.20.8

D-Galactose 9.00.4 9.00.3 0 0.70.3

L-Arabinose 8.10.2 9.50.1 0.80.2 1.30.3

a: Anaerobic cultures were allowed to consume each carbon source, initially at 0.25% w/v.
Concentrations of components resolved by UPLC were determined for duplicate cultures by
differential refractometry.


Table 2-2. Fermentation products of E. asburiae JDR-1 derived from MeGAXn. Acetate, ethanol
and format concentrations were determined in duplicate cultures with UPLC and
methanol with GC.

Products, mM, and yield (product/substrate)
Substrate (mM)
Acetate Ethanol Methanol Formate

Xylose (14.3) 9.50.2 (0.7) 9.4 0.7(0.7) ND 1.60.3

Glucuronate (11) 12.80.6 (1.1) 0 ND 2.80.4

MeGAX (4.0) 8.10.9 (2.0) 1.2 0.3(0.3) 4.31.0 (1.1) 0
MeGAXn acid hydrolysate <2.5' 0
(Xylose 13.7; MeGAX 1.75)a
Xylose(ll)+Glucuronate (11) 20.4 0.5(1.9) 5.3 0.4(0.5) ND 0

a: Composition of the acid hydrolysate was determined by UPLC and differential refractometry.
b: Due to background noise and very small product amounts, accurate data was not obtained for
quantification of methanol in the MeGAXn hydrolysate.
c: None detected.










Table 2-3. Distribution of 13C in fermentation products formed in anaerobic cultures of E.
asburiae JDR-1 and E. coli B grown with differentially 13C labeled xylose and
glucose .

Labeled products, mM, and (fraction) labeled with 13C
Fermentation Acetate Ethanol Lactate
CH3C*OOH CH3C*H20H CH3C*HOHCOOH+CH3C*HOHC*OOH
[2- C]xylose, 4.8 (0.40) 5.8 (0.43) 0.9 (0.45)
E. asburiae JDR-1
[2-13C]xylose
[2- 'yose, 3.0(0.26) 1.9(0.27) 2.8 (0.31)
E.coli B
C*H3COOH C*H3CH20H C*H3CHOHCOOH

[13C]g 2.3 (0.34) 4.6 (0.37) 4.8 (0.38)
E.asburiae JDR-1
[6-13C]glUCOSO
[6- C]gucose 1.9(0.28) 4.7 (0.35) 5.4 (0.40)
E.asburiae JDR-1
a: Carbons enriched in 13C in different fermentation products were determined and quantified by
13C-NMR (Fig. 2-5) and are noted by *. Total Products were quantified by HPLC. The fractions
of labeled products to their total products were calculated and noted parenthetically in the table.


Table 2-4. Anaerobic molar cell dry weight and ATP yield from different substrates calculated
based on estimated YATP, 8, for all substrates in E. asburiae JDR-1.

Fermentation substrates

Glucose Xylose Glucuronate MeGAX

YM-substrate (g/mole)a 20.51.4 10.20.7 10.40.3 32.01.1
Estimated ATP yield per mole of 2.6 1.3 1.3 4.0
substrate
a: YM-substrate: molar cell dry weight yields for different substrates, determined in triplicate with
indicated standard deviations.












(Xyl-Xyl-Xyl-Xyl-Xyl-Xyl-Xyl-Xyl-Xyl-Xyl-Xyl-Xyl-)n
MeGA MeGA
(Sweetgum Methylglucuronoxylan)

Dilute Acid
Hydrolysis


XylX+
Xy + MeGA
(Xylose) (MeGAX)
Figure 2-1. Scheme for the release of xylose and MeGAX by dilute acid hydrolysis of sweetgum
xylan.


















































Figure 2-2. Microscopic images of E. asburiae JDR1. A) Differential Interference Contrast (DIC)
microscopic photos of E. asburiae JDR-1 growing in MeGAX minimal medium at
exponential phase. B) Image of E. asburiae by negative stain electron microscopy.














2.5 1.6
21.4
2
E 1.2

S1.5 0 -
o 0.8
| 1 --0.6
S- 0.4
S0.5
1 -- 0.2
0' 0
0 10 20 30 40
Time (h)
B.
2.5 1.2

E 2 < 1

-\ 0.8 o

o a
0.6



o 0.5
o 0.2



0 10 20 30 40

Time (h)

Figure 2-3. Aerobic growth, substrate utilization, and formation ot products trom acid
hydrolysates of MeGAXn. by A) E. asburiae JDR-1 and B) E. coli B. Xylose
(diamonds), MeGAX (squares), and acetic acid (triangles) were determined in media
by HPLC. Growth was determined by measuring turbidity as OD600 (open circles).












10 1.4
1.2
8-
E 0
6 0.8

4 0.6
0.4
0 2 0.2
0 0.2

0 5 10 15
Time (h)

Figure 2-4. Aerobic growth of E. asburiae JDR-1 on different combinations of sugar substrates.
Concentrations of substrates and acetic acid as a product were determined by HPLC.
Growth was determined as turbidity (OD600). A) Growth on glucose (7.5 mM) and
xylose (7.5 mM). Concentrations of glucose (closed circles), xylose (diamonds) and
acetic acid (triangles); OD600 (open circles); B) Growth on glucuronic acid (10 mM)
and xylose (10.5 mM). Concentrations of glucuronic acid (open squares) and xylose
(diamonds); OD600 (open circles) C) Growth on MeGAX (6.5 mM). Concentrations
of MeGAX (squares): OD600 (open circles).












B 12 1

10 5 0.8

E8
0.6 O
-6
0.4 3
0 4 O

0 2 0.2

0 0
0 5 10 15

C. Time (h)

8 1


0.8


0.6 O


a)0.4 5

0 2
) 0.2


0 0
0 5 10 15
Time (h)

Figure 2-4. Continued












; 1, ;Z 1. 1-1 1 ?





o um
















200 150 100 50 0 PPM


Figure 2-5. Pathway determination for the metabolism of xylose and glucose by E. asburiae
JDR-1. Media from anaerobic cultures of E. asburiae JDR-1 and E. coli B grown
with xylose or glucose enriched with 13C in specific carbons were analyzed by 75.5
MHz 13C-NMR spectrometry. A) [2-13C]-xylose fermented by E. asburiae JDR-1; B)
[2-13C]-xylose fermented by E. coli B; C) [l-13C]-glucose fermented by E. asburiae
JDR-1; D) [6-13C]-glucose fermented by E. asburiae JDR-1.

























0E 0 i
U =

t '4









200 10 100 50 PPM
UU U U

a a
u u






ozo
99 ^ j86154






200 150 1 o0 50 0 PPM








CQO



*
UU0

*
U











200 150 lao 60 o PPM


Figure 2-5. Continued











V















Figure 2-5. Continued


Chapter 2 was adapted and modified from [APPLIED AND ENVIRONMENTAL
MICROBIOLOGY, Jan. 2009, p. 395-404], Copyright American Society for Microbiology.









CHAPTER 3
GENETIC ENGINEERING OF ENTEROBACTER ASBURIAE STRAIN JDR-1 FOR
EFFICIENT D(-) LACTIC ACID PRODUCTION FROM HEMICELLULOSE
HYDROLYSATE

Introduction

As one of three main components of lignocellulosics, hemicellulose contains

polysaccharides comprised of pentoses, hexoses and sugar acids that account for 20-35% of the

total biomass from different sources (Saha 2003). Depending on the source, these fractions

contain heterogeneous polymers comprised of pentoses, hexoses and sugar acids. The ability to

utilize all hemicellulosic sugars is important for efficient conversion of lignocellulosic materials

to fuel ethanol and other value-added products (McMillan 1997; Saha 2003; Shanmugam and

Ingram 2008). Bacteria, e.g. E. coli KO 11 and related strains, have been developed for the

efficient conversion of both pentoses and hexoses to ethanol and other products (Ingram et al.

1987; Ingram et al. 1997). Xylan, in the form of 4-O-methylglucuronoxylan, is a polysaccharide

found in all hemicelluloses, and is the predominant structural component of hardwood and

agricultural residues (Preston et al. 2003). This polysaccharide may consist of more than 70

P-xylopyronose residues, linked by 0-1,4-glycosidic bonds (Timell 1964). In hardwood and

softwood xylan, a 4-O-methylglucuronic acid has been reported to be attached to the 2'- position

of every 6th to 8th xylose residue (Kardosova et al. 1998; Jacobs et al. 2001). Dilute acid

hydrolysis is commonly used to render the monosaccharides of lignocellulose accessible for

fermentation (Lee et al. 1997; Saha et al. 2005a). However, the a-1,2 glucuronosyl linkage in

xylan is resistant to dilute acid hydrolysis, resulting in the release of methylglucuronoxylose

(MeGAX) along with free xylose. MeGAX is not fermented by bacterial biocatalysts currently

used to convert hemicellulose to ethanol, such as Escherichia coli KO 11 (Rodriguez et al. 2001).

In sweetgum xylan, as much 27% of the carbohydrate may reside in this unfermentable fraction









after dilute acid pretreatment. E. asburiae JDR-1 has the ability to efficiently ferment both

MeGAX and xylose, producing ethanol and acetate as primary fermentation products (Chapter

2).

In this study, we have demonstrated that E. asburiae JDR-1 can be genetically engineered

to produce optically pure D(-) lactate as the primary product by deleting genes (pflB and als)

essential for the production of ethanol, acetate and 2,3 butanediol. The engineered strain was

evaluated for its ability to produce lactate from dilute acid hydrolysates of sweetgum xylan.

Materials and Methods

Bacterial Strains, Media, and Growth Conditions

The bacterial strains constructed and used in these studies are listed in Table 3-1. The E.

asburiae JDR-1 served as a starting point for genetic engineering. During strain construction,

cultures were grown aerobically at 300C, 370C, or 390C in Luria broth (10 g 11 Difco tryptone, 5

g 11 Difco yeast extract, and 5 g 11 NaC1) containing either 2% (w/v) glucose, 5% sucrose or 3%

(w/v) arabinose. Ampicillin (50 mg 11), tetracycline (12.5 mg 11), kanamycin (20 mg 11 and 50

mg 1 ), apramycin (20 mg 11) or chloramphenicol (10 mg 11 and 40 mg 11) were added as

needed.

Sweetgum methylglucuronoxylan (MeGAXn) was prepared from sweetgum stem wood

(Liquidambar styraciflua) as previously described and characterized by C13-NMR (Hurlbert and

Preston 2001; Kardosova et al. 1998). Dilute acid hydrolysates of methyglucuronoxylan were

prepared by acid hydrolysis of 1% (w/v) sweetgum xylan with 0.1 N H2S04 at 121 'C for 60 min,

followed by neutralization with BaCO3. Total carbohydrate concentrations of substrate

preparations were determined by the phenol-sulfuric acid assay (Dubois et al. 1956) with xylose as

reference or by HPLC as previously described (Chapter 2). Fermentation media were

supplemented with Zucker and Hankin mineral salts (ZH salts) at pH 7.4 (Zucker and Hankin









1970). Growth media were buffered with 100 mM sodium phosphate buffer (pH 7.0) or 100 mM

3-(N-morpholino) propane sulfonic acid (MOPS) buffer (pH 7.0) when necessary.

Genetic Methods

Standard methods were used for most of the genetic manipulations. Qiagen kits were used

for genomic DNA and plasmid extraction (Qiagen, Valencia, CA). Polymerase chain reaction

(PCR) amplification was performed with an I-cycler thermal cycler (BioRad, Hecules, CA) with

primers synthesized by Operon (Huntsville, Alabama). Topo cloning kits were used for cloning

(Invitrogen, Carlsbad, CA). Electroporation was performed on Gene pulser Xcell (BioRad,

Hercules, CA). Restriction endonucleases were purchased from New England Biolabs (Ipswich,

MA). DNA sequencing was provided by the University of Florida Interdisciplinary Center for

Biotechnology Research.

Deletion of pflB and als genes in E. asburiae JDR-1

The method for gene deletion in E.coli was used as previously described (Zhang et al. 2007;

Janatama et al., 2008) with minor modifications applied to E. asburiae JDR-1. ThepflB gene in

E. asburiae JDR-1 was also selected as an integration site for the PET operon. Several sets of

primers were designed based on sequences ofpflB orthologs in other Enterobacter spp. to

amplify this gene fragment from E. asburiae JDR-1. Only one set derived from E. coli B was

found to amplify the E. asburiae JDR-1 pflB gene fragment. The amplified E. asburiae JDR-1

DNA sequence and E. coli Kl2pflB sequence were found to have 93% identity. The plasmids

constructed are listed in Table 1. The partial sequence of the E. asburiae JDR-1 pflB gene (gene

bank accession number: EU719655) was determined within a DNA fragment amplified by PCR

using specific primers based on the E. coli pflB sequence. The 3 kb cat-sacB cassette was

obtained by digesting pLOI4162 with Smal and Sfol, and used in subsequent ligations. The pflB

gene fragment amplified from E. asburiae JDR-1 was cloned into pCR 4-TOPO vector









(Invitrogen) to obtain a plasmid, pTOPOpfl. This plasmid was diluted 500-fold and served as

template for inside-out PCR amplification using the pfl inside-out primers. The resulting 5.5 kb

fragment containing the replicon was ligated to the blunt-end cat-sacB cassette from pLOI4162

to produce a new plasmid, pTOP04162pfl. This 5.5 kb fragment was also used to construct a

second plasmid, pTOPODpfl, by phosphorylation and self-ligation. Both pTOP04162pfl and

pTOPODpfl were then digested with XmnI, diluted 500-fold and used as templates for

amplification using the pfl primer set to produce linear DNA fragments for integration step 1

(pfl'-cat-sacB-pfl") and step 2 (pfl' pfl"), respectively. pfl' andpfl" represent twopfl gene

fragments. After electroporation of the step 1 fragment into E. asburiae JDR-1 containing

pLOI3240, cells were incubated for 2 hr at 300C. The recombinant candidates were selected for

chloramphenicol (20 mg 1-1) resistance in Luria broth plates after overnight incubation (15 h) at

390C. Colonies were patched on both kanamycin (50 mg 11) plates and chloramphenicol (40 mg

I 1) plates. Those colonies growing on chloramphenicol (40 mg 1 ) plates but not on kanamycin

(50 mg 11) plates were subjected for PCR confirmation. The confirmed mutant colonies were

transformed with pLOI3240, and prepared for electroporation with the step 2 fragment (pfl' -pfl').

After electroporation, cells were incubated at 300C for 4 h and then transferred into a 250-ml

flask containing 100 ml of LB without NaCl with 10% sucrose. Following an overnight

incubation (300C), colonies were streaked on LB minus NaCl plates containing 6% w/v sucrose

(390C, 16 h). Colonies were tested for loss of apramycin and chloramphenicol resistance and

confirmed by PCR. The resulting strain E. asburiae El had a disrupted pflB gene without

detectable heterologous DNA sequences.

E. asburiae El served as a starting strain for further deletion of als gene. A segment of E.

asburiae JDR-1 asl gene (FJ008982) was PCR amplified using degenerate primers designed based









on conserved sequences identified in homologous asl genes found in Enterobacter sp. 638,

Erwinia carotovora subsp. atroseptica SCRI1043, Yersinia enterocolitica subsp. enterocolitica

8081 and Serratiaproteamaculans 568. With the segment of the E. asburiae JDR-1 asl gene, this

gene was disrupted in E. asburiae El with the same method forpflB disruption.

Fermentation

Batch fermentations were carried out in 16- by 100-mm screw-cap tubes filled with

nitrogen and sealed with rubber stoppers. The tubes were set in a Glas-Col minirotator at 60 rpm

in a 300C incubator. Neutralized sweetgum xylan acid hydrolysate (0.5% w/v) was added to 2x

ZH salts directly as growth medium buffered by 100 mM phosphate buffer or MOPS buffer at

pH 7.0. Fermentations in hydrolysates were inoculated to an initial optical density at 600 nm of

1.0 (determined using a Beckman DU500 series spectrophotometer). For analysis of

fermentation products, cultures were centrifuged, and the supernatants were passed through 0.22

um filters and subjected to HPLC. Products were resolved on a Bio-Rad HPX-87H column with

0.01 N H2SO4 at 65 C. Samples were delivered with a 710B WISP automatic injector and

chromatography controlled with a Waters 610 solvent delivery system at a flow rate of 0.5

ml/min. Products were detected by differential refractometry with a Waters 2410 RI detector.

Data analysis was performed with Waters Millennium Software. A quantitative relationship was

determined between E. asburiae JDR-1 cell dry weight and culture OD at 600 nm. For

calculation of specific consumption rates and specific production rates, the cell dry weight was

determined based on the OD600 of the fermentation culture, which was 1.0 (0.51 g 1-1) initially

and did not appreciably change during the fermentation in 0.5% xylan hydrolysate.

Determination of Lactate Isomers Produced by E. asburiae LI

To determine the isomers of lactate formed, fermentation products were assayed with

D-lactate of L-lactate dehydrogenases (Taguchi and Ohta 1991). The conditions of the









colorimetric enzyme assays were similar to those used to measure lactate dehydrogenase activity

(Babson and Babson 1973). j-NAD was obtained from Research Products International Corp,

Chicago IL. All other reagents, substrates, and enzymes were obtained from Sigma Chemical Co.,

St. Louis MO. lodonitrotetrazolium chloride (40 mg), 100 mg NAD and 10 mg

phenylmethylsulfonyl fluoride were dissolved in 20 ml 0.2 M Tris-HCI (pH 8.2) to obtain the

colorimetric reagent. Reactions were initiated by adding 4 Kunitz units (1 [tmol/min) of either

L-lactate dehydrogenase (rabbit muscle, 140 U/mg protein) or D-lactate dehydrogenase

(Lactobacillus leichmanii, 232 U/mg protein) in 100 ul colorimetric reagent and 100 ul sample at

room temperature. The reduction of iodonitrotetrazolium dye was measured at room temperature

of a Beckman 520 spectrophotometer at 503 nm. Sodium salts of L and D-lactate (Sigma) were

used as standards to define enantiomer specificity of the reaction.

Results

Fermentation Characteristics of the Wild Type Strain E. asburiae JDR-1

When growing with either 0.8% glucose, 0.5% arabinose or 0.5% xylose as the sole carbon

source, the wild type strain produced several products including succinate, lactate, acetate,

2,3-butanediol and ethanol. Glucose fermentations resulted in the formation of 2,3-butanediol,

ethanol and acetate as major products. Larger amount of acetate and no 2,3-butanediol was

detected in 0.5% xylose and 0.5% arabinose fermentations (Table 3-2).

The initial concentrations of substrates in the medium containing 0.5% sweetgum

hemicellulose hydrolysate were determined by HPLC to be 20 mM xylose, 1.4 mM MeGAX and

a small amount of MeGAX2. Previous studies indicated that MeGAX was metabolized by E.

asburiae JDR-1 into methanol, glucuronate and xylose (Chapter 2). In these previous studies

glucuronate fermentation by E. asburiae JDR-1 generated acetate in nearly 100% yield,

indicating fermentation products more reduced than acetate could only be produced from the free









xylose and the xylose released from MeGAX in the hydrolysate. The theoretical maximum yield

of lactate from this hydrolysate medium was calculated to be 35.7 mM based on the total xylose

initially present. In the fermentation of methylglucuronoxylan hydrolysate, E. asburiae JDR-1

was able to utilize all of the MeGAX within 30 hours and xylose within 40 hours. Similar

amounts of ethanol (15.6 mM) and acetate (20 mM) were produced but no 2,3-butanediol or

lactate was detected (Table 3-2, Fig. 3-la). When supplemented with LB, E. asburiae JDR-1

fermented the 0.5% hydrolysate much more rapidly than with ZH minimal salts. Substrates were

utilized within 15 hours, producing 16.2 mM ethanol, 22 mM acetate, and 3.2 mM succinate,

again with no 2,3-butanediol or lactate detected. (Table 3-2, Fig. 3-1c).

Fermentation Characteristics of the Engineered Strains E. asburiae El and LI

The major competing pathway to lactate production initiates from the pyruvate format

lyase catalyzed reaction, which produces format and acetyl-CoA in the wild type strain E.

asburiae JDR-1. Both acetate and ethanol are produced from acetyl-CoA. In order to convert

more carbon flux from pyruvate to lactate, the pfIB gene of JDR-1 was deleted to obtain strain E.

asburiae El. Since 2,3-butanediol was also produced by E. asburiae El in the fermentation of

glucose (Table 3-2), the als gene which encodes acetolactate synthase was deleted in E. asburiae

El to eliminate 2,3-butanediol production (Moat at al. 2002). The resulting strain E. asburiae L1

was a double mutant lacking pflB and als genes (Fig 3-2).

Both E. asburiae El and LI were able to produce lactate as the predominant product in

glucose, xylose and arabinose fermentations. E. asburiae El produced 2.9 mM 2,3-butandiol in

0.8% glucose fermentation. The LI strain with an interrupted 2,3-butanediol-producing pathway

produced no 2,3-butanediol and achieved higher lactate yield (94.l1% of the theoretical

maximum). In xylose and arabinose fermentations, the L1 strain also achieved higher lactate

yield than El strain (Table 3-2).









The E. asburiae LI fermented slowly in the xylan hydrolysate with ZH minimal salts. In

60 hours, only a portion of free xylose in the hydrolysate was utilized and the MeGAX portion

was not utilized (Fig 3-1b). Within 100 hours, 22.2 mM lactate was produced (Table 3-2). The

low fementation rate of LI in hydrolysate medium may be due to a limiting activity of lactate

dehydrogenase. The absence of detectable lactate formation in the parent strain during

fermentation of xylan hydrolysates also indicates a limitation in lactate dehydrogenase activity of

E. asburiae JDR-1. The E. asburiae LI strain fermented more rapidly in the xylan hydrolysate

supplemented with LB, with the complete consumption of both MeGAX as well as xylose in 65

hours (Fig. 3-id) with the formation of 36.4 mM lactate as well as very small amount of acetate

and succinate (Table 3-2). Both El and LI were able to produce lactate at 100% of the

theoretical maximum yield. The small amounts of acetate were likely derived from the

glucuronate group of the 1.4 mM MeGAX present in the hydrolysate substrate.

The utilization of MeGAX by the LI strain required LB supplementation, while the

original isolate, E. asburiae JDR-1, utilized MeGAX in both minimal (Fig. 3-la) and LB

supplemented (Fig. 3-1c) media during the mixed acid fermentation that produced acetate and

lactate in nearly equal amounts (Table 3-2). Supplementation with LB resulted in a 2-fold

increase in the utilization rate of xylose and nearly 3-fold increase in the production rate of

lactate in the LI strain (Table 3-3).

D-Lactate Was Produced by E. asburiae LI

The optical enantiomer(s) of lactate produced by E. asburiae LI from the fermentation of

xylan hydrolysates was determined by measuring the oxidation of lactate catalyzed by D- or

L-lactate dehydrogenase with the reduction of iodonitrotetrazolium dye mediated via NADH

formation as described in the Materials and Methods section. A sample of medium containing

3.6 [mol of lactate (determined by UPLC) of an E. asburiae LI fermentation (72 hour) of 0.5%









xylan hydrolysate supplemented with LB resulted in an increase in A503 from 0 to 0.113 in 5

minutes when assayed with 4 units of D-lactate dehydrogenase When the same sample was

assayed under the same conditions with 4 units of L-lactate dehydrogenase, there was no

detectable increase in A503. Therefore the lactate produced by E. asburiae LI was D-lactate with

an apparent optical purity 100%.

Discussion

The fermentations of dilute acid hydrolysates of methylglucuronoxylan by E. asburiae

strains El and LI provide the first examples of lactate formation from the aldouronate as well as

the xylose present in these hydrolysates. The efficient formation of the D(-) entantiomer

demonstrates a metabolic potential for the efficient production of optically pure lactate from the

most predominant polysaccharide components in the hemicelluloses fractions derived from

woody biomass and agricultural residues. Although the relatively low production rate and

dependence on rich media limit direct application of E. asburiae L1, metabolic evolution by

adaptive culturing and further genetic engineering may overcome these limitations. The studies

here serve as a basis to further develop biocatalysts capable of aldouronate utilization to

efficiently ferment hemicellulose hydrolysates for production of alternative fuels and chemicals.













Table 3-1. Bacterial strains and plasmids.


Strain and plasmid Relevant characteristics Source or
reference
Strains
E. coli Top 10 For general cloning Invitrogen

E.asburiae JDR-1 Wild type Our lab

E.asburiae El E. asburiae JDR-\1ApflB This chapter

E.asburiae LI E. asburiae JDR- 1ApflB Aals This chapter
Plasmids

PLOI3240 Amr red, red recombinase protein (Wood et al.
2005)
pLOI4162 bla cat; cat-sacB cassette (Jantama et
al. 2008)
pCR 4-TOPO bla kan amp; TOPO TA cloning vector Invitrogen
pTOPOpfl pflB (PCR) amplified from E. asburiae JDR-1 This chapter
and cloned into PCR4-TOPO vector
pTOPO4162pfl cat-sacB cassette cloned into pflB in pTOPOpfl This chapter

pTOPODpfl PCR fragment amplified from pTOPOpfl, kinase This chapter
treated, and self-ligated
pTOPOals als (PCR) amplified from E. asburiae JDR-1 This chapter
and cloned into PCR4-TOPO vector
pTOPO4162als cat-sacB cassette cloned into als in pTOPOpfl This chapter
pTOPODals PCR fragment amplified from pTOPOals, kinase This chapter
treated, and self-ligated










Table 3-2. Comparing fermentation products of wild type and genetically engineered E. asburiae
JDR-1 strains".

Fermentation products (mM)
Lactate
Ethanol Acetate 2,3-Butanediol Succinate Lactate 0/ yieldb
E. asburiae JDR-1
0.8% glucose 26.8 11.5 12.9 5.2 3.9 4.6
0.5% xylose 20.9 17.5 0 4.1 1.0 1.9
0.5% arabinose 24.0 17.1 0 4.2 1.0 1.9
0.5% xylan hydrolysate 15.6 20.0 0 0 0 0
0.5% xylan hydrolysate with LB 16.2 22.0 0 3.2 0 0
E. asburiae El
0.8% glucose 5.6 0 2.9 2.7 77.0 91.7
0.5% xylose 3.4 2.8 0 3.2 46.7 89.8
0.5% arabinose 6.5 2.9 0 2.1 41.3 78.0
0.5% xylan hydrolysate with LB 0 2.0 0 0 36.2 100.4
E. asburiae L1
0.8% glucose 4.4 0 0 1.7 78.9 94.1
0.5% xylose 1.5 2.9 0 1.3 47.2 90.8
0.5% arabinose 5.0 2.8 0 2.1 49.6 93.6
0.5% xylan hydrolysate" 0 0 0 0 22.2 96.0
0.5% xylan hydrolysate with LB 0 3.0 0 1.0 36.4 101.2
a: Fermentations were completed within 72 hours with minimal media, or otherwise as indicated
footnote c. The initial concentration of 0.8% glucose, 0.5% xylose and 0.5% arabinose media
were determined by HPLC to be 42 mM, 31 mM and 31.5 mM respectively. The 0.5% xylan
hydrolysate medium was measured to contain 20 mM xylose and 1.4 mM MeGAX.
b: Percent of actual yield of lactate to theoretical maximum yield. Maximum yield is defined as
2 mole lactate/ mole glucose or 5 mole lactate/ 3 mole xylose.
c: This result was obtained after fermentation for 100 hours at which time 65% of the xylose in
the hydrolysate was utilized.












Table 3-3. Specific consumption rates and specific production rates of E. asburiae LI in 5g /liter
acid hydrolysate of sweetgum xylan.a


Strains q Xylose (g q MeGAX (g q Lactate (g lactate
xylose/g DCW/h) MeGAX/g DCW/h) /g DCW/h)
E.asburiae LI i 0.0670.006 0 0.0490.003
ZH salts
E.asburiae LI in 0.130.01 0.0190.002 0.130.005
0.12% LB

a: q xylose and q MeGAX: Xyose and MeGAX specific consumption rate respectively, as grams
of substrate consumed per gram dry cell weight per hour. q lactate: lactate specific production
rate, products generated per gram dry cell weight per hour.















40

35 a
5 30
E
25
o
2 20

15
10



35 b

30

25

20
( 15
0 10




35 C
S 30
E
C 25
0

O 15
0
3 10



35 d



0
20








Time (Hours)

Figure 3-1. Fermentation time course for different strains in media containing 0.5% sweetgum
xylan hydrolysate a) E. asburiae JDR-1 in minimal medium, b) E. asburiae L1 in
minimal medium, c) E. asburiae JDR-1 in LB, d) E. asburiae L1 in LB. Substrates

and fermentation products: xylose (+), MeGAX (m), acetic acid ( A), ethanol (0),

lactic acid (0).
lactic acid (0).




























2-
NAD

NAD


succinate


glucose,
xylose, PEP
arabinose PEP
ADP
ATP NADH NADI
Acetolactate Pyruvate- % Lactate
H | pflBF a t e

I* -!- Formate

4b t di


Acetyl-CoA Acetate


2NADH ) ADP ATP

2NAD+

Ethanol

Figure 3-2. Diagram to illustrate deletion of als andpflB genes modifying mixed-acid
fermentation of E. asburiae JDR-1 into a homolactate production pathway in E.
asburiae Ll. Deletion of pathways is indicated in the figure as symbol X.


S,*J-UU LaIICUIUI


213











CHAPTER 4
GENETIC ENGINEERING ENTEROBACTER ASBURIAE STRAIN JDR-1 FOR EFFICIENT
ETHANOL PRODUCTION FROM HEMICELLULOSE HYDROLYSATE

Introduction

Lignocellulosic resources, including forest and agricultural residues and evolving energy

crops, offer benign alternatives to petroleum-based resources for production of fuels and

chemicals. As renewable resources, these are expected to decrease dependence on exhaustible

supplies of petroleum and mitigate the net release of carbon dioxide into the atmosphere. The

development of economically acceptable bioconversion processes will require pretreatments that

release the maximal quantities of hexoses (predominantly glucose from cellulose), and pentoses

(arabinose and xylose) from hemicelluloses, and also require microbial biocatalysts that will

efficiently convert these to a single targeted product. MeGAXn are the predominant components

in the hemicellulose fractions of agricultural residues and energy crops, e.g corn stover,

sugarcane bagasse, poplar, and switchgrass (Dien et al. 2006; Pordesimo et al. 2005; Sun et al.

2004; Sun et al. 2001). In sweetgum xylan, as much as 27% of the carbohydrate may reside in

this unfermentable fraction, MeGAX, after dilute acid pretreatment (Chapter 1; Rodriguez et al

2001). Complete utilization of all hemicellulosic sugars can improve conversion efficiency of

lignocellulosic materials to fuel ethanol.

In this study, the PET operon containingpdc and adhB genes from Zymomonas mobilis

(Ingram at al 1987; Ingram and Conway, 1988) was incorporated into apff E. asburiae JDR-1

strain by plasmid transformation to construct a homoethanologenic strain. The resulting

recombinant strains were compared with wild type E. asburiae JDR-1 and the ethanologenic

strain E. coli KO11 to evaluate their efficiencies for ethanol production from dilute acid

hydrolysates of sweetgum methylglucuronoxylan.









Materials and Methods

Bacterial Strains, Media, and Growth Conditions

E. coli strains, E. asburiae strains and plasmids used are listed in Table 4-1. The E.

asburiae JDR-1 served as a starting point for genetic engineering. Batch fermentations were

carried out in medium saturated with nitrogen in tubes set in a Glas-Col minirotator at 60 rpm in

300C incubator. Fermentations in hydrolysates were inoculated to an initial optical density at 600

nm of 1.0. Fermentation products were resolved on a BioRad HPX-87H column with a Waters

HPLC system. The growth conditions and media were same with the ones applied in chapter 3.

Genetic methods

Standard methods were used for most of the genetic manipulations as described in chapter 3.

The plasmids used and constructed in this study are listed in Table 4-1.

Transformation of Plasmid Carrying PET Operon into E. asburiae JDR-1

E. asburiae JDR-1 was grown with one of several antibiotics at different concentrations in

LB and minimal media on agar plates or in liquid media to test its antibiotic resistance. Based

upon its sensitivity to chloramphenicol and tetracycline, plasmids pLOI555 ( cmR) and

pLOI297(tetR), both containing the PET operon, were transformed into E. asburiae JDR-1 or E.

asburiae El by electroporation in a 100 [l cuvette under the condition of 1.8kV, 25 [tF

capacitance and 2000 resistance. For electroporation, competent cells from 25 ml exponential

phase cultures were washed 3 times by suspension and centrifugation with cold 10% glycerol.

Cultures were plated on LB agar containing 2% glucose and tetracycline (12.5 mg 11) or

chloramphenicol (40 mg 11) to select E. asburiae JDR-1 and El carrying pLOI297 or pLOI555

respectively. Plasmids were extracted confirming their presence in E. asburiae cells.









Integration of PET Operon into E. asburiae JDR-1 Chromosome

Since E. asburiae JDR-1 was found to have high resistance to ampicillin, the red

recombinase plasmid pLOI3240 with apramycin resistance gene was transformed into E.

asburiae JDR-1. The rrlE gene fragment (gene bank accession number: EU719656 ) of this

strain was amplified using specific primers based on E. coli rrlE gene sequence, with E. asburiae

JDR-1 genomic DNA as template. The identity was found to be 96% between rrlE gene DNA

sequences of E. asburiae JDR-1 and E. coli K12. The high sequence similarity provided the

possibility of using E. coli rrlE fragment for recombination with E. asburiae JDR-1 rrlE gene

DNA. The fragment, with PET operon flanked by E. coli rrlE gene

(rrlE'-pdc-adhA-adhB-Km-rrlE"), was cut from pLOI4672 (constructed by Xueli, Zhang) with

AscI (New England Biolabs, US), and transformed into E. asburiae JDR-1 by electroporation.

The selected colonies from kanamycin (20 mg 1 ) medium were patched on kanamycin (50 mg

1 1) and chloramphenicol (40 mg 11) medium to differentiate correctly integrated colonies from

colonies with contaminated pLOI4672 plasmid. The colonies growing on kanamycin medium but

not on chloramphenicol medium were subjected to PCR confirmation.

The pflB gene in E. asburiae JDR-1 was also selected as an integration site for the PET

operon. Two sets of specific primers based on Enterobacter 638

(http://genome.jgi-psf org/finished microbes/ent_6/ent_6.home. html) pflB sequence and two

sets of specific primers based on E. colipflB sequence were employed to amplify pyruvate

formatelyase gene sequence from E. asburiae JDR-1. One set derived from E. coli was able to

amplify the pflB gene (gene bank accession number: EU71965 5) fragment from E. asburiae

JDR-1 genomic DNA. The amplified E. asburiae JDR-1 DNA sequence and the E. coli K12 pflB

sequence were found to have 93% identity. The subsequent chromosome integration process was

same as the previous one at rrlE site with the plasmid pLOI4666 constructed by Xueli, Zhang. A









primer set for the kan gene was used to confirm the integration. The confirmed recombinant

strains for both insertion sites were grown on the aldehyde detection plates (Conway et al. 1987).

The colonies with strongest red color indictive of aldehyde production were selected for further

fermentation analysis.

Plasmid Stability in E. asburiae JDR-1

E. asburiae JDR-1 harboring either pLOI555 or pLOI297 was serially transferred in Luria

broth containing 2% glucose without antibiotics for more than 72 generations at 300C. One

generation was defined as a 2-fold increase in culture turbidity. Appropriate dilutions of cultures

were plated on Luria agar with and without antibiotic; colonies formed were counted and

calculated to obtain the ratio of cells retaining antibiotic resistance to total cells. Ten colonies

retaining antibiotic resistance (and therefore presumed to retain pLOI555 or pLOI297) after 72

generations were subjected to fermentation to test their ethanol producing ability.

Assay of PDC Activity

Pyruvate decarboxylase activity was assayed in engineered E. asburiae JDR-1 strains by

monitoring the pyruvate-dependent oxidation of NADH with alcohol dehydrogenase as a

coupling enzyme (Conway et al. 1987; Ohta et al. 1991b). Exponential phase anaerobic cultures

were harvested and cells were disrupted using the FastPrep bead mill MP system (MP

Biomedicals, Irvine, CA) in 0.05 M phosphate buffer. The supernatant was collected after 15 min

centrifugation at 1800 rpm (Eppendorf centrifuge 5414). The entire process was carried out at

40C. Heat treatment for 15 min at 600C was used to inactivate competing native enzymes ofE.

asburiae JDR-1 which might affect quantitative measurements of PDC activities in

transformants. The enzyme activity assay of PDC was performed in the reaction mixture of 1.0

mM TPP (thiamine pyrophosphate), 1.0 mM MgCl2, 0.40 mM NADH, 20 mM sodium pyruvate

and 0.05 M sodium phosphate buffer, pH 6.5. The assay was started by adding 20 [l crude cell









extract. Protein concentration of the crude extract was determined with BCA protein assay

reagent kit (Pierce Chemical Co., Rockford, IL).

Results

Fermentation Characteristics of the Wild Type Strain E. asburiae JDR-1

E. asburiae JDR-1 performed a mixed-acid fermentation in low substrate concentration.

When growing in 2.5% (w/v) glucose or 2% (w/v) xylose, the wild type strain produced a wide

range of products, including succinate, lactate, acetate, format, 2,3-butanediol and ethanol

(Table 4-2). In glucose fermentation, succinate and acetate were produced at low concentrations,

approximately at 1 mM. Lactate was produced at approximately 10 mM, and the major products

were format, 2,3-butanediol and ethanol, each at approximately 40 mM. More acetate and less

2,3-butanediol were produced in xylose fermentation (Table 4-2). In both batch fermentations

buffered with 0.1 M sodium phosphate (pH 7.0), the wild type strain failed to utilize all the

substrates during the 48 h fermentation. Even in the buffered medium the pH after fermentation

decreased to 4.8, which suggested that acid production might be the main factor preventing the

cells from utilizing all the substrate.

The components in the medium containing 0.5% sweetgum hemicellulose hydrolysate

were determined by HPLC to be 20 mM xylose, 1.4 mM MeGAX and a small amount of

MeGAX2 (Fig 4-1). Previous studies suggested that MeGAX was metabolized by E. asburiae

JDR-1 into methanol, glucuronate and xylose. Glucuronate fermentation by E. asburiae JDR-1

generated acetate in nearly 100% yield, indicating more reduced fermentation products (ethanol

and lactate) could only come from the free xylose and the xylose released from MeGAX

( chapter 2). Therefore, the theoretical maximum yield of ethanol from this hydrolysate was

calculated to be 35.7 mM based on the total amount of xylose present in hydrolysate. E. asburiae

JDR-1 was able to completely utilize MeGAX in the 0.5% hydrolysate in about 12 hours and









xylose in 20 hours after a period of several hours for adaptation to the hydrolysate medium.

Similar amounts of ethanol (15.6 mM) and acetate (20 mM) were produced with small amount of

format and no detectable 2,3-butanediol; the ethanol yield was 44.2% of the theoretical

maximum (Table 4-3, Fig 4-1, Fig 4-2A). The specific consumption rates of xylose and MeGAX

in the hydrolysate and specific production rates of acetate and ethanol are included in Table 4-4.

Fermentation of E. asburiae 4666 and E. asburiae 4672 in Glucose

In order to increase ethanol yield, PET operon containing Zymomonas mobilispdc, adhA

and adhB genes were integrated into E. asburiae JDR-1 chromosome within either pflB or rrlE

gene, to obtain the engineered strains E. asburiae 4666 and E. asburiae 4672 respectively.

However, both strains could not completely utilize 2.5% glucose in the batch fermentations, and

the final pH decreased to 5. E. asburiae 4672 had similar fermentation products with the wild

type (Table 4-2), suggesting that the integrated PET operon might not be expressed efficiently. E.

asburiae 4666, with the PET operon inserted into pflB gene, showed a different profile of

fermentation products. This strain produced much less format (3.2mM) and much more lactate

(36.5mM) than the wild type, while kept other fermentation products at similar level including

ethanol. This suggested that insertion of PET operon into plfB gene inactivated pyruvate format

lyase, resulting in less format and acetyl-CoA production. More carbon flux was converted

from pyruvate to lactate through lactate dehydrogenase. Both engineered strains had a similar

ethanol yield as the wild type strain (Table 4-2).

Fermentation of E. asburiae JDR-1 (pLOI297) and E. asburiae JDR-1 (pLOI555)

Plasmids pLOI297 and pLOI555 were transformed into E. asburiae JDR-1 for

overexpression ofpdc and adh genes. Both transformed strains were able to completely utilize

2.5% (w/v) glucose or 2% (w/v) xylose within 48 hours, with ethanol as the predominant

fermentation product. The ethanol yields of glucose fermentation were 94.l1% and 95.3% for E.









asburiae JDR-1 (pLOI297) and E. asburiae JDR-1 (pLOI555), respectively (Table 2). E.

asburiae JDR-1 (pLOI555) was further tested in xylose fermentation, and the ethanol yield was

even higher, greater than 98% of theoretical. There were also other fermentation products present

at concentrations below 10 mM (Table 4-2).

E. asburiae JDR-1 (pLOI555) and JDR-1 (pLOI297) were tested for the fermentation of

dilute acid hyrolysates of sweetgum MeGAXn. Both strains consumed MeGAX as well as xylose

within 18 hr and fermentation was complete within 25 hr (Fig 4-2C for JDR-1 (pLOI555); data

for JDR-1 (pLOI297) was not shown). The xylose specific consumption rate of JDR-1 (pLOI555)

was similar to the parent strain but the MeGAX specific consumption rate was lower. Ethanol

was the major fermentation product, and the yield was much higher than the parent strain.

However, both strains produced substantial amount of acetate (approximately 10 mM) and had

lower yields of ethanol than with either xylose or glucose as substrates (Table 4-4).

Fermentation of Strain E. asburiae El (pLOI555) Compared with E. coli KO11 and Other
E. asburiae JDR-1 Derivatives

Neither 2,3-butanediol nor lactic acid was produced in the hydrolysate fermentation by

either E. asburiae JDR-1 (pLOI297) or JDR-1 (pLOI555). This result indicated that only the

acetate production pathway initiated from pyruvate format lyase competed for pyruvate and

lowered the ethanol yield. In order to direct greater carbon flux from pyruvate to ethanol, the

pJfB gene of E. asburiae JDR-1 was deleted to obtain strain E. asburiae El, followed by

pLOI555 transformation. When testing this strain in hydrolysate fermentations, no formic acid

was produced, and only small amount of acetate was produced (4.5 mM). After several hours of

adaption, the MeGAX portion was consumed in 12 hr and the xylose portion was consumed in

20 hr (Fig 4-2D). While the specific consumption rates of the substrates were close to the parent

strain and JDR-1 (pLOI555), E. asburiae El (pLOI555) had a much higher specific production









rate of the ethanol (0.1 10.01g ethanol/g DCW/h) and a much lower specific production rate of

the acetate (0.0220.003 g ethanol/g DCW/h). Most of the carbon sources in the hydrolysates

were converted to ethanol, achieving 99% of maximal theoretical yield (Table 4-3, Table 4-4,

Fig 4-1).

The E. coli KO11, which was reported to produce 0.54 gram ethanol per gram glucose

(Ohta et al. 1991a), could only produce ethanol at 63% of the theoretical maximum in the

sweetgum xylan hydrolysate medium, and accumulated a substantial amount (10.60.3 mM) of

acetate (Fig 1, Fig 2C). The sum of ethanol and acetate was 33.1 mM forE. co/i KOll, and 40.2

mM for JDR-1 (pLOI555), 39.9 mM for JDR-1 (pLOI297) and 40.5 mM for El (pLOI555)

(Table 4-3). This result indicated that E. coli KO 11 utilized less substrate in the hydrolysate than

the 3 engineered E. asburiae strains and produced lower quantities of products as a result of the

inability of E. coli KO 1 Ito utilize MeGAX in the hydrolysate (Fig 4-1, Fig. 4-2B). The ethanol

specific production rate of E. coli KO 11 (0.0740.006 g ethanol/g DCW/h) was much lower than

E. asburiae El (pLOI555) (0.110.01 g ethanol/g DCW/h) (Table 4-4). Compared with E. coli

KOll 1, E. asburiae El (pLOI555) utilized more substrate in sweetgum hydrolysate and was able

to produce 57.8% more ethanol at higher rate.

PDC Activity in E. asburiae Strains

The PDC enzyme activity produced as a result of expression of heterologous gene pdc in

engineered E. asburiae strains (Table 5). Because of the relative thermal stability of PDC

encoded by thepdc gene of Zymomonas mobilis, a heat treatment at 650C for 15 minutes was

used to inactivate competing native enzymes, e.g., activities associated with the pyruvate

dehydrogenase complex, could affect measurements of PDC activity (Conway et al. 1987; Ohta

et al. 1991b). While crude extracts from both strains showed pyruvate-dependent NADH oxidase

activity before heat treatment (data not shown), the wild type strains were unable to oxidize









NADH after the heat treatment. However, all three strains carrying plasmid with the PET operon

showed substantial PDC activities after heat treatment, showing the presence of PDC encoded by

pdc genes derived from Zymomonas mobilis.This result indicates that the successesful expression

the heterologous pdc is the key factor for the homogeneous production of ethanol in engineered

E. asburiae strains.

Plasmid Stability in E. asburiae JDR1

The pLOI297 transformant was relatively unstable, with only 10.7% of transformed E.

asburiae JDR-1 cells retaining tetracycline resistance after cultivation for 72 generations without

antibiotic selection pressure. The pLOI555 transformant, however, was quite stable, with 98.1%

of pLOI555 transformed E.asburiae JDR-1 cells retaining chloramphenicol resistance after

growth for 72 generations in the absence of antibiotic (Table 4-6). Fermentation analysis of 10

descendent colonies retaining antibiotic resistance from strains carrying pLOI297 and pLOI555

was also performed to confirm that strains with retained antibiotic resistance also retained the

homoethanologenic phenotype.

Discussion

A wild type Enterobacter asburiae strain with limited knowledge of its genetic and

physiological properties was genetically engineered for a new metabolic potential. The

methodology and protocols developed in this study may provide reference value for engineering

other wild type Enterobacter spp. While E. asburiae JDR-1 was determined to be relatively

resistant to ampicillin and probably other p-lactam antibiotics, it was sensitive to tetracycline

(12.5 mg 11), kanamycin (20 mg 11), apramycin (20 mg 11) and chloramphenicol (10 mg 1 ).

To determine if a plasmid-based system developed for use in E. coli could be maintained and

function in E. asburiae JDR-1, pCR4-TOPO plasmid with a small insertion was electroporated

into the competent cells and the transformants were able to be selected on a kanamycin (50 mg









1 1) plate. The transformed pCR4-TOPO plasmid in E. asburiae JDR-1 was qualitatively

determined by DNA gel electrophoresis to have a lower concentration than in E.coli Top 10 host

(data not shown). These results show that plasmids may be developed in E. coli for the

introduction of genes encoding a desired metabolic potential in E. asburiae JDR-1, and may have

applications in the metabolic transformation of other Enterobacter spp. as well.

With these transformation systems, E. asburiae JDR-1 (pLOI297) and E. asburiae JDR-1

(pLOI555), developed with the strategy successfully employed to genetically engineer the first

ethanologenic strains of E. coli (Ingram et al. 1987), were able to produce ethanol at 94.1% and

95.3% of theoretical yield in glucose, but failed to achieve such high yield in the dilute acid

hydrolysates of methylglucuronxylan. The dilute acid hydrolysate of sweetgum xylan used as

fermentation substrate is a mixture of different substrates. While xylose and MeGAX could be

quantified and very small amount of MeGAX2 could be detected by HPLC, there were other

chemical compounds too little to be detected. A variety of biological toxins are generated during

the dilute acid treatment. These include acetic acid, degradation products of sugars such as

furfural (dehydration product of pentoses) and 5-hydroxymethylfurfural (dehydration product of

hexoses), and soluble aromatics from lignin (aromatic alcohols, acids, and aldehydes). These

toxins retard the fermentation of hemicellulose syrups by yeast or other biocatalysts (Palmqvist

and Hahn-Hagerdal 2000a; Palmqvist and Hahn-Hagerdal 2000b). Although it was reported that

some of the aldehydes did not decrease ethanol yield in fermentation by E.coli based biocatalysts

(Zaldivar et al. 1999); bacteria usually utilize reducing power, e.g. NADPH, to reduce some of

the aldehydes to decrease their toxicity (Gutierrez et al. 2002). In the minimal medium of

hydrolysate used in this paper, with low substrate concentration and no supplemental nutrition,

the NADH/NAD+ ratio might be decreased by reducing the aldehydes (Miller at al. 2009). As a









result, more pyruvate might be converted to acetate, because no NADH was required in this

pathway. This might be the reason that E. asburiae JDR-1 (pLOI297) and E. asburiae JDR-1

(pLOI555) had lower ethanol yield in the hydrolysate.

To decrease the formation of organic acids, acetate and format, thepflB gene was then

deleted. The convenient one-step gene inactivation method successfully applied to E. coli

(Datsenko and Wanner 2000) failed to knock out the pfB gene in E. asburiae JDR-1, requiring

the development of a different protocol. An alternative gene deletion method used PCR

fragments with several hundred bases of homologous sequence at both ends instead of 40 bp

used by the one-step method (Jantama et al. 2008). Longer homologous sequences are expected

to increase the rate of homologous recombination (Puchta and Hohn 1991), and may delay

complete degradation of the linear DNA by exonucleases, thus increasing the probability of

recombination events. Recombinants were not selected on the plates containing levels of

antibiotics used for selection of E.coli recombinants and required lower concentrations,

kanamycin (20 mg 11) and chloramphenicol (10 mg 11) to be used. This is likely the basis for

growth of non-recombinant as well as recombinant colonies and required a second selection that

was achieved by patching colonies onto kanamycin (50 mg 11) and chloramphenicol (40 mg 11)

plates. By maximizing DNA concentration to approximately 5tg/[l and cell concentrations of

1010 cells/100[l in electroporation transformation, usually 3 to 6 E. asburiae JDR-1

recombinants could be obtained by this process. The methodology developed here might also be

applied to engineer other Enterobacter spp. with genetic manipulations developed for E.coli.

The E. asburiae strain with a genomicpflB deletion was transformed with a plasmid,

pLOI555, to obtain E. asburiae El (pLOI555), a strain capable of efficiently converting the

xylose residues derived from methyglucuronoxylan to ethanol, achieving a yield at 99% of the









theoretical maximum. In this respect it has been able to outperform E. coli KO11 in medium of

sweetgum xylan hydrolysate, which has been developed as a commercial ethanologenic

biocatalyst.

The specific PDC activities measured in transformed E. asburiae strains were noticeably

lower than those measured in the engineered K. oxytoca M5A1 (Ohta et al. 1991a), possibly due

to lower copy number of the plasmids pLOI297 and pLOI555 in E. asburiae JDR-1. However, as

found with engineered K. oxytoca strains, E. asburiae JDR-1 pLOI297 had higher activity than

pLOI555, which may be due to the presence of the ColEl replicon in pLOI297 resulting in a

higher copy number than in the strain transformed with pLOI555. It was found that E. asburiae

El (pLOI555) with highest ethanol yield in hydrolysate had the lowest PDC activity in the

glucose culture.

The contribution of the adh gene from pLOI1555 is likely critical to homoethanol

production in E. asburiae El as it was in initially generating the ethanologenic strains in E. coli

(Ingram and Conway, 1988; Ingram et al. 1987). When selected genes were deleted in E.

asburiae JDR-1 to produce lactate as the predominant product from E. asburiae L1,

fermentations were slow and incomplete without supplementation with Luria Bertani medium

(chapter 3), supporting the conclusion that efficient fermention to a targeted product requires

high level of expression of the gene encoding the oxido-reductase responsible for generating that

final fermentation product during the reoxidation ofNADH.

Plasmid stability is critical for biocatalysts engineered with genes conferring a desired

metabolic potential confined within a plasmid, as consistent traits are required for long-term

applications. The plasmid pLOI297, containing colEl replicon, was present in high copy numbers

in E. coli strains, but was unstable in K. oxytoca M5A1. pLOI555 derived from cryptic









low-copy-number plasmids in E. coli B (ATCC 11303), however, was very stable in K. oxytoca

M5Al(Ohta et al. 1991b). Similar to the studies in K. oxytoca, pLOI555 plasmids were found to

be more stable than pLOI297 in E. asburiae JDR-1. The relative stability of the plasmid in E.

asburiae El (pLOI555) recommend it for further development, perhaps through introduction of

the pdc and adh genes into the chromosome as has been achieved for the successful development

of E. coli KO 11 and its derivatives as ethanologenic biocatalysts (Jarboe et al. 2007).











Table 4-1. Bacterium strains and plasmids for engineering ethanologenic E. asburiae.
Strain and Relevant characteristics Source or reference
plasmid
strains
E.coli Top 10 For general cloning Ivitrogen
E.coli KO11 pfl::(pdc-adhB-cat) Afrd (Ohta et al., 1991a)
E. asburiae Wild type This paper
JDR-1
E. asburiae 4666 rrlE:: (pdc-adhA-adhB-kan) This chapter
E. asburiae 4672 pfl:: (pdc-adhA-adhB-kan) This chapter
E. asburiae El Enterobacter asburiae JDR-1ApflB This chapter
Plasmids
PLOI3240 Amr red, red recombinase protein (Wood et al. 2005)
pLOI297 Tcrpdc+ adhB+ (Ingram et al., 1989)
pLOI555 Cmrpdc adhB+ (Ohta et al., 1991b)
pLOI4666 Kan cat; pflB '-pdc-adhA-adhB-FRT-Kan-FRT-pflB" This chapter by Zhang
pLOI4672 Kan cat; rrlE '-pdc-adhA-adhB-FRT-Kan-FRT-rrlE" This chapter by Zhang
pLOI4162 bla cat; cat-sacB cassette (Jantama et al., 2008)
pCR 4-TOPO bla kan amp; TOPO TA cloning vector Invitrogen
pTOPOpfl pflB (PCR) amplified from E. a. JDR-1 and cloned Chapter 3
into PCR4-TOPO vector
pTOP04162pfl cat-sacB cassette cloned into pflB in pTOPOpfl Chapter 3
pTOPODpfl PCR fragment amplified from pTOPOpfl, kinase Chapter 3
treated, and self-ligated












Table 4-2. Comparison of sugar fermentation products of wild type and genetically engineered E.
asburiae JDR-1.a
Fermentation products (mM)
Fermentations 2,3-butane Ethanol yield(%
Succinate Lactate Formate Acetate Ethanol of theoret
diol oftheoretical)
2.5% glucose
E. asburiae JDR-1c 2.0 9.6 39.1 1.0 45.9 45.0 25.6
E. asburiae 4666' 0.3 36.5 3.2 4.0 15.3 31.8 28.7
E. asburiae 4672c 9.2 11.4 10.6 16.0 21.5 35.6 27.9
E. asburiae
E1297) 1.8 4.7 9.4 3.8 0 261.6 94.1
JDR-1(pLOI297)
E. asburiae
E1r-p 5) 1.6 2 7.7 3.4 0 265 95.3
JDR-1(pLOI555)
2%xylose
E. asburiae JDR-1c 12.7 5.6 15.0 25.2 13.4 42.6 35.6
E. asburiae
E.Dbri5 2.2 1.2 3.6 4.2 0 217.4 98.0
JDR-l(pLOI555)
a) Fermentations were carried out at 300C in ZH minimal media for 48 hours as described in the
Materials and Methods section.
b) Percentage of amount of ethanol produced to a theoretical maximal amount. A yield of 100%
is defined as 2 mole ethanol/ mole glucose or 5 mole ethanol/ 3 mole xylose.
c) E. asburiae JDR-1, E. asburiae 4666 and E. asburiae 4672 did not completely utilize the
substrates within 48 hours.

Table 4-3. Fermentation products from acid hydrolysates of sweetgum xylan."
Fermentation products (mM)
Formic Acetic Ethanol yield(% of
Ethanol b
acid acid theoretical)b
E. asburiae JDR-1 4.90.4 20.00.7 15.60.8 442
E.coli KO 11 5.91.0 10.60.3 22.50.2 631
E. asburiae 4.00.4 13.50.5 26.71.0 753
(pLOI55)
E. asbuiae 3.80.3 9.90.3 30.01.5 845
(pLOI297)
E.as iaEl 0 4.50.2 35.51.1 993
(pLOI555)
a) Fermentations were carried out at 300C in ZH minimal media for 48 hours as described in the
Materials and Methods section. Results are averages of 3 experiments.
b) percentage of amount of ethanol produced relative to the theoretical maximum. A yield of
100% is defined as 2 mole ethanol/ mole glucose or 5 mole ethanol/ 3 mole xylose.











Table 4-4. Specific consumption rates and specific production rates in acid hydrolysates of
sweetgum xylan (5g /liter)a.

Strains q Xylose q MeGAX q Acetate q Ethanol

E. asburiae JDR-1 0.330.04 0.0870.012 0.130.01 0.0600.009
E.coliKO11 0.380.04 0 0.110.01 0.0740.006
E. asburiae JDR-1 (pLOI555) 0.290.03 0.0580.012 0.140.02 0.0520.004
E. asburiae El (pLOI555) 0.320.03 0.0770.013 0.0220.003 0.11+0.01
a) q xylose is defined as consumed g xylose /g DCW(dry cell weight) /h; q MeGAX is defined as
consumed g MeGAX /g DCW(dry cell weight) /h; q acetate is defined as produced g acetate /g
DCW(dry cell weight) /h; q ethanol is defined as produced g ethanol /g DCW(dry cell weight)/h.
Results are averages of 3 experiments.


Table 4-5. Specific activity of PDC in cell crude extract from E. asburiae JDR-1 derived strains.a
Strains Specific Activity(Ub/mg of cell protein)


E. asburiae JDR-1 0


E. asburiae 4666 0
E. asburiae 4672 0
E. asburiae JDR-1 (pLOI297) 1.020.12
E. asburiae JDR-1 (pLOI555) 0.770.13
E. asburiae El (pLOI555) 0.530.10
a) Results are averages of 3 experiments.
b) One U is defined as the amount of the enzyme that catalyzes the conversion of 1 Itmole of
substrate per minute at room temperature.


Table 4-6. Plasmid stability of pLOI297 and pLOI555 in E. asburiae JDR-1. Results were
averages of 3 experiments.
Plasmids %cells retaining antibiotic resistance
After 36 generations After 72 generations
pLOI297 29.51.3 10.72.6
pLOI555 100.02.8 98.1+11.8


















2000Initial medium


100.00


S0.00


S -E. asburiae JDR-1
50.00


) 0.00
E. coli KO11

0.00

-E. asburiae El
(POL1555)




5.00 1000 1500 20.00 25.00 30.00
Mnrtes

Figure 4-1. HPLC profile of fermentation products of E. asburiae JDR-1, E.coi KO 11 and E.
asburiae El (pLOI555) in 0.5% sweetgum xylan hydrolysate with MOPS. (The
unlabeled peaks with retention time of 11 minutes and 21 minutes were for salts and
buffers.)
































0 10 20 30 40 50 60 70


Time (hours)


E
C 15


10
0
o

0 5


0


0 10 20 30 40
Time (hours)


50 60 70


Figure 4-2. Fermentation time course of E. asburiae JDR-1 A), E.coli KO 11 B), E. asburiae
JDR-1 (pLOI555) C) and E. asburiae El (pLOI555) D) in media of buffered
sweetgum xylan hydrolysate. Substrates and fermentation products: xylose (4),
MeGAX (m), acetic acid ( A), formic acid (o), ethanol (o).













30

25

E 20

S15

o 10
0
0
5

0


0 10 20 30 40
Time (hours)


40

35

S30
E
25
o


8 15

0 10
5

0


50 60 70


0 10 20 30 40
Time(hours)


50 60 70


Figure 4-2. Continued











CHAPTER 5
GENETIC DEFINITION OF THE METHYLGLUCURONOXYLOSE UTILIZATION
PATHWAY IN ENTEROBACTER ASBURIAE JDR-1

Introduction

Currently an accepted hypothetical process of degradation of MeGAX3 is that MeGAX3

enters the cell via a specific transporting system (Gilson et al. 1988, Shulami et al. 1999). Inside

the cell, glucoronoxylotriose is hydrolyzed at the a-1,2 glucuronosyl linkage by an

a-glucuronidase, releasing xylotriose and D-glucuronic acid. Xylotriose is hydrolysed to xylose

by P-xylosidase. When and how the methyl group is released from MeGAX3 is not known

(Shulami et al. 1999). All the a-glucuronidases share homology in amino acid sequences,

belonging to GH family 67 (CAZy database, http://afmb.cnrs-mrs.fr/CAZY/). Paenibacillus

JDR-2 was determined to utilize methylglucuronoxylotriose (MeGAX3),

methylglucuronoxylobiose (MeGAX2) and MeGAX, probably by the same metabolic pathways.

The a-glucuronidase from Paenibacillus sp. JDR2 has been characterized and found to have

hydrolase activity with MeGAX, as well as MeGAX2 and MeGAX3 (Nong 2009). Whether E.

asburiae strain JDR-1 uses a similar or a novel pathway, and what enzymes and transporters

constitute the pathway is the main subject of this chapter.

Materials and Methods

Degenerate Primer Method to Determine Presence of GH 67 genes

Standard methods were used for most of the genetic manipulations as described in chapter 3.

Several a-glucuronidase genes from GH family 67 (CAZy database,

http://afmb.cnr-smrs.fr/CAZY/), were selected and input into Codehop

(http://blocks.fhcrc.org/blocks/code-hop.html ), a web based software to find conserved domains

from several homologous sequences and design degenerate primer sets. These sets were applied









to genomic DNA of E. asburiae strain JDR-1 to amplify sequences homologous to the select set

of a-glucuronidase genes with the Touch Down PCR conditions (Schunck et al. 1995).

Determination of Genes Involved in MeGAX Utilization Process with Mutant Library
Method

Tn5 transposomes (EZ-Tn5TM TransposomesTM, Epicentre, US) were transformed into E.

asburiae JDR-1 with electroporation method to generate a mutant library of about ten thousand

colonies. An antibiotic counter enrichment method using D-cycloserine was employed to

accumulate mutants that were unable to grow on MeGAX but could grow on xylose or mannitol

minimal media. Enriched cultures were streaked on xylose or mannitol minimal medium and

copied to MeGAX minimal agar plates to identify MeGAX defective mutants. The selected

defective mutants were inoculated in liquid MeGAX minimal medium to confirm the phenotype.

The genomic DNA of selected MeGAX defective mutants was extracted, fragmented and

self-ligated. The ligated DNA circle contained Tn5 insertion cartridge including replicator and

kanamycin resistant genes, which constituted a plasmid capable of replication. These plasmids

were transformed into E.coli and sequenced to determine the genes that were interrupted by the

Tn5 transposon. The Blastx program of NCBI was used to obtain Amino acid sequences most

similar to the defective genes in MeGAX defective mutants to identify their function. The

relative growth rates of the E. asburiae JDR-1 ptsT mutant were measured in 0.1% MeGAX,

mannitol, maltose, rhamnose, mannose, glucuronate and xylose minimal media with or without

ImM cAMP, and compared with wild type E. asburiae JDR-1. Two of the MeGAX defective

mutants were determined to be xylA- and xylB mutants. The MeGAX minimal medium growing

with xylA- was analyzed with HPLC after 96 hours.









Determination of MeGAX Utilization Genes by Transforming E. coli with Cosmid Library
and Topo Plasmid Library

Genomic DNA was extracted from 15 ml overnight cultures of E. asburiae JDR-1. The

cell culture was washed and lysed; then DNA was precipitated with CTAB (Delsal et al. 1989).

Phenol/chloroform was used to extract genomic DNA, which was pushed through Hamilton 50

1tl syringe with 22s needle 30 times to shear and generate fragments. After an end repairing

reaction for the sheared fragments with Klenow enzyme (NEB), 1% low melting point agarose

gel was used to separate the DNA fragments with Field Inversion Gel Electropherisis (FIGE)

(Chow 2007, Guang 2009). The 30 to 40 kb fragments were cut out and ligated with pWEB-TNC

vetor, packaged with the MaxPlax Packaging Extracts and trasfected E.coli EPI100-T1R

(EPICENTRE Biotechnologies, USA). The obtained transformants were selected on 50 [g/ml

Ampicillin LB plates. Similarly, an E. asburiae JDR-1 genomic DNA library of 4kb to 6 kb

fragments ligated to pCR4-Topo (Invitrogen, USA) were established in E.coli TOP 10. The E.coli

TOP 10 transformed with E. asburiae strain JDR-1 genomic fragments were selected to obtain

transformants able to grow in MeGAX minimal medium.

Obtaining Candidate Genes by Sequencing Cosmid Containing P-Xylosidase Gene and
Genome Comparison

E. asburiae JDR-1 cosmid library consists of E.coli carrying pCC1 vector with 33 to 45 kb

fragments were incubated with 4-methylumbelliferyl 3-xylopyranoside (MUX) on LB agar plates.

After exposure to UV, the clones (obtained by Guang Nong in 2003) that degrade MUX and

showing fluorescence were identified to possess P-xylosidase gene (Christakopoulos et al. 1996).

The cosmid containing P-xylosidase was sequenced by the shotgun method and resulted in 14

contigs after assembly. NCBI Blastx program was used to suggest functions of genes in these

contigs.









Enterobacter 638 was a newly sequenced Enterobacter strain

(http://genome.j gi-psf org/finishedmicrobes/ent_6/ent_6.home. html), the sequences of which

showed highest similarity among all the bacteria with the known sequences of E. asburiae JDR-1.

Because of the similarity, we compared its genome to genome of E. coli K12 to obtain hydrolase

genes. About 50 hydrolase genes present in the annotated Enterobacter 638 genome but not in

E.coli K12 along with several transporter genes were obtained. The annotated functions of these

genes were evaluated with CAZy to select candidate genes for deletion to confirm involvement

in MeGAX utilization process.

Deletion of Candidate Genes to Confirm Their Involvement in MeGAX Utilization Process

The presence of candidate genes in E. asburiae JDR-1 was determined by PCR with

degenerate primers or PCR with specific primers. Partial sequences of the candidate genes were

cloned into Topo vector and ligated to cat-sacB fragment to construct DNA fragments to delete

their counterpart genes in E. asburiae JDR-1 genome using the gene deletion method in chapter

3. The deletion strains were grown in MeGAX minimal medium to determine whether the

specific gene deletion affects the MeGAX utilization phenotype.

Results

MeGAX Defective Mutants ofE. asburiae JDR-1

No PCR product was obtained with degenerate primers for a-glucuronidase genes. This

result suggested the absence of a-glucuronidases in GH family 67 in E. asburiae JDR-1.

Anohter supporting evidence is the phenotype that E. asburiae JDR-1 is only able to utilize

MeGAX but not MeGAX2 or MeGAX3, since all GH67 a-glucuronidases that have been

characterized are able to hydrolyze the a-1,2 glycosidic linkage in MeGAX3. Therefore, E.

asburiae JDR-1 may contain a novel enzyme with activity to hydrolyze MeGAX or a new

process to degrade MeGAX.









To determine genes required in the process of MeGAX utilization, more than 10,000

insertion mutants with Tn5 transposon of E. asburiae JDR-1 were obtained to select for MeGAX

defective mutants. After repeating experiments of rounds of counter enrichment using xylose and

MeGAX with D-cycloserine, about 700 mutants were screened for their ability to grow on

MeGAX minimal agar plates, of which 52 MeGAX utilization defective mutants were selected

and screened in liquid MeGAX minimum medium to confirm the phenotype. As much as 16

individual MeGAX defective mutants were isolated. Of those mutants, 15 were sequenced to be

phosphotransferase system enzyme I defective mutants (pstT mutant), in which 14 mutants had

same insertion site within ptsl (Genebank accession number: FJ527306) and 1 mutant had a

different insertion site within the same gene. One of the defective mutants was determined to

have an insertion mutation in a gene with suggested function as an oxireductase domain protein.

However, no characterized gene was homologous to this oxi gene (Genebank accession number:

FJ527305). Therefore, the specific function of this gene is unknown.

Phosphotransferase system enzyme I (PTS EI) is also named phosphoenolpyruvate-protein

kinase, which transfers a phosphate group from PEP to itself then to HPr. HPr transfers the

phosphate group to phospho-III ( or II-A), which is specific to different substrates. PTS El is also

a key enzyme in the catabolite repression system. In the phosphorylated form, II-A-P activates

adenylcyclase and produces cAMP, which binds to CAP (catabolite activator protein), and the

CAP-cAMP complex activates multiple operons encoding enzymes utilizing different non-PTS

sugars. In dephosphorylated form, II-A binds to permeases and blocks the transportation of some

substrates (Figure 5-1) (Saier et al. 1976).

Three possible reasons may explain why E. asburiea strain JDR-1 ptsT mutants were

defective in the utilization of MeGAX: 1) MeGAX may be transported by the PTS system, and









its specific EII was unable to obtain the phosphate group from El in theptsT mutant, in which

case MeGAX is classified as a PTS sugar. 2) The permease for MeGAX may be inhibited by

binding of the non-phpsphorylated form of the II-A protein, in which case MeGAX is a non-PTS

sugar. 3) One or more genes encoding proteins required for utilization of MeGAX may be under

positive regulation of CAP-cAMP complex, and the phosphorylated form of IIA required to

activate of the adenylcyclase was not available in the ptsT mutant. MeGAX is a non-PTS sugar

in this case. To determine the transportation and regulation of MeGAX in E. asburiea JDR-1, the

growth rate of ptsT mutant was measured in MeGAX, mannitol, maltose, rhamnose, mannose,

glucuronate and xylose minimal media with or without ImM cAMP, and compared with wild

type (Table 5-1). Mannitol is a PTS sugar in E. coli. With or without cAMP, E. asburiea JDR-1

ptsl mutant was not able to grow with mannitol as sole carbon source, suggesting that mannitol

was also imported with PTS in E. asburiea JDR-1. Maltose is a non-PTS sugar in E. coli and

regulated by CAP-cAMP complex. The PtsJ mutant grew normally with ImM cAMP in maltose

while it grew slowly in the absence of cAMP, suggesting maltose was a non-PTS sugar in E.

asburiea JDR-1 and under regulation of CAP-cAMP. The growth patterns of the ptsT mutant in

maltose and in MeGAX were similar. The ptsT mutant demonstrated a poor growth rate in the

medium of MeGAX with ImM cAMP, but did not grow without cAMP. These results suggested

that in E. asburiea JDR-1, MeGAX was a non-PTS sugar, the utilization of which was repressed

by PTS sugars with the CAP-cAMP catabolite regulation.

To obtain MeGAX defective mutants other than the dominant ptsT mutants, mannitol was

used as growth substrate with MeGAX to counter enrich MeGAX defective mutants with

D-cycloserine. Two genetically different mutants were obtained; the first one had a Tn5

transposon insertion at xylose isomerase (xylA) and the second one at xylulokinase (xylB), both









of which were actually xylose defective mutants. These two mutants were unable to utilize

MeGAX due to their inability to utilize xylose. Their inability to utilize xylose was presumably

responsible for their slow growth in MeGAX. The xylA- mutant accumulated 4.1 mM xylose

(quantified by HPLC) after 72 hr growth in 3.9mM MeGAX minimal medium (Fig 5-2). This

result suggested that all xylose released from MeGAX in the process of MeGAX utilization was

not used by xylA- mutant and accumulated in the medium. Although the methylglucuronate part

was utilized by the mutant, the accumulation of xylose slowed the process and decreased the

growth rate. This result, along with the analysis of fermentation products and the substrate

preferential utilization experiment, indicated that MeGAX was degraded into xylose and

methylglucuronate (MeGA).

Candidate Genes Determined From Cosmid and Plasmid Library Transformation Method

E. coli TOP 10 host was transformed by the 4 to 6 kb TopoPCR4 library. The transformants

were selected upon MeGAX minimal medium, no E.coli TOP 10 transformant able to grow was

obtained Similarly, no E. coli EPI100 transformants carrying the pWEB-TNC cosmid with 30

to 40 kb fragment was able to grow in MeGAX minimal medium. These results suggest that the

essential genes involved in MeGAX utilization might be located at different positions in the

genome ofE. asburiae strain JDR-1 separated by significant distance, so that the 30 to 40 kb

fragment could not contain all the essential genes that might enable E. coli to utilize MeGAX.

Alternatively, unknown regulation factors may be missing that are required for expression of the

E. asburiae genes in E. coli EPI100.

Obtaining Candidate Genes for Function Confirmation by Gene Deletion

14 sequence contigs were assembled after shotgun sequencing of the cosmid containing

P-xlosidase. Some candidate genes with functions related to xylose oligomer utilization,

especially MeGAX utilization, were expected to be obtained from the contigs. However, only an









interesting gene was identified in one contig, which is a sequence of about 600 base pairs

(Genebank accession number: FJ527309) immediately located before the P-xlosidase gene. The

highest homologous gene for this fragment is a sugar (Glycoside-Pentoside-Hexuronide)

transporter gene from Enterobacter 638, which could possibly be the transporter for MeGAX.

An interesting gene that Enterobacter 638 contains but absent in E. coli K12 is a GH88

family gene listed in CAZy. A characterized enzyme in this family is a A -4,5 unsaturated

-glucuronyl hydrolase (EC 3.2.1.-). This enzyme recognizes A -4,5 unsaturated uronic acid at

the nonreducing terminal of polysaccharide and hydrolyzes the a-linkage between the

unsaturated uronic acid and sugars at the reducing end (Fig 5-3) (Hashimoto et al. 1999). In the

compound MeGAX, if the methyl group was removed by a 3-elimination reaction from the

glucuronic acid residue, a A -4,5 unsaturated-glucuronyl-xylose would be produced. A A -4,5

unsaturated -glucuronyl hydrolase from GH88 may recognize the A -4,5 unsaturated-glucuronyl

residue and hydrolyze the a-1,2 linkage between the A -4,5 unsaturated-glucuronyl residue and

xylose residue. PCR with degenerate primers was able to amplify a GH88 family gene

(Genebank accession number: FJ527307) from E. asburiae JDR1 genomic DNA.

Two more glycoside hydrolase genes, a lytic transglycosylase (Genebank accession

number: FJ527310) and a glycotransferase (Genebank accession number: FJ527308), were

amplified from E.asburiae JDR-1 genomic DNA with degenerate primers based on GH67

(a-glucuronidases) genes. However, both genes were not homologous with GH67 genes

indicated by Blastx program of NCBI.

The four genes were individually deleted in E. asburiae JDR-1. However, all the

knock-out mutants maintained the ability to utilize MeGAX in minimal medium, which









suggested these genes may not be involved in MeGAX utilization or other genes in E. asburiae

JDR-1 may back up the function of these genes.

Discussion

To define the MeGAX utilization pathway in E. asburiae JDR1, several genetic methods

was employed. Four genes, oxi, ptsTl, xylA-, xylB- were determined to be involved by screening

the single insertion mutant library of E. asburiae JDR1. Along with previous physiological

research results, such as rates of substrate utilization, product formation and molar growth yields,

the genetic information indicated that MeGAX was transported into the cell and hydrolyzed to

release methanol, glucuronic acid and xylose. The drawback of this method is the low efficiency

to obtain hydrolase genes, since too many genes were involved in MeGAX utilization process

and a specific dominant gene deletion would be obtained most often after the counter enrichment.

The possibility to obtain all genes or the specific glycohydrolyse gene by this method is very

low.

Assuming E. asburiae JDR-1 genes were able to be expressed in E. coli with their original

transcription and translation elements, the cosmid and plasmid library transformation method

would be a simple and effective method. However, the results turned out to be otherwise. Except

for the possibility of poor gene expression, the essential genes for MeGAX utilization may be

located at different positions in the genome and can not be contained within single 30 to 40 kb

fragments cloned into the cosmid library.

No E. asburiae JDR1 strain with selected candidate gene deletion was defective in

MeGAX utilization. This result did not necessarily mean that these genes were not involved in

this process. For example, the genomic information of Enterobacer 638 indicated that this

bacterium had two GH88 genes located at different positions. Thus some of these deleted









candidate genes may have an allele which could complement its function to enable E. asburiae

JDR1 to utilize MeGAX.

The complexity ofE. asburiae JDR1 genome made it difficult to obtain the whole pathway

of MeGAX utilization or the essential genes. The future plan is to compare the changes in

transcriptome of E. asburiae JDR1 growing in MeGAX and other substrartes. The genes

up-regulated when E. asburiae JDR1 utilizes MeGAX as main carbon source may represent the

genes which constitute the pathway of MeGAX utilization and genes associated with MeGAX

utilization.


100












Table 5-1. Growth rate of ptsT in different substrate with and without cAMP
Source of carbon cAMP WT growth rate PTSTGrowth rate
ImM ++++a +
MeGAX ++++b

MatI1mM ++++ 0
Mannitol
0 ++++ 0
1 mM ++++ ++++
Maltose m .. .
0 ++++ +
Rhamnose 0 ++++ ++++
Mannose 0 ++++ ++++
Glucuronate 0 ++++ ++
Xylose 0 ++++ ++++
a: Relative growth rate was indicated with "+", "++++" indicates the fastest and "+" indicates the
slowest.
b: "0" indicates no growth.





























vaton of at Factor x


Activation of catabokc genes -


Nature Reviews I Microbiology


Figure 5-1. Function and regulation of PTS system (Boris Gorkel & Jorg Stilke 2008) with
permission from Nature Publishing Group. Copyright @ Nature Publishing Group


100.oo- v0"^/\ LH salts
ZH al




ZH salts
SZHsalts


MeGAX minimal
medium


MeGAX minimal medium
after aerobic growth of
E. asburiae JDR-1



MeGAX minimal medium
after aerobic growth of
E.asburiae JDR-1 xylA


Figure 5-2. HPLC profile of aerobic culturing media of E. asburiae JDR-1 xylA.







102


s .Do


ro.oo Is.oo
ular


2n.on 2s.00


30-D D


:j Ih,.



























Figure 5-3. Enzyme activity of GH88 enzyme unsaturated -glucuronyl hydrolase (EC 3.2.1.-)
(Hashimoto et al., 1999) with permission from Elsevier Ltd. Copyright @ Elsevier
Ltd











CHAPTER 6
CONCLUSION

Dilute acid pretreatment is a well established method for hydrolyzing the

methylglucuronoxylans in the hemicellulose fraction from lignocellulosics to release fermentable

xylose. Xylose substituted with a-1,2-methylglucuronate is released as methylglucuronoxylose

(MeGAX), which cannot be fermented by biocatalysts currently used to produce biofuels and

chemicals. E. asburiae JDR-1, isolated from colonized hardwood (sweetgum), efficiently

ferments both methylglucuronoxylose and xylose, producing predominantly ethanol and acetate.

13C-NMR studies defined the Embden-Meyerhof pathway for metabolism of glucose and the

pentose phosphate pathway for xylose metabolism. Rates of substrate utilization, product

formation and molar growth yields indicate methylglucuronoxylose is transported into the cell

and hydrolyzed to release methanol, xylose and hexauronate for further catabolism. Enterobacter

asburiae stain JDR-1 is the first microorganism described that ferments methylglucuronoxylose

generated along with xylose by the acid mediated saccharification of hemicellulose.

Deletion ofpflB and als genes in this bacterium modified the native mixed acid

fermentation pathways into pathway for homolactate production. The resulting strain, E.

asburiae L1, completely utilized both xylose and MeGAX in a dilute acid hydrolysate of

sweetgum xylan and produced optically pure D(-)lactate at a yield approximating 100% of the

theoretical maximum. With the ability to utilize MeGAX in xylan hydrolysates, as well as the

other saccharides released during dilute acid pretreatment, the engineered strain provides a

uniquely efficient bacterial biocatalyst for lactate production from the hemicellulose fractions of

lignocellulosic biomass.


104









The PET operon including Zymomonas mobilis pyruvate decarboxylase (pdc) and alcohol

dehydrogenase B (adhB) genes was transferred into E. asburiae JDR-1, converting the native

mixed acid fermentation pathways to homoethanol production. Integration of PET operon into

JDR-1 chromosome did not increase the ethanol yield, which may be due to insufficient PDC

expression. In contrast, expressing pdc and adhB in a stable plasmid pLOI555 significantly

increased the ethanol production. Deletion of the pyruvate formatelyase (pflB) gene further

increased the ethanol yield, resulting E. asburiae El (pLOI555). This strain completely utilizes

both xylose and MeGAX in dilute acid hydrolysate of sweetgum xylan and produced ethanol

with a yield 99% of the theoretical maximum at the rate of 0.11 g ethanol/g DCW/h, which was

1.57 the yield and 1.48 the rate obtained with ethanologenic Escherichia coli KO 11. This

engineered strain of E. asburiae JDR-1, which is able to ferment the predominant hexoses and

pentoses derived from both hemicellulose and cellulose fractions, offers a promising subject for

development as an ethanologenic biocatalyst for production of fuels and chemicals from

agricultural residues and energy crops.

To define the genetic basis in MeGAX utilization, an E. asburiae JDR-1 mutant library

generated by Tn5 transposon mutagenesis was screened for MeGAX-defective mutants. The

mutant strains obtained were identified as oxi, ptsl, xylA-, xylB-. Growth status ofptsT in

different carbon sources with and without 1 mM cAMP suggested MeGAX was transported in E.

asburiae JDR-1 through a non-PTS pathway and under regulation by the cAMP-CAP catabolism

repression system. The slow growth of xylose defective mutants, xylA- and xylB-, in MeGAX

with accumulation of free xylose in the growth medium suggests that xylose is released in the

process of MeGAX utilization. Along with previous physiological studies, such as rates of

substrate utilization, product formation and molar growth yields, the genetic information









indicates MeGAX is transported into the cell and hydrolyzed to release methanol, glucuronic

acid and xylose. Several other genetic methods, such as the PCR by degenerate primers, positive

gene selection and specific deletion of candidate gene, were also applied to define the MeGAX

utilization pathway in E. asburiae JDR-1. However, these methods failed to identify specific

genes involved in MeGAX utilization. The experiment to analyze transcriptome changes when E.

asburiae JDR1 utilizes MeGAX as sole carbon source will be performed for obtaining

information of genes constituting the pathway of MeGAX utilization.


106









REFERENCE LIST


Asis CA, Adachi K (2004) Isolation of endophytic diazotroph Pantoea agglomerans and
nondiazotroph Enterobacter asburiae from sweetpotato stem in Japan. Letters in Applied
Microbiology 38:19-23

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 (Xyll)
from the xylose-assimilating yeast Pichia stipitis. Gene 109:89-97

Babson AL, Babson SR (1973) Kinetic colorimetric measurement of serum lactate
dehydrogenase activity. Clin Chem 19:766-769

Bauchop T, Elsden SR (1960) The growth of micro-organisms in relation to their energy supply.
J Gen Microbiol 23:457-469

Beg QK, Kapoor M, Mahajan L, Hoondal GS (2001) Microbial xylanases and their industrial
applications: a review. Appl Microbiol Biotechnol 56:326-338

Boris Gorkel & Jorg Stiulkel (2008) Carbon catabolite repression in bacteria: many ways to
make the most out of nutrients. Nat Rev Microbiol 6: 613-624

Brenner DJ, Mcwhorter AC, Kai A, Steigerwalt AG, Farmer JJ (1986) Enterobacter asburiae
Sp-Nov, a new speciesfound in clinical specimens, and reassignment of Erwinia dissolvens
and Erwinia nimipressuralis to the genus Enterobacter as Enterobacter dissolvens
Comb-Nov and Enterobacter nimipressuralis Comb-Nov. J Clin Microbiol 23:1114-1120

Chow V, Nong G, Preston JF (2007) Structure, function, and regulation of the aldouronate
utilization gene cluster from Paenibacillus sp. strain JDR-2. J Bacteriol 189: 8863-8870

Christakopoulos P, Mamma D, Nerinckx W, Kekos D, Macris B, Claeyssens M (1996)
Production and partial characterization of xylanase from Fusarium oxysporum. Bioresour
Technol 58:115-119

Converse AO, Kwarteng IK, Grethlein HE, Ooshima H (1989) Kinetics of thermochemical
pretreatment of lignocellulosic materials. Appl Biochem Biotechnol 20-1:63-78

Conway T, Sewell GW, Osman YA, Ingram LO (1987) Cloning and sequencing of the alcohol
dehydrogenase II gene from Zymomonas mobilis. J Bacteriol 169:2591-2597

Datsenko KA, Wanner BL (2000) One-step inactivation of chromosomal genes in Escherichia
coli K-12 using PCR products. Proc Nat Acad Sci U.S.A. 97:6640-6645

de Vries RP, Visser J (2001) Aspergillus enzymes involved in degradation of plant cell wall
polysaccharides. Microbiol Molecul Bio Rev 65:497-522









Delsal G, Manfioletti G, Schneider C (1989) The CTAB DNA precipitation method a common
mini-scale preparation of template DNA from phagemids, phages or plasmids suitable for
sequencing. Biotech 7:514-520

Dien BS, Cotta MA, Jeffries TW (2003) Bacteria engineered for fuel ethanol production: current
status. Appl Microbiol Biotechnol 63:258-266

Dien BS, Jung HJG, Vogel KP, Casler MD, Lamb JFS, Iten L, Mitchell RB, Sarath G (2006)
Chemical composition and response to dilute-acid pretreatment and enzymatic
saccharification of alfalfa, reed canarygrass, and switchgrass. Biomass and Bioenergy
30:880-891.

Dombek KM, Ingram LO (1987) Ethanol production during batch fermentation with
Saccharomyces cerevisiae changes in glycolytic enzymes and internal Ph. Appl Environ
Microbiol 53:1286-1291

Dubois M, Gilles KA, Hamilton JK, Rebers PA, Smith F (1956) Colorimetric method for
determination of sugars and related substances. Anal Chem 28:350-356

Eggeman T, Elander RT (2005) Process and economic analysis of pretreatment technologies.
Bioresour Technol 96:2019-2025

Elbeltagy A, Nishioka K, Sato T, Suzuki H, Ye B, Hamada T, Isawa T, Mitsui H et al. (2001)
Endophytic colonization and in plant nitrogen fixation by a Herbaspirillum sp. isolated
from wild rice species. Appl Environ Microbiol 67: 5285-5293.

Eissen M, Metzger JO, Schmidt E, Schneidewind U (2002) 10 years after Rio Concepts on the
contribution of chemistry to a sustainable development. Angewandte Chemie International
Edition 41:414-436

Esteghlalian A, Hashimoto AG, Fenske JJ, Penner MH (1997) Modeling and optimization of the
dilute-sulfuric-acid pretreatment of corn stover, poplar and switchgrass. Bioresour Technol
59:129-136

Freer SN, Detroy RW (1983) Characterization of cellobiose fermentations to ethanol by Yeasts.
Biotechnol Bioeng 25:541-557

Galbe M, Zacchi G (2002) A review of the production of ethanol from softwood. Appl Microbiol
Biotechnol 59:618-628

Gao K, Mckinley KR (1994) Use of macroalgae for marine biomass production and Co2
remediation a review. J Appl Phycol 6:45-60

Gilson E, Alloing G, Schmidt T, Claverys JP, Dudler R, Hofnung M (1988) Evidence for
high-affinity binding protein dependent transport systems in gram positive bacteria and in
Mycoplasma. Embo J 7:3971-3974

Gray KA, Zhao LS, Emptage M (2006) Bioethanol. Curr Opin Chem Biol 10:141-146









Gunsalus IC, Horecker BL, Wood WA (1955) Pathways of carbohydrate metabolism in
microorganisms. Bacteriol Rev 19:79-128

Gunsalus IC (1961) Energy yielding metabolism in bacteria. In: Gunsalus IC and Stanier RY (ed)
the Bacteria. Academic Press Inc., New York, pp 1-58

Gutierrez T, Buszko ML, Ingram LO, Preston JF (2002) Reduction of furfural to furfuryl alcohol
by ethanologenic strains of bacteria and its effect on ethanol production from xylose. Appl
Biochem Biotechnol 98:327-340

Hahn-Hagerdal B, Galbe M, Gorwa-Grauslund MF, Liden G, Zacchi G (2006) Bio-ethanol the
fuel of tomorrow from the residues of today. Trends Biotechnol 24:549-556

Hallmann J, QuadtHallmann A, Mahaffee WF, Kloepper JW (1997) Bacterial endophytes in
agricultural crops. Canadian Journal of Microbiology 43:895-914

Hashimoto W, Kobayashi E, Nankai H, Sato N, Miya T, Kawai S, Murata K (1999) Unsaturated
glucuronyl hydrolase of Bacillus sp GL1: Novel enzyme prerequisite for metabolism of
unsaturated oligosaccharides produced by polysaccharide lyases. Arch Biochem Biophy
368:367-374

Hasona A, Kim Y, Healy FG, Ingram LO, Shanmugam KT (2004) Pyruvate format lyase and
acetate kinase are essential for anaerobic growth of Escherichia coli on xylose. J Bacteriol
186:7593-7600

Himmel ME, Ding SY, Johnson DK, Adney WS, Nimlos MR, Brady JW, Foust TD (2007)
Biomass recalcitrance: Engineering plants and enzymes for biofuels production. Science
315:804-807

Hinman ND, Schell DJ, Riley CJ, Bergeron PW, Walter PJ (1992) Preliminary estimate of the
cost of ethanol production for SSF technology. Appl Biochem Biotechnol 34-5:639-649

Hoffert MI, Caldeira K, Benford G, Criswell DR, Green C, Herzog H, Jain AK, Kheshgi HS,
Lackner KS, Lewis JS, Lightfoot HD, Manheimer W, Mankins JC, Mauel ME, Perkins LJ,
Schlesinger ME, Volk T, Wigley TML (2002) Advanced technology paths to global climate
stability: Energy for a greenhouse planet. Science 298:981-987

Hoffmann H, Stindl S, Ludwig W, Stumpf A, Mehlen A, Heesemann J, Monget D, Schleifer KH,
Roggenkamp A (2005) Reassignment of Enterobacter dissolvens to Enterobacter cloacae as
E. cloacae subspecies dissolvens comb. nov and emended description of Enterobacter
asburiae and Enterobacter kobei. System Appl Microbiol 28:196-205

Hong J, Ye XH, Zhang YHP (2007) Quantitative determination of cellulose accessibility to
cellulase based on adsorption of a nonhydrolytic fusion protein containing CBM and GFP
with its applications. Langmuir 23:12535-12540

Hu ZH, Wen ZY (2008) Enhancing enzymatic digestibility of switchgrass by microwave-assisted
alkali pretreatment. Biochem Eng J 38:369-378









Hugouvieuxcottepattat N, Robertbaudouy J (1987) Hexuronate catabolism in Erwinia
c hil1 %(i/Ihini. J Bacteriol 169:1223-1231

Hurlbert JC, Preston JF (2001) Functional characterization of a novel xylanase from a corn strain
of Erwinia c hil) inheiiin. J Bacteriol 183:2093-2100

Ingram LO, Aldrich HC, Borges ACC, Causey TB, Martinez A, Morales F, Saleh A, Underwood
SA, Yomano LP, York SW, Zaldivar J, Zhou SD (1999) Enteric bacterial catalysts for fuel
ethanol production. Biotechnol Pro 15:855-866

Ingram LO, Alterthum F, Ohta K (1989) Efficient ethanol production from xylose, lactose and
glucose by recombinant Escherichia coli. Appl Environ Microbiol 55:1943-1948

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

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

Jacobs A, Larsson PT, Dahlman 0 (2001) Distribution of uronic acids in xylans from various
species of soft- and hardwood as determined by MALDI mass spectrometry.
Biomacromolecules 2:979-990

Jarboe LR, Grabar TB, Yomano LP, Shanmugan KT, Ingram LO (2007) Development of
ethanologenic bacteria. Biofuels 108:237-261

Jantama K, Haupt MJ, Svoronos SA, Zhang XL, Moore JC, Shanmugam KT, Ingram LO (2008)
Combining metabolic engineering and metabolic evolution to develop nonrecombinant
strains of Escherichia coli C that produce succinate and malate. Biotechnol Bioeng
99:1140-1153

Jeffries TW (2005) Ethanol fermentation on the move. Nat Biotechnol 23:40-41

Jeffries TW, Jin YS (2004) Metabolic engineering for improved fermentation of pentoses by
yeasts. Appl Microbiol Biotechnol 63:495-509

Jeuland N, Montagne X, Gautrot X (2004) Potentiality of ethanol as a fuel for dedicated engine.
Oil & Gas Science and Technology-Revue de 1 Institut Francais du Petrole 59:559-570

Jones JK, Purves CB, Timell TE (1961) Constitution of A 4-0-Methylglucuronoxylan from
Wood of Trembling Aspen (Populus TremuloidesMichx). Can J of Chem 39:1059-1066

Kardosova A, Matulova M, Malovikova A (1998) (4-0-methyl-alpha-D-glucurono)-D-xylan
from Rudbeckiafulgida, var. sullivantii (Boynton et Beadle). Carbohydr Res 308:99-105

Kheshgi HS, Prince RC, Marland G (2000) The potential of biomass fuels in the context of
global climate change: Focus on transportation fuels. Ann Rev Ener Environ 25:199-244









Kim Y, Ingram LO, Shanmugam KT (2007) Construction of an Escherichia coli K-12 mutant for
homoethanologenic fermentation of glucose or xylose without foreign genes. Appl Environ
Microbiol 73:1766-1771

Lal R (2004) Soil carbon sequestration impacts on global climate change and food security.
Science 304:1623-1627

Lawford HG, Rousseau JD (1991) Ethanol production by recombinant Escherichia coli carrying
genes from Zymomonas mobilis. Appl Biochem Biotechnol 28-9:221-236

Lawford HG, Rousseau JD (2002) Performance testing of Zymomonas mobilis metabolically
engineered for cofermentation of glucose, xylose, and arabinose. Appl Biochem Biotechnol
98:429-448

Lee YY, lyer P, Torget RW (1999) Dilute-acid hydrolysis of lignocellulosic biomass. Adv
Biochem Eng Biotechnol 65:93-115

Lin Y, Tanaka S (2006) Ethanol fermentation from biomass resources: current state and
prospects. Appl Microbiol Biotechnol 69:627-642

Linton KJ, Higgins CF (1998) The Escherichia coli ATP-binding cassette (ABC) proteins.
Molecul Microbiol 28:5-13

Lynd LR, Wyman CE, Gerngross TU (1999) Biocommodity engineering. Biotechnol Pro
15:777-793

Matsushika A, Sawayama S (2008) Efficient bioethanol production from xylose by recombinant
Saccharomyces cerevisiae requires high activity of xylose reductase and moderate
xylulokinase activity. J Biosci Bioeng 106: 306-309.

Matsushika A, Watanabe S, Kodaki T, Makino K, Inoue H, Murakami K, Takimura 0,
Sawayama S (2008). Expression of protein engineered NADP+-dependent xylitol
dehydrogenase increase ethanol production from xylose in recombinant Saccharomyces
cerevisiae. Appl Microbiol Biotechnol 81: 243-255

McInroy, J.A. and Kloepper, J.W. (1995) Survey of indigenous bacterial endophytes from cotton
and sweet corn. Plant and Soil 173: 337-342.

McMillan JD (1997) Bioethanol production: Status and prospects. Renewable Energy
10:295-302

Meshartree M, Dale BE, Craig WK (1988) Comparison of steam and ammonia pretreatment for
enzymatic hydrolysis of cellulose. Appl Microbiol Biotechnol 29:462-468

Miller EN, Jarboe LR, Yomano LP, York SW, Shanmugam KT, Ingram LO (2009) Silencing of
NADPH-dependent oxidoreductase genes (yqhD and dkgA) in furfural-resistant
ethanologenic Escherichia coli. Appl Environ Microbiol 75: 4315-4323









Moat AG, Foster JW, Spector MP (2002) Microbial Physiology, 4th ed. Wiley-Liss, New York

Monties B (2005) Biological variability of lignin. Cellul Chem Technol 39:341-367

Nong G, Chow V, Rice JD, St John FJ, Preston JF (2009) Aldouronate utilization in
Paenibacillus sp. strain JDR-2: physiological and enzymatic evidence for coupling of
extracellular depolymerization and intracellular metabolism. Appl Environ Microbiol. in
press

Ohta K, Beall DS, Mejia JP, Shanmugam KT, Ingram LO (1991a) 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

Ohta K, Beall DS, Mejia JP, Shanmugam KT, Ingram LO (1991b) Metabolic engineering of
Klebsiella oxytoca M5A1 for ethanol production from xylose and glucose. Appl Environ
Microbiol 57:2810-2815

Ostergaard S, Olsson L, Nielsen J (2000) Metabolic engineering of Saccharomyces cerevisiae.
Microbiol Molecul Biol Rev 64:34-50

Palmqvist E, Hahn-Hagerdal B (2000b) Fermentation of lignocellulosic hydrolysates. II:
inhibitors and mechanisms of inhibition. Bioresour Technol 74:25-33

Palmqvist E, Hahn-Hagerdal B (2000a) Fermentation of lignocellulosic hydrolysates. I:
inhibition and detoxification. Bioresour Technol 74:17-24

Patel MA, Ou MS, Harbrucker R, Aldrich HC, Buszko ML, Ingram LO, Shanmugam KT (2006)
Isolation and characterization of acid-tolerant, thermophilic bacteria for effective
fermentation of biomass-derived sugars to lactic acid. Appl Environ Microbiol
72:3228-3235

Pordesimo LO, Hames BR, Sokhansanj S, Edens WC (2005). Variation in corn stover
composition and energy content with crop maturity. Biomass and Bioenergy 28:366-374.

Prasad S, Singh A, Joshi HC (2007) Ethanol as an alternative fuel from agricultural, industrial
and urban residues. Resour Conserv Recycl 50:1-39

Preston JF, Hurlbert JC, Rice JD, Ragunathan A, St John FJ (2003) Microbial strategies for the
depolymerization of glucuronoxylan: leads to biotechnological applications of
endoxylanases. Applications of Enzymes to Lignocellulosics 855:191-210

Preston JF, Rice JD, Ingram LO, Keen NT (1992) Differential depolymerization mechanisms of
pectate lyases secreted by Erwinia c h1iyiwh'\iitiniui Ecl6. J Bacteriol 174:2039-2042

Puchta H, Hohn B (1991) A transient assay in plant cells reveals a positive correlation between
extrachromosomal recombination rates and length of homologous overlap. Nucleic Acids
Res 19:2693-2700









Puls J (1997) Chemistry and biochemistry of hemicelluloses: Relationship between
hemicellulose structure and enzymes required for hydrolysis. Macromol Symp 120:183-196

Qian Y, Yomano LP, Preston JF, Aldrich HC, Ingram LO (2003) Cloning, characterization, and
functional expression of the Klebsiella oxytoca xylodextrin utilization operon (xynTB) in
Escherichia coli. Appl Environ Microbiol 69:5957-5967

Quadt-Hallmann A, Kloepper JW (1996) Immunological detection and localization of the cotton
endophyte Enterobacter asburiae JM22 in different plant species. Canadian Journal of
Microbiology 42, 1144-1154.

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, Tschaplinski T (2006)
The path forward for biofuels and biomaterials. Science 311:484-489

Reddy N, Yang Y (2005) Biofibers from agricultural byproducts for industrial applications.
Trends Biotechnol 23:22-27

Robertbaudouy J, Portalier R, Stoeber F (1981) Regulation of hexuronate system genes in
Escherichia Coli K-12 multiple regulation of the uxu operon by exur and uxur gene
products. J Bacteriol 145:211-220

Rodriguez M, Martinez EA, York SW, Zuobi-Hasona K, Ingram LO, Shanmugam KT and
Preston JF (2001) Properties of the hemicellulose fractions of lignocellulosic biomass
affecting bacterial ethanol production. Abstracts of the 101st National Meetings of the
American Society of Microbiology meetings in Orlando, FL, USA. P 535 (2001).

Rogers PL, Jeon YJ, Lee KJ, Lawford HG (2007) Zymomonas mobilis for fuel ethanol and
higher value products. Biofuels 108:263-288

Russell JB, Cook GM (1995) Energetics of bacterial growth balance of anabolic and catabolic
reactions. Microbiol Rev 59:48-62

Saha BC (2003) Hemicellulose bioconversion. J Ind Microbiol Biotechnol 30:279-291

Saha BC, Iten LB, Cotta MA, Wu YV (2005b) Dilute acid pretreatment, enzymatic
saccharification, and fermentation of rice hulls to ethanol. Biotechnol Progr 21:816-822

Saha BC, Iten LB, Cotta MA, Wu YV (2005a) Dilute acid pretreatment, enzymatic
saccharification and fermentation of wheat straw to ethanol. Process Biochem 40:3693-3700

Saier MH, Simoni RD, Roseman S (1976) Sugar transport. Properties of mutant bacteria
defective in proteins of phosphoenolpyruvate: sugar phosphotransferase system. J Biol
Chem 251:6584-6597

Saloheimo A, Rauta J, Stasyk OV, Sibimy AA, Penttila M, Ruohonen L (2007).Xylose transport
studies with xylose-utilizing Saccharomyces cerevisiae strains expressing heterologous and
homologous permeases. Appl Microbiol Biotechnol 74: 1041-1052









SanFrancisco MJD, Keenan RW (1993) Uptake of galacturonic acid in Erwinia ch inywnthl/wini
Ecl6. JBacteriol 175:4263-4265

Schunck B, Kraft W, Truyen U (1995) A simple touch-down polymerase chain reaction for the
detection of canine parvovirus and feline Panleukopenia virus in feces. Journal of
Virological Methods 55:427-433

Scott AI, Baxter RL (1981) Applications of 13C NMR to metabolic studies. Ann Rev Biophy
Bioeng 10:151-174

Shanmugam KT, Ingram LO (2008) Engineering biocatalysts for production of commodity
chemicals. J Mol Microbiol Biotechnol 15:8-15

Sharma V, Kumar V, Archana G, Kumar GN (2005) Substrate specificity of glucose
dehydrogenase (GDH) of Enterobacter asburiae PSI3 and rock phosphate solubilization
with GDH substrates as C sources. Can J Microbiol 51:477-482

Shulami S, Gat G, Sonenshein AL, Shoham Y (1999) The glucuronic acid utilization gene
cluster from Bacillus stearothermophilus T-6. J Bacteriol 181:3695-3704

Simoni RD, Roseman S, Saier MH, Jr. (1976) Sugar transport. Properties of mutant bacteria
defective in proteins of the phosphoenolpyruvate: sugar phosphotransferase system. J Biol
Chem 251:6584-6597

Slininger PJ, Bothast RJ, Okos MR, Ladisch MR (1985) Comparative evaluation of ethanol
production by xylose fermenting yeasts presented high xylose concentrations. Biotechnol
Lett 7:431-436

Smalley AJ, Jahrling P, Van Demark PJ (1968) Molar growth yields as evidence for xxidative
phosphorylation in Streptococcus faecalis strain 10C1. J Bacteriol 96:1595-1560

Sprenger GA (1996) Carbohydrate metabolism in Zymomonas mobilis: A catabolic highway with
some scenic routes. FEMS Microbiol Lett 145:301-307

St John FJ, Rice JD, Preston JF (2006) Paenibacillus sp strain JDR-2 and XynA(1): a novel
system for methylglucuronoxylan utilization. Appl Environ Microbiol 72:1496-1506

Sticklen M (2006) Plant genetic engineering to improve biomass characteristics for biofuels.
Curr Opin Biotechnol 17:315-319

Sun JX, Sun XF, Sun RC, Su YQ (2004) Fractional extraction and structural characterization of
sugarcane bagasse hemicelluloses. Carbohydrate Polymers 56:195-204.

Sunna A, Antranikian G (1997) Xylanolytic enzymes from fungi and bacteria. Crit Rev
Biotechnol 17:39-67


114









Sun RC, Fang JM, Tomkinson J, Geng ZC, Liu JC (2001) Fractional isolation, physico-chemical
characterization and homogeneous esterification of hemicelluloses from fast-growing poplar
wood. Carbohydrate Polymers 44:29-39.

Taguchi H, Ohta T (1991) D-lactate dehydrogenase is a member of the D-isomer-specific
2-hydroxyacid dehydrogenase family cloning, sequencing, and expression in Escherichia
coli of the D-Lactate dehydrogenase gene of Lactobacillusplantarum. J Biol Chem
266:12588-12594

Tarkow H, Feist WC (1969) A mechanism for improving digestibility of lignocellulosic
materials with dilute alkali and liquid ammonia. advan chem Ser 197-218

Teymouri F, Laureano-Perez L, Alizadeh H, Dale BE (2004) Ammonia fiber explosion treatment
of corn stover. Appl Biochem Biotechnol 113-16:951-963

Timell TE (1964) Wood Hemicelluloses. Adv Carbohydr Chem 19:247-302

Vlasenko EY, Ding H, Labavitch JM, Shoemaker SP (1997) Enzymatic hydrolysis of pretreated
rice straw. Bioresour Technol 59:109-119

von Blottnitz H, Curran MA (2007) A review of assessments conducted on bio-ethanol as a
transportation fuel from a net energy, greenhouse gas, and environmental life cycle
perspective. Journal of Cleaner Production 15:607-619

vonSivers M, Zacchi G (1996) Ethanol from lignocellulosics: A review of the economy.
Bioresour Technol 56:131-140

vonSivers M, Zacchi G (1995) A technoeconomic comparison of 3 processes for the production
of ethanol from pine. Bioresour Technol 51:43-52

Wesseler J (2007) Opportunities ('costs) matter: A comment on Pimentel and Patzek "Ethanol
production using corn, switchgrass, and wood; biodiesel production using soybean and
sunflower". Energy Policy 35:1414-1416

Whitesides GM, Crabtree GW (2007) Don't forget long-term fundamental research in energy.
Science 315:796-798

Wood BE, Ingram LO (1992) Ethanol production from cellobiose, amorphous cellulose, and
crystalline cellulose by recombinant Klebsiella oxytoca containing chromosomally
integrated Zymomonas mobilis genes for ethanol production and plasmids expressing
thermostable cellulase genes from Clostridium thermocellum. Appl Environ Microbiol
58:2103-2110

Wood BE, Yomano LP, York SW, and Ingram LO (2005) Development of industrial medium
required elimination of the 2,3-butanediol fermentation pathway to maintain ethanol yield in
an ethanologenic strain of Klebsiella oxytoca. Biotechnol Prog 21:1366-1372









Wyman CE (1999) Biomass ethanol: Technical progress, opportunities, and commercial
challenges. Ann Rev Energy Environ 24:189-226

Wyman CE (2003) Potential synergies and challenges in refining cellulosic biomass to fuels,
chemicals, and power. Biotechnol Prog 19:254-262

Xiao B, Sun XF, Sun RC (2001) Chemical, structural, and thermal characterizations of
alkali-soluble lignins and hemicelluloses, and cellulose from maize stems, rye straw, and
rice straw. Polym Degrad Stab 74:307-319

Yomano LP, York SW, Ingram LO (1998) Isolation and characterization of ethanol-tolerant
mutants of Escherichia coli KO 11 for fuel ethanol production. J Ind Microbiol Biotechnol
20:132-138

Yomano LP, York SW, Zhou S, Shamugam KT, Ingram LO (2008) Re-engineering Escherichia
coli for fuel ethanol production. Biotechnol Lett 20:132-138

York SW, Ingram LO (1996) Ethanol production by recombinant Escherichia coli KO 11 using
crude yeast autolysate as a nutrient supplement. Biotechnol Lett 18:683-688

Zaldivar J, Ingram LO (1999) Effect of organic acids on the growth and fermentation of
ethanologenic Escherichia coli LY01. Biotechnol Bioeng 66:203-210

Zaldivar J, Martinez A, Ingram LO (1999) Effect of selected aldehydes on the growth and
fermentation of ethanologenic Escherichia coli. Biotechnol Bioeng 65:24-33

Zaldivar J, Martinez A, Ingram LO (2000) Effect of alcohol compounds found in hemicellulose
hydrolysate on the growth and fermentation of ethanologenic Escherichia coli. Biotechnol
Bioeng 68:524-530

Zhang X, Jantama K, Moore JC, Shanmugam KT, Ingram LO (2007) Production of L-alanine by
metabolically engineered Escherichia coli. Appl Microbiol Biotechnol 77:355-366

Zhang YHP (2008) Reviving the carbohydrate economy via multi-product lignocellulose
biorefineries. J Ind Microbiol Biotechnol 35:367-375

Zhou SD, Davis FC, Ingram LO (2001) Gene integration and expression and extracellular
secretion of Erwinia c/ 1i)iy\hiini endoglucanase CelY (celY) and CelZ (celZ) in
ethanologenic Klebsiella oxytoca P2. Appl Environ Microbiol 67:6-14

Zhou SG, Ingram LO (2000) Synergistic hydrolysis of carboxymethyl cellulose and acid-swollen
cellulose by two endoglucanases (CelZ and CelY) from Erwinia c hiliny\1wheini. J Bacteriol
182:5676-5682

Zucker M, Hankin L (1970) Regulation of pectate lyase synthesis in Pseudomonasfluorescens
and Erwinia carotovora. J Bacteriol 104:13-18


116









Zuobi-Hasona K, St John FJ, Rice JD, Preston JF (2001) Oligosaccharides containing
glucuronoxylose as substrates for selecting bacteria for depolymerization of hemicellulose.
Abstracts of the 10 1st National Meetings of the American Society of Microbiology meetings
in Orlando, FL, USA. p 534.









BIOGRAPHICAL SKETCH

Changhao Bi was born in Qujing, China in 1979. After attending the Kunming Second

Middle School and the Kuming First High School, he was admitted in Nankai University. During

his student career there, he received his bachelor's degree in biochemistry in 2001 and master's

degree in molecular biology and biochemistry in 2004. In the same year, he was accepted to the

Ph.D program in the Department of Microbiology and Cell Science at the University of Florida,

then chose to work in Dr. Preston's laboratory until received his Ph.D. from the University of

Florida in August 2009.





PAGE 1

1 METABOLIC CHARACTERIZATIO N AND ENGINEERING OF ENTEROBACTER ASBURIAE STRAIN JDR-1 TO DEVELOP MICROBIAL BIOCATALYSTS FOR EFFICIENT HEMICELLULOSE UTILIZATION By Changhao Bi A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2009

PAGE 2

2 2009 Changhao Bi

PAGE 3

3 To my parents and my wife

PAGE 4

4 ACKNOWLEDGMENTS I thank my family, for without their support, I would not have been able to complete this work. And I would like to thank my supervis or, Dr. Preston and the rest of my committee. In the past 5 years, I have gained much knowledge and developed many skills in the field of microbiology and molecular biology with the rele ntless help of my supervisor Dr. Preston and my committee members. However, what has a ffected me most is the serious attitude for scientific research my supervisor and his fe llow professors demonstr ated. Also, I thank my lab colleagues for the kind help and support to my research.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........8 LIST OF FIGURES................................................................................................................ .........9 ABSTRACT....................................................................................................................... ............10 CHAPTER 1 LITERATURE REVIEW.......................................................................................................12 Advantages of Utilizing Lignocellulosic Bi omass to Produce Ethanol and Chemicals.........12 National Energy Security................................................................................................12 Economic Advantages.....................................................................................................13 Environmental Advantages.............................................................................................15 Constituents of Lignocellulose and Hemicellulose................................................................16 Lignocellulose.................................................................................................................16 Hemicellulose..................................................................................................................17 Pretreatment Processes of Lignocellulose..............................................................................18 Pretreatment Methods......................................................................................................18 Dilute Acid Pretreatment.................................................................................................19 Cost of Pretreatment Methods.........................................................................................21 Development of Ethanol Produc ing Biocatalysts toward He micellulose Utilization.............21 Escherichia coli ...............................................................................................................22 Zymomonas mobilis .........................................................................................................23 Klebsiella oxytoca ...........................................................................................................24 Yeasts ...............................................................................................................................25 Objectives of this Research....................................................................................................26 2 COMPLETE FERMENTATION OF XYLOSE AND METHYLGLUCURONOXYLOSE DERIVED FROM METHYLGLUCURONOXYLAN BY Enterobacter asburiae STRAIN JDR-1..................29 Introduction................................................................................................................... ..........29 Materials and Methods.......................................................................................................... .30 Preparation of Substrat es and Culture Media..................................................................30 Isolation and Identification of E. asburiae JDR-1...........................................................32 Substrate Utilization and Fermentation Product Analysis...............................................32 Determination of Metabolic Pathways by 13C-NMR......................................................33 Determination of Molar Cell Dry Weight Yield.............................................................34 Results........................................................................................................................ .............35 Identification and Characterization of E. asburiae JDR-1..............................................35 Utilization of Acid Hydrolysates of Methylgluronoxylan by E. asburiae JDR-1...........36 Substrate Preference of E. asburiae JDR-1.....................................................................36 Fermentation Characteristics...........................................................................................37 Central Metabolic Pathways Determined by 13C-NMR..................................................38

PAGE 6

6 Growth and Projected ATP Yields with Different Substrates.........................................40 Discussion..................................................................................................................... ..........41 Fermentation of MeGAX by E. asburiae JDR-1.............................................................41 Central Metabolic Pathways Used by E. asburiae JDR-1...............................................41 Possible Pathway of MeGAX Metabolism......................................................................43 Bioenergetics of E. asburiae JDR-1................................................................................45 Role of E. asburiae JDR-1 in Soil Ecology and Bioprocessing......................................46 Biotechnological Applications........................................................................................46 3 GENETIC ENGINEERING OF Enterobacter asburiae STRAIN JDR-1 FOR EFFICIENT D(-)LACTIC ACID PR ODUCTION FROM HEMICELLULOSE HYDROLYSATE...................................................................................................................58 Introduction................................................................................................................... ..........58 Materials and Methods.......................................................................................................... .59 Bacterial Strains, Media, and Growth Conditions...........................................................59 Genetic Methods..............................................................................................................60 Deletion of pflB and als Gene in E. asburiae JDR-1......................................................60 Fermentation................................................................................................................... .62 Determination of Lactate Isomers Produced by E. asburiae L1.....................................62 Results........................................................................................................................ .............63 Fermentation Characteristics of the Wild type Strain E. asburiae JDR-1......................63 Fermentation Characteristics of the Engineered Strains E. asburiae E1 and L1............64 D-Lactate Was Produced by E. asburiae L1...................................................................65 Discussion..................................................................................................................... ..........66 4 GENETIC ENGINEERING Enterobacter asburiae STRAIN JDR-1 FOR EFFICIENT ETHANOL PRODUCTION FROM HEMICELLULOSE HYDROLYSATE...................................................................................................................72 Introduction................................................................................................................... ..........72 Materials and Methods.......................................................................................................... .73 Bacterial Strains, Media, and Growth Conditions...........................................................73 Genetic methods..............................................................................................................73 Transformation of Plasmid Carrying PET Operon into E. asburiae JDR-1....................73 Integration of PET Operon into E. asburiae JDR-1 Chromosome.................................74 Plasmid Stability in E. asburiae JDR-1...........................................................................75 Assay of PDC Activity....................................................................................................75 Results........................................................................................................................ .............76 Fermentation Characteristics of the Wild Type Strain E. asburiae JDR-1.....................76 Fermentation of E. asburiae 4666 and E. asburiae 4672 in Glucose..............................77 Fermentation of E. asburiae JDR-1 (pLOI297) and E. asburiae JDR-1 (pLOI555)....................................................................................................................77 Fermentation of Strain E. asburiae E1 (pLOI555) Compared with E. coli KO11 and Other E. asburiae JDR-1 Derivatives...................................................................78 PDC Activity in E. asburiae Strains................................................................................79 Plasmid Stability in E. asburiae JDR1............................................................................80 Discussion..................................................................................................................... ..........80 5 GENETIC DEFINITION OF THE MEGAX UTILIZATION PATHWAY IN Enterobacter asburiae JDR-1.................................................................................................91

PAGE 7

7 Introduction................................................................................................................... ..........91 Materials and Methods.......................................................................................................... .91 Degenerate Primer Method to Dete rmine Presence of GH 67 genes..............................91 Determination of Genes Involved in MeGAX Utilization Process with Mutant Library Method............................................................................................................92 Determination of MeGAX Utilization Genes by Transforming E. coli with Cosmid Library and Topo Plasmid Library.................................................................93 Obtaining Candidate Genes by Sequencing Cosmid Containing -xylosidase Gene and Genome Comparison...................................................................................93 Deletion of Candidate Genes to C onfirm Their Involvement in MeGAX Utilization Process.......................................................................................................94 Results........................................................................................................................ .............94 MeGAX Defective Mutants of E. asburiae JDR-1.........................................................94 Candidate Genes Determined From Cosmid and Plasmid Library Transformation Method...............................................................................................97 Obtaining Candidate Genes for Function Confirmation by Gene Deletion....................97 Discussion..................................................................................................................... ..........99 6 CONCLUSION.....................................................................................................................104 REFERENCE LIST................................................................................................................. ....107 BIOGRAPHICAL SKETCH.......................................................................................................118

PAGE 8

8 LIST OF TABLES Table page 2-1. Fermentation products formed by E. asburiae JDR-1 from monosaccharides derived from hemicellulosea...........................................................................................................48 2-2. Fermentation products of E. asburiae JDR-1 derived from MeGAXn. Acetate, ethanol and formate concentrations were determined in duplicate cultures with HPLC and methanol with GC............................................................................................48 2-3. Distribution of 13C in fermentation products formed in anaerobic cultures of E. asburiae JDR-1 and E. coli B grown with differentially 13C labeled xylose and glucosea..............................................................................................................................49 2-4. Anaerobic molar cell dry weight and ATP yi eld from different substrates calculated based on estimated YATP, 8, for all substrates in E. asburiae JDR-1..............................49 3-1. Bacterial strains and plasmids............................................................................................ ....67 3-2. Comparing fermentation products of w ild type and genetically engineered E. asburiae JDR-1 strainsa.....................................................................................................68 3-3. Specific consumption rates a nd specific production rates of E. asburiae L1 in 5g/liter acid hydrolysate of sweetgum xylan.a...................................................................69 4-1. Bacterium strains and plasmids for engineering ethanologenic E. asburiae ..........................85 4-2. Comparison of sugar fermentation products of wild type and genetically engineered E. asburiae JDR-1. Fermentations were ca rried out at 30C in ZH minimal media for 48 hours as described in the Materials and Methods section.......................................86 4-3. Fermentation products from acid hydrolys ates of sweetgum xylan. Fermentations were carried out at 30C in ZH minima l media for 48 hours as described in the Materials and Methods section. Results were averages of 3 experiments.........................86 4-4. Specific consumption rates and specific production rates in acid hydrolysates of sweetgum xylan (5g /liter)a. Results were averages of 3 experiments...............................87 4-5. Specific activity of PDC in cell crude extract from E. asburiae JDR-1 derived strains. Results were averages of 3 experiments................................................................87 4-6. Plasmid stability of pLOI297 and pLOI555 in E. asburiae JDR-1. Results were averages of 3 experiments..................................................................................................87 5-1. Growth rate of ptsIin different substrate with and without cAMP.....................................101

PAGE 9

9 LIST OF FIGURES Figure page 1-1. Model for corn stover cell walls.......................................................................................... ...28 2-1. Scheme for the release of xylose and MeGAX by dilute acid hydrolysis of sweetgum xylan.......................................................................................................................... .........50 2-2. Microscopic images of E. asburiae JDR1..............................................................................51 2-3. Aerobic growth, substrate utilizati on, and formation of products from acid hydrolysates of MeGAXn...................................................................................................52 2-4. Aerobic growth of E. asburiae JDR-1 on different combinations of sugar substrates...........53 2-5. Pathway determination for the me tabolism of xylose and glucose by E. asburiae JDR-1.......................................................................................................................... .......55 3-1. Fermentation time course for different strains in media containing 0.5% sweetgum xylan hydrolysate.............................................................................................................. .70 3-2. Diagram to illustrate deletion of als and pfl B genes modifying mixed-acid fermentation of E. asburiae JDR-1 into a homolactate production pathway in E. asburiae L1........................................................................................................................71 4-1. HPLC profile of fe rmentation products of E. asburiae JDR-1, E.coli KO11 and E. asburiae E1 (pLOI555) in 0.5% sweetgum xylan hydrolysate with MOPS......................88 4-2. Fermentation time course of E. asburiae JDR-1 A), E.coli KO11 B), E. asburiae JDR-1 (pLOI555) C) and E. asburiae E1 (pLOI555) D) in media of buffered sweetgum xylan hydrolysate..............................................................................................89 5-1. Function and regulation of PTS............................................................................................102 5-2. HPLC profile of ae robic culturing media of E. asburiae JDR-1 xylA-. ................................102 5-3. Enzyme activity of GH88 enzyme uns aturated -glucuronyl hydrolase (EC 3.2.1.-) (Hashimoto et al., 1999)...................................................................................................103

PAGE 10

10 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy METABOLIC CHARACTERIZATIO N AND ENGINEERING OF ENTEROBACTER ASBURIAE STRAIN JDR-1 TO DEVELOP MICROBIAL BIOCATALYSTS FOR EFFICIENT HEMICELLULOSE UTILIZATION By Changhao Bi August 2009 Chair: James F. Preston Major: Microbiology and Cell Science Acid pretreatment is commonly used to release pentoses from the hemicellulose fraction of cellulosic biomass for bioconve rsion. The predominant pentose in the hemicellulose fraction of hardwoods and crop residues is xylose in the polysaccharide methylglucuronoxylan, in which as many as 1 in 6 of the -1,4 linked xylopyranose residues are substituted with -1,2-linked 4-O-methylglucuronopyranose. Resistance of the -1,2-methylglucuronosyl linkages to acid hydrolysis results in release of the aldobiuronate 4-O-methylglucuronoxylose (MeGAX) which is not fermented by bacterial biocatalysts currently used for bioco nversion of hemicellulose. Enterobacter asburiae JDR-1, isolated from colonized hardwood (sweetgum), was s hown to efficiently ferment both MeGAX and xylose in acid hydrolysates of sweetgum xylan, producing predomin antly ethanol and acetate. 13C-NMR studies defined the Embden-Meyerhof pathway for metabolism of glucose and the pentose phosphate pathway for xylose metabolis m. MeGAX-defective mutants isolated after Tn5 transposon mutagenesis led to identification of oxi ptsI xylA and xylB genes required for efficient MeGAX metabolism. Based on both gene tic and physiological studies, including rates of substrate utilization, product formation and molar gr owth yields, MeGAX utilization by E. asburiae JDR-1 requires transporta tion into the cell and hydr olysis by a novel pathway to release methanol, gl ucuronic acid and xylose.

PAGE 11

11 E. asburiae JDR-1 with deletion of the pflB gene encoding pyruvate formate lyase ( E. asburiae E1) was transformed with a plasmid (pLOI555) carrying pdc and adhB genes derived from Zymomonas mobilis to generate a recombinant strain capable of completely fermenting both xylose and MeGAX in a hydrolysat e of sweetgum xylan to ethanol with a yield at 99% of the theoretic al maximum. A second strain, E. asburiae L1 was engineered by deletion of pflB and als genes to produce D(-)lactate at a yield approximately 100% of the theoretical maximum in hydrolysate of sw eetgum xylan. The engineered strains of E. asburiae JDR-1 provide novel microbial biocatalys ts for maximal conversion of carbohydrate compounds in acid hydrolysates of hemicellulo se into ethanol and other fermentation products.

PAGE 12

12 CHAPTER 1 LITERATURE REVIEW Advantages of Utilizing Lignocellulosic Bi omass to Produce Ethanol and Chemicals The main driving forces to develop renewa ble supplies of energy are derived from 1) National energy security: to decr ease the dependence of imported fu els from certain areas with political instability and prepare for depletion of fossil fuels; 2) Economic advantages: to develop less expensive renewable fuels in large am ounts without increasing food prices; and 3) Environmental advantages: to alleviate the greenhouse effect and make use of industrial, agricultural and municipal wastes. National Energy Security Developing biobased sustainabl e fuel production is crucial for national energy security. Societies are mainly sustained by petroleum-ba sed energy currently, which is a non-renewable resource with limited reserves. As the human population increased 4-fold in the twentieth century, global energy consumption increased 16-fold to reach 13 tr illion watts. When the population reaches 8 to11 billion by 2050, the ener gy consumption is expected to be 27 to 42 trillion watts (Whitesides and Crabtree 2007; Hoffer t et al. 2002). The estimated amount of oil in determined reservoirs is approximately 1 trilli on barrels while the total possible reserves might be as high as 2 trillion barrels with new disc overies, expansion of existing fields and new extracting technologies. Since curr ent oil production and consumpti on is about 30 billion barrels per year, the global oil reserve might become extremely limited within 40 to 50 years and exhausted in 30 to 70 years (Wesseler 2007). Al though some reports present a more optimistic estimation of the oil reserve, the growth and increase in global economy is expected to accelerate the rate of petroleum depletion. The United States consumes approximately 30 billion barrels oil per year and about 30 % of total oil produced wo rldwide. As much as 60% of the oil U.S.

PAGE 13

13 consumed is imported (Gray et al. 2006). The de pendence of oil from foreign sources negatively affects the U.S. national energy security. Development of a renewable transportation fuel is urgently needed to reduce the dependence on imported oil. Biomass is the only known renewable bioresour ce that can be used for production of liquid transportation fuel (Wyman 2003). Domistic production of ethanol from biomass will reduce the dependence on imported fuels (Lin and Tanaka 2006). The U.S. Department of Energy has proposed an objective of producing 30% of the U.S. transportation fuels from biomass by 2030, based on the availability of ligno cellulose in the U.S. (Himmel et al. 2007; Hahn-Hagerdal et al. 2006). The DOE and USDA reported that the Un ited States can produce approximately 1.3 billion dry tons of lignocellulose per year, in cluding 933 million tons from agricultural and 368 million tons from forest resources (Zhang 2008). On e billion dry tons of lignocellulosic biomass contain sufficient carbohydrate to produce 80-130 billion gallons of bioethanol. Therefore the 1.3 billion dry tons of lignocellulose produced per year in the United States is capable of producing approximately 60 billion gallons of bioethanol to meet the goal of repl acing 30% tr ansportation fuel in the U.S. In addition, the ocean also ha s a great potential for bi ofuel production with its lage surface area for marine biomass producti on. Microalgae and macroalgae, resources for biofuel production, are reported to produce 81~150 dr y ton/ha/year, if saturating levels of CO2 are provided (Gao and Mckinley 1994). Economic Advantages Lignocellulosic biomass is abundant, renewabl e and relatively inexpensive. Development of biomass as an energy resource has the poten tial for production of cheap er alternative fuels. Production of biomass-based fuel does not compet e with food sources and does not increase food prices. Furthermore, land not fit for food crops can be used to culture energy crops, albeit will

PAGE 14

14 lower annual yields. Production of large amounts of bioethanol may help st ablelize the price of transportation fuels. Biomass is produced by photosynthesis, duri ng which light energy is converted to chemical energy stored as fixed CO2 in biomass. Lignocellulosic biomass is the most abundant renewable biological resource, with a production of approximately 2 x 1011 dry tons per year (Reddy and Yang 2005). Most lignocellulosic biomass is available at lower pr ices than crude oil. For example, the crude oil ($40 to 80$ per barrel) costs about $7.1 to $14.2/GJ, which is much higher than the price of lignoce llulose (up to $3/GJ), based on equivalent energy content. Although the values of lignocelluloses are variab le according to feedstock composition, the price of lignocellulose-based products is expected to be competitive with fossil resource-derived products if processing costs of lignocelluloses are reduced. When municipal waste is used, the reduction of waste removal fees brings a dditional economic advantages (Wyman 2003). In 2006, fermentation of corn starch accounted for more than 93 % of the total ethanol production in the U.S. (Lin and Tanaka 2006). Incr easing ethanol production in the near future will require more corn to be diverted away fr om food and feed sources, which will inevitably increase the price of food and other downstream products of cor n. For example, during the years 2004 to 2006, corn kernel price ro se by more than 70% from $2.25 per bushel to $4 per bushel. The Chinese government even banned building new ethanol production facilities based on grains in 2006 (Zhang 2008). However, using lignocellulose as feedstock to produce ethanol does not compete with food and feed sources, or increas e the food prices which negatively impacts the society welfare (Gray et al. 2006; Sticklen 2006). Furthermore, bioenergy crops do not compete with agricultural crops for land committed to cul tivation, since land of lower quality can be used for energy crops production, albeit the lower annual yield. In addition, cultivation of crops for

PAGE 15

15 energy production generally require s less fertilizer and energy inputs compared to food crops (Wyman 1999). Environmental Advantages Using biomass-based fuel has the potential to alleviate the greenhouse effect and reduce the pollution of industrial, agricultural and munici pal wastes. Fossil fuels, the only sources of net CO2 production, are the primary sources of the greenhouse gases. Fossil fuel combustion is responsible for approximately 82% of net green house gas emission, as much as about 7.0 billion tons of carbon per year (Lal 2004; Eissen et al. 2002). Since the i ndustrial revolution in nineteenth century, nearly 270 billon tons of car bon from the combustion of fossil fuels have accumulated in the atmosphere and the atmospheric CO2 level has increased from 275 to 380 ppm. Consequently, the global average near-s urface temperature has risen by 0.6.2C in the twentieth century (Lal 2004). The CO2 level is expected to pass 550 ppm by the middle of this century and cause further increase of temperature according to current momentum (Kheshgi et al. 2000; Galbe and Zacchi 2002). Because CO2 generated from ethanol is recycled into plants and algae by photosynthesis, no net CO2 is produced with utilizati on of ligcellulosic ethanol. Therefore development of ligcellulosic ethanol as an alternative fuel has the potential to reduce accumulation of CO2 gas and reduce the greenhouse effect Ethanol is also a potent fuel oxygenate. When blended with gasoline, ethanol is able to improve engine combustion and reduce CO2 emissions (Jeuland et al. 2004) Additionally, utilization of ethanol also reduces the amount of ozone forming compounds, such as SOx, NOx, and CO (von Blottnitz and Curran 2007). Large volumes of waste material are produced every year. Although portions of the waste can be sold as animal feed or burned to generate electricity, most waste material that can not be utilized is usually buried in land fills. There are environmenta l concerns about the large amount

PAGE 16

16 of hazardous matter produced by buried wastes Household and office lignocellulosic waste (paper, yard waste, etc.) also contribute to the material currently be ing placed in land fills (Prasad et al. 2007). With development of bioma ss-based ethanol, most of these lignocellulosic wastes are able to be processed as feedstock to produce ethanol, thus largely reducing the amount of land fills, removing disposal problem of wastes and reducing pollution from wastes. Constituents of Lignocellulose and Hemicellulose Lignocellulose In general, lignocellulosic biomass is the mixt ure of structural components of plant matter as opposed to the sugar storage compounds, starch and sucrose. In other words, lignocellulosic biomass is the inedible portion of plant material including wood, plants, gr asses, crop residues, and ligocellulosic wastes. The primary components of plant cell walls are cellulose, hemicellulose and lignin. Cellulose is the most abundant carbohydrate polymer in the world and is the major structural component of plants, ranging from 36% to 50% of total plant mass. Cellulose in plant walls is constituted by associated crystalline fibers c onsisting of many indivi dual cellulose strands assembled by hydrogen bonds and Van der Waals fo rces. These single cellulose strands are -1,4-linked glucose polymers (Lynd et al. 1999). Quantitative data of cellulose accessibility clearly suggests that on ly a small fraction of -glucosidic bonds of cr ystalline cellulose are accessible to cellulases (Hong et al. 2007). The s econd most abundant constituent in biomass is the hemicellulose. Polymer methylglucuronoa rabinoxylan is the major component of hemicellulose in graminaceous plants representi ng up to 35% of total plant mass (Lynd et al. 1999). In hardwoods hemicelullose is compos ed primarily of methylglucuronoxylan ranging from 20% to 35% of total lignocellulose (Pul s 1997). Lignin is composed of phenyl-propane

PAGE 17

17 derivatives, randomly linked through carboncarbon bonds by an enzymatic dehydrogenation process, and associated to hemicellulo se to form a network (Monties 2005). Hemicellulose The major hemicellulosic component of hard wood, methylglucuronoxylan, is a -1,4 linked xylose chain polymer, in which th e xylose groups are randomly substituted with -1,2 linked 4-O-methylglucurona te residues at certain frequenc y. Hemicelluloses are located among the lignin and the cellulose fibers and covalently linked at many points of their main chain and substitution group with the lignin network. For exam ple, two types of covalent cross-links are identified between hemicellulose and lignin: di ferulic acid bridges and ester linkages between lignin and glucuronic acid attach ed to xylans (de Vries and Vi sser 2001). Hemicelluloses also associate with cellulose strands by hydrogen bonds to form layers of network around them. So that the xylan layer has both covalent linkage s to lignin and non-cova lent interaction with cellulose to maintain the plant cell wall and he lps the associated cellulose fibers against degradation by cellulase s (Beg et al. 2001). Xylans from different sources, such as gras ses, cereals, softwood, and hardwood, differ in composition. In various sources, the frequency of methylglucuronate substitution is different, ranging from 1 of every 6 xylose residues to 1 of every10 xylose residues (Jacobs et al. 2001, Sunna and Antranikian 1997). The 13C-NMR studies of methyl glucuronoxylan from the hardwood sweetgum illustrate the frequency of me thylglucuronate substitution is 1 in every 6 xylose residues (Preston et al 2003; Puls 1997). Methylglucuronoa rabinoxylan in softwood has a higher 4-O-methylglucuronate content than hard wood xylans, but can contain arabinose at a frequency of about one for ev ery six xylose residues. The -L-arobinofuranose units are linked by -1,3-glycosidic bonds to the C-3 position of the xylose groups in xylan main chain. Softwood xylans are shorter than the hardwood xylans, with a de gree of polymerization from 70

PAGE 18

18 to 130 xylose groups and less branched (Xiao et al. 2001; Sunna and Antranikian 1997). A model for the corn stover cell wall is shown in Fig 1-1 (Saha 2003). The xylans and lignins, are highly cross-linked by diferulic linkages, and constitute a network where the cellulose fibers fit in underneath. Pretreatment Processes of Lignocellulose Pretreatment Methods In lignocellulose bioref ineries, biological c onversion of lignocellulose generally has three main steps: 1) lignocellulose pret reatment, which converts the origin al lignocellulose structure to enzyme-reactive cellulose and pe ntoses; 2) enzymatic cellulose hydrolysis, by which cellulases hydrolyze reactive cellulose to fe rmentable sugars; and 3) ferm entation, which produces ethanol or other bio-based chemicals from released sugars (e.g., lactic acid, succinic acid) (Ragauskas et al. 2006). Conversion of lignocellulosic biomass to fermentable sugars is the most challenging step in the process of utilizati on, due to the complex nature of li gnocelluloses. Util ization of the carbohydrates of lignocellulosic biomass require s a physico-chemical pretreatment. Following the first pretreatment step, most methods require an enzymatic cellulose hydrolysis step, in which celluases from fungi are supplemented to degrade the cellulose compartment into glucose. Various pretreatment options ar e available to fractionate, sepa rate, solubilize and hydrolyze cellulose, hemicellulose, and lignin components. These methods include d ilute sulfuric acid, flow through, pH controlled water, ammonia fiber explosion (AFEX), ammonia recycle percolation (ARP) and lime pretreatment. Except for AFEX, all the other pretreatment methods function primarily by removing hemicellulose an d lignin. Since the hemicellulose and lignin fractions are associated with in the cellulose matrix, the methods which remove either hemicellulose or lignin increase cellulose accessi bility and make it susceptible to enzymatic hydrolysis (Saha 2003; Teymouri et al. 2004).

PAGE 19

19 AFEX is a physico-chemical pretreatment in which lignocellulosic materials are exposed to liquid ammonia at high temperature and pressure for a period of time, and then the pressure is quickly reduced to induce the ammonia to expl ode. The explosion significantly improves the enzymatic saccharification rates of various herb aceous crops and grasses. However, the AFEX pretreatment does not significantl y solubilize hemicellulose compar ed to acid pretreatment and acid catalyzed steam explosion (Meshart ree et al. 1988; Vlasenko et al. 1997). The effects of lime pretreatment depend on th e lignin content of the materials (Hu and Wen 2008). The mechanism of this pretreatment method is saponification of intermolecular ester bonds which crosslink xylan hemicellulose and other components. By removal of these crosslinks, the enzymatic accessibility of the tr eated lignocellulosic materials increases. Dilute NaOH pretreatment of lignocellulosic materials ca uses swelling of materi als, which leads to internal surface area increase, crystallinity decrease, linkage separation between lignin and carbohydrates and lignin structure di sruption (Tarkow and Feist 1969). Concentrated sulfuric and hydrochloric acid s have been used to treat lignocellulosic materials. Although they are powerful agents for cellulose hydrolysis, concentrated acids are toxic, corrosive and hazardous. Pretreatments with concentrated acids also require expensive corrosion-resistant reactors. In addition, this me thod is not very economical, due to the higher cost of concentrated acids (vonSivers and Zacchi 1995). Dilute Acid Pretreatment Dilute acid pretreatment is currently developed as a pref erred pretreatment method (Saha et al. 2005a; Saha et al. 2005b). This met hod employs mild acid conditions (0.5-3.0% H2SO4) with temperatures from 130C to 200C and pressu res from 3 atm to15 atm. There are two major types of dilute acid pretreatment processes: high temperature (>160 C), continuous flow process for low solid loading (5 10%) (Converse et al., 1989) ; and low temperature (<160 C), batch

PAGE 20

20 process for high solid loading (10 0%) (Esteghlalian et al. 1997). In dilute acid hydrolysis, the acid, temperature and pressure f unction together to release single sugars and small oligomers from the hemicellulose portion almost completely. In addition, dilute acid treatment disrupts and exposes the cellulose portion for effective enzymatic hydrolysis. The dilute acid treatment was reported to convert approximately 83% of the to tal xylose content to fr ee xylose and slightly more than 2% to xylooligomers. The release of hemicellulose proportion may be the major factor for the nearly complete hydrolysis of cellulose (~92%) with the subseque nt cellulase treatment (Eggeman and Elander 2005). In this process, the high xylan to xylose conversion yields is also economically favorable because xylan accounts for up to a third of the total carbohydrate in most lignocellulosic materials (Hinman et al. 1992). There are some drawbacks and limitations of the dilute acid pretreatment method. The most detrimental feature is the formation of aci d hydrolysis side products such as furfural and -1,2-methylglucuronoxylose (Jones et al. 1961; Zaldivar et al. 1999). Furfural forms from the acid and heat catalyzed dehydration of xylose. The formation of this side product inhibits microbial growth and fermentation and reduces the net convertible xylose concentration (Zaldivar et al. 1999). The aldouronate, -1,2-methylglucuronoxylos e (MeGAX), is produced from acid hydrolysis of methylglucuronoxylan and methylglucuronoarabinoxylan due to the stability of the -1,2 glycosyl linkage betw een methylglucuronate and xylose. This linkage is thought to form an internal lactone between the carboxylate residue on the glucuronic acid and a hydroxyl on the xylose in main chain under acidic conditions (Jones et al. 1961). While the arabinose and acetyl substitutions are released, the substitu ted aldouronate, MeGAX, is not hydrolyzed in the pretreatment. It is also unable to be ut ilized by current ethanologenic biocatalysts (Rodriquez et al 2001). Considering the fr equency of substitution of

PAGE 21

21 methylglucuronate is from 1 for every 10 to 1 for every 6 xylose residues, current bioconversion process can only recover 83% to 90% of the to tal xylose fraction. This suggests a biocatalyst capable of utilizing the current ly unusable MeGAX portion of the dilute acid hydrolysate will increase the efficiency of bioconversion started with dilu te acid pretreated biomass. Cost of Pretreatment Methods Research demonstrated none of these pret reatment methods had a clear advantage economically (Eggeman and Elander 2005). The direct fixed capital for th e dilute acid, AFEX, ARP, and lime methods are approximately the sa me. As for the dilute acid method, the primary cost is associated with equipments required to handle corrosive conditions. Thus the pretreatment reactors account for most of the cost, and a mi nor cost is associated with chemical supply requirements. AFEX pretreatment, on the other hand, requires costly pure ammonia, but less expensive reactors and equipments. Dilute aci d, AFEX and lime pretreatments result in the lowest total fixed capital per gallon of annual ethanol production capacity making these the most preferred methods for large scale pretreatment cu rrently. The cost for both dilute acid and AFEX pretreatment is calculated to be $3.72 pe r gallon ethanol (Eggem an and Elander 2005). Development of Ethanol Producing Biocatal ysts toward Hemicellulose Utilization The development of industrially suitable mi croorganisms for converting biomass into fuel ethanol is a major technical roa dblock for a mature bioethanol industry. Over the last several decades, microorganisms are developed as biocatal ysts to convert lignocellulosic sugars into ethanol (Ingram et al. 1999b; Je ffries 2005). Microorganisms capable of fermenting sugars not fermentable by brewers' yeast are needed to utilize complex single carbohydrates or sugar oligomers derived from lignocellulosic biomass. The greatest successes have been achieved in the engineering of Gr am-negative bacteria: Escherichia coli Klebsiella oxytoca and Zymomonas mobilis (Dien et al. 2003).

PAGE 22

22 Escherichia coli E. coli has several advantages as a biocatalys t for ethanol production. These advantages include the ability to ferment a wide range of s ugars, no requirements for complex growth factors, well established genetic manipula tion methods and prior industrial usage. The construction of E. coli strains to selectively produce ethanol is one of the first successful app lications of microbial metabolic engineering. Deve lopment of ethanologenic E. coli includes a combination of genetic engineering and metabolic adaption evolution. The adh B and pdc genes were co-expressed under the control of the native lac promoter, and the construct was named the PET (production of ethanol) operon (Ingram et al. 1987). The PET ope ron was inserted into the pyruvate formate lyase gene ( pfl ) to eliminate this enzyme competing fo r pyruvate. The integration recombinants were further screened for increased chloram phenicol resistance in order to enhance the expression level of PET operon gene products. In addition, the fumarate reductase gene ( frd ) in the succinate fermentation pathway was disrupted to eliminate succinate production. The final strain, E. coli KO11, is able to convert glucose and xylos e to ethanol at yiel ds 103 to 106% of theoretical maximum in rich media (Ingram et al. 1999). Long-term adaptation evolution method was used by Yomano et al in 1998 on medium supplemented with ethanol to increase the et hanol tolerance of KO11 by approximately 10%. E. coli LY01A was an isolate from an adapted cultu re. The adaptation reduced time to ferment 140 g/l xylose from 120 h by E. coli KO11 to 96 h by E. coli LY01A (Yomano et al. 1998). E. coli LY01 also has a higher tolerance to hydrolysat e-associated inhibito rs, including aldehydes, alcohols and organic acids (Zaldiva r and Ingram 1999; Zaldivar et al. 1999; Zaldivar et al., 2000). While both KO11 and LY01 achieve high ethanol yiel ds and titers in rich media, both strains perform poorly in minimal media (Jarboe et al 2007; York and Ingram 1996). To enable ethanologenic E.coli to ferment in inexpensive minimal media, a new strain was constructed

PAGE 23

23 from strain SZ110, which was developed from KO11 for lactate production in mineral salts medium. Strain LY160, with the restored pflB gene and PET operon inserted down stream of the rrlE gene promoter of SZ110, has higher yield of ethanol in minimal medium than previously constructed ethanologenic E.coli (Yomano et al. 2007). Non-recombinant ethanologic E.coli was also developed by is olating mutant strains capable of anaerobic growth derived from ldhA pflB double deletion strains. The selected strain SE2378 ferments glucose and xylose to etha nol with an 82% yield (Kim et al. 2007). It was found this strain had an essential mutati on within the pyruvate de hydrogenase operon to enable this enzyme to be ac tive in anaerobic condition. With th is enzyme, an additional NADH is produced compared to pyruvate formatelyase from one pyruvate. With activation of pyruvate dehydrogenase, a homo-ethanologenic pathway in E. coli is oxi-reductively balanced. E. coli based biocatalysts are still under development to enhance their traits for better industrial application. Zymomonas mobilis Zymomonas mobilis ( Z. mobilis ) is a Gram-negative microorganism with appealing properties as a biocatalyst fo r ethanol production. It has a hom oethanologenic fermentation pathway and high ethanol tolerance. Even compared with Saccharomyces sp. it has a higher ethanol yield and specific ethano l productivity (Rogers et al. 2007). Zymomonas is also the only microorganism that metabolizes glucose an aerobically using the Entner-Doudoroff (ED) pathway but not the Embden-Meyerhoff (EM) or pentosephosphate pathway. Because less ATP is produced by the ED pathway, Zymomonas produces relatively less bi omass and more carbon is converted to fermentation products than bacteria using EM pathway. However, Z. mobilis has a major disadvantage for biomass conversion. This or ganism has a limited substrate range, glucose, fructose and sucrose (Sprenger 1996).

PAGE 24

24 Strains of Z. mobilis capable of fermenting xylose and arabinose were developed to extend their substrate range. The genes of xyl ose isomerase and xylulose kinase from E.coli were expressed on a plasmid using strong constitutive promoter from Z. mobilis (Lawford and Rousseau 1991). The transformed strain CP4 (pZB5) is able to ferment on xylose with an ethanol yield of 86%. A newer strain, Z. mobilis AX101, ferments both arabinose and xylose with necessary recombinant genes integrated in chromosomal DNA (Lawford and Rousseau 2002). However, AX101 ferments arabinose more slowly than xylose, and the fermentations are often incomplete. This strain also has low tolerance to acetic acid, especially in the presence of ethanol. Z. mobilis continues to be developed in many labs and companies as a potential ethanologenic biocatalyst. Klebsiella oxytoca Klebsiella oxytoca is an enteric bacterium isolated fr om paper and pulp streams and some other sources of wood. This microorganism is capab le of growing at a pH as low as 5.0 and at 35C. K. oxytoca has the potential to be developed as a biocatalyst for cellulose bioconversion, since it grows on a wide variety of sugars includ ing hexoses and pentoses, as well as cellobiose and cellotriose (Freer and Detr oy 1983; Wood and Ingram 1992). Si milar to the construction of E. coli KO11, the PET operon was integrated in the pfl gene of K. oxytoca chromosome along with a cat marker. The best isolate from recombinants, K. oxytoca P2, ferments glucose (100 g/l) or cellobiose (100 g/l) to ethanol with yields of 44 g/l within 48 h (Wood and Ingram 1992). K. oxytoca P2 was further engineered to expres s heterogeneous cellulases for reducing enzyme cost. Two extracellular endoglucanase genes (CelZ and CelY) from Erwinia chrysanthemi were integrated into the chromosomal DNA of K. oxytoca P2 with the required auxiliary transporter gene e xpressed on a plasmid (pCPP2006). However, the engineered

PAGE 25

25 cellulolytic K. oxytoca SZ21 strain still utilized cellulo se (Sigmacell 50) poorly without supplemention of additional cellulases (Zhou et al. 2001; Zhou and Ingram 2000). Yeasts The traditional organism used for ethanol production, yeast, is also an important engineering target for developi ng biocatalysts for li gnocellulose c onversion. Since most yeasts do not ferment xylose, the prevailing attempts are to engineer yeast to utilize xylose. With the introduction of genes encoding xylose reductase (XR) and xylitol dehydrogenase (XDH) from Pichia stipitis into Saccharomyces the recombinant strain still produces ethanol at low yield from xylose with low growth rate (Amore et al 1991; Slininger et al 1985). There are more problems in genetically engineer ing yeast than bacteria due to the biological gap between prokaryotes and eukaryotes, such as the difference in internal pH between bacteria and yeasts, incorrect protein folding, incorrect post-translati onal modification and cofactor imbalance. These problems were reported to cause low ethanol yield, and low specific growth rate for yeast as biocatalysts for fuel ethanol production from lignocellulosic bi omass (Dombek and Ingram 1987; Jeffries and Jin 2004; Ostergaard et al. 2000). However, improvements in ethanol productivity and yield were achieved recently by several research groups. Matsushika and Sawayama engineered Saccharomyces strains by optimizing the XR/XDH/ xylulokinase (XK) ratio (Matsushika and Sawayama, 2008). They also inr oduced an XDH with altered coenzyme affinity into and XR and XK expressing Saccharomyces strain to enhance th e xylose fermentation (Matsushika et al., 2008a,b). Xylose transporters in Trichoderma reesei (Xlt1) were heterologously expressed in Saccharomyces strains expressing XR, XDH and XK (Hector et al., 2008) to increase xylose utiliza tion. Future improvements in et hanol productivity are possible with better understanding of the xylose utilization mechanism in yeast and new metabolic engineering methods.

PAGE 26

26 Objectives of this Research The goal of this study is to develop biocatalys ts to utilize the curre ntly unusable MeGAX in the dilute acid hydrolysate of hemicellulose to increase the efficiency of hemicellulose bioconversion. A new Enterobacter asburiae strain JDR-1( E. asburiae JDR-1) capable of MeGAX utilization was isolated by our lab for the purpose of MeGAX utiliz ation research and metabolic engineering. Most E. asburiae species isolated are short motil e rods with peritrichous flagella (Hoffmann et al. 2005). Some of the early isolated Enterobacter asburiae species were characterized as human pathogens (Brenner et al. 1986). However, some other strains of E. asburiae were isolated from soil environments with implicated activity to mobilize phosphate from calcium phosphate for plant nutrition (Sharma et al. 2005). Enterobacter is also one of the most commonly isolated bacterial genuses as plant endophytes. E. asburiae was found as an endophyte in cotton, cucumber, common bean, rice and sweetpotato (McInroy and Kloepper 1995; Elbeltagy et al. 2001; Asis and Adachi 2004). Enterobacter asburiae colonizes different plant species and establishes endophytic populations in various tissues. For example a systemic colonization of E. asburiae JM22 in cotton plants was reported to specifically locate in the root surfaces, within epidermal cells, a nd inside intercellula r spaces of the root cortex close to the conducting elements (Quadt-Hallmann and Kloeppe r 1996). Besides gaining entrance to plants through natural openings or wounds endophytic bacteria appear to actively penetrate plant tissues using hydrolytic enzymes like cellulases an d pectinases. However, once inside the plants, endophytes seem to utilize exudates that enter the in tercellular spaces as s ubstrates rather than break down the more complex cell wall materials as the major food source (Hallmann et al. 1997).

PAGE 27

27 Scope of this work is : 1) Evaluate the potential of E. asburiae JDR-1 for efficient conversion dilute acid hydrolysat es and characterize its metabo lic properties to understand the MeGAX utilization process. 2) Genetically engineer E. asburiae JDR-1 to homogeneously produce ethanol or other fermentation products 3) Define the MeGAX utilization process through identification of gene s involved, which is also for engineering current Enterobacteraciae biocatalysts to utilize MeGAX in hemicellulose hydrolysate.

PAGE 28

28 Figure 1-1. Model for corn stover cell walls (Saha 2003) With permission from Springer Science + Business Media. Copyright @ Springer Science + Business Media

PAGE 29

29 CHAPTER 2 COMPLETE FERMENTATION OF XYLOSE AND METHYLGLUCURONOXYLOSE DERIVED FROM METHYLGLUCURONOXYLAN BY ENTEROBACTER ASBURIAE STRAIN JDR-1 Introduction The quest for alternatives to petroleum res ources for production of fuels and chemicals has been motivated in part by economic incentives associated with limited and diminishing supply, and from the recognition of the contribution that increasing leve ls of carbon dioxide derived from these resources have contributed to gl obal warming (Kheshgi et al. 2000). Microbial fermentation products derived from resources re newable through photosynthes is are particularly attractive alternatives that may be generated with existing technologies (McMillan 1997). The development of yeast and bacteria l biocatalysts has been applie d to the commercial production of ethanol as an alternative fuel from starch and sucrose derived from commodity crops, e.g. corn and sugarcane (Dien et al. 2003; Jarboe et al. 2007). To e xpand production of ethanol and chemical feedstocks from renewable resour ces that do not economically impact these commodities, lignocellulosic resources, including forest and agricultural residues, have become targets for bioconversion of cellulose and hemicellu lose to fermentable suga rs (Jarboe et al. 2007; Shanmugam and Ingram 2008; vonSivers and Zacchi 1996). Current industrial methods for pretreatment of lignocellulose for bioconversion to ethanol solubilize the hemicellulose fraction by dilute acid hydrolysis, releasing the pentoses for fermentation (Preston et al. 2003). The predominan t structural polymer in the hemicellulose fraction of hardwoods and crop residu es is methylglucuronoxylan (MeGAXn), a -1,4 linked xylan in which xylose residues are periodically substituted with -1,2-linked 4-O-methyl-glucuronic acid (MeGA) (P reston et al. 2003). Resistance of the -1,2 glucuronosyl linkages to dilute acid hydr olysis results in the rel ease of the aldobiuronate,

PAGE 30

30 methylglucuronoxylose (MeGAX), which is not ferm ented by bacterial biocatalysts currently used to convert hemicellulose -derived xylose to ethanol, e.g., E. coli KO11(Preston et al. 2003; Rodriguez 2001; Ingram et al. 1999). The fr equency of MeGA substitutions on the xylose residues of methylglucuronoxylan ranges from less than one in ten in crop residues to one in six to seven in hardwoods, sweetgum and yellow po plar, and as much as 18% to 27% of the carbohydrate may reside in this unfermentable fr action (primarily as MeGAX) following dilute acid pretreatment (Rodriguez 2001). A scheme fo r the release of xylose and MeGAX by dilute acid hydrolysis of sweetgum xylan is depicted in Fig. 2-1 below. In this chapter, I describe the isola tion and characterization of a bacterium, E. asburiae JDR-1, with the ability to ferment MeGAX as the sole carbon source. E. asburiae JDR-1 was defined with respect to its meta bolic potential for the conversi on of MeGAX, xylose and glucose to specific fermentation products. As a Gram-negative member of the Enterobacteriaceae the genetic basis for this metabolic potential may be applied to the further development of bacterial biocatalysts for conversion of hemicellulose-derived pentoses to biobased products. Materials and Methods Preparation of Substrates and Culture Media Sweetgum methylglucuronoxylan (MeGAXn) was prepared from sweetgum stem wood ( Liquidambar styraciflua ) as previously described and characterized by 13C-NMR (Hurlbert and Preston 2001; Kardosova et al. 1998). Dilute acid hydrolysates of methylglucuronoxylan were prepared by hydrolysis with 0.1 N H2SO4 (4 g methylglucuronoxylan in 400 ml 0.1 N H2SO4) at 121 C for 60 min, followed by neutralization with BaCO3. Anion exchange resin (Bio-Rad AG2-X8) in the acetate form was used to adso rb the charged aldouronate s; the uncharged xylose and xylooligosaccharides, mainly small amounts of xylobiose, were eluted with water. The aldouronates were then eluted with 20 % (v/v) acetic acid. After concen tration under vacuum at

PAGE 31

31 50 C, aldouronates were separated on a 2.5 cm x 160 cm BioGel P-2 column (BioRad, Hercules, CA) with 50 mM formic acid as the eluent. The formic acid was removed from the purified sugar sample fractions by lyophilization. MeGAX and MeGAX2 were identified by thin layer chromatography (TLC) analysis using MeGAX and MeGAX2 standards structurally defined by 13C and 1H-NMR spectrometry (Zuobi-Hasona 2001). Xylobiose and xylotri ose were obtained and purified from MeGAXn digested with Paenibacillus sp. strain JDR-2 XynA1 catalytic domain (CD), a recombinant GH10 endoxylanase XynA1 CD overexpressed in E. coli The substrate containing 30 mg/ml MeGAXn was prepared with 10 mM sodium phosphate buffer, pH 6.5. Digestions were initiated by the addition of 3.5 U of XynA1 CD into 50 ml substrate and incubated with rocking at 30C for 24 h. An additional 1 U was added after 24 h and incubation was continued for 40 h. Aldouronates, xylobiose, and xylotrios e were separated with the P2 column and identified by TLC (St John et al. 2006). Total carbohydrate conc entrations related to substrate preparations were determined by th e phenol-sulfuric acid as say (Dubois et al. 1956), with xylose as the reference. The conditions of acid hydrolysis generated mostly MeGAX and a small amount of MeGAX2 from MeGAXn, with no aldouronates larger than MeGAX2 detected. MeGAX3 was prepared from GH10 endoxylanase-ca talyzed depolymerization of sweetgum MeGAXn and then purified by gel filtration on BioGel P4 (St John et al. 2006). Minimal medium containing the substrates described above wa s prepared upon mixing sterile substrate solutions (2x c oncentration) with the same vol ume of a 2x solution of Zucker and Hankin mineral salts (ZH salts) at pH 7.4 (Zucker and Hankin 1970). Neutralized MeGAXn acid hydrolysate (0.5% w/v) was also added to ZH salts directly as a growth substrate. Where indicated, some media preparati ons were supplemented with 0.1% yeast extract (YE medium).

PAGE 32

32 Isolation and Identification of E. asburiae JDR-1 E. asburiae JDR-1 was isolated from di scs of sweetgum stem wood ( Liquidambar styraciflua ) buried, soon after cutting, about one inch below the soil surface in a sweetgum stand for approximately three weeks. Discs were su spended in 50 ml sterile deionized water and sonicated in a 125 Watt Branson Ultrasonic Clea ner water bath for 10 min. The sonicate was inoculated into 0.2% (w/v) MeGAX YE medium a nd incubated at 37C. Cultures were streaked on MeGAX minimal medium agar plat es. Isolated colonies were passed several times between MeGAX broths and agars until pure. Exponen tial phase cultures growing on 0.2 % MeGAX minimal media were cryostored in 25% sterile glycerol at -70C. The purified isolate was submitted to MIDI Labs (http://www.midilabs.com) for partial 16s rRNA sequencing and FAME analysis. BBL EnterotubeTM II (Becton, Dickinson and Company, USA) inoculation was also used to identify the isolate based upon metabolic capability using the standard protocol. The Differential Interfer ence Contrast (DIC) microscopic photos of E. asburiae JDR-1 growing in MeGAX minimal medium at exponential phase were obtained with a Zeiss DIC microscope at 40x15-fold magnification. The nega tive stain electron microscopic photo was taken with a Zeiss EM10A electron microscope. Substrate Utilization and Fermentation Product Analysis Growth and substrate utilization analysis wa s performed in cultures actively aerated by shaking. For preparing inocula, cultures of E. coli B (ATCC 11303) and E. asburiae JDR-1 from cryostored samples were directly streaked on Lu ria-Bertani (LB) agar plates. After overnight incubation at 37C, isolated co lonies were picked to inocul ate liquid media specified for a particular experiment. Growth studies were performed at 37C in 16 mm x 100 mm test tubes containing 6 ml medium. Optical densities of cultures were measured at 600 nm (OD600) with a Beckman DU500 series spectrophotometer. The correlation of cell density and OD600 is CDW/L

PAGE 33

33 (g cell dry weight/L)=0.49*OD600+0.02 was experimentally determined. Sample dilutions were made to obtain OD600 readings between 0.2 and 0.8 abso rbance units which, corrected for dilution factors, provided turbidity values for gr owth studies. Individual 6 ml cultures for study were inoculated with 12 l (0.2% volume) of overnight cultures and maintained at 37C with constant shaking (Eberbach shaker set at low). Batch fermentations under anaerobic conditi ons at 37C were conduc ted in 13 mm x 100 mm screw cap tubes containing 3.0 ml medium. Inocula (0.5% [v /v]) were from overnight aerobic cultures grown in the same medium. After inoculation, nitrogen gas was used to flush and saturate the sealed batch culture. The tubes we re set in a Glas-Col mi nirotator at 60 rpm. For analysis of substrates and fermentation products, the culture supernatants were passed through 0.22 um filters and subjected to HPLC an alysis. Products were resolved on a Bio-Rad HPX-87H column with 0.01 N H2SO4 as the eluent at 65C. Samp les were delivered with a 710B WISP automated injector and chromatography c ontrolled with a Waters 610 solvent delivery system at flow rate of 0.5 ml/min. Products we re detected by differential refractometry with a Waters 2410 RI detector. Data analysis was pe rformed with Waters Millennium Software. To determine and quantify methanol, unfiltered supernatants from fermentation cultures were also analyzed by gas chromatography (6890N Networ k GC system, Agilent Technologies), using isopropanol as an internal standard. This de tection method was used since diffusion during HPLC precluded quantitative detection of methanol by differential refractometry. Determination of Metabolic Pathways by 13C-NMR The central metabolic pathways utilized by E. asburiae JDR-1 during glucose and xylose fermentation were evaluated with 13C-NMR (Scott and Baxter 1981). Cultures were grown in LB medium to mid-exponential phase at 37C. Cult ures (0.5 ml) were centrifuged and the cells washed with 2x ZH salts so lution. The cell pellets were suspended in 1.0 ml 0.5% [2-13C]xylose

PAGE 34

34 (99% enrichment; Omicron Biochemicals In c, IN) in ZH minimal medium. Similar fermentations were also prepared with 1.0 ml 0.5% [1-13C]glucose, or 1.0 ml 0.5% [6-13C]glucose ZH minimal medium using D-[1-13C]glucose or D-[6-13C]glucose (99% enrichment; Cambridge Isotope Laboratories, An dover, MA). Fermentations were carried out under anaerobic conditions at 37C for 8 hours. Cells were removed by centrifugation, and the supernatants analyzed by HP LC (after filtration) and 13C-NMR spectrometry. NMR spectra were obtained using a VXR300 NMR spectrometer (NMR facility of the Department of Chemistry, University of Florida) operati ng in the Fourier tran sform mode as follows: 75.46 MHz; excitation pulse width, 7.0 s; pulse repe tition delay, 0 s; spectral wi dth, 16502; 256 acquisitions. Acetone (30 l) containing 13C at natural abundance in 700 l sample was used as an internal reference of 31.07 ppm for the 13C methyl carbon (Kardosova et al. 1998). Individual carbon atoms for fermentation products were iden tified by shift assignments and quantified by comparison with standards (13C at natural abundance) of known concentrations. Determination of Molar Cell Dry Weight Yield For molar growth yield experiments (Bauc hop and Elsden 1960; Smalley et al. 1968; Gunsalus 1961), anaerobic growth was performe d in 50 ml minimal medium containing either 0.26% glucose, 0.36% xylose, 0.35% glucuronate or 0.2% MeGAX as sole carbon source with the fermentation conditions described above. After 24 hours of growth and complete utilization of the carbon source, cells were harvested by centrifugation and the resulting pellets were washed twice with deionized water. The pellets were dried to constant weight in a Sargent vacuum dryer at 60C for up to 36 hours. The cu lture supernants were analyzed by HPLC to determine substrate consumption. The molar cell dry weight yield was calculated as cell dry weight (gram) divided by consumed substrate (mole).

PAGE 35

35 Results Identification and Characterization of E. asburiae JDR-1 A novel bacterial strain able to grow on MeGAX minimal medium was obtained by John D Rice of our lab (Bi et al 2009) a nd subsequently identified with three tests. The partial 16S rRNA sequence (accession number EU117142, Gene Bank, NCBI), amplified using primers corresponding to E. coli 16S rRNA positions 005 and 531 (526bp), provided an alignment with 99.5% identity within the sequence of Enterobacter asburiae (MIDI Aerobic Bacteria Database version 4.0, January 1999). Results of FAME (fatty acid methyl ester) analysis indicated this strain had the greatest similarity index with Enterobacter asburiae species (0.766) compared with any other entry in the MIDI database. A biocode of 32061, obtained from the Enterotube II (BBL) test, also corresponded to Enterobacter asburiae species. Based upon these three criteria, the isolate was identified within Enterobacter asburiae species and designated as Enterobacter asburiae strain JDR-1. The strain has been depos ited with the Agriculture Research Service Patent Culture Collection of the USDA, Pe oria, IL., under NRRL number NRRL B-S0074. When exponential phase cultures were observed by optical DIC microscopy, E. asburiae JDR-1 appeared as short motile rods (Fig 2-2 A). Negative stain electr on microscopy revealed 3 m x 1 m cells with peritrichous flagella (Fig 2-2 B). The microscopy study was done the Laboratory for Electron Microsc opy and Imaging, Department of Microbiology and Cell Science and kindly assisted by Ms. Donna Williams. These morphological characteristics were similar to those of other isolates of Enterobacter asburiae (Hoffmann et al. 2005). When grown on LB agar plates, colonies of E. asburiae JDR-1 were morphologically indistinguishable from E. coli colonies.

PAGE 36

36 Utilization of Acid Hydrolysates of Methylgluronoxylan by E. asburiae JDR-1 The unique ability of E. asburiae JDR-1 to grow on the aldob iuronate MeGAX as the sole carbon source suggested a potential for the comp lete metabolism of the carbohydrates generated by the dilute acid pretreatment currently applied for the releas e and fermentation of xylose in hemicellulose fractions. To evaluate this potential, E. asburiae JDR-1 was grown aerobically in minimal medium comprised of neutralized MeGAXn acid hydrolysate and Zucker and Hankin mineral salts. Based upon HPLC analysis of me dia samples taken during different stages of growth, E. asburiae JDR-1 utilized MeGAX completely in minimal media with MeGAXn hydrolysate after it depleted xylose (Fi g. 2-3 A). Biphasic growth occurred as E. asburiae JDR-1 switched from utilizing xylose to MeGAX. In contrast to E. asburiae JDR-1, E. coli B consumed only the free xylose with the MeGAX concentr ation in the medium remaining constant. Concentrations of xylose and MeGAX in hydrolysate medium of sweetgum MeGAXn, as determined by HPLC, were 0.206% w/v a nd 0.036% w/v, respectively. Therefore, E. asburiae JDR-1 utilized 17.5% more substrate (mass amount) than E. coli B, which was unable to utilize MeGAX (Fig. 2-3 B). Under aerobic conditions, both E. asburiae JDR-1 and E. coli B formed acetic acid during exponential growth phase that was metabolized upon comp lete utilization of the carbon sources in the MeGAXn hydrolysates. E. asburiae JDR-1 was also able to grow in xylobiose and xylotriose minimal medium, which E. coli B could not utilize. However, E. asburiae JDR-1 was unable to utilize MeGAX2 and MeGAX3 (data not shown). Substrate Preference of E. asburiae JDR-1 E. asburiae JDR-1 was found to grow aerobically in minimal medium containing different sole carbon sources, such as glucose, xylose, mannitol, maltose, rhamnose, mannose, glucuronate and glycerol. As noted above, it was able to quantitatively me tabolize MeGAX, but was unable to utilize MeGAX2 generated by acid hydrolysis, or MeGAX3 generated by a GH10

PAGE 37

37 endoxylanase. When growing in a minimal medium containing an eqimolar mixture of glucose and xylose, E. asburiae JDR-1 displayed a diauxic growth pattern typical of species of Enterobacteraceae (Fig. 2-4 A). Glucose (8 mM) was consumed within approximately 8 hours, while xylose utilization began when glucose was almost entirely consumed and was depleted in 14 hours. To study the process by which MeGAX was utilized, E. asburiae JDR-1 was grown in minimal medium containing both xylose and glucur onate, products that might be generated from MeGAX. A single phase growth curve was observed in which both substrates were consumed by 15 hours (Fig. 2-4 B). This is similar to its si ngle phase growth curve on MeGAX, in which the 6.5 mM substrate was depleted in about 11 hours (F ig. 2-4 C). The similarity in growth pattern with MeGAX and the combination of xylose and glucuronate as carbon s ources supports the possibility that free glucurona te and free xylose may be rel eased during the metabolism of MeGAX. Fermentation Characteristics Fermentation experiments were perfor med to evaluate the potential of E. asburiae JDR-1 as a biocatalyst for the production of biobased produc ts, and define the processes involved in the metabolism of MeGAX. Using limiting amounts (0.25% w/v) of substrates and cultivation under anaerobic standing conditions, E. asburiae JDR-1 was able to ferment all major sugars constituting hemicellulose, including D-gluc ose, D-xylose, D-mannose, L-arabinose and D-galactose. The major products from xylose an d galactose fermentation were acetic acid and ethanol present in similar molar quantities. Ace tic acid, ethanol and small amounts of lactic acid were produced from glucose, mannose and arabin ose (Table 2-1). Small amounts of formic acid and very small amounts of fumaric and succinic aci ds were detected in most fermentations. The

PAGE 38

38 HPLC profiles indicate that E. asburiae JDR-1 performs mixed acid fermentation as does E. coli but with preferential formation of acetate and ethanol over lactate. With sweetgum MeGAXn hydrolysate as substrate, E. asburiae JDR-1 consumed 99% of the substrate when the pH was maintained above 5, giving the major prod ucts acetic acid and ethanol (Table 2-2). With gl ucuronic acid as carbon source, acetic acid was the major fermentation product. To study the process of MeGAX metabolism, the presumed degradation products of MeGAX, xylose and glucuronate, both at 11 mM, we re used as substrates. The predominant products were 20.4 mM acetate and 5.25 mM ethanol. Major fermentation products from 4.0 mM MeGAX were 8.1 mM acetic aci d, 1.2 mM ethanol, and 4.3 mM methanol. Central Metabolic Pathways Determined by 13C-NMR The total quantities of ethanol acetate and lactate were determined by HPLC and the quantities of 13C labeled products were quantified from integration of differentially labeled compounds detected in the 13C-NMR spectra. This allowed determination of the fraction of each fermentation product that was differentially labeled with 13C, which helped to illustrate the central metabolic pathways E. asburiae JDR-1 uses. The quantitie s of each product and the fractions labeled with 13C are presented in Table 2-3. To determine the primary pathway of xylose metabolism by E. asburiae JDR-1, comparisons were made for the fermentation of [2-13C]xylose with cultures of E. coli B. For E. coli B, employing only the pentose-phosphate pathwa y to metabolize xylose, the prominent shift signals in the 13C-NMR spectrum of the fermentation products were assigned to [1-13C]ethanol at 57.6 ppm, [2-13C]lactate at 68.8 ppm, and [1-13C]acetate at 181.0 ppm. Sh ift signals at 71.0 ppm and 74.5 ppm were assigned to the and anomers of unused [2-13C]xylose, and the signal at 30.6 ppm to the methyl carbons of the acetone standa rd (Fig. 2-5 B). Fractions of labeled versus total acetate, ethanol, and lactate with E. coli B were 0.26, 0.27, and 0.31, respectively, which

PAGE 39

39 was slightly less than the theoretical fr action 0.4 expected for metabolism through the pentose-phosphate pathway (Table 2-3). The lower quantities of la beled products as fractions of the total found for E. coli may reflect accuracy limitations for integration against the 13C-acetone standard, as these products all showed similar fractions (0.26-0.31) were labeled. When E. asburiae JDR-1 fermented [2-13C]xylose, a 13C-NMR spectrum for fermentation products was obtained with prominent signals for [1-13C]ethanol, [2-13C]lactate, and [1-13C]acetate at expected shift posit ions (Fig. 2-5 A). The fractions of labeled ethanol, labeled acetate and labeled lactate to their total amount s were 0.43, 0.4 and 0.45, respectively (Table 2-3), and nearly identical to the theoretical fraction of 0.4. Moreover, the fractions of labeled acetate and ethanol were not higher than the fraction of labeled lactate. These results establish that the pentose phosphate pathway is the main me tabolic pathway for xylose utilization in E. asburiae JDR-1. To determine the primary pathway E. asburiae JDR-1 utilizes to metabolize glucose, [1-13C]glucose and [6-13C]glucose were used as fermentation substrates. Similar 13C-NMR spectra of fermentation products were obtained from [6-13C]glucose and [1-13C]glucose (Fig. 2-5 C, 2-5 D). Shift signals at 92.4 a nd 96.2 ppm were assigned to the and anomers of unused [1-13C]glucose (Fig 2-5 C); signals at 60.9 and 60.1 ppm were assigned to the and anomers of unused [6-13C]glucose (Fig. 2-5 D). The signal at 30.6 ppm was assigned to the methyl carbons of the acetone standard. Excepting the shif t signals for reference and unused substrates, the prominent signals in both spectra were for [2-13C]ethanol at 17.1 ppm, [2-13C]acetate at 22.2 ppm and [3-13C]lactate at 20.3 ppm with similar distribut ions for both substrates. The absence of [1-13C]lactate indicates that no [1-13C]glucose was metabolized through the Entner-Douderoff (ED) pathway. Moreover, the fractions of all labe led products of their total amounts were similar

PAGE 40

40 for fermentation of [6-13C]glucose and [1-13C]glucose; and these fractions for [6-13C]glucose were not higher than those found for [1-13C]glucose (Table 2-3), indicating little or no [1-13C]glucose went through the pentose-phosphate pa thway. Collectively, these results establish that the Embden-Meyerhof (EM) pathway is the main metabolic pathway for glucose utilization in E. asburiae JDR-1. Growth and Projected ATP Yields with Different Substrates To understand the bioenergetics in the process of MeGAX fermentation by E. asburiae JDR-1, molar cell dry weight yiel ds were determined after 24 hours of growth with glucose, xylose, glucuronate and MeGAX as sole carbon sources in Zuck er-Hankin minimal medium. The experiment was performed three times and the average approximate YM values were about 10 g per mole of substrate for growth on xylose and gl ucuronate, 20 g for growth on glucose, and 30 g for growth on MeGAX (Table 2-4). The experimental YATP in anaerobic growth has been reported in the range of 8 to 12 gram cell dry we ight per mole of ATP fo r bacteria (Russell and Cook 1995). An estimated YATP value at the lower end of this range, 8, was used here since this is for anaerobic growth in batc h cultures in minimal medium w ith a relatively low concentration of carbon source (BAUCHOP and Elsden 1960; Guns alus 1961). The apparent ATP yields per mole of substrate were calculated based on the estimated YATP of 8 as 1.3 mole of ATP produced from either xylose or glucuronate, 2.6 from glucose and 4.0 from MeGAX (Table 2-4). These apparent ATP yields allow an estimate of the relative ATP yields obt ained for the different substrates without considerations of maintena nce energy or overflow metabolism (Russell and Cook 1995), providing insight into the metabolism of MeGAX. The ratios of the molar growth yields obtained with xylose, glucuronate, and MeGAX as carbon sources are 1:1:3, indicating that the requirement for MeGAX transport is less than that for separate transport of xylose and glucuronate.

PAGE 41

41 Discussion Fermentation of MeGAX by E. asburiae JDR-1 E. asburiae JDR-1 is the first bacterium that ha s been described to ferment MeGAX. The complete utilization of MeGAX by E. asburiae JDR-1 clearly indicates a metabolic potential not shared with E. coli B, as well as the ethanologenic strains E. coli KO11 (Rodriguez 2001) and Klebsiella oxytoca P-2 (Qian et al. 2003). Based upon the studies presented here, E. asburiae JDR-1 uses the EM glycolytic pathway to metabolize glucose and the pe ntose phosphate pathway to metabo lize xylose. This is the first definitive study to determine the role of these pathways in any Enterobacter spp., although the findings are not unexpected as th e role of these pathways have been well defined in other Enterobacteriaceae (Moat 2003). The roles of these pa thways in central metabolism are supported by genome annotations in Enterobacter sp. 638 (Complete sequence of chromosome of Enterobacter sp. 638, NC_009436, NCBI genome database). The formation of equal molar amounts of ethanol and acetic acid from xylose under anaerobic conditions suggests that pyruvate serving as a precur sor for acetic acid and ethanol might be converted to acetyl CoA by pyruvate formate lyase (Moat 2003). The ability of E. asburiae JDR-1 to completely convert MeGAX to acet ic acid, ethanol an d methanol suggests that demethylation may occur as an early event in the metabolic processing. The metabolic potential of E. asburiae JDR-1 to ferment mannose, galactose, and glucose, as well as xylose and arabinose, defines its potential to convert all of the pentoses and hexoses that comprise hemicellulosic biomass to fermentation products. Central Metabolic Pathways Used by E. asburiae JDR-1 In E. coli the xylulose-5-phosphate enters th e pentose-phosphate pathway, yielding fructose-6-phosphate and glycer aldehyde-3-phosphate (Moat 2003). Fructose-6-phosphate can be

PAGE 42

42 further metabolized to glyceraldehyde-3-phosphate. [2-13C]xylose metabolism by the phosphoketolase pathway is expected to yield [1-13C]acetate or [1-13C]ethanol and unlabeled pyruvate; thus no label will be foun d in lactate (Gunsalus et al. 1955). However, if organisms use the pentose-phosphate pathway, the [1,2-13C] pyruvate produced is expected to account for one-fifth of the total pyruvate, with [2-13C] pyruvate accounting for anot her one-fifth of the total pyruvate. The remaining three fifths of the pyruvate do not carry the 13C label (Patel et al. 2006). [1,2-13C] pyruvate can be reduced to [1,2-13C]lactate or converted to [1-13C] ethanol or [1-13C]acetate by pyruvate formate lyase and subsequent enzymes; [2-13C] pyruvate can be reduced to [2-13C]lactate or converted to [1-13C]ethanol and [1-13C]acetate. Therefore, if an organism uses only the pentose-pho sphate pathway to metabolize [2-13C]xylose, both [1-13C]ethanol and [1-13C]acetate would account for two-fifths of total ethanol and acetate produced, whereas [1,2-13C]lactate and [2-13C]lactate together would account for two-fifths of total lactate. The fraction of la beled carbon two in lactate would be two-fifths of the total carbon two in lactate, the same fracti on as labeled acetate and ethanol to total acetate and ethanol. If both the phosphoketolase and pentose phosphate pa thways were used, label would be found in the lactate, but the ratio of labe led ethanol or acetate to total et hanol or acetate would be higher than ratio of labeled lactate to total lactate. Th e labeling patterns obtained in this study indicate that the pentose-phosphate pathway is the main if not exclusive metabolic pathway for xylose utilization in E. asburiae JDR-1. Three major pathways, Embden-Meyerhof (EM), pentose-phosphate (PP), and Entner-Doudoroff (ED) pathways, are used by bacter ia to catabolize glucose or other sugars into pyruvate (Moat 2003). To determine the primary pathway E. asburiae JDR-1 utilizes to metabolize glucose, [1-13C]glucose and were used as fermenta tion substrates. For either the EM

PAGE 43

43 or ED pathway, labeled carbon derived from [6-13C]glucose would be found in carbon three of the intermediate pyruvate. Following subsequent metabolic processing, [2-13C]ethanol, [2-13C]acetate and [3-13C]lactate would be produced. No labe led carbons would be found at the other carbon positions of these compounds. However, different pathways gi ve different labeled fermentation products for [1-13C]glucose. If [1-13C]glucose were utilized by the EM pathway, all labeled carbon would also be found at carbon th ree position of pyruvate and the amounts of labeled products and their ratios to total products would be the sa me as those derived from that of [6-13C]glucose. If the pentose-phosphate pa thway were utilized, carbon one of [1-13C]glucose would be oxidized to the carboxyl group in 6phosphogluconate (6PG) which would then be oxidized to CO2 by glucose-6-phosphate dehydrogenase (Moat 2003). After the two oxidation steps, no other labeled fermentation products would be obtained except for labeled CO2. If the ED pathway is used, carbon one of [1-13C]glucose would become carbon one in one of the two pyruvate molecules generated from 2-keto-3 -deoxy-6-phosphogluconate. In the subsequent fermentation process, [1-13C]pyruvate would be cleaved by pyruvate formate lyase, yielding acetate or ethanol without the 13C label, and [1-13C]pyruvate would be reduced to [1-13C]lactate. The HPLC and NMR data (Fig. 2-5 C and D, and Table 2-3) showed that both [6-13C]glucose and [1-13C]glucose gave similar products fermented by E. asburiae JDR-1 with similar distributions of labeled carbons. These results establish that the EM pathway is the main metabolic pathway for glucose utilization in E. asburiae JDR-1. Possible Pathway of MeGAX Metabolism The processing of aldouronates formed by the action of GH10 endoxylanases, e.g. MeGAX3, has been shown to involve the removal of the -1,2-linked 4-O-methylglucuronosyl moiety from the reducing terminus of -1,4-linked xylotriose by the action of a GH67

PAGE 44

44 -glucuronidase (Shulami et al. 1999). -Glucuronidases (GH67) from some bacteria have been shown to act on MeGAX as well as MeGAX3 (Nong at al. 2005; Shulami at al. 1999). Consensus primers based on GH67 family genes (http://www.cazy.org/fam/GH67.html) may be used to amplify aguA genes from bacterial genomic DNA prio r to partial sequencing and cloning. The application of this strategy was not successf ul in identifying a gene encoding a GH67 -glucuronidase in E. asburiae JDR-1 (G. Nong and C. Bi, unpub lished). Either this strain synthesizes an -glucuronidase completely different fr om the GH67 that has been well defined, or a novel metabolic process may be used in wh ich the cleavage of the glycosidic bond follows other reactions. The fermentation products derived from xylose and glucuronate were similar to those obtained from MeGAX, although th e ethanol yield from MeGAX was lower. This result suggested that xylose and glucur onate might be released in the process of MeGAX utilization. This possibility was also supported by the similarity in the growth patterns of E. asburiae JDR-1 growing in MeGAX minimal medium and the gluc uronate plus xylose minimal medium (Fig. 2-4). Due to the small amount of methanol observed re lative to system noise of the GC system, accurate quantitative data for methanol as a fermentation product was not obtained in media containing acid hydrolysates of MeGAXn. Methanol was quantitatively detected in stoichiometric equivalence with MeGAX during the metabolism of MeGAX, indicting me thanol is derived from the methyl group released fr om methylglucuronate as an ear ly event in the metabolism of MeGAX (Table 2-2). Based on the results obta ined from this study, the process of MeGAX catabolism by E. asburiae JDR-1 may involve the release of methanol, glycosidic bond cleavage to release glucuronate and xylos e, and catabolism of these carbohydr ates to generate ethanol and

PAGE 45

45 acetic acid as predominant fermentation products An elimination reaction for the release of methanol prior to glycosidic bond cleavage is a possibility, in wh ich case the product of glycosidic bond cleavage w ould be delta-4,5-hexuronate. Bioenergetics of E. asburiae JDR-1 Glucose was shown to be metabolized in E. asburiae JDR-1 by the EM pathway. If the PTS system is used to import glucose as in E. coli (Simoni et al. 1976), the net formation of 3 moles of ATP result from 1 mole of glucose when major fermen tation products are ethanol and acetate. For xylose utilization via the pentosephosphate pathway, presumably requiring one mole ATP to import xylose with an ABC transporter as in E. coli (Linton and Higgins 1998), 1.5 moles of ATP would be produced per mole of xy lose when fermentation products were ethanol and acetate (Hasona et al. 2004). Gl ucuronate has been shown to be transported into cells with the UxuT transporter in E. coli This transporter bears significant homology to ExuT in Erwinia chrysanthemi which has been defined as responsible fo r the transport of glucuronate as well as galacturonate through an active process in wh ich energy is consumed. Considering energy consumed in active transport, th e maximal ATP yield in the pro cess of glucuronate fermentation may be 2 moles or less per mole of glucuronate when the major fermentation product is acetic acid (Hugouvieuxcottepattat a nd Robertbaudouy 1987; Robertbaudouy et al. 1981; SanFrancisco and Keenan 1993). The theoretical ATP yields fro m glucose, xylose and gl ucuronate were close to the calculated ATP yields based on the estimated YATP, 8, suggesting that the presumed transportation systems and metabolism of these substrates are correctly interpreted. The ratios of the molar growth yields obtained with xylos e, glucuronate, and MeGAX as carbon sources are 1.0:1.0:3.2 (Table 2-4), clearly indicating that the bioenerg etic requirement for MeGAX transport is less than that for separate transport of xylose and glucuronate.

PAGE 46

46 Role of E. asburiae JDR-1 in Soil Ecology and Bioprocessing Most Enterobacter asburiae species have been identified and characterized as human pathogens (Brenner et al. 1986). However, strains of E. asburiae have been isolated from soil environments and implicated in the mobilizatio n of phosphate for plant nutrition from calcium phosphate (Sharma et al. 2005). The ability of E. asburiae JDR-1 to utilize MeGAX, as well as xylobiose and xylotriose, may refl ect its evolution in a soil environment with hemicellulose providing a carbon resource. With the exception of Erwinia species, the Enterobacteriaceae are not known for their ability to secrete endoxylanases and endoglucanases. For Erwinia spp., the secreted endoxylanases are members of glycohydrol ase family 5 that generate oligosaccharides all of which are substituted w ith MeGA residues, and are not further metabolized by the cell (Hurlbert and Preston 2001; Preston et al. 1992). The role of these enzymes in Erwinia species may be related to their colonization and macer ation of plant tissues, and thus may serve a virulence function rela ted to pathogenesis. Enterobacter spp. have been found associated with plant roots and may provide commensal, if not sy mbiotic, relationships that contribute to positive development of the plant (Sharma et al. 2005). However, none have been identified which secrete endoxylanases. E. asburiae JDR-1 is unable to utilize th e aldotetrauronate product of GH10 endoxylanase, MeGAX3, which serves as a substrate for several gram-positive bacteria that secrete GH10 endoxylanases. At present, the generation of the aldobiuronate MeGAX is only known to result from the acid hydrolysis of methylglucuronoxylans. Since MeGAX is the only aldouronate this bacterium will use, this metabolic potential has yet to be linked to the biological generation of aldouronates from hemicellulosic biomass. Biotechnological Applications The predominant structural polymer in the hemicellulose fraction of hardwoods and crop residues is methylglucuronoxylan (P reston et al. 2003), and depending on the source, as much as

PAGE 47

47 27% of the carbohydrate may occur in the form of MeGAX following dilute acid pretreatment (Rodriguez 2001). Capturing the xylose, including that found in MeGAX, in acid hydrolysates of hemicellulose will increase the efficiency of converting lignocellulosic biomass to targeted products. Our current resear ch on the metabolism of E. asburiae JDR-1 has established its potential for the engineering of gram-negative bacteria for bioconversion of aldouronates as well xylose and other pentoses and hexoses that may be derived from the hemicellulose fraction of any lignocellulosic source. The definition of the genetic basis for MeGAX metabolism may allow its application to the engineering of established bacterial biocatalysts, e.g. E. coli KO11, for the efficient conversion of cellulosic biomass to biofuels and chemical feedstocks.

PAGE 48

48 Table 2-1. Fermentation products formed by E. asburiae JDR-1 from monosaccharides derived from hemicellulosea. Fermentation products (mM) Substrate (0.25% w/v) Acetic AcidEthanolLactic Acid Formic acid D-Xylose 10.1.1 10.2.70 1.6.3 D-Glucose 7.2.3 9.7.51.8.1 1.6.4 D-Mannose 7.5.1 9.2.20.9.2 3.2.8 D-Galactose 9.0.4 9.0.30 0.7.3 L-Arabinose 8.1.2 9.5.10.8.2 1.3.3 a: Anaerobic cultures were allowed to cons ume each carbon source, initially at 0.25% w/v. Concentrations of components resolved by HPLC were determined for duplicate cultures by differential refractometry. Table 2-2. Fermentation products of E. asburiae JDR-1 derived from MeGAXn. Acetate, ethanol and formate concentrations were determin ed in duplicate cultures with HPLC and methanol with GC. a: Composition of the acid hydrolysate was determ ined by HPLC and differential refractometry. b: Due to background noise and very small product amounts, accurate data was not obtained for quantification of methanol in the MeGAXn hydrolysate. c: None detected. Products, mM, and yield (product/substrate) Substrate (mM) Acetate Ethanol Methanol Formate Xylose (14.3) 9.5.2 (0.7) 9.4 0.7(0.7) ND 1.6.3 Glucuronate (11) 12.8.6 (1.1)0 ND 2.8.4 MeGAX (4.0) 8.1.9 (2.0) 1.2 0.3(0.3) 4.3.0 (1.1) 0 MeGAXn acid hydrolysate (Xylose 13.7; MeGAX 1.75)a 14.3.2 9.8.1 <2.5b 0 Xylose (11) +Glucuronate (1 1) 20.4 0.5(1.9)5.3 0.4(0.5) ND 0

PAGE 49

49 Table 2-3. Distribution of 13C in fermentation products formed in anaerobic cultures of E. asburiae JDR-1 and E. coli B grown with differentially 13C labeled xylose and glucosea. a: Carbons enriched in 13C in different fermentation products were determined and quantified by 13C-NMR (Fig. 2-5) and are noted by *. Total Pr oducts were quantified by HPLC. The fractions of labeled products to their total products were calculated and noted parenthetically in the table. Table 2-4. Anaerobic molar cell dry weight and ATP yield from diffe rent substrates calculated based on estimated YATP, 8, for all substrates in E. asburiae JDR-1. a: YM-substrate: molar cell dry weight yields for different substrates, determined in triplicate with indicated standard deviations. Labeled products, mM, and (fraction) labeled with 13C Acetate Ethanol Lactate Fermentation CH3C*OOH CH3C*H2OHCH3C* HOHCOOH+CH3C*HOHC*OOH [2-13C]xylose, E.asburiae JDR-1 4.8 (0.40) 5.8 (0.43) 0.9 (0.45) [2-13C]xylose, E.coli B 3.0 (0.26) 1.9 (0.27) 2.8 (0.31) C*H3COOH C*H3CH2OHC*H3CHOHCOOH [1-13C]glucose, E.asburiae JDR-1 2.3 (0.34) 4.6 (0.37) 4.8 (0.38) [6-13C]glucose, E.asburiae JDR-1 1.9 (0.28) 4.7 (0.35) 5.4 (0.40) Fermentation substrates GlucoseXylose Glucuronate MeGAX YM-substrate (g/mole)a 20.5.410.2.7 10.4.3 32.0.1 Estimated ATP yield per mole of substrate 2.6 1.3 1.3 4.0

PAGE 50

50 Figure 2-1. Scheme for the release of xylose a nd MeGAX by dilute acid hydrolysis of sweetgum xylan. Dilute Acid Hydrolysis(Xyl-Xyl-Xyl-Xyl-Xyl-Xyl-Xyl-Xyl-Xyl-Xyl-Xyl-Xyl-)nMeGA MeGA Xyl+(SweetgumMethylglucuronoxylan)Xyl MeGA (Xylose) (MeGAX)

PAGE 51

51 A B Figure 2-2. Microscopic images of E. asburiae JDR1. A) Differential Inte rference Contrast (DIC) microscopic photos of E. asburiae JDR-1 growing in MeGAX minimal medium at exponential phase. B) Image of E. asburiae by negative stain electron microscopy.

PAGE 52

52 A. B. Figure 2-3. Aerobic growth, subs trate utilization, and format ion of products from acid hydrolysates of MeGAXn. by A) E. asburiae JDR-1 and B) E. coli B. Xylose (diamonds), MeGAX (squares), and acetic acid (triangles) were determined in media by HPLC. Growth was determined by measur ing turbidity as OD600 (open circles). 0 0.5 1 1.5 2 2.5 010203040 Time (h)Concentration (mg/ml)0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6Culture OD 0 0.5 1 1.5 2 2.5 010203040 Time (h)Concentration (mg/ml0 0.2 0.4 0.6 0.8 1 1.2Culture OD

PAGE 53

53 A Figure 2-4. Aerobic growth of E. asburiae JDR-1 on different combinations of sugar substrates. Concentrations of substrates and acetic ac id as a product were determined by HPLC. Growth was determined as turbidity ( OD600). A) Growth on glucose (7.5 mM) and xylose (7.5 mM). Concentrations of glucos e (closed circles), xylose (diamonds) and acetic acid (triangles); OD600 (open circles) ; B) Growth on glucuronic acid (10 mM) and xylose (10.5 mM). Concentrations of gl ucuronic acid (open squares) and xylose (diamonds); OD600 (open circles) C) Gr owth on MeGAX (6.5 mM). Concentrations of MeGAX (squares): OD600 (open circles). 0 2 4 6 8 10 051015 Time (h)Concentration (mM)0 0.2 0.4 0.6 0.8 1 1.2 1.4Culture OD

PAGE 54

54 B C. Figure 2-4. Continued 0 2 4 6 8 051015 Time (h)Concentration (mM ) 0 0.2 0.4 0.6 0.8 1Culture OD 0 2 4 6 8 10 12 051015Time (h)Concentrationn (mM)0 0.2 0.4 0.6 0.8 1Culture OD

PAGE 55

55 Figure 2-5. Pathway determination for the metabolism of xylose and glucose by E. asburiae JDR-1. Media from an aerobic cultures of E. asburiae JDR-1 and E. coli B grown with xylose or glucose enriched with 13C in specific carbons were analyzed by 75.5 MHz 13C-NMR spectrometry. A) [2-13C]-xylose fermented by E. asburiae JDR-1; B) [2-13C]-xylose fermented by E. coli B; C) [1-13C]-glucose fermented by E. asburiae JDR-1; D) [6-13C]-glucose fermented by E. asburiae JDR-1. A CH3C*OOH CH3C*H2OH CH3C*HOHCOOH

PAGE 56

56 B. C. Figure 2-5. Continued CH3C*OOH CH3CHH C*OOH CH3C*H2OH CH3C*HOHCOOH C*H3CH2OH C*H3CHOHCOOH C*H3COOH

PAGE 57

57 D. Figure 2-5. Continued Chapter 2 was adapted and modified from [APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Jan. 2009, p. 395], Copyright American Society for Microbiology. C*H3COOH C*H3CHOHCOOH C*H3CH2OH

PAGE 58

58 CHAPTER 3 GENETIC ENGINEERING OF ENTEROBACTER ASBURIAE STRAIN JDR-1 FOR EFFICIENT D(-) LACTIC ACID PR ODUCTION FROM HEMICELLULOSE HYDROLYSATE Introduction As one of three main components of li gnocellulosics, hemice llulose contains polysaccharides comprised of pentoses, hexoses and sugar acids that account for 20-35% of the total biomass from different sources (Saha 2003). Depending on the source, these fractions contain heterogeneous polymers comprised of pent oses, hexoses and sugar acids. The ability to utilize all hemicellulosic sugars is important for efficient conversion of li gnocellulosic materials to fuel ethanol and other value-added products (McMillan 1997; Saha 2003; Shanmugam and Ingram 2008). Bacteria, e.g. E. coli KO11 and related strains, ha ve been developed for the efficient conversion of both pentoses and hexoses to ethanol and other products (Ingram et al 1987; Ingram et al 1997). Xylan, in the form of 4-O-me thylglucuronoxylan, is a polysaccharide found in all hemicelluloses, and is the predom inant structural component of hardwood and agricultural residues (Preston et al. 2003). This polysaccharide may consist of more than 70 -xylopyronose residues, linked by -1,4-glycosidic bonds (Time ll 1964). In hardwood and softwood xylan, a 4-O-methylglucuron ic acid has been reported to be attached to the 2position of every 6th to 8th xylose residue (Kardosova et al. 1998; Jacobs et al 2001). Dilute acid hydrolysis is commonly used to render the m onosaccharides of lignocellulose accessible for fermentation (Lee et al. 1997; Saha et al 2005a). However, the -1,2 glucuronosyl linkage in xylan is resistant to dilute aci d hydrolysis, resulting in the release of methylglucuronoxylose (MeGAX) along with free xylose. MeGAX is not fe rmented by bacterial biocatalysts currently used to convert hemicellulose to ethanol, such as Escherichia coli KO11 (Rodriguez et al 2001). In sweetgum xylan, as much 27% of the carbohydrat e may reside in this unfermentable fraction

PAGE 59

59 after dilute acid pretreatment. E. asburiae JDR-1 has the ability to efficiently ferment both MeGAX and xylose, producing ethanol and acetate as primary fermentation products (Chapter 2). In this study, we have demonstrated that E. asburiae JDR-1 can be genetically engineered to produce optically pure D (-) lactate as the primary product by deleting genes ( pflB and als ) essential for the production of et hanol, acetate and 2,3 butanediol. The engineered strain was evaluated for its ability to pr oduce lactate from dilute acid hydrolysates of sweetgum xylan. Materials and Methods Bacterial Strains, Media, and Growth Conditions The bacterial strains constructed and used in these studies are listed in Table 3-1. The E. asburiae JDR-1 served as a starting point for gene tic engineering. During strain construction, cultures were grown aerobically at 30C, 37C, or 39C in Luria broth (10 g l 1 Difco tryptone, 5 g l 1 Difco yeast extract, and 5 g l 1 NaCl) containing eith er 2% (w/v) glucose, 5% sucrose or 3% (w/v) arabinose. Ampicillin (50 mg l 1), tetracycline (12.5 mg l 1), kanamycin (20 mg l 1 and 50 mg l 1), apramycin (20 mg l 1) or chloramphenicol (10 mg l 1 and 40 mg l 1) were added as needed. Sweetgum methylglucuronoxylan (MeGAXn) was prepared from sweetgum stem wood ( Liquidambar styraciflua ) as previously describe d and characterized by C13-NMR (Hurlbert and Preston 2001; Kardosova et al 1998). Dilute acid hydrolysates of methyglucuronoxylan were prepared by acid hydrolysis of 1% (w/v) sweetgum xylan with 0.1 N H2SO4 at 121 C for 60 min, followed by neutralization with BaCO3. Total carbohydrate concen trations of substrate preparations were determined by the phenol-sulfuric acid assay (Dubois et al 1956) with xylose as reference or by HPLC as previously descri bed (Chapter 2). Fermentation media were supplemented with Zucker and Hankin mineral salts (ZH salts) at pH 7.4 (Zucker and Hankin

PAGE 60

60 1970). Growth media were buffered with 100 mM sodium phosphate buffer (pH 7.0) or 100 mM 3-(N-morpholino) propane sulfonic acid (MO PS) buffer (pH 7.0) when necessary. Genetic Methods Standard methods were used for most of the genetic manipulations. Qiagen kits were used for genomic DNA and plasmid extraction (Qiagen, Valencia, CA). Polymerase chain reaction (PCR) amplification was performed with an I-cycler thermal cycl er (BioRad, Hecules, CA) with primers synthesized by Operon (Huntsville, Alabam a). Topo cloning kits were used for cloning (Invitrogen, Carlsbad, CA). Electroporation wa s performed on Gene pulser Xcell (BioRad, Hercules, CA). Restriction endonucleases were purchased from New England Biolabs (Ipswich, MA). DNA sequencing was provided by the University of Florida Interdisciplinary Center for Biotechnology Research. Deletion of pflB and als genes in E. asburiae JDR-1 The method for gene deletion in E.coli was used as previously described (Zhang et al. 2007; Janatama et al., 2008) with mi nor modifications applied to E. asburiae JDR-1. The pflB gene in E. asburiae JDR-1 was also selected as an integration site for th e PET operon. Several sets of primers were designed based on sequences of pflB orthologs in other Enterobacter spp. to amplify this gene fragment from E. asburiae JDR-1. Only one set derived from E. coli B was found to amplify the E. asburiae JDR-1 pflB gene fragment. The amplified E. asburiae JDR-1 DNA sequence and E. coli K12 pflB sequence were found to have 93% identity. The plasmids constructed are listed in Tabl e 1. The partial sequence of the E. asburiae JDR-1 pflB gene (gene bank accession number: EU719655) was determined within a DNA fragment amplified by PCR using specific primers based on the E. coli pflB sequence. The 3 kb catsacB cassette was obtained by digesting pLOI4162 with Sma I and Sfo I, and used in subsequent ligations. The pflB gene fragment amplified from E. asburiae JDR-1 was cloned into pCR 4-TOPO vector

PAGE 61

61 (Invitrogen) to obtain a plasmid, pTOPOpfl. This plasmid was diluted 500-fold and served as template for inside-out PCR amplification using the pfl inside-out primers. The resulting 5.5 kb fragment containing the replic on was ligated to the blunt-end catsacB cassette from pLOI4162 to produce a new plasmid, pTOPO4162pfl. This 5.5 kb fragment was also used to construct a second plasmid, pTOPODpfl, by p hosphorylation and self-ligation. Both pTOPO4162pfl and pTOPODpfl were then digested with Xmn I, diluted 500-fold and used as templates for amplification using the pfl primer set to produce linear DNA fr agments for integration step 1 ( pfl catsacBpfl ) and step 2 ( pfl pfl ), respectively. pfl and pfl represent two pfl gene fragments. After electroporati on of the step 1 fragment into E. asburiae JDR-1 containing pLOI3240, cells were incubated for 2 hr at 30C. The recombinant candidates were selected for chloramphenicol (20 mg l 1) resistance in Luria broth plates after overnight incubation (15 h) at 39C. Colonies were patched on both kanamycin (50 mg l 1) plates and chloramphenicol (40 mg l 1) plates. Those colonies growi ng on chloramphenicol (40 mg l 1) plates but not on kanamycin (50 mg l 1) plates were subjected for PCR confirma tion. The confirmed mutant colonies were transformed with pLOI3240, and prepared for el ectroporation with the step 2 fragment ( pfl pfl ). After electroporation, cells were incubated at 30C for 4 h and then transferred into a 250-ml flask containing 100 ml of LB without NaCl with 10% sucrose. Following an overnight incubation (30C), colonies were streaked on LB minus NaCl plates c ontaining 6% w/v sucrose (39C, 16 h). Colonies were tested for loss of apramycin and chloramphenicol resistance and confirmed by PCR. The resulting strain E. asburiae E1 had a disrupted pflB gene without detectable heterologous DNA sequences. E. asburiae E1 served as a starting stra in for further deletion of als gene. A segment of E. asburiae JDR-1 asl gene (FJ008982) was PCR amplified using degenerate primers designed based

PAGE 62

62 on conserved sequences id entified in homologous asl genes found in Enterobacter sp. 638, Erwinia carotovora subsp. atroseptica SCRI1043, Yersinia enterocolitica subsp. enterocolitica 8081 and Serratia proteamaculans 568. With the segment of the E. asburiae JDR-1 asl gene, this gene was disrupted in E. asburiae E1 with the same method for pflB disruption. Fermentation Batch fermentations were carried out in 16by 100-mm screw-cap tubes filled with nitrogen and sealed with rubber stoppers. The tube s were set in a Glas-Col minirotator at 60 rpm in a 30C incubator. Neutrali zed sweetgum xylan acid hydrolysat e (0.5% w/v) was added to 2x ZH salts directly as growth medium buffere d by 100 mM phosphate buffer or MOPS buffer at pH 7.0. Fermentations in hydrolysates were inoculat ed to an initial optica l density at 600 nm of 1.0 (determined using a Beckman DU500 series spectrophotometer). For analysis of fermentation products, cultures were centrifuged, and the supernatants were passed through 0.22 um filters and subjected to HPLC. Products were resolved on a Bio-Rad HPX-87H column with 0.01 N H2SO4 at 65 C. Samples were delivered with a 710B WISP automatic injector and chromatography controlled with a Waters 610 solv ent delivery system at a flow rate of 0.5 ml/min. Products were detected by differential refractometry with a Waters 2410 RI detector. Data analysis was performed with Waters Mi llennium Software. A quantitative relationship was determined between E. asburiae JDR-1 cell dry weight and culture OD at 600 nm. For calculation of specific consumpti on rates and specific production ra tes, the cell dry weight was determined based on the OD600 of the fermentation culture, which was 1.0 (0.51 g l-1) initially and did not appreciably change during the fermentation in 0.5% xylan hydrolysate. Determination of Lactate Isomers Produced by E. asburiae L1 To determine the isomers of lactate forme d, fermentation products were assayed with D-lactate of L-lactate dehydrogenases (Taguc hi and Ohta 1991). The conditions of the

PAGE 63

63 colorimetric enzyme assays were similar to t hose used to measure lactate dehydrogenase activity (Babson and Babson 1973). -NAD was obtained from Research Products International Corp, Chicago IL. All other reagents, s ubstrates, and enzymes were obtain ed from Sigma Chemical Co., St. Louis MO. Iodonitrotetrazolium ch loride (40 mg), 100 mg NAD and 10 mg phenylmethylsulfonyl fluoride were dissolved in 20 ml 0.2 M Tris-HCl (pH 8.2) to obtain the colorimetric reagent. Reactions were initiated by adding 4 Kunitz units (1 mol/min) of either L-lactate dehydrogenase (rabbit muscle, 140 U /mg protein) or Dlactate dehydrogenase ( Lactobacillus leichmanii 232 U/mg protein) in 100 l colorimetric reagent and 100 l sample at room temperature. The reduction of iodonitrotetr azolium dye was measured at room temperature of a Beckman 520 spectrophotometer at 503 nm. Sodi um salts of L and D-lactate (Sigma) were used as standards to define enan tiomer specificity of the reaction. Results Fermentation Characteristics of the Wild Type Strain E. asburiae JDR-1 When growing with either 0.8% glucose, 0.5% arabinose or 0.5% xylose as the sole carbon source, the wild type strain produced several pr oducts including succinate, lactate, acetate, 2,3-butanediol and ethanol. Glucose fermentations resulted in the formation of 2,3-butanediol, ethanol and acetate as major products. Larger amount of acetate and no 2,3-butanediol was detected in 0.5% xylose and 0.5% ar abinose fermentations (Table 3-2). The initial concentrations of substrates in the medium containing 0.5% sweetgum hemicellulose hydrolysate were determined by HPLC to be 20 mM xylose, 1.4 mM MeGAX and a small amount of MeGAX2. Previous studies indicated that MeGAX was metabolized by E. asburiae JDR-1 into methanol, glucuronate and xylose (Chapter 2). In these previous studies glucuronate fermentation by E. asburiae JDR-1 generated acetate in nearly 100% yield, indicating fermentation products more reduced th an acetate could only be produced from the free

PAGE 64

64 xylose and the xylose released from MeGAX in th e hydrolysate. The theoretical maximum yield of lactate from this hydrolysate medium was cal culated to be 35.7 mM based on the total xylose initially present. In the fermentati on of methylglucur onoxylan hydrolysate, E. asburiae JDR-1 was able to utilize all of the MeGAX within 30 hours and xyl ose within 40 hours. Similar amounts of ethanol (15.6 mM) and acetate (20 mM) were produced but no 2,3-butanediol or lactate was detected (Table 3-2, Fig. 3-1a). When supplemented with LB, E. asburiae JDR-1 fermented the 0.5% hydrolysate much more rapidly than with ZH minimal salts. Substrates were utilized within 15 hours, producing 16.2 mM et hanol, 22 mM acetate, and 3.2 mM succinate, again with no 2,3-butanediol or lactat e detected. (Table 3-2, Fig. 3-1c). Fermentation Characteristics of the Engineered Strains E. asburiae E1 and L1 The major competing pathway to lactate pr oduction initiates from the pyruvate formate lyase catalyzed reaction, which produces formate and acetyl-CoA in the wild type strain E. asburiae JDR-1. Both acetate and ethanol are produ ced from acetyl-CoA. In order to convert more carbon flux from pyruvate to lactate, the pfl B gene of JDR-1 was deleted to obtain strain E. asburiae E1. Since 2,3-butanediol was also produced by E. asburiae E1 in the fermentation of glucose (Table 3-2), the als gene which encodes acetolact ate synthase was deleted in E. asburiae E1 to eliminate 2,3-butanediol production (Moat at al. 2002). Th e resulting strain E. asburiae L1 was a double mutant lacking pflB and als genes (Fig 3-2) Both E. asburiae E1 and L1 were able to produce la ctate as the predominant product in glucose, xylose and arabinose fermentations. E. asburiae E1 produced 2.9 mM 2,3-butandiol in 0.8% glucose fermentation. The L1 strain with an interrupted 2,3-butan ediol-producing pathway produced no 2,3-butanediol and achieved highe r lactate yield (94.1% of the theoretical maximum). In xylose and arabinose fermentations the L1 strain also achieved higher lactate yield than E1 strain (Table 3-2).

PAGE 65

65 The E. asburiae L1 fermented slowly in the xylan hydrolysate with ZH minimal salts. In 60 hours, only a portion of free xylose in the hydrolysate was utilized and the MeGAX portion was not utilized (Fig 3-1b). Within 100 hours, 22.2 mM lactate was produced (Table 3-2). The low fementation rate of L1 in hydrolysate medium may be due to a lim iting activity of lactate dehydrogenase. The absence of detectable lact ate formation in the parent strain during fermentation of xylan hydrolysates also indicates a limitation in lactate de hydrogenase activity of E. asburiae JDR-1. The E. asburiae L1 strain fermented more ra pidly in the xylan hydrolysate supplemented with LB, with the complete cons umption of both MeGAX as well as xylose in 65 hours (Fig. 3-1d) with the formatio n of 36.4 mM lactate as well as very small amount of acetate and succinate (Table 3-2). Both E1 and L1 we re able to produce lactate at 100% of the theoretical maximum yield. The small amounts of acetate were likely derived from the glucuronate group of the 1.4 mM MeGAX pres ent in the hydrolysate substrate. The utilization of MeGAX by the L1 strain required LB supplementation, while the original isolate, E. asburiae JDR-1, utilized MeGAX in both minimal (Fig. 3-1a) and LB supplemented (Fig. 3-1c) media during the mixe d acid fermentation that produced acetate and lactate in nearly equal amounts (Table 3-2). Supplementation w ith LB resulted in a 2-fold increase in the utilization rate of xylose and nearly 3-fold in crease in the production rate of lactate in the L1 strain (Table 3-3). D-Lactate Was Produced by E. asburiae L1 The optical enantiomer(s) of lactate produced by E. asburiae L1 from the fermentation of xylan hydrolysates was determined by measuring the oxidation of lactate catalyzed by Dor L-lactate dehydrogenase with the reduction of iodonitrotetrazolium dye mediated via NADH formation as described in the Materials and Me thods section. A sample of medium containing 3.6 mol of lactate (det ermined by HPLC) of an E. asburiae L1 fermentation (72 hour) of 0.5%

PAGE 66

66 xylan hydrolysate supplemented with LB resulted in an increase in A503 from 0 to 0.113 in 5 minutes when assayed with 4 units of D-lact ate dehydrogenase When the same sample was assayed under the same conditions with 4 uni ts of L-lactate dehydr ogenase, there was no detectable increase in A503. Therefore the lactate produced by E. asburiae L1 was D-lactate with an apparent optical purity 100%. Discussion The fermentations of dilute acid hyd rolysates of methylglucuronoxylan by E. asburiae strains E1 and L1 provide the first examples of lactate formation from the aldouronate as well as the xylose present in these hydrolysates. The efficient formation of the D(-) entantiomer demonstrates a metabolic potential for the effici ent production of optically pure lactate from the most predominant polysaccharide components in the hemicelluloses fractions derived from woody biomass and agricultu ral residues. Although the relatively low production rate and dependence on rich media lim it direct application of E. asburiae L1, metabolic evolution by adaptive culturing and further genetic engineering may overcome these limitations. The studies here serve as a basis to further develop bio catalysts capable of al douronate utilization to efficiently ferment hemicellulose hydrolysates fo r production of alternative fuels and chemicals.

PAGE 67

67 Table 3-1. Bacterial strains and plasmids. Strain and plasmid Relevant characteristics Source or reference Strains E.coli Top10 For general cloning Invitrogen E.asburiae JDR-1 Wild type Our lab E.asburiae E1 E. asburiae JDR-1 pflB This chapter E.asburiae L1 E. asburiae JDR-1 pflB als This chapter Plasmids PLOI3240 Amr red red recombinase protein (Wood et al. 2005) pLOI4162 bla cat; cat-sacB cassette (Jantama et al. 2008) pCR 4-TOPO bla kan amp ; TOPO TA cloning vector Invitrogen pTOPOpfl pflB (PCR) amplified from E. asburiae JDR-1 and cloned into PCR4-TOPO vector This chapter pTOPO4162pfl cat-sacB cassette cloned into pflB in pTOPOpfl This chapter pTOPODpfl PCR fragment amplified from pTOPOpfl, kinase treated, and self-ligated This chapter pTOPOals als (PCR) amplified from E. asburiae JDR-1 and cloned into PCR4-TOPO vector This chapter pTOPO4162als cat-sacB cassette cloned into als in pTOPOpfl This chapter pTOPODals PCR fragment amplified from pTOPOals, kinase treated, and self-ligated This chapter

PAGE 68

68 Table 3-2. Comparing fermentation products of wild type and genetically engineered E. asburiae JDR-1 strainsa. a: Fermentations were completed within 72 hours w ith minimal media, or otherwise as indicated footnote c. The initial concentr ation of 0.8% glucose, 0.5% xylose and 0.5% arabinose media were determined by HPLC to be 42 mM, 31 mM and 31.5 mM respectively. The 0.5% xylan hydrolysate medium was measured to contain 20 mM xylose and 1.4 mM MeGAX. b: Percent of actual yield of lactate to theore tical maximum yield. Maximu m yield is defined as 2 mole lactate/ mole glucose or 5 mole lactate/ 3 mole xylose. c: This result was obtained af ter fermentation for 100 hours at wh ich time 65% of the xylose in the hydrolysate was utilized. Fermentation products (mM) EthanolAcetate2,3-Butanediol Succinate Lactate Lactate %yieldb E. asburiae JDR-1 0.8% glucose 26.8 11.5 12.9 5.2 3.9 4.6 0.5% xylose 20.9 17.5 0 4.1 1.0 1.9 0.5% arabinose 24.0 17.1 0 4.2 1.0 1.9 0.5% xylan hydrolysate 15.6 20.0 0 0 0 0 0.5% xylan hydrolysate with LB 16.2 22.0 0 3.2 0 0 E. asburiae E1 0.8% glucose 5.6 0 2.9 2.7 77.0 91.7 0.5% xylose 3.4 2.8 0 3.2 46.7 89.8 0.5% arabinose 6.5 2.9 0 2.1 41.3 78.0 0.5% xylan hydrolysate with LB 0 2.0 0 0 36.2 100.4 E. asburiae L1 0.8% glucose 4.4 0 0 1.7 78.9 94.1 0.5% xylose 1.5 2.9 0 1.3 47.2 90.8 0.5% arabinose 5.0 2.8 0 2.1 49.6 93.6 0.5% xylan hydrolysatec 0 0 0 0 22.2 96.0 0.5% xylan hydrolysate with LB 0 3.0 0 1.0 36.4 101.2

PAGE 69

69 Table 3-3. Specific consumption rate s and specific production rates of E. asburiae L1 in 5g /liter acid hydrolysate of sweetgum xylan.a a: q xylose and q MeGAX: Xyose and MeGAX specifi c consumption rate respectively, as grams of substrate consumed per gram dry cell weight per hour. q lactate: la ctate specific production rate, products generated per gram dry cell weight per hour. Strains q Xylose (g xylose/g DCW/h) q MeGAX (g MeGAX/g DCW/h) q Lactate (g lactate /g DCW/h) E.asburiae L1 in ZH salts 0.0670.006 0 0.0490.003 E.asburiae L1 in 0.12% LB 0.13.01 0.0190.002 0.13.005

PAGE 70

70 Figure 3-1. Fermentation time course for differe nt strains in media containing 0.5% sweetgum xylan hydrolysate a) E. asburiae JDR-1 in minimal medium, b) E. asburiae L1 in minimal medium, c) E. asburiae JDR-1 in LB, d) E. asburiae L1 in LB. Substrates and fermentation products: xylose ( ), MeGAX ( ), acetic acid ( ), ethanol ( ), lactic acid ( ). Time (Hours)Concentration (mM) 0 5 10 15 20 25 30 35 40Concentration (mM) Concentration (mM) 0 5 10 15 20 25 30 35Concentration (mM)a b c d

PAGE 71

71 Figure 3-2. Diagram to illustrate deletion of als and pfl B genes modifying mixed-acid fermentation of E. asburiae JDR-1 into a homolactate production pathway in E. asburiae L1. Deletion of pathwa ys is indicated in th e figure as symbol X. Acetyl-CoAADPATP 2NADH 2NAD+Pyruvate PEP pflB ldh ADP ATP succinate LactateNADHNAD+ als2-Acetolactate NADH NAD+ 2,3-butanediolX Acetate Formate EthanolX Acetyl-CoAADPATP 2NADH 2NAD+Pyruvate PEP pflB ldh ADP ATP succinate LactateNADHNAD+ als2-Acetolactate NADH NAD+ 2,3-butanediolX Acetate Formate EthanolX glucose, xylose, arabinose

PAGE 72

72 CHAPTER 4 GENETIC ENGINEERING ENTEROBACTER ASBURIAE STRAIN JDR-1 FOR EFFICIENT ETHANOL PRODUCTION FROM HE MICELLULOSE HYDROLYSATE Introduction Lignocellulosic resources, incl uding forest and agricultura l residues and evolving energy crops, offer benign alternatives to petroleu m-based resources for production of fuels and chemicals. As renewable resources, these are ex pected to decrease de pendence on exhaustible supplies of petroleum and mitigate the net rele ase of carbon dioxide into the atmosphere. The development of economically acceptable bioconversi on processes will require pretreatments that release the maximal quantities of hexoses (predomi nantly glucose from ce llulose), and pentoses (arabinose and xylose) from hemicelluloses, and al so require microbial bi ocatalysts that will efficiently convert these to a single targeted product. MeGAXn are the predom inant components in the hemicellulose fractions of agricultura l residues and energy crops, e.g corn stover, sugarcane bagasse, poplar, and switchgrass (Dien et al. 2006; Pordesimo et al. 2005; Sun et al. 2004; Sun et al. 2001). In sweetgum xylan, as much as 27% of the carbohydrate may reside in this unfermentable fraction, MeGAX, after dilute aci d pretreatment (Chapter 1; Rodriguez et al 2001). Complete utilization of all hemicellulosic sugars can improve conversion efficiency of lignocellulosic materials to fuel ethanol. In this study, the PET operon containing pdc and adhB genes from Zymomonas mobilis (Ingram at al 1987; Ingram and Conw ay, 1988) was incorporated into a pflE. asburiae JDR-1 strain by plasmid transformation to constr uct a homoethanologenic strain. The resulting recombinant strains were compared with wild type E. asburiae JDR-1 and the ethanologenic strain E. coli KO11 to evaluate their efficiencies for ethanol production from dilute acid hydrolysates of sweetgum methylglucuronoxylan.

PAGE 73

73 Materials and Methods Bacterial Strains, Media, and Growth Conditions E. coli strains, E. asburiae strains and plasmids used are listed in Table 4-1. The E. asburiae JDR-1 served as a starting point for gene tic engineering. Batch fermentations were carried out in medium saturated w ith nitrogen in tubes set in a Gl as-Col minirotator at 60 rpm in 30C incubator. Fermentations in hydrolysates were inoculated to an initial optical density at 600 nm of 1.0. Fermentation products were resolved on a BioRad HPX-87H column with a Waters HPLC system. The growth conditions and media we re same with the ones applied in chapter 3. Genetic methods Standard methods were used for most of the ge netic manipulations as de cribed in chapter 3. The plasmids used and constructed in this study are listed in Table 4-1. Transformation of Plasmid Carrying PET Operon into E. asburiae JDR-1 E. asburiae JDR-1 was grown with one of several anti biotics at different concentrations in LB and minimal media on agar plates or in liqui d media to test its antibiotic resistance. Based upon its sensitivity to chloramphenicol and tetracycline, plasmids pLOI555 ( cmR) and pLOI297(tetR), both containing the PET operon, were transformed into E. asburiae JDR-1 or E. asburiae E1 by electroporation in a 100 l cuvette under the condition of 1.8kV, 25 F capacitance and 200 resistance. For electroporation, comp etent cells from 25 ml exponential phase cultures were washed 3 times by suspen sion and centrifugation with cold 10% glycerol. Cultures were plated on LB agar containi ng 2% glucose and tetr acycline (12.5 mg l 1) or chloramphenicol (40 mg l 1) to select E. asburiae JDR-1 and E1 carrying pLOI297 or pLOI555 respectively. Plasmids were extrac ted confirming their presence in E. asburiae cells.

PAGE 74

74 Integration of PET Operon into E. asburiae JDR-1 Chromosome Since E. asburiae JDR-1 was found to have high re sistance to ampicillin, the red recombinase plasmid pLOI3240 with apramycin resistance gene was transformed into E. asburiae JDR-1. The rrlE gene fragment (gene bank accession number: EU719656 ) of this strain was amplified using specific primers based on E. coli rrlE gene sequence, with E. asburiae JDR-1 genomic DNA as template. The id entity was found to be 96% between rrlE gene DNA sequences of E. asburiae JDR-1 and E. coli K12. The high sequence similarity provided the possibility of using E. coli rrlE fragment for recombination with E. asburiae JDR-1 rrlE gene DNA. The fragment, with PET operon flanked by E. coli rrlE gene ( rrlE-pdc-adhA-adhB-Km-rrlE ), was cut from pLOI4672 (constr ucted by Xueli, Zhang) with AscI (New England Biolabs, US), and transformed into E. asburiae JDR-1 by electroporation. The selected colonies from kanamycin (20 mg l 1) medium were patched on kanamycin (50 mg l 1) and chloramphenicol (40 mg l 1) medium to differentiate correctly integrated colonies from colonies with contaminated pLOI4672 plasmid. Th e colonies growing on kanamycin medium but not on chloramphenicol medium were subjected to PCR confirmation. The pflB gene in E. asburiae JDR-1 was also selected as an integration site for the PET operon. Two sets of specific primers based on Enterobacter 638 (http://genome.jgi-psf.org/finished _microbes/ent_6/ent_6.home. html) pflB sequence and two sets of specific primers based on E. coli pflB sequence were employed to amplify pyruvate formatelyase gene sequence from E. asburiae JDR-1. One set derived from E. coli was able to amplify the pflB gene (gene bank accession num ber: EU719655) fragment from E. asburiae JDR-1 genomic DNA. The amplified E. asburiae JDR-1 DNA sequence and the E. coli K12 pflB sequence were found to have 93% identity. The s ubsequent chromosome integration process was same as the previous one at rrlE site with the plasmid pLOI4666 constructed by Xueli, Zhang. A

PAGE 75

75 primer set for the kan gene was used to confirm the in tegration. The confirmed recombinant strains for both insertion sites were grown on th e aldehyde detection plat es (Conway et al. 1987). The colonies with strongest red color indictive of aldehyde produc tion were selected for further fermentation analysis. Plasmid Stability in E. asburiae JDR-1 E. asburiae JDR-1 harboring either pLOI555 or pLOI297 was serially transferred in Luria broth containing 2% glucose without antibiotics for more than 72 generations at 30C. One generation was defined as a 2-fold increase in cu lture turbidity. Appropriate dilutions of cultures were plated on Luria agar with and without antibiotic; colonies formed were counted and calculated to obtain the ratio of cells retaining an tibiotic resistance to total cells. Ten colonies retaining antibiotic resistance (and therefore presumed to retain pLOI555 or pLOI297) after 72 generations were subjected to fermentation to test thei r ethanol producing ability. Assay of PDC Activity Pyruvate decarboxylase activity was assayed in engineered E. asburiae JDR-1 strains by monitoring the pyruvate-dependent oxidation of NADH with alcohol dehydrogenase as a coupling enzyme (Conway et al. 1987; Ohta et al. 1991b). Exponential phase anaerobic cultures were harvested and cells were disrupted us ing the FastPrep bead mill MP system (MP Biomedicals, Irvine, CA) in 0.05 M phosphate buffe r. The supernatant was collected after 15 min centrifugation at 1800 rpm (Eppendor f centrifuge 5414). The entire process was carried out at 4C. Heat treatment for 15 min at 60C was us ed to inactivate competing native enzymes of E. asburiae JDR-1 which might affect quantitativ e measurements of PDC activities in transformants. The enzyme activity assay of PD C was performed in the reaction mixture of 1.0 mM TPP (thiamine pyrophosphate), 1.0 mM MgCl2, 0.40 mM NADH, 20 mM sodium pyruvate and 0.05 M sodium phosphate buffer, pH 6.5. The assay was started by adding 20 l crude cell

PAGE 76

76 extract. Protein concentration of the crude extract was determined with BCA protein assay reagent kit (Pierce Chemical Co., Rockford, IL). Results Fermentation Characteristics of the Wild Type Strain E. asburiae JDR-1 E. asburiae JDR-1 performed a mixed-acid fermenta tion in low substrate concentration. When growing in 2.5% (w/v) glucose or 2% (w/v) xylose, the wild type strain produced a wide range of products, including su ccinate, lactate, acetate, fo rmate, 2,3-butanediol and ethanol (Table 4-2). In glucose fermentation, succinate and acetate were produced at low concentrations, approximately at 1 mM. Lactate was produced at approximately 10 mM, and the major products were formate, 2,3-butanediol and ethanol, each at approximately 40 mM. More acetate and less 2,3-butanediol were produced in xylose fermentati on (Table 4-2). In both batch fermentations buffered with 0.1 M sodium phosphate (pH 7.0), the w ild type strain failed to utilize all the substrates during the 48 h fermentation. Even in the buffered medium the pH after fermentation decreased to 4.8, which suggested that acid produc tion might be the main factor preventing the cells from utilizing a ll the substrate. The components in the medium containi ng 0.5% sweetgum hemicellulose hydrolysate were determined by HPLC to be 20 mM xyl ose, 1.4 mM MeGAX and a small amount of MeGAX2 (Fig 4-1). Previous studies sugge sted that MeGAX was metabolized by E. asburiae JDR-1 into methanol, glucuronate a nd xylose. Glucuronate fermentation by E. asburiae JDR-1 generated acetate in nearly 100% yield, indicating more reduced fermentation products (ethanol and lactate) could onl y come from the free xylose and the xylose released from MeGAX ( chapter 2). Therefore, the theoretical maximu m yield of ethanol from this hydrolysate was calculated to be 35.7 mM based on the total amount of xylos e present in hydrolysate. E. asburiae JDR-1 was able to completely utilize MeGAX in the 0.5% hydrolysate in about 12 hours and

PAGE 77

77 xylose in 20 hours after a peri od of several hours for adaptation to the hydrolysate medium. Similar amounts of ethanol (15.6 mM) and acetat e (20 mM) were produced with small amount of formate and no detectable 2,3-butanediol; the ethanol yield was 44.2% of the theoretical maximum (Table 4-3, Fig 4-1, Fig 4-2A). The specific consumption rates of xylose and MeGAX in the hydrolysate and specific production rates of acetate and ethanol ar e included in Table 4-4. Fermentation of E. asburiae 4666 and E. asburiae 4672 in Glucose In order to increase ethanol yield, PET operon containing Zymomonas mobilis pdc adhA and adhB genes were integrated into E. asburiae JDR-1 chromosome within either pflB or rrlE gene, to obtain the engineered strains E. asburiae 4666 and E. asburiae 4672 respectively. However, both strains could not completely utilize 2.5% glucose in the batch fermentations, and the final pH decreased to 5. E. asburiae 4672 had similar fermentation products with the wild type (Table 4-2), suggesting that the integrated PET operon might not be expressed efficiently. E. asburiae 4666, with the PET ope ron inserted into pflB gene, showed a different profile of fermentation products. This strain produced much less formate (3.2mM) and much more lactate (36.5mM) than the wild type, while kept other fermentation products at similar level including ethanol. This suggested that insertion of PET operon into plfB gene inactivated pyruvate formate lyase, resulting in less formate and acetyl-C oA production. More carbon flux was converted from pyruvate to lactate through lactate dehydrogenase. Both engi neered strains had a similar ethanol yield as the wild type strain (Table 4-2). Fermentation of E. asburiae JDR-1 (pLOI297) and E. asburiae JDR-1 (pLOI555) Plasmids pLOI297 and pLOI555 were transformed into E. asburiae JDR-1 for overexpression of pdc and adh genes. Both transformed strains were able to completely utilize 2.5% (w/v) glucose or 2% (w/v) xylose within 48 hours, with ethanol as the predominant fermentation product. The ethanol yields of glucose fermentation were 94.1% and 95.3% for E.

PAGE 78

78 asburiae JDR-1 (pLOI297) and E. asburiae JDR-1 (pLOI555), respectively (Table 2). E. asburiae JDR-1 (pLOI555) was further tested in xyl ose fermentation, and the ethanol yield was even higher, greater than 98% of theoretical. There were also other fermentation products present at concentrations below 10 mM (Table 4-2). E. asburiae JDR-1 (pLOI555) and JDR-1 (pLOI297) were tested for the fermentation of dilute acid hyrolysates of sweetgum MeGAXn. Both strains consumed MeGAX as well as xylose within 18 hr and fermentation was complete with in 25 hr (Fig 4-2C for JDR-1 (pLOI555); data for JDR-1 (pLOI297) was not shown). The xylose specific consumption ra te of JDR-1 (pLOI555) was similar to the parent stra in but the MeGAX specific consump tion rate was lower. Ethanol was the major fermentation product, and the yiel d was much higher than the parent strain. However, both strains produced substantial am ount of acetate (approximately10 mM) and had lower yields of ethanol than with either xyl ose or glucose as subs trates (Table 4-4). Fermentation of Strain E. asburiae E1 (pLOI555) Compared with E. coli KO11 and Other E. asburiae JDR-1 Derivatives Neither 2,3-butanediol nor lactic acid was produced in the hydrolysate fermentation by either E. asburiae JDR-1 (pLOI297) or JDR1 (pLOI555). This result i ndicated that only the acetate production pathway initiated from pyruvate formate lyase competed for pyruvate and lowered the ethanol yield. In or der to direct greater carbon flux from pyruvate to ethanol, the pfl B gene of E. asburiae JDR-1 was deleted to obtain strain E. asburiae E1, followed by pLOI555 transformation. When tes ting this strain in hydrolysate fermentations, no formic acid was produced, and only small amount of acetate wa s produced (4.5 mM). Af ter several hours of adaption, the MeGAX portion was consumed in 12 hr and the xylose portion was consumed in 20 hr (Fig 4-2D). While the specific consumption ra tes of the substrates we re close to the parent strain and JDR-1 (pLOI555), E. asburiae E1 (pLOI555) had a much higher specific production

PAGE 79

79 rate of the ethanol (0.11.01g ethanol/g DCW/h) and a much lower specific production rate of the acetate (0.022.003 g ethanol/g DCW/h). Most of the carbon sources in the hydrolysates were converted to ethanol, achieving 99% of maximal theoretical yield (Table 4-3, Table 4-4, Fig 4-1). The E. coli KO11, which was reported to produce 0.54 gram ethanol per gram glucose (Ohta et al. 1991a), could only produce ethanol at 63% of the theoretical maximum in the sweetgum xylan hydrolysate medium, and accumula ted a substantial amount (10.6.3 mM) of acetate (Fig 1, Fig 2C). The sum of ethanol and acetate was 33.1 mM for E. coli KO11, and 40.2 mM for JDR-1 (pLOI555), 39.9 mM for JDR-1 (pLOI297) a nd 40.5 mM for E1 (pLOI555) (Table 4-3). This result indicated that E. coli KO11 utilized less substrate in the hydrolysate than the 3 engineered E. asburiae strains and produced lower quantitie s of products as a result of the inability of E. coli KO11to utilize MeGAX in the hydrolysate (Fig 4-1, Fig. 4-2B). The ethanol specific production rate of E. coli KO11 (0.074.006 g ethanol/g DCW/h) was much lower than E. asburiae E1 (pLOI555) (0.11.01 g ethanol/g DC W/h) (Table 4-4). Compared with E. coli KO11, E. asburiae E1 (pLOI555) utilized more substrat e in sweetgum hydrolysate and was able to produce 57.8% more ethanol at higher rate. PDC Activity in E. asburiae Strains The PDC enzyme activity produced as a resu lt of expression of heterologous gene pdc in engineered E. asburiae strains (Table 5). Because of the relative thermal stability of PDC encoded by the pdc gene of Zymomonas mobilis a heat treatment at 65C for 15 minutes was used to inactivate competing native enzymes, e.g. activities associated with the pyruvate dehydrogenase complex, could affect measurements of PDC activity (Conway et al. 1987; Ohta et al. 1991b). While crude extracts from both st rains showed pyruvate-dependent NADH oxidase activity before heat treatment (d ata not shown), the wild type strains were unable to oxidize

PAGE 80

80 NADH after the heat treatment. However, all th ree strains carrying plasmid with the PET operon showed substantial PDC activities after heat tr eatment, showing the presence of PDC encoded by pdc genes derived from Zymomonas mobilis .This result indicates that the successesful expression the heterologous pdc is the key factor for the homogeneous production of ethanol in engineered E. asburiae strains. Plasmid Stability in E. asburiae JDR1 The pLOI297 transformant was relatively uns table, with only 10.7% of transformed E. asburiae JDR-1 cells retaining tetracycline resistan ce after cultivation for 72 generations without antibiotic selection pressure. The pLOI555 transf ormant, however, was quite stable, with 98.1% of pLOI555 transformed E.asburiae JDR-1 cells retaining chloramphenicol resistance after growth for 72 generations in the absence of antibiotic (Table 46). Fermentation analysis of 10 descendent colonies retaining antibiotic resistance from st rains carrying pLOI297 and pLOI555 was also performed to confirm that strains with retained antibiotic resist ance also retained the homoethanologenic phenotype. Discussion A wild type Enterobacter asburiae strain with limited knowledge of its genetic and physiological properties was genetically engi neered for a new metabolic potential. The methodology and protocols developed in this study may provide re ference value for engineering other wild type Enterobacter spp. While E. asburiae JDR-1 was determined to be relatively resistant to ampicillin and probably other -lactam antibiotics, it was sensitive to tetracycline (12.5 mg l 1), kanamycin (20 mg l 1), apramycin (20 mg l 1) and chloramphenicol (10 mg l 1). To determine if a plasmid-based system developed for use in E. coli could be maintained and function in E. asburiae JDR-1, pCR4-TOPO plasmid with a small insertion was electroporated into the competent cells and the transformants we re able to be selected on a kanamycin (50 mg

PAGE 81

81 l 1) plate. The transformed pCR4-TOPO plasmid in E. asburiae JDR-1 was qualitatively determined by DNA gel electrophoresis to have a lower concentration than in E.coli Top10 host (data not shown). These results show that plasmids may be developed in E. coli for the introduction of genes encoding a desired metabolic potential in E. asburiae JDR-1, and may have applications in the metabolic transformation of other Enterobacter spp. as well. With these transformation systems, E. asburiae JDR-1 (pLOI297) and E. asburiae JDR-1 (pLOI555), developed with the strategy successfully employed to genetically engineer the first ethanologenic strains of E. coli (Ingram et al. 1987), were able to produce ethanol at 94.1% and 95.3% of theoretical yield in glucose, but failed to achieve such high yield in the dilute acid hydrolysates of methylglucuronxylan The dilute acid hydrolysate of sweetgum xylan used as fermentation substrate is a mixt ure of different substrates. Wh ile xylose and MeGAX could be quantified and very small amount of MeGAX2 could be detected by HPLC, there were other chemical compounds too little to be detected. A variety of biologi cal toxins are generated during the dilute acid treatment. These include acetic ac id, degradation products of sugars such as furfural (dehydration product of pentoses) and 5-hydroxymethylfu rfural (dehydration product of hexoses), and soluble aromatics from lignin (aromatic alcohols, acids, and aldehydes). These toxins retard the fermentation of hemicellulose sy rups by yeast or other biocatalysts (Palmqvist and Hahn-Hagerdal 2000a; Palmqvist and Hahn-Ha gerdal 2000b). Although it was reported that some of the aldehydes did not decrea se ethanol yield in fermentation by E.coli based biocatalysts (Zaldivar et al. 1999); bacteria usually utilize reducing power, e.g. NADPH, to reduce some of the aldehydes to decrease their toxicity (Gutie rrez et al. 2002). In th e minimal medium of hydrolysate used in this paper, with low substrate concentra tion and no supplemental nutrition, the NADH/NAD+ ratio might be decreased by reducing the aldehydes (Miller at al. 2009). As a

PAGE 82

82 result, more pyruvate might be converted to acetate, because no NADH was required in this pathway. This might be the reason that E. asburiae JDR-1 (pLOI297) and E. asburiae JDR-1 (pLOI555) had lower ethanol yield in the hydrolysate. To decrease the formation of organi c acids, acetate and formate, the pfl B gene was then deleted. The convenient one-step gene inac tivation method successfully applied to E. coli (Datsenko and Wanner 2000) failed to knock out the pfl B gene in E. asburiae JDR-1, requiring the development of a different protocol. An alternative gene deletion method used PCR fragments with several hundred bases of homo logous sequence at both ends instead of 40 bp used by the one-step method (Jantama et al. 2008). Longer homologous se quences are expected to increase the rate of homologous recombin ation (Puchta and Hohn 1991), and may delay complete degradation of the linear DNA by e xonucleases, thus increa sing the probability of recombination events. Recombinants were not selected on the plates containing levels of antibiotics used for selection of E.coli recombinants and required lower concentrations, kanamycin (20 mg l 1) and chloramphenicol (10 mg l 1) to be used. This is likely the basis for growth of non-recombinant as well as recombinan t colonies and required a second selection that was achieved by patching colonies onto kanamycin (50 mg l 1) and chloramphenicol (40 mg l 1) plates. By maximizing DNA concentration to approximately 5 g/ l and cell concentrations of 1010 cells/100 l in electroporation transf ormation, usually 3 to 6 E. asburiae JDR-1 recombinants could be obtained by this process. The methodology develope d here might also be applied to engineer other Enterobacter spp. with genetic manipulations developed for E.coli The E. asburiae strain with a genomic pflB deletion was transformed with a plasmid, pLOI555, to obtain E. asburiae E1 (pLOI555), a strain capable of efficiently converting the xylose residues derived from methyglucuronoxylan to ethanol, achieving a yield at 99% of the

PAGE 83

83 theoretical maximum. In this respec t it has been able to outperform E. coli KO11 in medium of sweetgum xylan hydrolysate, which has been developed as a comme rcial ethanologenic biocatalyst. The specific PDC activities measured in transformed E. asburiae strains were noticeably lower than those measured in the engineered K. oxytoca M5A1 (Ohta et al. 1991a), possibly due to lower copy number of the plasmids pLOI297 and pLOI555 in E. asburiae JDR-1. However, as found with engineered K. oxytoca strains, E. asburiae JDR-1 pLOI297 had higher activity than pLOI555, which may be due to the presence of the ColE1 replicon in pLOI297 resulting in a higher copy number than in the strain transformed with pLOI555. It was found that E. asburiae E1 (pLOI555) with highest ethanol yield in hydrolysate had the lowest PDC activity in the glucose culture. The contribution of the adh gene from pLOI1555 is lik ely critical to homoethanol production in E. asburiae E1 as it was in initially genera ting the ethanologenic strains in E. coli (Ingram and Conway, 1988; Ingram et al. 1987 ). When selected genes were deleted in E. asburiae JDR-1 to produce lactate as the predominant product from E. asburiae L1, fermentations were slow and incomplete wit hout supplementation with Luria Bertani medium (chapter 3), supporting the conclu sion that efficient fermention to a targeted product requires high level of expression of the gene encoding th e oxido-reductase responsi ble for generating that final fermentation product duri ng the reoxidation of NADH. Plasmid stability is critical for biocatalysts engineered with gene s conferring a desired metabolic potential confined within a plasmid, as consistent traits are required for long-term applications. The plasmid pLOI297, containing colEl replicon, was presen t in high copy numbers in E. coli strains, but was unstable in K. oxytoca M5A1. pLOI555 derived from cryptic

PAGE 84

84 low-copy-number plasmids in E. coli B (ATCC 11303), however, was very stable in K. oxytoca M5A1(Ohta et al. 1991b). Si milar to the studies in K. oxytoca pLOI555 plasmids were found to be more stable than pLOI297 in E. asburiae JDR-1. The relative stab ility of the plasmid in E. asburiae E1 (pLOI555) recommend it for further de velopment, perhaps through introduction of the pdc and adh genes into the chromosome as has been achieved for the successful development of E. coli KO11 and its derivatives as ethanologeni c biocatalysts (Jar boe et al. 2007).

PAGE 85

85 Table 4-1. Bacterium strains and plas mids for engineering ethanologenic E. asburiae Strain and plasmid Relevant characteristics Source or reference strains E.coli Top10 For general cloning Ivitrogen E.coli KO11 pfl :: (pdc-adhB-cat) frd (Ohta et al., 1991a) E. asburiae JDR-1 Wild type This paper E. asburiae 4666 rrlE :: (pdc-adhA-adhB-kan) This chapter E. asburiae 4672 pfl :: (pdc-adhA-adhB-kan) This chapter E. asburiae E1 Enterobacter asburiae JDR-1 pfl B This chapter Plasmids PLOI3240 Amr red red recombinase protein (Wood et al. 2005) pLOI297 Tcr pdc+ adhB+ (Ingram et al., 1989) pLOI555 Cmr pdc+ adhB+ (Ohta et al., 1991b) pLOI4666 Kan cat ; pflB-pdc-adhA-adhB-FRT-Kan-FRT-pflB This chapter by Zhang pLOI4672 Kan cat ; rrlE-pdc-adhA-adhB-FRT-Kan-FRT-rrlE This chapter by Zhang pLOI4162 bla cat; cat-sacB cassette (Jantama et al., 2008) pCR 4-TOPO bla kan amp ; TOPO TA cloning vector Invitrogen pTOPOpfl pflB (PCR) amplified from E. a. JDR-1 and cloned into PCR4-TOPO vector Chapter 3 pTOPO4162pfl cat-sacB cassette cloned into pflB in pTOPOpfl Chapter 3 pTOPODpfl PCR fragment amplif ied from pTOPOpfl, kinase treated, and self-ligated Chapter 3

PAGE 86

86 Table 4-2. Comparison of sugar fermentation products of wild type and genetically engineered E. asburiae JDR-1.a Fermentation products (mM) Fermentations Succinate Lactate FormateAcetate 2,3-butane diol Ethanol Ethanol yield(% of theoretical)b 2.5% glucose E. asburiae JDR-1c 2.0 9.6 39.1 1.0 45.9 45.0 25.6 E. asburiae 4666c 0.3 36.5 3.2 4.0 15.3 31.8 28.7 E. asburiae 4672c 9.2 11.4 10.6 16.0 21.5 35.6 27.9 E. asburiae JDR-1(pLOI297) 1.8 4.7 9.4 3.8 0 261.6 94.1 E. asburiae JDR-1(pLOI555) 1.6 2 7.7 3.4 0 265 95.3 2%xylose E. asburiae JDR-1c 12.7 5.6 15.0 25.2 13.4 42.6 35.6 E. asburiae JDR-1(pLOI555) 2.2 1.2 3.6 4.2 0 217.4 98.0 a) Fermentations were carried out at 30C in ZH minimal media for 48 hours as described in the Materials and Methods section. b) Percentage of amount of ethanol produced to a theoretical maximal amount. A yield of 100% is defined as 2 mole ethanol/ mole gluc ose or 5 mole ethanol/ 3 mole xylose. c) E. asburiae JDR-1, E. asburiae 4666 and E. asburiae 4672 did not completely utilize the substrates within 48 hours. Table 4-3. Fermentation products from acid hydrolysates of sweetgum xylan.a Fermentation products (mM) Formic acid Acetic acid Ethanol Ethanol yield(% of theoretical) b E. asburiae JDR-1 4.9.420.0.7 15.6.8 44 E.coli KO11 5.9.010.6.3 22.5.2 63 E. asburiae JDR-1 (pLOI555) 4.0.413.5.5 26.7.0 75 E. asburiae JDR-1 (pLOI297) 3.8.39.9.3 30.0.5 84 E. asburiae E1 (pLOI555) 0 4.5.2 35.5.1 99 a) Fermentations were carried out at 30C in ZH minimal media for 48 hours as described in the Materials and Methods section. Results are averages of 3 experiments. b) percentage of amount of ethanol produced rela tive to the theoretical maximum. A yield of 100% is defined as 2 mole ethanol/ mole gl ucose or 5 mole ethanol/ 3 mole xylose.

PAGE 87

87 Table 4-4. Specific consumption rates and specifi c production rates in acid hydrolysates of sweetgum xylan (5g /liter)a. Strains q Xylose q MeGAX q Acetate q Ethanol E. asburiae JDR-1 0.33.04 0.087.0120.13.01 0.060.009 E.coli KO11 0.38.04 0 0.11.01 0.074.006 E. asburiae JDR-1 (pLOI555) 0.29.03 0.058.0120.14.02 0.052.004 E. asburiae E1 (pLOI555) 0.32.03 0.077.0130.022.0030.11.01 a) q xylose is defined as consumed g xylose /g DCW(dry cell weight) /h ; q MeGAX is defined as consumed g MeGAX /g DCW(dry cell weight) /h; q acetate is defined as produced g acetate /g DCW(dry cell weight) /h; q ethanol is defined as produced g ethanol /g DCW(dry cell weight)/h. Results are averages of 3 experiments. Table 4-5. Specific activity of PD C in cell crude extract from E. asburiae JDR-1 derived strains.a Strains Specific Activity(Ub/mg of cell protein) E. asburiae JDR-1 0 E. asburiae 4666 0 E. asburiae 4672 0 E. asburiae JDR-1 (pLOI297) 1.02.12 E. asburiae JDR-1 (pLOI555) 0.77.13 E. asburiae E1 (pLOI555) 0.53.10 a) Results are averages of 3 experiments. b) One U is defined as the amount of the enzyme that catalyzes the conversion of 1 mole of substrate per minute at room temperature. Table 4-6. Plasmid stability of pLOI297 and pLOI555 in E. asburiae JDR-1. Results were averages of 3 experiments. Plasmids %cells retaining antibiotic resistance After 36 generationsAfter 72 generations pLOI297 29.5.3 10.7.6 pLOI555 100.0.8 98.11.8

PAGE 88

88 Figure 4-1. HPLC profile of fermentation products of E. asburiae JDR-1, E.coli KO11 and E. asburiae E1 (pLOI555) in 0.5% sweetgum xylan hydrolysate with MOPS. (The unlabeled peaks with retenti on time of 11 minutes and 21 minutes were for salts and buffers.) 0.00 M inutes 5.00 10.00 15.00 20.00 25.00 30.00 0.00 100.00 200.00 0.00 50.00 0.00 Initial medium E asburiae JDR-1 E coli KO11 E asburiae E1 ( POLI555 ) MeGAX Acetae Ethanol Xylose Refretive index mV

PAGE 89

89 A. B. Figure 4-2. Fermentation time course of E. asburiae JDR-1 A), E.coli KO11 B), E. asburiae JDR-1 (pLOI555) C) and E. asburiae E1 (pLOI555) D) in media of buffered sweetgum xylan hydrolysate. Substrates and fermentation products: xylose ( ), MeGAX ( ), acetic acid ( ), formic acid ( ), ethanol ( ). 0 5 10 15 20 25 010203040506070 Time (hours)Concentration (mM) 0 5 10 15 20 25 010203040506070 Time (hours)Concentration (mM)

PAGE 90

90 Figure 4-2. Continued C D 0 5 10 15 20 25 30 010203040506070 Time (hours)Concentration (mM) 0 5 10 15 20 25 30 35 40 010203040506070 Time(hours)Concentration (mM)

PAGE 91

91 CHAPTER 5 GENETIC DEFINITION OF THE M ETHYLGLUCURONOXYLOSE UTILIZATION PATHWAY IN ENTEROBACTER ASBURIAE JDR-1 Introduction Currently an accepted hypothetical process of degradation of MeGAX3 is that MeGAX3 enters the cell via a specific transporting syst em (Gilson et al. 1988, Shul ami et al. 1999). Inside the cell, glucoronoxylotriose is hydrolyzed at the -1,2 glucuronosyl linkage by an -glucuronidase, releasing xylotri ose and D-glucuronic acid. Xylotr iose is hydrolysed to xylose by -xylosidase. When and how the met hyl group is released from MeGAX3 is not known (Shulami et al. 1999). All the -glucuronidases share homology in amino acid sequences, belonging to GH family 67 (CAZy database, http://afmb.cnrs-mrs.fr/CAZY/). Paenibacillus JDR-2 was determined to utili ze methylglucuron oxylotriose (MeGAX3), methylglucuronoxylobiose (MeGAX2) and MeGAX, probably by the same metabolic pathways. The -glucuronidase from Paenibacillus sp. JDR2 has been characterized and found to have hydrolase activity with MeGAX, as well as MeGAX2 and MeGAX3 (Nong 2009). Whether E. asburiae strain JDR-1 uses a similar or a novel pa thway, and what enzymes and transporters constitute the pathway is the main subject of this chapter. Materials and Methods Degenerate Primer Method to Det ermine Presence of GH 67 genes Standard methods were used for most of the ge netic manipulations as de cribed in chapter 3. Several -glucuronidase genes from GH family 67 (CAZy database, http://afmb.cnr-smrs.fr/CAZY/), were selected and i nput into Codehop (http://blocks.fhcrc.org/blocks/code-hop.html ), a web based software to find conserved domains from several homologous sequences and design degenerate primer sets. These sets were applied

PAGE 92

92 to genomic DNA of E. asburiae strain JDR-1 to amplify sequences homologous to the select set of -glucuronidase genes with the Touch Down PCR conditions (Schunck et al. 1995). Determination of Genes Involved in MeGAX Utilization Process with Mutant Library Method Tn5 transposomes (EZ-Tn5 Transposomes, Epicentre, US) were transformed into E. asburiae JDR-1 with electroporation method to generate a mutant library of about ten thousand colonies. An antibiotic counter enrichment method using D-cycloserine was employed to accumulate mutants that were unable to grow on MeGAX but could grow on xylose or mannitol minimal media. Enriched cultures were streak ed on xylose or mannitol minimal medium and copied to MeGAX minimal agar plates to iden tify MeGAX defective mutants. The selected defective mutants were inoculated in liquid MeGAX minimal medium to confirm the phenotype. The genomic DNA of selected MeGAX defective mutants was extracted, fragmented and self-ligated. The ligated DNA circ le contained Tn5 insertion cartr idge including replicator and kanamycin resistant genes, which constituted a plasmid capable of replication. These plasmids were transformed into E.coli and sequenced to determine the ge nes that were interrupted by the Tn5 transposon. The Blastx program of NCBI wa s used to obtain Amino acid sequences most similar to the defective genes in MeGAX def ective mutants to identify their function. The relative growth rates of the E. asburiae JDR-1 ptsImutant were measured in 0.1% MeGAX, mannitol, maltose, rhamnose, mannose, glucuronat e and xylose minimal media with or without 1mM cAMP, and compared with wild type E. asburiae JDR-1. Two of the MeGAX defective mutants were determined to be xylAand xylBmutants. The MeGAX mi nimal medium growing with xylAwas analyzed with HP LC after 96 hours.

PAGE 93

93 Determination of MeGAX Utilization Genes by Transforming E. coli with Cosmid Library and Topo Plasmid Library Genomic DNA was extracted from 15 ml overnight cultures of E. asburiae JDR-1. The cell culture was washed and lysed; then DNA wa s precipitated with CTAB (Delsal et al. 1989). Phenol/chloroform was used to extract genom ic DNA, which was pushed through Hamilton 50 l syringe with 22s needle 30 times to shear a nd generate fragments. After an end repairing reaction for the sheared fragments with Klenow enzyme (NEB), 1% low melting point agarose gel was used to separate the DNA fragments w ith Field Inversion Gel Electropherisis (FIGE) (Chow 2007, Guang 2009). The 30 to 40 kb fragments were cut out and ligated with pWEB-TNC vetor, packaged with the MaxPlax Packaging Extracts and trasfected E.coli EPI100-T1R (EPICENTRE Biotechnologies, USA). The obtaine d transformants were selected on 50 g/ml Ampicillin LB plates. Similarly, an E. asburiae JDR-1 genomic DNA library of 4kb to 6 kb fragments ligated to pCR4-Topo (Invitr ogen, USA) were established in E.coli TOP10. The E.coli TOP10 transformed with E. asburiae strain JDR-1 genomic fragment s were selected to obtain transformants able to grow in MeGAX minimal medium. Obtaining Candidate Genes by Sequencing Cosmid Containing -Xylosidase Gene and Genome Comparison E. asburiae JDR-1 cosmid library consists of E.coli carrying pCC1 vector with 33 to 45 kb fragments were incubated with 4-methylumbelliferyl -xylopyranoside (MUX) on LB agar plates. After exposure to UV, the clones (obtained by Guang Nong in 2003) that degrade MUX and showing fluorescence were identified to possess -xylosidase gene (Chris takopoulos et al. 1996). The cosmid containing -xylosidase was sequenced by the shotgun method and resulted in 14 contigs after assembly. NCBI Blastx program was used to suggest functions of genes in these contigs.

PAGE 94

94 Enterobacter 638 was a newly sequenced Enterobacter strain (http://genome.jgi-psf.org/finished_microbes/en t_6/ent_6.home.html), the sequences of which showed highest similarity among all the bacteria with the known sequences of E. asburiae JDR-1. Because of the similarity, we compared its genome to genome of E. coli K12 to obtain hydrolase genes. About 50 hydrolase genes present in the annotated Enterobacter 638 genome but not in E.coli K12 along with several transporter genes were obtained. Th e annotated functions of these genes were evaluated with CAZy to select candi date genes for deletion to confirm involvement in MeGAX utilization process. Deletion of Candidate Genes to Confirm Th eir Involvement in MeGAX Utilization Process The presence of candidate genes in E. asburiae JDR-1 was determined by PCR with degenerate primers or PCR with specific primers. Partial sequences of th e candidate genes were cloned into Topo vector and ligated to cat-sac B fragment to construct DNA fragments to delete their counterpart genes in E. asburiae JDR-1 genome using the gene deletion method in chapter 3. The deletion strains were grown in MeGAX minimal medium to determine whether the specific gene deletion affects the MeGAX utilization phenotype. Results MeGAX Defective Mutants of E. asburiae JDR-1 No PCR product was obtained with degenerate primers for -glucuronidase genes. This result suggestted the absence of -glucuronidases in GH family 67 in E. asburiae JDR-1. Anohter supporting evidence is the phenotype that E. asburiae JDR-1 is only able to utilize MeGAX but not MeGAX2 or MeGAX3, since all GH67 -glucuronidases that have been characterized are ab le to hydrolyze the -1,2 glycosidic linkage in MeGAX3. Therefore, E. asburiae JDR-1 may contain a novel enzyme with activity to hydrolyze MeGAX or a new process to degrade MeGAX.

PAGE 95

95 To determine genes required in the process of MeGAX utilization, more than 10,000 insertion mutants with Tn5 transposon of E. asburiae JDR-1 were obtained to select for MeGAX defective mutants. After repeating experiments of rounds of counter enri chment using xylose and MeGAX with D-cycloserine, about 700 mutants we re screened for thei r ability to grow on MeGAX minimal agar plates, of which 52 MeGAX u tilization defective muta nts were selected and screened in liquid MeGAX minimum medium to confirm the phenotype. As much as 16 individual MeGAX defective mutant s were isolated. Of those mutants, 15 were sequenced to be phosphotransferase system enzyme I defective mutants ( pstImutant), in which 14 mutants had same insertion site within ptsI (Genebank accession number: FJ527306) and 1 mutant had a different insertion site within the same gene. One of the defective mutants was determined to have an insertion mutation in a gene with sugge sted function as an oxireductase domain protein. However, no characterized gene was homologous to this oxi gene (Genebank accession number: FJ527305). Therefore, the specific func tion of this gene is unkown. Phosphotransferase system enzyme I (PTS EI) is also named phosphoenolpyruvate-protein kinase, which transfers a phosphate group from PE P to itself then to HPr. HPr transfers the phosphate group to phospho-III ( or II-A ), which is specific to differe nt substrates. PTS EI is also a key enzyme in the catabolite repression syst em. In the phosphorylated form, II-A-P activates adenylcyclase and produces cAMP, which binds to CAP (catabolite activator protein), and the CAP-cAMP complex activates multiple operons encoding enzymes utilizing different non-PTS sugars. In dephosphorylated form, II-A binds to permeases and blocks the transportation of some substrates (Figure 5-1) (Saier et al. 1976). Three possible reasons may explain why E. asburiea strain JDR-1 ptsImutants were defective in the utilizat ion of MeGAX: 1) MeGAX may be tr ansported by the PTS system, and

PAGE 96

96 its specific EII was unable to obtain the phosphate group from EI in the ptsImutant, in which case MeGAX is classified as a PTS sugar. 2) The permease for MeGAX may be inhibited by binding of the non-phpsphorylated form of the II -A protein, in which case MeGAX is a non-PTS sugar. 3) One or more genes encoding protei ns required for utilization of MeGAX may be under positive regulation of CAP-cAMP complex, and the phosphorylated form of IIA required to activate of the adenylcyclase was not available in the ptsImutant. MeGAX is a non-PTS sugar in this case. To determine the tran sportation and regulation of MeGAX in E. asburiea JDR-1, the growth rate of ptsImutant was measured in MeGAX, mannitol, maltose, rhamnose, mannose, glucuronate and xylose minimal media with or without 1mM cA MP, and compared with wild type (Table 5-1). Mannit ol is a PTS sugar in E. coli With or without cAMP, E. asburiea JDR-1 ptsImutant was not able to grow with mannitol as sole carbon source, suggesting that mannitol was also imported with PTS in E. asburiea JDR-1. Maltose is a non-PTS sugar in E. coli and regulated by CAP-cAMP complex. The PtsImutant grew normally with 1mM cAMP in maltose while it grew slowly in th e absence of cAMP, suggesting maltose was a non-PTS sugar in E. asburiea JDR-1 and under regulation of CAPcAMP. The growth patterns of the ptsImutant in maltose and in MeGAX were similar. The ptsImutant demonstrated a poor growth rate in the medium of MeGAX with 1mM cAMP, but did not grow without cAMP. These results suggested that in E. asburiea JDR-1, MeGAX was a non-PTS sugar, the utilization of which was repressed by PTS sugars with the CAPcAMP catabolite regulation. To obtain MeGAX defective mutants other than the dominant ptsImutants, mannitol was used as growth substrate with MeGAX to counter enrich MeGAX defective mutants with D-cycloserine. Two genetically different muta nts were obtained; the first one had a Tn5 transposon insertion at xylose isomerase ( xylA ) and the second one at xylulokinase ( xylB ), both

PAGE 97

97 of which were actually xylose defective muta nts. These two mutants were unable to utilize MeGAX due to their inability to utilize xylose. Th eir inability to utilize xylose was presumably responsible for their slow growth in MeGAX. The xylAmutant accumulated 4.1 mM xylose (quantified by HPLC) after 72 hr growth in 3.9mM MeGAX minimal medium (Fig 5-2). This result suggested that all xylos e released from MeGAX in the process of MeGAX utilization was not used by xylAmutant and accumulated in the medium Although the methylglucuronate part was utilized by the mutant, the accumulation of xylose slowed the process and decreased the growth rate. This result, along with the analys is of fermentation products and the substrate preferential utilization expe riment, indicated that MeGAX wa s degraded into xylose and methylglucuronate (MeGA). Candidate Genes Determined From Cosmid and Plasmid Library Tr ansformation Method E. coli TOP10 host was transformed by the 4 to 6 kb TopoPCR4 library. The transformants were selected upon MeGAX minimal medium, no E.coli TOP10 transformant able to grow was obtained Similarly, no E. coli EPI100 transformants carrying the pWEB-TNC cosmid with 30 to 40 kb fragment was able to grow in MeGAX mi nimal medium. These resu lts suggest that the essential genes involved in MeGAX utilization might be located at different positions in the genome of E. asburiae strain JDR-1 separated by significant distance, so that the 30 to 40 kb fragment could not contain all the essential genes that might enable E. coli to utilize MeGAX. Alternatively, unknown regulation fa ctors may be missing that are required for expression of the E. asburiae genes in E. coli EPI100. Obtaining Candidate Genes for Function Confirmation by Gene Deletion 14 sequence contigs were assembled after s hotgun sequencing of the cosmid containing -xlosidase. Some candidate genes with functio ns related to xylose oligomer utilization, especially MeGAX utilization, were expected to be obtained from the contigs. However, only an

PAGE 98

98 interesting gene was identified in one conti g, which is a sequence of about 600 base pairs (Genebank accession number: FJ527309) immediately located before the -xlosidase gene. The highest homologous gene for this fragment is a sugar (Glycoside-P entoside-Hexuronide) transporter gene from Enterobacter 638, which could possibly be the transporter for MeGAX. An interesting gene that Enterobacter 638 contains but absent in E. coli K12 is a GH88 family gene listed in CAZy. A characterized enzyme in this family is a -4,5 unsaturated -glucuronyl hydrolase (EC 3.2.1.-). This enzyme recognizes -4,5 unsaturated uronic acid at the nonreducing terminal of polys accharide and hydrolyzes the -linkage between the unsaturated uronic acid and sugars at the reducing end (Fig 5-3) (Hashimoto et al. 1999). In the compound MeGAX, if the methyl group was removed by a -elimination reaction from the glucuronic acid residue, a -4,5 unsaturated-glucuronyl-x ylose would be produced. A -4,5 unsaturated -glucuronyl hydrolas e from GH88 may recognize the -4,5 unsaturated-glucuronyl residue and hydrolyze the -1,2 linkage between the -4,5 unsaturated-glucuronyl residue and xylose residue. PCR with degenerate primer s was able to amplify a GH88 family gene (Genebank accession number: FJ527307) from E. asburiae JDR1 genomic DNA. Two more glycoside hydrolase genes, a lytic transglycosylase (Genebank accession number: FJ527310) and a glycotransferase (Genebank accession number: FJ527308), were amplified from E.asburiae JDR-1 genomic DNA with degenerate primers based on GH67 ( -glucuronidases) genes. However, both ge nes were not homologous with GH67 genes indicated by Blastx program of NCBI. The four genes were individually deleted in E. asburiae JDR-1. However, all the knock-out mutants maintained the ability to utilize MeGAX in minimal medium, which

PAGE 99

99 suggested these genes may not be involve d in MeGAX utilization or other genes in E. asburiae JDR-1 may back up the f unction of these genes. Discussion To define the MeGAX utilization pathway in E. asburiae JDR1, several genetic methods was employed. Four genes, oxi-, ptsI-, xylA-, xylBwere determined to be involved by screening the single insertion mutant library of E. asburiae JDR1. Along with pr evious physiological research results, such as rates of substrate utilization, product form ation and molar growth yields, the genetic information indicated that MeGAX wa s transported into the cell and hydrolyzed to release methanol, glucuronic acid and xylose. The drawback of this method is the low efficiency to obtain hydrolase genes, si nce too many genes were involve d in MeGAX utilization process and a specific dominant gene dele tion would be obained most ofte n after the counter enrichment. The possibility to obtain all genes or the spec ific glycohydrolyse gene by this method is very low. Assuming E. asburiae JDR-1 genes were able to be expressed in E. coli with their original transcription and translation elements, the co smid and plasmid library transformation method would be a simple and effective method. However, the results turned out to be otherwise. Except for the possibility of poor gene expression, th e essential genes for Me GAX utilization may be located at different positions in the genome a nd can not be contained within single 30 to 40 kb fragments cloned into the cosmid library. No E. asburiae JDR1 strain with selected candida te gene deletion was defective in MeGAX utilization. This result did not necessarily mean that these genes were not involved in this process. For example, the genomic information of Enterobacer 638 indicated that this bacterium had two GH88 genes located at differe nt positions. Thus some of these deleted

PAGE 100

100 candidate genes may have an allele which could complement its function to enable E. asburiae JDR1 to utilize MeGAX. The complexity of E. asburiae JDR1 genome made it difficult to obtain the whole pathway of MeGAX utilization or the essential genes. Th e future plan is to compare the changes in transcriptome of E. asburiae JDR1 growing in MeGAX and other substrartes. The genes up-regulated when E. asburiae JDR1 utilizes MeGAX as main carbon source may represent the genes which constitute the pathway of MeGAX utilization and genes associated with MeGAX utilization.

PAGE 101

101 Table 5-1. Growth rate of ptsIin different substrate with and without cAMP Source of carbon cAMP WT growth rate PTSI-Growth rate 1mM ++++a + MeGAX 0 ++++ 0b 1mM ++++ 0 Mannitol 0 ++++ 0 1mM ++++ ++++ Maltose 0 ++++ + Rhamnose 0 ++++ ++++ Mannose 0 ++++ ++++ Glucuronate 0 ++++ ++ Xylose 0 ++++ ++++ a: Relative growth rate was indicated with +, ++++ indicates the fastest and + indicates the slowest. b: indicates no growth.

PAGE 102

102 Figure 5-1. Function and regulati on of PTS system (Boris Grk e1 & Jrg Stlke 2008) with permission from Nature Publishing Gr oup. Copyright @ Nature Publishing Group Figure 5-2. HPLC profile of aerobic culturing media of E. asburiae JDR-1 xylA-. xylose ZH salts ZH salts MeGAX minimal medium after aerobic growth of E.asburiae JDR-1 MeGAX minimal medium after aerobic growth of E.asburiae JDR-1 xylAZH salts ZH salts ZH salts MeGAX minimal medium ZH salts MeGAX

PAGE 103

103 Figure 5-3. Enzyme activity of GH88 enzyme unsaturated -glucuronyl hydrolase (EC 3.2.1.-) (Hashimoto et al., 1999) with permission fr om Elsevier Ltd. C opyright @ Elsevier Ltd

PAGE 104

104 CHAPTER 6 CONCLUSION Dilute acid pretreatment is a well established method for hydrolyzing the methylglucuronoxylans in the hemicellulose fraction from lignocellulosics to release fermentable xylose. Xylose substituted with -1,2-methylglucuronate is rel eased as methylglucuronoxylose (MeGAX), which cannot be fermented by biocatalys ts currently used to produce biofuels and chemicals. E. asburiae JDR-1, isolated from colonized hardwood (sweetgum), efficiently ferments both methylglucuronoxylose and xylose, producing predominantly ethanol and acetate. 13C-NMR studies defined the Embden-Meyerhof pathway for metabolism of glucose and the pentose phosphate pathway for xylose metabolis m. Rates of substrate utilization, product formation and molar growth yields indicate me thylglucuronoxylose is tr ansported into the cell and hydrolyzed to release methanol, xylose and hexauronate for further catabolism. Enterobacter asburiae stain JDR-1 is the first microorganism de scribed that ferments methylglucuronoxylose generated along with xylose by the acid medi ated saccharification of hemicellulose. Deletion of pflB and als genes in this bacterium m odified the native mixed acid fermentation pathways into pathway for homolactate production. The resulting strain, E. asburiae L1, completely utilized both xylose a nd MeGAX in a dilute acid hydrolysate of sweetgum xylan and produced optically pure D(-) lactate at a yield approximating 100% of the theoretical maximum. With the ability to utilize MeGAX in xylan hydrolysates, as well as the other saccharides released during dilute acid pr etreatment, the engineer ed strain provides a uniquely efficient bacterial biocatalyst for lactat e production from the hemicellulose fractions of lignocellulosic biomass.

PAGE 105

105 The PET operon including Zymomonas mobilis pyruvate decarboxylase ( pdc) and alcohol dehydrogenase B ( adhB) genes was transferred into E. asburiae JDR-1, converting the native mixed acid fermentation pathways to homoethanol production. Integrati on of PET operon into JDR-1 chromosome did not increase the ethanol yield, which may be due to insufficient PDC expression. In contrast, expressing pdc and adhB in a stable plasmid pLOI555 significantly increased the ethanol production. Deleti on of the pyruvate formatelyase ( pflB ) gene further increased the ethanol yield, resulting E. asburiae E1 (pLOI555). This strain completely utilizes both xylose and MeGAX in dilute acid hydrolys ate of sweetgum xylan and produced ethanol with a yield 99% of the theoretical maximum at the rate of 0.11 g ethanol/g DCW/h, which was 1.57 the yield and 1.48 the rate obtained with ethanologenic Escherichia coli KO11. This engineered strain of E. asburiae JDR-1, which is able to ferment the predominant hexoses and pentoses derived from both hemicellulose and ce llulose fractions, offers a promising subject for development as an ethanologenic biocatalys t for production of fuels and chemicals from agricultural residues and energy crops. To define the genetic basis in MeGAX utilization, an E. asburiae JDR-1 mutant library generated by Tn5 transposon mutagenesis was screened for MeGAX-defective mutants. The mutant strains obtained were identified as oxi-, ptsI-, xylA-, xylB-. Growth status of ptsIin different carbon sources with and without 1 mM cAMP suggested MeGAX was transported in E. asburiae JDR-1 through a non-PTS pathway and under regulation by the cA MP-CAP catabolism repression system. The slow growth of xylose defective mutants, xylAand xylB-, in MeGAX with accumulation of free xylose in the growth medium suggests that xylose is released in the process of MeGAX utilization. Alo ng with previous physiological studies, such as rates of substrate utilization, product formation and mo lar growth yields, the genetic information

PAGE 106

106 indicates MeGAX is transported into the cell an d hydrolyzed to release methanol, glucuronic acid and xylose. Several other genetic methods, su ch as the PCR by degenerate primers, positive gene selection and specific deletion of candidate gene, were also app lied to define the MeGAX utilization pathway in E. asburiae JDR-1. However, these methods failed to identify specific genes involved in MeGAX utilization. The experi ment to analyze transcriptome changes when E. asburiae JDR1 utilizes MeGAX as sole carbon s ource will be performed for obtaining information of genes constituting the pathway of MeGAX utilization.

PAGE 107

107 REFERENCE LIST Asis CA, Adachi K (2004) Isolation of e ndophytic diazotroph Pantoea agglomerans and nondiazotroph Enterobacter asbur iae from sweetpotato stem in Japan. Letters in Applied Microbiology 38:19-23 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 Babson AL, Babson SR (1973) Kinetic colori metric measurement of serum lactate dehydrogenase activity. Clin Chem 19:766-769 Bauchop T, Elsden SR (1960) The growth of micr o-organisms in relation to their energy supply. J Gen Microbiol 23:457-469 Beg QK, Kapoor M, Mahajan L, Hoondal GS (2001) Microbial xylanases and their industrial applications: a review. Appl Microbiol Biotechnol 56:326-338 Boris Grke1 & Jrg Stlke1 (2008) Carbon cata bolite repression in bacteria: many ways to make the most out of nutrients Nat Rev Microbiol 6: 613-624 Brenner DJ, Mcwhorter AC, Kai A, Steigerwalt AG, Farmer JJ (1986) Enterobacter asburiae Sp-Nov, a new speciesfound in clinical specimens, and reassignment of Erwinia dissolvens and Erwinia nimipressuralis to the genus Enterobacter as Enterobacter dissolvens Comb-Nov and Enterobacter nimipressuralis Comb-Nov. J Clin Microbiol 23:1114-1120 Chow V, Nong G, Preston JF (2007) Structure, function, and regulation of the aldouronate utilization gene cluster from Paenibacillus sp. strain JDR-2. J Bacteriol 189: 8863 Christakopoulos P, Mamma D, Nerinckx W, Kekos D, Macris B, Claeyssens M (1996) Production and partia l characterization of xylanase from Fusarium oxysporum Bioresour Technol 58:115-119 Converse AO, Kwarteng IK, Grethlein HE, Ooshim a H (1989) Kinetics of thermochemical pretreatment of lignocellulosic material s. Appl Biochem Biotechnol 20-1:63-78 Conway T, Sewell GW, Osman YA, Ingram LO (1987) Cloning and sequ encing of the alcohol dehydrogenase II gene from Zymomonas mobilis J Bacteriol 169:2591-2597 Datsenko KA, Wanner BL (2000) One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products. Proc Nat Acad Sci U.S.A. 97:6640-6645 de Vries RP, Visser J (2001) Aspergillus enzymes involved in degrad ation of plant cell wall polysaccharides. Microbiol Molecul Bio Rev 65:497-522

PAGE 108

108 Delsal G, Manfioletti G, Schneider C (1989) The CTAB DNA precip itation method a common mini-scale preparation of template DNA from phagemids, phages or plasmids suitable for sequencing. Biotech 7:514-520 Dien BS, Cotta MA, Jeffries TW (2003) Bacteria engineered for fuel ethanol production: current status. Appl Microbio l Biotechnol 63:258-266 Dien BS, Jung HJG, Vogel KP, Casler MD, Lamb JFS, Iten L, Mitchell RB, Sarath G (2006) Chemical composition and response to d ilute-acid pretreatme nt and enzymatic saccharification of alfalfa, reed canarygr ass, and switchgrass. Biomass and Bioenergy 30:880-891. Dombek KM, Ingram LO (1987) Ethanol pr oduction during batch fermentation with Saccharomyces cerevisiae changes in glycolytic enzymes and internal Ph. Appl Environ Microbiol 53:1286-1291 Dubois M, Gilles KA, Hamilton JK, Rebers PA, Smith F (1956) Colorimetric method for determination of sugars and rela ted substances. Anal Chem 28:350-356 Eggeman T, Elander RT (2005) Process and econo mic analysis of pretreatment technologies. Bioresour Technol 96:2019-2025 Elbeltagy A, Nishioka K, Sato T, Suzuki H, Ye B, Hamada T, Isawa T, Mitsui H et al. (2001) Endophytic colonization and in planta nitrogen fixation by a Herbaspirillum sp. isolated from wild rice species. A ppl Environ Microbiol 67: 5285. Eissen M, Metzger JO, Schmidt E, Schneidewind U (2002) 10 years after Rio Concepts on the contribution of chemistry to a sustainable de velopment. Angewandte Chemie International Edition 41:414-436 Esteghlalian A, Hashimoto AG, Fenske JJ, Penne r MH (1997) Modeling a nd optimization of the dilute-sulfuric-acid pretreatment of corn st over, poplar and switchg rass. Bioresour Technol 59:129-136 Freer SN, Detroy RW (1983) Characterization of ce llobiose fermentations to ethanol by Yeasts. Biotechnol Bioeng 25:541-557 Galbe M, Zacchi G (2002) A review of the produc tion of ethanol from softwood. Appl Microbiol Biotechnol 59:618-628 Gao K, Mckinley KR (1994) Use of macr oalgae for marine biomass production and Co2 remediation a review. J Appl Phycol 6:45-60 Gilson E, Alloing G, Schmidt T, Claverys JP, Dudler R, Hofnung M (1988) Evidence for high-affinity binding protein dependent transport systems in gram positive bacteria and in Mycoplasma Embo J 7:3971-3974 Gray KA, Zhao LS, Emptage M (2006) Bi oethanol. Curr Opin Chem Biol 10:141-146

PAGE 109

109 Gunsalus IC, Horecker BL, Wood WA (1955) Pathways of carbohydrate metabolism in microorganisms. Bacteriol Rev 19:79-128 Gunsalus IC (1961) Energy yielding metabolism in b acteria. In: Gunsalus IC and Stanier RY (ed) the Bacteria. Academic Press Inc., New York, pp 1-58 Gutierrez T, Buszko ML, Ingram LO, Preston JF (2002) Reduction of furfural to furfuryl alcohol by ethanologenic strains of bact eria and its effect on ethano l production from xylose. Appl Biochem Biotechnol 98:327-340 Hahn-Hagerdal B, Galbe M, Gorwa-Grauslund MF Liden G, Zacchi G (2006) Bio-ethanol the fuel of tomorrow from the residues of today. Trends Biotechnol 24:549-556 Hallmann J, QuadtHallmann A, Mahaffee WF, Kloepper JW (1997) Bacterial endophytes in agricultural crops. Canadian J ournal of Microbiology 43:895-914 Hashimoto W, Kobayashi E, Nankai H, Sato N, Miya T, Kawai S, Murata K (1999) Unsaturated glucuronyl hydrolase of Bacillus sp GL1: Novel enzyme prer equisite for metabolism of unsaturated oligosaccharides produced by pol ysaccharide lyases. Arch Biochem Biophy 368:367-374 Hasona A, Kim Y, Healy FG, Ingram LO, Shan mugam KT (2004) Pyruvate formate lyase and acetate kinase are essential for anaerobic growth of Escherichia coli on xylose. J Bacteriol 186:7593-7600 Himmel ME, Ding SY, Johnson DK, Adney WS, Nimlos MR, Brady JW, Foust TD (2007) Biomass recalcitrance: Engineering plants and enzymes for biofue ls production. Science 315:804-807 Hinman ND, Schell DJ, Riley CJ, Bergeron PW, Wa lter PJ (1992) Preliminary estimate of the cost of ethanol production for SSF technol ogy. Appl Biochem Biotechnol 34-5:639-649 Hoffert MI, Caldeira K, Benford G, Criswell DR, Green C, Herzog H, Jain AK, Kheshgi HS, Lackner KS, Lewis JS, Lightfoot HD, Manheim er W, Mankins JC, Mauel ME, Perkins LJ, Schlesinger ME, Volk T, Wigley TML (2002) Advanced technology path s to global climate stability: Energy for a gree nhouse planet. Science 298:981-987 Hoffmann H, Stindl S, Ludwig W, Stumpf A, Me hlen A, Heesemann J, Monget D, Schleifer KH, Roggenkamp A (2005) Reassignment of Enterobacter dissolvens to Enterobacter cloacae as E. cloacae subspecies dissolvens comb. nov and emended description of Enterobacter asburiae and Enterobacter kobei System Appl Microbiol 28:196-205 Hong J, Ye XH, Zhang YHP (2007) Quantitative de termination of cellulose accessibility to cellulase based on adsorption of a nonhydrolyt ic fusion protein containing CBM and GFP with its applications Langmuir 23:12535-12540 Hu ZH, Wen ZY (2008) Enhancing enzymatic dige stibility of switchgrass by microwave-assisted alkali pretreatment. Biochem Eng J 38:369-378

PAGE 110

110 Hugouvieuxcottepattat N, Robertbaudouy J (1987) Hexuronate catabolism in Erwinia chrysanthemi J Bacteriol 169:1223-1231 Hurlbert JC, Preston JF (2001) F unctional characterization of a nove l xylanase from a corn strain of Erwinia chrysanthemi J Bacteriol 183:2093-2100 Ingram LO, Aldrich HC, Borges ACC, Causey TB, Martinez A, Morales F, Saleh A, Underwood SA, Yomano LP, York SW, Zaldivar J, Zhou SD (1999) Enteric bacteria l catalysts for fuel ethanol production. Bi otechnol Pro 15:855-866 Ingram LO, Alterthum F, Ohta K (1989) Efficien t ethanol production from xylose, lactose and glucose by recombinant Escherichia coli Appl Environ Microbiol 55:1943-1948 Ingram LO, Conway T (1988) Expression of diff erent levels of ethanol ogenic enzymes from Zymomonas Mobilis in recombinant strains of Escherichia coli Appl Environ Microbiol 54:397-404 Ingram LO, Conway T, Clark DP Sewell GW, Preston JF (1987) Ge netic engineering of ethanol production in Escherichia coli Appl. Environ Microbiol 53:2420-2425 Jacobs A, Larsson PT, Dahlman O (2001) Distribution of uronic acids in xylans from various species of softand hardwood as de termined by MALDI mass spectrometry. Biomacromolecules 2:979-990 Jarboe LR, Grabar TB, Yomano LP, Shanmuga n KT, Ingram LO (2007) Development of ethanologenic bacter ia. Biofuels 108:237-261 Jantama K, Haupt MJ, Svoronos SA, Zhang XL, Moore JC, Shanmugam KT, Ingram LO (2008) Combining metabolic engineering and metabol ic evolution to develop nonrecombinant strains of Escherichia coli C that produce succinate and malate. Biotechnol Bioeng 99:1140-1153 Jeffries TW (2005) Ethanol fermentati on on the move. Nat Biotechnol 23:40-41 Jeffries TW, Jin YS (2004) Metabolic engineerin g for improved fermentation of pentoses by yeasts. Appl Microbio l Biotechnol 63:495-509 Jeuland N, Montagne X, Gautrot X (2004) Potentia lity of ethanol as a fuel for dedicated engine. Oil & Gas Science and Technology-Revue de l Institut Francais du Petrole 59:559-570 Jones JK, Purves CB, Timell TE (1961) Consti tution of A 4-0-Methylglucuronoxylan from Wood of Trembling Aspen ( Populus Tremuloides Michx ). Can J of Chem 39:1059-1066 Kardosova A, Matulova M, Malovikova A ( 1998) (4-O-methyl-alpha-D-glucurono)-D-xylan from Rudbeckia fulgida, var. sullivantii (Boynton et Beadle). Carbohydr Res 308:99-105 Kheshgi HS, Prince RC, Marland G (2000) The potential of biomass fuels in the context of global climate change: Focus on transportation fuels. Ann Rev Ener Environ 25:199-244

PAGE 111

111 Kim Y, Ingram LO, Shanmugam KT (2007) Construction of an Escherichia coli K-12 mutant for homoethanologenic fermentation of glucose or xy lose without foreign genes. Appl Environ Microbiol 73:1766-1771 Lal R (2004) Soil carbon sequestration impacts on global climate change and food security. Science 304:1623-1627 Lawford HG, Rousseau JD (1991) Ethanol production by recombinant Escherichia coli carrying genes from Zymomonas mobilis Appl Biochem Bi otechnol 28-9:221-236 Lawford HG, Rousseau JD (2002) Performance testing of Zymomonas mobilis metabolically engineered for cofermentation of glucose, xyl ose, and arabinose. Appl Biochem Biotechnol 98:429-448 Lee YY, Iyer P, Torget RW (1999) Dilute-aci d hydrolysis of lignocellulosic biomass. Adv Biochem Eng Biotechnol 65:93-115 Lin Y, Tanaka S (2006) Ethanol fermentation from biomass re sources: current state and prospects. Appl Microbiol Biotechnol 69:627-642 Linton KJ, Higgins CF (1998) The Escherichia coli ATP-binding cassette (ABC) proteins. Molecul Microbiol 28:5-13 Lynd LR, Wyman CE, Gerngross TU (1999) Bi ocommodity engineering. Biotechnol Pro 15:777-793 Matsushika A, Sawayama S (2008) Efficient bioethanol producti on from xylose by recombinant Saccharomyces cerevisiae requires high activity of xyl ose reductase and moderate xylulokinase activity. J Biosci Bioeng 106: 306. Matsushika A, Watanabe S, Kodaki T, Maki no K, Inoue H, Murakami K, Takimura O, Sawayama S (2008). Expressi on of protein engineered NADP+-dependent xylitol dehydrogenase increase ethanol produc tion from xylose in recombinant Saccharomyces cerevisiae Appl Microbiol Biotechnol 81: 243 McInroy, J.A. and Kloepper, J.W. (1995) Survey of indigenous bacteria l endophytes from cotton and sweet corn. Plant and Soil 173: 337. McMillan JD (1997) Bioetha nol production: Status and prospects. Renewable Energy 10:295-302 Meshartree M, Dale BE, Craig WK (1988) Comparison of steam and ammonia pretreatment for enzymatic hydrolysis of cellulose. Appl Microbiol Biotechnol 29:462-468 Miller EN, Jarboe LR, Yomano LP, York SW, Sh anmugam KT, Ingram LO (2009) Silencing of NADPH-dependent oxidoreductase genes ( yqhD and dkgA ) in furfural-resistant ethanologenic Escherichia coli Appl Environ Microbiol 75: 4315-4323

PAGE 112

112 Moat AG, Foster JW, Spector MP (2002) Microbial Physiology, 4th ed. Wiley-Liss, New York Monties B (2005) Biological variability of lignin. Cellul Chem Technol 39:341-367 Nong G, Chow V, Rice JD, St John FJ, Pr eston JF (2009) Aldouronate utilization in Paenibacillus sp. strain JDR-2: physiological an d enzymatic evidence for coupling of extracellular depolymerization and intracellu lar metabolism. Appl Environ Microbiol. in press Ohta K, Beall DS, Mejia JP, Shanmugam KT, Ingram LO (1991a) 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 Ohta K, Beall DS, Mejia JP, Shanmugam KT, Ingram LO (1991b) Metabolic engineering of Klebsiella oxytoca M5A1 for ethanol production from xylose and glucose. Appl Environ Microbiol 57:2810-2815 Ostergaard S, Olsson L, Nielsen J (2000) Metabolic engineering of Saccharomyces cerevisiae Microbiol Molecul Biol Rev 64:34-50 Palmqvist E, Hahn-Hagerdal B (2000b) Fermen tation of lignocellulosi c hydrolysates. II: inhibitors and mechanis ms of inhibition. Bioresour Technol 74:25-33 Palmqvist E, Hahn-Hagerdal B (2000a) Fermen tation of lignocellulosic hydrolysates. I: inhibition and detoxification. Bioresour Technol 74:17-24 Patel MA, Ou MS, Harbrucker R, Aldrich HC Buszko ML, Ingram LO, Shanmugam KT (2006) Isolation and characterization of acid-tolerant, thermophilic bacteria for effective fermentation of biomass-derived sugars to lactic acid. Appl Environ Microbiol 72:3228-3235 Pordesimo LO, Hames BR, Sokhansanj S, Eden s WC (2005). Variation in corn stover composition and energy content with crop ma turity. Biomass and Bioenergy 28:366-374. Prasad S, Singh A, Joshi HC (2007) Ethanol as an alternative fuel from agricultural, industrial and urban residues. Resour Conserv Recycl 50:1-39 Preston JF, Hurlbert JC, Rice JD Ragunathan A, St John FJ (2003) Microbial strategies for the depolymerization of glucuronoxylan: leads to biotechnological applications of endoxylanases. Applications of Enzy mes to Lignocellulosics 855:191-210 Preston JF, Rice JD, Ingram LO, Keen NT (1992) Differential depolymerization mechanisms of pectate lyases secreted by Erwinia chrysanthemi Ec16. J Bacteriol 174:2039-2042 Puchta H, Hohn B (1991) A transient assay in pl ant cells reveals a posit ive correlation between extrachromosomal recombination rates and length of homologous overlap. Nucleic Acids Res 19:2693-2700

PAGE 113

113 Puls J (1997) Chemistry and biochemistry of hemicelluloses: Relationship between hemicellulose structure and enzymes requi red for hydrolysis. M acromol Symp 120:183-196 Qian Y, Yomano LP, Preston JF, Aldrich HC, In gram LO (2003) Cloning, characterization, and functional expression of the Klebsiella oxytoca xylodextrin utiliza tion operon (xynTB) in Escherichia coli Appl Environ Microbiol 69:5957-5967 Quadt-Hallmann A, Kloepper JW (1996) Immunologi cal detection and local ization of the cotton endophyte Enterobacter asburiae JM22 in different plant spec ies. Canadian Journal of Microbiology 42, 1144. 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, Tschaplinski T (2006) The path forward for biofuels a nd biomaterials. Science 311:484-489 Reddy N, Yang Y (2005) Biofibers from agricultur al byproducts for indu strial applications. Trends Biotechnol 23:22-27 Robertbaudouy J, Portalier R, Stoeber F (1981) Regulation of hexuronate system genes in Escherichia Coli K-12 multiple regulation of the uxu operon by exur and uxur gene products. J Bacteriol 145:211-220 Rodriguez M, Martinez EA, York SW, Zuobi-Hasona K, Ingram LO, Shanmugam KT and Preston JF (2001) Properties of the hemicellu lose fractions of li gnocellulosic biomass affecting bacterial ethanol production. Abstracts of the 101st National Meetings of the American Society of Microbiol ogy meetings in Orlando, FL, USA P 535 (2001). Rogers PL, Jeon YJ, Lee KJ, Lawford HG (2007) Zymomonas mobilis for fuel ethanol and higher value products. Biofuels 108:263-288 Russell JB, Cook GM (1995) Energetic s of bacterial growth balan ce of anabolic and catabolic reactions. Microbiol Rev 59:48-62 Saha BC (2003) Hemicellulose bioconvers ion. J Ind Microbiol Biotechnol 30:279-291 Saha BC, Iten LB, Cotta MA, Wu YV ( 2005b) Dilute acid pretreatment, enzymatic saccharification, and fermentation of rice hu lls to ethanol. Biotechnol Progr 21:816-822 Saha BC, Iten LB, Cotta MA, Wu YV ( 2005a) Dilute acid pretreatment, enzymatic saccharification and fermentation of wheat st raw to ethanol. Process Biochem 40:3693-3700 Saier MH, Simoni RD, Roseman S (1976) Sugar transport. Properties of mutant bacteria defective in proteins of phosphoenolpyruvate : sugar phosphotransferase system. J Biol Chem 251:6584-6597 Saloheimo A, Rauta J, Stasyk OV, Sibirny AA, Penttil M, Ruohonen L (2007).Xylose transport studies with xylose-utilizing Saccharomyces cerevisiae strains expressing heterologous and homologous permeases. Appl Mi crobiol Biotechnol 74: 1041

PAGE 114

114 SanFrancisco MJD, Keenan RW (1993) Uptake of galacturonic acid in Erwinia chrysanthemi Ec16. J Bacteriol 175:4263-4265 Schunck B, Kraft W, Truyen U (1995) A simple touch-down polymerase chain reaction for the detection of canine parvovirus and feline Panleukopenia virus in feces. Journal of Virological Methods 55:427-433 Scott AI, Baxter RL (1981) Applications of 13C NMR to metabolic studies. Ann Rev Biophy Bioeng 10:151-174 Shanmugam KT, Ingram LO (2008) Engineerin g biocatalysts for production of commodity chemicals. J Mol Microbiol Biotechnol 15:8-15 Sharma V, Kumar V, Archana G, Kumar GN (2005) Substrate specificity of glucose dehydrogenase (GDH) of Enterobacter asburiae PSI3 and rock phosphate solubilization with GDH substrates as C sour ces. Can J Microbiol 51:477-482 Shulami S, Gat G, Sonenshein AL, Shoham Y ( 1999) The glucuronic acid utilization gene cluster from Bacillus stearothermophilus T-6. J Bacteriol 181:3695-3704 Simoni RD, Roseman S, Saier MH, Jr. (1976) Suga r transport. Properties of mutant bacteria defective in proteins of th e phosphoenolpyruvate: sugar phosphot ransferase system. J Biol Chem 251:6584-6597 Slininger PJ, Bothast RJ, Okos MR, Ladisch MR (1985) Comparative evaluation of ethanol production by xylose fermenting yeasts presen ted high xylose concentrations. Biotechnol Lett 7:431-436 Smalley AJ, Jahrling P, Van Demark PJ (1968) Mo lar growth yields as evidence for xxidative phosphorylation in Streptococcus faecalis strain 10Cl. J Bacteriol 96:1595-1560 Sprenger GA (1996) Carbohydrate metabolism in Zymomonas mobilis : A catabolic highway with some scenic routes. FEMS Microbiol Lett 145:301-307 St John FJ, Rice JD, Preston JF (2006) Paenibacillus sp strain JDR-2 and XynA(1): a novel system for methylglucuronoxylan utilizati on. Appl Environ Microbiol 72:1496-1506 Sticklen M (2006) Plant genetic engineering to improve biomass characteristics for biofuels. Curr Opin Biotechnol 17:315-319 Sun JX, Sun XF, Sun RC, Su YQ (2004) Fractional extraction and structural characterization of sugarcane bagasse hemicelluloses. Carbohydrate Polymers 56:195-204. Sunna A, Antranikian G (1997) Xylanolytic enzymes from fungi a nd bacteria. Crit Rev Biotechnol 17:39-67

PAGE 115

115 Sun RC, Fang JM, Tomkinson J, Geng ZC, Liu JC (2001) Fractional isolation, physico-chemical characterization and homogeneous esterification of hemicelluloses from fast-growing poplar wood. Carbohydrate Polymers 44:29-39. Taguchi H, Ohta T (1991) D-la ctate dehydrogenase is a member of the D-isomer-specific 2-hydroxyacid dehydrogenase family clon ing, sequencing, and expression in Escherichia coli of the D-Lactate dehydrogenase gene of Lactobacillus plantarum J Biol Chem 266:12588-12594 Tarkow H, Feist WC (1969) A mechanism fo r improving digestibility of lignocellulosic materials with dilute alkali and liq uid ammonia. advan chem Ser 197-218 Teymouri F, Laureano-Perez L, Alizadeh H, Dale BE (2004) Ammonia fiber explosion treatment of corn stover. Appl Bi ochem Biotechnol 113-16:951-963 Timell TE (1964) Wood Hemicelluloses. Adv Carbohydr Chem 19:247-302 Vlasenko EY, Ding H, Labavitch JM, Shoemaker SP (1997) Enzymatic hydrolysis of pretreated rice straw. Bioresour Technol 59:109-119 von Blottnitz H, Curran MA (2007) A review of assessments conducted on bio-ethanol as a transportation fuel from a net energy, greenhouse gas, a nd environmental life cycle perspective. Journal of Cl eaner Production 15:607-619 vonSivers M, Zacchi G (1996) Ethanol from li gnocellulosics: A review of the economy. Bioresour Technol 56:131-140 vonSivers M, Zacchi G (1995) A technoeconomic comparison of 3 processes for the production of ethanol from pine. Bi oresour Technol 51:43-52 Wesseler J (2007) Opportunities ('costs) matter: A comment on Pimentel and Patzek "Ethanol production using corn, switchgrass, and wood; biodiesel producti on using soybean and sunflower". Energy Policy 35:1414-1416 Whitesides GM, Crabtree GW (2007) Don't forget long-term fundamental research in energy. Science 315:796-798 Wood BE, Ingram LO (1992) Ethanol production from cellobiose, amorphous cellulose, and crystalline cellulose by recombinant Klebsiella oxytoca containing chromosomally integrated Zymomonas mobilis genes for ethanol production and plasmids expressing thermostable cellulase genes from Clostridium thermocellum Appl Environ Microbiol 58:2103-2110 Wood BE, Yomano LP, York SW, and Ingram LO (2005) Development of industrial medium required elimination of the 2,3-butanediol fermen tation pathway to mainta in ethanol yield in an ethanologenic strain of Klebsiella oxytoca Biotechnol Prog 21:1366-1372

PAGE 116

116 Wyman CE (1999) Biomass ethanol: Technical progress, opportunities, and commercial challenges. Ann Rev En ergy Environ 24:189-226 Wyman CE (2003) Potential synergies and challe nges in refining cellulosic biomass to fuels, chemicals, and power. Biotechnol Prog 19:254-262 Xiao B, Sun XF, Sun RC (2001) Chemical, st ructural, and thermal characterizations of alkali-soluble lignins and hemicelluloses, and cellulose from maize stems, rye straw, and rice straw. Polym Degrad Stab 74:307-319 Yomano LP, York SW, Ingram LO (1998) Isola tion and characterization of ethanol-tolerant mutants of Escherichia coli KO11 for fuel ethanol producti on. J Ind Microbiol Biotechnol 20:132-138 Yomano LP, York SW, Zhou S, Shamugam KT, Ingram LO (2008) Re-engineering Escherichia coli for fuel ethanol producti on. Biotechnol Lett 20:132-138 York SW, Ingram LO (1996) Ethanol production by recombinant Escherichia coli KO11 using crude yeast autolysate as a nutrien t supplement. Biotechnol Lett 18:683-688 Zaldivar J, Ingram LO (1999) Effect of or ganic acids on the growth and fermentation of ethanologenic Escherichia coli LY01. Biotechnol Bioeng 66:203-210 Zaldivar J, Martinez A, Ingram LO (1999) Eff ect of selected aldehydes on the growth and fermentation of ethanologenic Escherichia coli Biotechnol Bioeng 65:24-33 Zaldivar J, Martinez A, Ingram LO (2000) Eff ect of alcohol compounds found in hemicellulose hydrolysate on the growth and fermentation of ethanologenic Escherichia coli Biotechnol Bioeng 68:524-530 Zhang X, Jantama K, Moore JC, Shanmugam KT, Ingram LO (2007) Production of L-alanine by metabolically engineered Escherichia coli Appl Microbiol Biotechnol 77:355-366 Zhang YHP (2008) Reviving the carbohydrate economy via multi-product lignocellulose biorefineries. J Ind Micr obiol Biotechnol 35:367-375 Zhou SD, Davis FC, Ingram LO (2001) Gene integration and expression and extracellular secretion of Erwinia chrysanthemi endoglucanase CelY (celY) and CelZ (celZ) in ethanologenic Klebsiella oxytoca P2. Appl Environ Microbiol 67:6-14 Zhou SG, Ingram LO (2000) Synergistic hydrolysis of carboxymethyl cellulose and acid-swollen cellulose by two endoglucanas es (CelZ and CelY) from Erwinia chrysanthemi J Bacteriol 182:5676-5682 Zucker M, Hankin L (1970) Regulati on of pectate lyase synthesis in Pseudomonas fluorescens and Erwinia carotovora J Bacteriol 104:13-18

PAGE 117

117 Zuobi-Hasona K, St John FJ, Rice JD, Prest on JF (2001) Oligosaccharides containing glucuronoxylose as substrates for selecting ba cteria for depolymerization of hemicellulose. Abstracts of the 101st National Meetings of the American Society of Microbiology meetings in Orlando, FL, USA. p 534.

PAGE 118

118 BIOGRAPHICAL SKETCH Changhao Bi was born in Qujing, China in 1979. After attending the Kunming Second Middle School and the Kuming Fi rst High School, he was admitted in Nankai University. During his student career there, he rece ived his bachelors degree in biochemistry in 2001 and masters degree in molecular biology and biochemistry in 2004. In the same year, he was accepted to the Ph.D program in the Department of Microbiology and Cell Science at the University of Florida, then chose to work in Dr. Prestons laboratory until received his Ph.D. from the University of Florida in August 2009.