_.+..Y.x.... v 33:?
. A .i

. H
“but

1;.

fl . l n l .
A; .‘\ 2P» .‘
1‘ 1.?0HH... mer
. 2,.qr)L.V..IVJ
I}
{tuned-1.1 V.

($3.2...

.51.... :
4 . ‘1} A
A}:

n

. ‘vii .

.III‘GEi

33.2%.. .
‘31.. $.23:
2.2.2.. 3.:
‘52... c.
‘1 Egg: .

343%..

13115552.:
(ail\ola .

E.
yuuiulwxr .
u .1. 0. n . ‘5‘
a§..s...§:..i§¢i.n

:2‘32it.(x..?zl ($35.13... {123} at
V 3.3.5, 1.4.1251} . lg“.
Eliyzilzvu‘ILA-thvl. .

1.3.; £32....llfli...

1.5.3:
5.2: :3. i fighting! .n...
.. $11.5!!! 13...}.3 Eg‘ifl
I? it...
.. .

I...
. .31.... “til: . if
‘3 . 19am, ‘ - 5.

3:5“.“351‘ 2:33 it!) .31....

.2: .2... L. .ifivffl.
$31.? Wanna h .2!th

i 1:131; 1!.l.

 

' nsnARY
‘ Michigan State
L University

 

This is to certify that the
dissertation entitled

X-RAY CRYSTALLOGRAPHIC STUDIES OF BRANCHING
ENZYME/POLYSACCHARIDE COMPLEX AND RNA
POLYMERASE Ill TRANSCRIPTION FACOTOR TFIIIB COMPLEX

presented by

LeiFeng

has been accepted towards fulfillment
of the requirements for the

Ph.D. degree in Chemistry

 

 

QM

MaJ‘OrProfessor’s Signature
August 26, 2009

 

Date

MSU is an Affirmative Action/Equal Opportunity Employer

4 A_ . __ _-—c-o-o-o-o-o-o‘o-o-c-n-o-o-o-I-o--o-o--n-o-o-o—u-n-o-o-o---o-a—.-.—-o-o-o-o.

PLACE IN RETURN BOX to remove this checkout from your record.

To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 K:/Proleoc&PrelelRC/DataDue.indd

 

 

X-RAY CRYSTALLOGRAPHIC STUDIES OF BRANCHING
ENZYME/POLYSACCHARIDE COMPLEX AND RNA POLYMERASE III
TRANSCRIPTION FACOTOR TFIIIB COMPLEX

By

Lei F eng

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Chemistry

2009

ABSTRACT
X-RAY CRYSTALLOGRAPHIC STUDIES OF BRANCHING
ENZYME/POLYSACCHARIDE COMPLEX AND RNA POLYMERASE III
TRANSCRIPTION FACOTR TFIIIB COMPLEX
By

Lei Feng

Glycogen and starch are major carbon sources and the carbohydrate storage molecule
in living organisms. These two highly branched polysaccharides play a very important
role in glucose cycles in nature. Branching enzyme is one of the biosynthetic pathway
enzymes. It cleaves the (1-1,4 glucosidic bond and transfers the oligosaccharide to the
(it-1,6 position to form (1-1, 6 branch points. This action determines the final structure of
glycogen and starch, which is very important in nature and industrial application.

The E. coli branching enzyme has been crystallized and its structure has been solved.
In order to gain insight into its branching mechanism, the binding between the enzyme
and its substrate needs to be investigated. Linear and cyclic oligosaccharides were used to
bind with E. coli branching enzyme. The three dimensional crystal structures were
obtained by substrate soaking experiments and x-ray diffraction. From the data it is clear
that E.coli branching enzyme binds with a, B and y-cyclodextrins on the surface of the
protein. Four binding sites were found and the residues involved were identified. The
cyclodextrins bind with the protein through hydrogen bonds and aromatic stacking

interactions at the binding sites. Furthermore, the binding between Ecoli branching

enzyme and linear oligosaccharides such as maltohexaose and maltoheptase identified
binding sites near the catalytic region, and lead to a hypothesis for the mechanism of
substrate and branch chain specificity.

Branching enzyme 11 from maize endo sperm was also purified and screened for
crystallographic study. The full length protein did not crystallize. Further study unveiled a
stable truncated version of the protein. Crystallization attempts on this variant are
ongoing.

The structure of the TFIIIB/DNA transcription initiation complex can lead to the
understanding of how transcription initiates. This is an extremely important step in the
transfer of genetic information. The TBP, BRF and B” protein are the three subunits of
TFIIIB, the central RNA polymerase Ill cofactor. Based on previous studies, different
modified TATA box containing promoters were selected for the study. Different
constructs of the subunits were designed and purified to make a variety of
TBP/BRF/B”/DNA complexes in vitro. The quaternary complexes were screened for

crystallization, but crystals were not obtained.

Copyright by
LEI FENG

2009

To my dear mother, Mrs. Ren, Sumei; my dear father, Mr. Feng, Mengxue; my

lovely daughter Helen and my wife

ACKNOWLEDGMENTS

First of all, I would like to thank my advisors, Professor James H. Geiger for
mentoring me in my adventure into science research. He showed enormous patience and
support when I got lost in my projects. Without his help and encouragement I could not
have finished my graduate study at Michigan State University. His knowledge and
passion in science guided me through the difficult times. I sincerely thank him for all his
help.

I also would like to thank our collaborator, Professor Jack Preiss. As a well-known
pioneer in the area of the starch biosynthesis pathway, his support for our project was
tremendous. I got lots of invaluable help from discussions with Dr. Preiss about the
branching enzyme project.

I owe a very special thank to Dr. Stacy Hovde. She is the one who taught me
everything in the lab and helped me on all my research. She actually worked closely with
me for long time in Geiger’s lab before she went to the Department of Biochemistry and
Molecular Biology. Without her kind help, I could not achieve anything. Thank you,
Stacy!

I also want to say thank you to my friends and group members Xiaofei J ia, Xiangshu
Jin, Fang Sheng, and Susan Wortas. Xiaofei helped me with everything in my study and

life; I learned a lot from the discussion with Fang. Their help and friendship have been

vi

the source of my energy, courage and happiness.

I also would like to thank our former group members Aimee Brooks and Katie Strong
for their efforts in the project.

Finally I am also in great debt to my beloved parents and my wife. Their love and
support has always been with me. My parents’ unconditional love, support and
understanding is the most important encouragement to me when I encounter all kind of

challenges in the pursuit of my Ph.D degree.

vii

TABLE OF CONTENTS

 

LIST OF TABLES ................................................................................... xi
LIST OF FIGURES ................................................................................. xii
LIST OF ABBREVIATIONS .................................................................... xvii
CHAPTER 1: INTRODUCTION ..... - - - - 1
1.1 BRANCHING ENZYME ................................................................................................. 1
1.1.1 Glycogen ........................................................................................................ 1
1.1.2 Starch ............................................................................................................. 2
1.1.3 Glycogen / Starch Biosynthesis ..................................................................... 6
1.1.4 Branching Enzyme (BE) ................................................................................ 9
1.1.5 Escherichia coli branching enzyme ( E. coli BE) ......................................... 13

1.2 TRANSCRIPTIONAL COMPLEX TFIIIB/DNA ............................................................. 15

1 .2. 1 Transcription ................................................................................................ 1 5

1 .22 RNA polymerase III promoters .................................................................... 16
1.2.3 TFIIIB transcription factor ........................................................................... 18
1.2.4 Architecture of TFIIIB ................................................................................................. 20
1.2.4.1 TBP protein ............................................................................................... 21
1.2.4.2 Brfl protein .............................................................................................. 23
1.2.4.3 B” protein ................................................................................................. 25

1.2.5 Oligonucleotide sequence ............................................................................ 28
REFERENCE CITED .......................................................................................................... 30

CHAPTER 2: X-RAY CRYSTALLOGRAPHIC STUDY OF E. COLI BE AND

 

OLIGOSACCHARIDE COMPLEX - -- . ..... - - - - _ 41
2.1 E. COLI BRANCHING ENZYME(BE) AND POLYSACCHARIDE ........................................ 41
2.2 WHY WERE THESE OLIGOSACCHARIDES SELECTED? ................................................. 43
2.3 MATERIALS AND METHODS ....................................................................................... 47

2.3.1 Experimental design .......................................................................................... 47
2.3.2 Protein over-expression and purification ..................................................... 47
2.3.3 Protein crystallization .................................................................................. 50
2.3.4 Attempts at co-crystallization of BE with oligosaccharides ........................ 51
2.3.5 Soaking BE crystals with oligosaccharide solutions .................................... 51
2.3.6 X-ray diffraction data collection and processing ......................................... 53

viii

2.4 THREE DIMENSIONAL STRUCTURE OF E. COLI BE / CYCLODEXTRIN COMPLEXES ....... 54

2.4.1 Three dimensional structure of E. coli BE / a- cyclodextrin complex. .............. 55
2.4.2 Three dimensional structure of E. coli BE / B- cyclodextrin complex .......... 63
2.4.3 Three dimensional structure of E. coli BE/ 7- cyclodextrin complex ........... 68
2.4.5 Overlay of three cyclodextrin bound BE structures .......................................... 72
2.4.6 Overall binding of cyclodextrins with E.coli BE .............................................. 79
2.5 THREE DIMENSIONAL STRUCTURES OF E. COLI BE AND LINEAR OLIGOSACCHARIDES 81
2.5.1 Three dimensional structure of E. coli BE / maltoheptaose complex ................ 83
2.5.2 Three dimensional structure of E. coli BE / maltohexaose complex ................. 93
2.5.3 Overlay of E. coli BE/a-CD structure and E. coli BE/M7 structure .............. 96
2.5.4 Overall composite of E. coli BE / oligosaccharide binding structure ........... 99
2.6 CONCLUSION AND HYPOTHESIS .............................................................................. 100
2.6.1 Hypothesis .................................................................................................. 100
2.6.2 Verification of the hypothesis .................................................................... 106
REFERENCE CITED ........................................................................................................ 108

CHAPTER 3: OVEREXPERESSION, PURIFICATION AND LIMIT
PROTEOLYSIS OF BRANCHING ENZYME 11 FROM MAIZE ENDOSPERM 112

3.1 INTRODUCTION To MAIZE STARCH BRANCHING ENZYME II ..................................... 112
3.2 MATERIALS AND METHOD ....................................................................................... 114
3.2.] Protein over-expression and purification ........................................................ 114
3.2.2 Crystallization screen on maize starch branching enzyme 11 (SBEII) ............ 117
3.3 FURTHER EXPERIMENT PLAN .................................................................................. 121
REFERENCE CITED ........................................................................................................ 122

CHAPTER 4 X-RAY CRYSTALLOGRAPHIC STUDY OF RNA POLYMERASE

 

III TRANSCRIPTION FACOTOR TFIIIB COMPLEX - - 124
4.1 OUR RESEARCH OBJECTIVE .................................................................................. 124
4.2 MATERIALS AND METHODS .................................................................................... 125

4.2.1 Over-expression and purification of the S. cerevisiae TBP protein ........... 125
4.2.1.1 Over-expression of the TBP constructs .................................................. 126
4.2.1.2 Ni-NTA affinity chromatography purification of the TBP constructs 127
4.2.1.3 Histidine tag removal ............................................................................. 128
4.2.1.4 Chromatographic purification by heparin column ................................. 129

4.2.2 Over-expression and purification of the Brf protein .................................. 130
4.2.2.1 Over-expression of the Brf constructs .................................................... 133
4.2.2.2 Ni—NTA affinity chromatography purification of the Brf proteins ......... 133
4.2.2.3 Protein refolding ..................................................................................... 135
4.2.2.4 Removal of cleavable histidine tag from the S. cerevisiae Brf protein .. 135
4.2.2.5 FPLC ion exchange chromatography purification ................................. 136

ix

4.2.2.6 New S.cerevisiae Brf construct design and purification attempt ........... 138

4.2.3 Over-expression and purification of the B” protein ................................... 138
4.2.3.1 Over-expression of the B” (240-520) ..................................................... 139
4.2.3.2 Ni-NTA purification of the B” protein ................................................... 140
4.2.3.3 FPLC ion exchange chromatography purification ................................. 140
4.2.3.4 Cleavage of poly His tag from the B” (240-540) and B” (265-540) ...... 141

4.2.4 Purification of oligonucleotides ................................................................. 142

4.2.4 Making quaternary complexes and crystallization screening .................... 143

4.2.5 SEC chromatographic purification of quaternary complexes .................... 146

4.3 CONCLUSION AND FUTURE PLAN ON THIS PROJECT ................................................. 149
REFERENCE CITED ........................................................................................................ 150
Appendixes ......................................................................................... 152

Appendix 1 X-ray data collection statistics .............................................. 153

Appendix 2 Structure refinement statistics ............................................... 154

Appendix 3 Ramachandran plot of BE/alpha CD structure ........................... 155

Appendix 4 Ramachandran plot of BE/beta CD structure ............................ 156

Appendix 5 Ramachandran plot of BE/gamma CD structure ........................ 157

Appendix 6 Ramachandran plot of BE/M7 structure .................................. 158

Appendix 7 Ramachandran plot of BE/M6 structure .................................. 159

Appendix 8 Protein-ligand contact between BE and alpha CD ...................... 160

Appendix 9 Protein-ligand contact between BE and beta CD ........................ 164

Appendix 10Protein-ligand contact between BE and gamma CD .................... 167

Appendix 11 Protein-ligand contact between BE and M7 .............................. 170

Appendix 12Protein-ligand contact between BE and M6 .............................. 175

LIST OF TABLES

Table 2. 1 Dissociation constants for different -glucan and Potato tuber starch branching
enzyme I (SBEI) ............................................................................................... 46

Table 2. 2 Summary of substrates used in soaking experiment and soaking conditions .. 52

Table 2. 3 Detailed binding Sites in E. coli BE /cyclodextrin complex Structures ............. 54
Table 2. 4 Binding sites in E. coli BE /linear oligosacchardie complex structures ........... 82
Table 4. 1 Different Brf protein constructs used ............................................................. 131
Table 4. 2 DNA used in our project ................................................................................ 142

Appendix 1 X-ray data collection statistics of the ligand-bound E. coli BE protein .. 153

Appendix 2 Structure refinement statistics of the ligand-bound E. coli BE protein... 154

Appendix 8 Protein-ligand close contacts between BE and alpha CD. ..................... 160
Appendix 9 Protein-ligand close contacts between BE and beta CD. ....................... 164
Appendix 10 Protein-ligand close contacts between BE and gamma CD. .................. 167

Appendix 11 Protein-ligand close contacts between BE and maltoheptaose (M7). 170

Appendix 12 Protein-ligand close contacts between BE and maltohexaose (M6). ..... 175

Xi

LIST OF FIGURES

Images in this thesis/dissertation are presented in color

Figure 1. l Granules of wheat starcht .............................................................................. 3
Figure 1. 2 Schematic diagram of starch granule structure .............................................. 4
Figure 1. 3 Starch is made of two distinct polysaccharide components: ......................... 5
Figure 1. 4 The scale drawing of the structure of the amylose. ....................................... 5
Figure 1. 5 Overview of the industrial processing of starch ............................................ 7
Figure 1. 6 Biosynthetic pathway of glycogen/ starch ..................................................... 8
Figure 1. 7 The reactions catalyzed by the members of a-amylase family. ................... 10
Figure 1.8 Conserved catalytic residues in E. coli branching enzyme ........................... 12

Figure 1. 9 Ribbon depiction of the x-ray crystal structure of E. coli BE truncated at

amino acid 113 ........................................................................................................... 14
Figure 1. 10 Different types of RNA pol III promoters ................................................. 17
Figure 1. 11 A schematic model of TFIIIC-dependent TFIIIB binding. ........................ 19
Figure 1. 12 A schematic representation of TFIIIB transcription complexes ................ 21
Figure 1. 13 Amino acid sequence of S. cerevisiae TBP protein. .................................. 22
Figure 1. 14 A schematic graph of S. Cerevisiae Brfl ................................................... 23
Figure 1. 15 The Similarity in the human and S. cerevisiae B” sequence ...................... 26
Figure l. 16 The amino acid sequence of S. cerevisiae B". ............................................ 27

xii

Figure 2.

l

E. coli BE (PDB databank # 1M7X) overall structure ............................... 42

Figure 2. 2 Structures of (It-cyclodextrin, B-cyclodextrin and y-cyclodextrin and crystal
of amylose ................................................................................................. 45
Figure 2. 3 SEC Chromatograph of truncated E. coli BE protein ................................. 49
Figure 2. 4 SDS-PAGE gel of pure truncated E. coli BE protein ................................. 49
Figure 2. 5 Crystals of E. coli branching enzyme ......................................................... 50
Figure 2. 6 2Fo-Fc electron density map of (II-cyclodextrin binding with E. coli BE 55
Figure 2. 7 (II-cyclodextrin binding with binding Site I ................................................ 57
Figure 2. 8 (it-cyclodextrin binds with binding Site II ................................................... 58
Figure 2. 9 Binding between (II-cyclodextrin and E. coli BE binding site 11, another
angle ........................................................................................................................... 59
Figure 2. 10 (II-cyclodextrin binds with binding site 111 ................................................. 60
Figure 2. 11 2F o-Fc electron density map of loop 213-215 on D chain of E. coli BE
/a-cyclodextrin ........................................................................................................... 6 1
Figure 2. 12 E. coli BE binds with a-CD at binding Site IV ........................................... 62
Figure 2. 13 B- CD binds with E. coli BE at binding site II ............... I ............................. 64
Figure 2. 14 The interaction between E. coli BE binding Site IV and B-CD .................. 65
Figure 2. 15 The detailed interaction at binding Site I of molecule A with B-CD .......... 66
Figure 2. 16 The binding between B-CD and E. coli BE binding Site I from top angle. 67
Figure 2. 17 7- CD binds with E. coli BE at binding Site II. ........................................... 69
Figure 2. 18 7- CD binds with E. coli BE at binding Site 111 ........................................... 70
Figure 2. 19 7- CD binds with E. coli BE at binding site Iv .......................................... 71

xiii

Figure 2.

2O

Overlay of (II—CD, B-CD, y-CD and their bound residues at binding site II

.................................................................................................................................... 72

Figure 2.

21

Overlay of (l-CD, B-CD, y-CD and their bound residues at binding site II

.................................................................................................................................... 73

Figure 2.

Figure 2.

Figure 2.

Figure 2.

22

23

24

25

Overlay of a—CD, y-CD and their bound residues at binding Site III ........ 74

Overlay of a—CD, B-CD, y-CD and their bound residues at binding site IV

.................................................................................................................................... 75
Overlay of a—CD, B-CD and their bound residues at binding site I .......... 76
Overlay of (II—CD, B-CD and their bound residues at binding site I .......... 77
A composite of the crystal structures of E. coli BE bound to (l-CD ............ 80

Figure 2.

Figure 2.

VI.

Figure 2.
Figure 2.
Figure 2.
Figure 2.
Figure 2.
Figure 2.
Figure 2.
Figure 2.
Figure 2.
Figure 2.

Figure 2.

26

27

28

29

3O

31

32

33

34

35

36

37

38

the Fo-Fc and 2Fo-Fc electron density map of the M7 at binding site V and

...................................................................................................................... 84
Maltoheptaose binds with E. coli BE at binding sites V and VI ................ 85
Maltoheptaose binds with E. coli BE at binding Site V and VI .................. 86

Maltoheptaose binds with E. coli BE at binding Site V in molecule A ...... 88

Maltoheptaose binds with E. coli BE at binding Site VII ........................... 89
Maltoheptaose binds with E. coli BE at binding Site III. ............................ 91
Maltoheptaose binds with E. coli BE at binding site IV. ............................ 91
A composite view of the B molecule Of E. coli BE /M7 ............................ 92
Maltohexaose binds with E. coli BE at binding Site V ............................... 94
Maltohexaose binds with E. coli BE at binding Site III. ............................. 95
Maltohexaose binds with E. coli BE at binding Site IV. ............................. 95
The overlay of BE/a-CD and BE/M7 structures at binding site III ........... 96

xiv

Figure 2. 39 The overlay Of BE/a-CD and BE/M7 structures at binding site IV .......... 98
Figure 2. 40 The overall composite binding sites on the E. coli BE surface .................. 99
Figure 2. 41 Overlay of 1WPC, 1HXO, lM7X and 1UA3 ........................................ 102
Figure 2. 42 Oligosaccharide model (blue color) binds to binding Site V and the active
Site simultaneously ................................................................................................... 104
Figure 3. 1 SEC chromatograph and SDS-PAGE of SBEII protein ............................ 116
Figure 3. 2 Two final SBEII protein samples for crystallization screening. ................ 116
Figure 3. 3 Trypsin limit proteolysis test on SBEII at 4 °C. ........................................ 118
Figure 3. 4 SEC Chromatograph and SDS-PAGE of proteolysis product of SBEII ..... 119
Figure 3. 5 SDS-PAGE of a proteolysis experiment .................................................... 120
Figure 4. 1 Crystal structure of the TBP-TATA ........................................................... 126
Figure 4. 2 A SDS-PAGE sample of the TBP protein after Ni-NTA column purification
.................................................................................................................................. 128
Figure 4. 3 Histidine tag removal in the TBP purification ........................................... 129
Figure 4. 4 SDS-PAGE samples of TBP protein .......................................................... 130
Figure 4. 5 A schematic representation Of the S. Cerevisiae and K. lactis Brf .......... 132
Figure 4. 6 A SDS-PAGE of the K. lactis Brf (302-501) after Ni-NTA purification. 134
Figure 4. 7 The cleavage of His-tag on Brf(420-531) ................................................. 136
Figure 4. 8 A SDS-PAGE sample of purified K. lactis Brf proteins ............................ 137
Figure 4. 9 A Gel mobility Shift assay using U6 promoter probe, and TBPc and
indicated amount of Brf and B” ............................................................................... 137

XV

Figure 4. 10 A SDS-PAGE sample of B”(240-540) His-tag cleavage ......................... 141
Figure 4. 11 SDS-PAGE of two quaternary complexes ............................................... 144
Figure 4. 12 UV spectrum of TFIIIB protein, DNA and pure quaternary complex ..... 144

Figure 4. 13 Photos of putative quaternary complex crystals ...................................... 145

Figure 4. 14 SEC chromatograph of quaternary complex TBP(61-240)/ Brf(435-551)
B”(240-520)/DNA1 . ................................................................................................ 147

Figure 4. 15 The SEC chromatograph and SDS-PAGE of a quaternary complex

TBP(61-240) /Brf (407-531) /B”(265-540)/DNA1 (see Table 4.2) ......................... 148
Appendix 3 Ramachandran plot of the BE/alphaCD structure .................................... 155
Appendix 4 Ramachandran plot of the BE/betaCD structure ...................................... 156
Appendix 5 Ramachandran plot of the BE/gamma CD structure ................................ 157
Appendix 6 Ramachandran plot of the BE/M7 structure ............................................. 158
Appendix 7 Ramachandran plot of the BE/M6 structure ............................................. 159

xvi

Amino Acids

Ala, A
Arg, R
Asn, N
Asp, D
Cys, C
Gln, Q
Glu, E
Gly, G
His, H
Ile, 1
Leu, L
Lys, K
Met, M
Phe, F
Pro, P
Ser, S
Thr, T

Trp, W

LIST OF ABBREVIATIONS

Alanine
Arginine
Asparagine
Aspartic acid
Cysteine
Glutamine
Glutamic acid
Glycine
Histidine
Isoleucine
Leucine
Lysine
Methionine
Phenylanaline
Proline

Serine
Threonine

Tryptophan

xvii

Tyr, Y

Val, V

Tyrosine

Valine

Other Symbols and Abbreviations

BE

WT

CD

E. coli

LB

UV

Vis

APS

DNA

F PLC

branching enzyme

wild type

cyclodextrin

Escherichia coli

Luria broth

Ultraviolet light

visible light

Advanced Photon Source

Base pair

Dalton

Ribonucleic acid
Deoxyribonucleic acid

Fast protein liquid chromatography
Hour

Dissociation constant

Wavelength of the maximum absorption

AngstrOm

xviii

mm

MW

nM

DEAE

IPTG

Tris

DTT

PEG

PCR

PDB

CCP4

R-factor

Rmsd

SDS-PAGE

SEC

Liter

Molar

microliter

milliliter

Nanometer

millimeter

Molecular weight

Nanomolar

Diethylaminoethyl cellulose

Isopropyl- 1 -thio-B-D-galactopyranoside
2-Amino-2-(hydroxymethyl)— 1 ,3 -propanediol
Dithiothreitol

Polyethylene glycol

Polymerase chain reaction

Protein Data Bank

Collaborative computational project, number 4
Reliability factor

root mean square deviation

Sodium dodecyl sulfate — polyacrylamide gel electrophoresis

Size exclusion chromatography

xix

CHAPTER 1: INTRODUCTION
1.1 Branching Enzyme

1.1.1 Glycogen

Glycogen is the major carbohydrate storage molecule in bacteria and animal cells. It
is widely found in the muscles and livers of all higher animals. In fact, it is called animal
starch because it is the primary reserve polysaccharide for animals and very close to
starch(1 -3).

Glycogen is usually found in the form of granules in the cytosol in cells. Glycogen
forms an energy reserve that can be quickly broken down to meet a sudden need for
glucose when the blood glucose level is low. Glycogen from liver can be degraded to
glucose under the action of glycogen phosphorylase SO that it can meet the need of the
rest of the body. Abnormal glycogen storage can cause various diseases(4).

Glycogen is a highly branched non-reducing water-soluble polysaccharide. In 1936,
the structures of glycogen from different organism were determined (5-6). Depending on
the source of glycogen, it consists of more than sixty thousand glucose units with
molecular weights up to 107 Daltons with 8-12% branches. Chemistry studies Showed
glycogen is a polysaccharide composed of a-1,4-linked glucans and branched by
(II-1,6-glycosidic linkages (1, 6- 7). In glycogen, the branches usually occur at intervals of

8-10 glucose units. The branched polysaccharide can be accumulated in cells in a more

efficient way, which makes the glycogen granules denser.

1.1.2 Starch

Different from glycogen, starch is another form of energy and carbohydrate storage
found only in photosynthetic eukaryotes or their non-photosynthetic derivatives. It is the
second most abundant polysaccharide in nature after cellulose. Starch is also the primary
nutrition source in the human diet and is becoming increasingly important as a renewable
industrial biomaterial (Figure 1.1). It is widely found in higher plants such as wheat,
maize, potato and tapioca (8-9).

Starch is synthesized in plants as a result Of photosynthesis, a process during which
energy from sunlight is converted into chemical energy and stored in starch (10). It is
synthesized in both plastids of leaves and amyloplasts in tubers, seeds and roots of plants.
Large amounts of starch accumulate in the latter case for long-term energy storage.

Unlike glycogen, which is basically homologous and water—soluble, starch is
essentially an insoluble semi-crystalline material. Starch is a huge ( from 0.1 to over 50
u m in diameter) complex made of two distinct polysaccharide fractions : amylopectin
and amylose ( Figure 1.3). Native starch granules typically contain around 20-25%
amylose and 75-80% amylopectin (10). Amylopectin , the major compound, is composed
of moderate size of (1-1,4 linked glucans that are clustered together and hooked to longer
spacer glucans by (II-1,6 linkages. Nevertheless amylose consists almost exclusively of

linear a-1,4-linked polymers ( with less than 0.6% branches). Compared with glycogen,

starch has significantly less branches on average. The (II-1,6 linkages occur only every
24-30 glucose units.

The semi-crystalline structure of the starch granule is represented by Figure 1. 2. The
Short amylopectin chains form double helices and associate into clusters. Then the
clusters pack together to form a structure of alternating crystalline and amorphous
lamellar composition (11 -12).

The linear arnylose is found to exist in a left-handed helical structure in its crystalline
state (13-15). In each turn there are 6 glucose units, which make it very similar to

a-cyclodextn'n. Figure 1.4 is the schematic depiction of the helical structure(2).

 

Figure l. 1 Granules of wheat starch, Stained with iodine, photographed through a light
microscope(1 6)

b)

 

Semicrystalline Amorphous
A chain
growth ring A backgrouq .
morphous _
Crystamne Iamellae
3’ 9W“ ””9 _

I

Arno . ous ame ae

 

 

 

 

 

Figure 1. 2 Schematic diagram of starch granule structure. (a) A single granule, comprising
concentric rings of alternating amorphous and semicrystalline composition. The semicrystalline
growth rings contain stacks of amorphous and crystalline Iamellae. (b) Predicting the scattering
for a model structure allows us to fit the SAXS data. The model consists of a paracrystalline
lamellar stack, embedded in an amorphous background medium. (c) The currently accepted
racemose structure for amylopectin within the semicrystalline grth ring. A chain sections of
amylopectin form double helices, which are regularly packed into crystalline Iamellae. B chains
of amylopectin provide intercluster connections. Branching points for both A and B chains of
amylopectin are predominantly located within the amorphous lamellae(12).

 

3)
O
OH OH OH OH OH
O O 0 O o
OH
OH I\ OH OH n OH H ,

reducing

01- 1 ,4 Amylase end

 

 

 

b)

“20H CH20H CH20H CHZOH Amylopectin
O
OH OH OH OH
OL-l 6
o 0 o ’
H OH O/
n OH H\
CH20H CH2 CH20H CHZOH H20H
0 O
OH 0 0H 0 OH OH H
o O
O OH
OH OH OH 0H 11 H

Figure l. 3 Starch is made of two distinct polysaccharide components:

 

 

 

 

amylose a) and b) amylopectin (17)

 

Figure l. 4 The scale drawing of the structure of the amylose. The glucose units assume a
tightly coiled helical structure. Each turn of the helix has 6 glucose units(2).

1.1.3 Glycogen I Starch Biosynthesis

Glycogen and starch as major forms of energy storage have also been drawing more
and more interest in industrial usage. Starch is one of the most important products
synthesized by plants that are used in industrial processes. It is widely used in food,
papermaking and the pharmaceutical industries to produce starch hydrolysate, glucose
syrups, fructose and cyclodextrins. It is also being researched to replace non-renewable
petroleum based energy sources and materials (18). But the use of starch is still limited,
because its properties are not ideal for a variety of applications. So altering the properties
of starch is an important research subject. Although some of the properties can be
changed by chemical modification, using chemical processes is not desired or successful
under certain circumstances. AS an alternative, bioengineering methods provide a proven
and fertile avenue for the achievement of these goals (19-21). For example, the
production of glucose from starch has been converted from an acid hydrolysis method to
enzymatic treatment with several different enzymes ( Figure 1.5) (22). Furthermore,
the modification of the starch biosynthetic enzymes has enabled the genetic modification
of crops in a rational manner to produce novel starches with improved functionality(23).

By altering the starch biosynthetic enzymes, it is possible to change the amylose /
amylopectin ratio, improve the freeze-thaw property and alter the phosphate content of
starch. It can also improve the yield and make the production of starch cheaper for use by

industry, or find new applications for the modified starches (23-24).

 

'_ STARCH
C
o l \
B Thermostable
g aflmylase Thermostable
:1 ¢ C GTase
2' \
" DEXTRIN

: branched + linear oligo’s - CYCLODEXTRIN S
C

Fungal Amyloglucosidase Pullulanase
a-amylase Pullulanase fl-amylase
/ I
2 MALTOSE GLUCOSE MALTOSE
MALTOTRIOSE MALTOSE SYRUP
3 HIGH GLUCOSE SYRUP HIGH
_ / \
Glucose C 11' ti
isomerase rysta '23 on
FRUCTOSE CRYSTALLINE
SYRUP SUGAR

Figure l. 5 Overview of the industrial processing of starch into cyclodextrins,
maltodextrins, glucose or fructose syrups and crystalline sugars (22)

Understanding how the biosynthesis of glycogen / starch works in vivo was the major
challenge. It was found that three enzymes are involved in the three steps of the
biosynthetic pathway of bacterial glycogen and starch (Figure 1.6). The first step is the
formation of ADP-glucose catalyzed by ADP-Glucose Pyrophosphorylase. This is the rate
determining step. The formation of the (II-1,4 glucosidic linkages occurs first by synthesis
Of ADP-glucose from ATP and alpha glucose l-P and then transfer of the glucose moiety
from the sugar nucleotide to a pre-existing glucan primer(25). The ADP-glucose
molecule iS the glucose donor in the bacterial and plant cells. Its formation is catalyzed
by ADP-Glucose pyrophosphorylase (ADP-Glc PPase), which is an allosteric enzyme

(26-28). The ADP-Glc PPase from E. coli and plants (especially potato tuber) are well

characterized (26, 29-42). The activation and the inhibition of the enzyme were well
studied. The crystal structure of the enzyme was obtained (35).

The elongation of the polysaccharide chain is the second step. It is catalyzed by
glycogen or starch synthase. It forms (II-1,4 glucosidic bonds between the primer glucose

unit and the sugar nucleotide. Branching enzyme catalyzes the last step to make the (II-1,6

linked branches.
PP-
fwni4 H0 HQOH
0P03 HO OH ADP
H-ADP -glucose pyrophosphorylase
(ADP-Py) ADP-Glucose
0
HO HO
H tarc synt ase HO
O CHZOH I 0H0
CHZOH
+ HO OH
ADP GI OH HO
- ucose OH OH

HO CHZOH
Branching Enzyme R0 CHZOHO
CHZOH H O O
OH O

HO
OH
Figure 1. 6 Biosynthetic pathway of glycogen/ starch. There are three enzymes
involved in the three consecutive steps: ADP-Glucose pyrophosphorylase, Glycogen or
Starch Synthase and Branching Enzyme

Research on how the enzymes work in the pathway will shed light on the
fundamental mechanism of glycogen / starch biosynthesis. Furthermore, it will give us

more information towards potential starch bioengineering modification strategies.

1.1.4 Branching Enzyme (BE)

Branching Enzyme (1,4-a-glucan 6-glucosyltransferase; EC2.4.1.18) (BE) is found in
all organisms that make either starch or glycogen. In the biosynthetic pathway of
glycogen/starch, branching enzyme plays an important role in the determination of the
final structure of either glycogen or starch. The rearrangement of the linear (1-1,4 linked
polysaccharide chain is achieved by the cleavage of the a-l,4 glucosidic bond and the
subsequent transferring of the oligosaccharide to the a-l,6 position. Forming (II-1,6
branch points gives more non-reducing ends to the polysaccharide, thus making it more
reactive to synthesis and digestion. Branches are also critical for maintaining its solubility
in the cell: exclusively linear glycogen / starch is prone to precipitate in cell while the
branched product can maintain appropriate solubility.

BE belongs to the a-amylase family which consists of the amylases, isoamylase,
cyclodextrin glucanotransferase (CGT), the pullulanases and branching enzyme. They all
act on III-glycosidic bonds and hydrolyze this bond to produce a-anomeric mono- or
oligosaccharides (hydrolysis), form a,1-4 or 1-6 glycosidic linkages ( transglycosylation) ,
or a combination of both activities(10) (Figure 1.7)(17). All these enzymes share a

common TIM-barrel fold consisting of an 8 B-sheet barrel surrounded 8 a-helices (10).

a) Hon ['1on H20” Hon Hon 1"on
- l
H H M, H H + OH
O
H OH HO H H H
H H
OH , OH H

b) CH20H CHZOH CHZOH CHZOH Hon
OH OH OH OH OH
H
isoamylase H H
H

HCHZOH HH:/§\O E:ZO:HH

 

CGT

 

} Cyclic sugar chain

   

”2 CH20H
OH OH H
O OH
H H H

Figure l. 7 The reactions catalyzed by the members of the (II-amylase family. a)
(II-amylase hydrolyses (II-1,4 bonds. b) isoamylase cleaves a-1,6 bonds. 0) COT
catalyzes the formation of cyclodextrins and (1) BE catalyzes the formation of a-1,6
branches (17).

10

In addition to the central fold containing the catalytic domain, BE also has three
other domains not shared with most of the other enzymes of this family, two n-terminal
domains and a c-terminal domain (10, 43).

BE exists as a single polypeptide in most glycogen-producing organisms (bacteria,
fimgi, higher mammals, etc.) and at least two forms of the enzyme are found in most
plants, starch branching enzyme 1 (SBEI) and starch branching enzyme II (SBEII)(44-4 7).
These enzymes differ in their substrate and branch. chain length specificity. SBEII
preferentially transfers shorter chains predominantly of lengths of 7 and 11, though there
are significant populations of chains between these values (48). SBEl on the other hand,
transfers much longer chains. For example, Maize branching enzyme I (MBEI)
predominantly transfers polysaccharide chain of 10-13 glucose units (49-50). Studies of
chimeric forms of SBEl and SBEII from maize indicated that the c-terminal domain is
involved in substrate preference and catalytic capacity, while the n-terminal domain is
involved in determining the size of the chain transferred (43, 51).

From the amino acid sequence aligmnent, Escherichia coli (E. coli) branching
enzyme and starch branching enzyme share some critical residues that are involved in the

catalytic function ( Figure 1.8)(43)

11

 

 

 

 

 

E.coli 297 s w G ”Y— O P T 330 L N v I
human 248 s F G Y O I T 231 I I v L
mBEI 303 s F G Y H v T 336 L R v L
mBEll 277 s F G Y H v T 310 L L v L
isoa 253 Y w G Y M T E 237 I K v Y
Map 79 Y H G Y w N o 112 M Y L M
a-Por 59 w E R Y O P v 91 v R I Y
CGT 97 Y H G i. w A R 130 I K v I
E. ml 398 G I D A L R v o A v A 454
human 350 R F o o F R F o G v T 403
mBEl 403 M F o G F R F D G v T 466
mBEll 379 K F o G F R F o G v T 437
Isoa 368 G v o G F R F o L A s 431
a-Asp 199 s I o G L R I o T v K 226
a~Por 190 G v A G F R L o A s K 229
CGT 222 G I o G I R M L9; A v K 253
E.coli 520VLPLSHDEV
human 475 AYAESHOOA
mBEl 475 AYAESHDQA
mBEll 503TYAESHDQA
isoa 504 N F I D v H o G M
a-Asp 291TFVENHDNP
a—Por 294 VFVDNHDNQ
CGT 322 TFIONHDME

 

 

 

Figure 1.8 Conserved catalytic residues in E.coli branching enzyme, human branching
enzyme and maize branching enzyme I and II. Some other enzymes from the a-amylase
family are also aligned: isoa is isoamylase from Pseudomonas amyloderamosa, a-Asp
and a-Por are (II-amylase from Aspergillus Oryzae and Porcine Pancreatic. CGT is
cyclodextrin glucanotransferase from Bascillus Circulans(43). The residues involved in
catalysis are highlighted in boxes.

12

1.1.5 Escherichia coli branching enzyme ( E.coli BE)

The fill] length wild type E.coli BE was over-expressed and purified. It has 728
amino acid residues with a molecular weight of 84KDa. In order to gain insight into the
structure, mechanism and specificity of E. coli BE, crystallization was attempted on the
wild type full length enzyme without success. Further study indicated a 7OKDa limited
proteolysis product still retained 40-60% of the branching enzyme activity(52). The
sequence of the truncated enzyme was determined. It lacks the first 113 amino acid
residues due to the limited proteolysis. This construct of enzyme (named as N113BE) was
purified and crystallized successfully. The three dimensional structure of the protein was
reported (43, 53).

The determination of the structure of E. coli BE was a milestone in the study of
glycogen / starch biosynthetic pathway enzymes. It was the first structure of a BB from
any organism and also was the first structure of any of the enzymes that make up the
starch biosynthetic pathway.(43) Figure 1.9 shows the ribbon depiction of the truncated E.
coli BE structure with the second n-terminal domain in red and the c-terminal domain in
blue. Studies of the E. coli BE have shown that the enzyme predominantly transfers
shorter chains of between 7-15 glucose units in length. However, removal of the
n-terminal 113 amino acid residues had only a moderate effect on catalytic activity but it
significantly altered the transfer-chain preference to longer chains containing 15-25

glucose units. Substantial amounts of chains as long as 30-40 residues were transferred

13

with this mutant, indicating that the far n-terminus of E. coli BE has a significant effect
on branch chain specificity(54-55). The overall structure of the enzyme was discussed in
detail; and an acarbose molecule was modeled into the active site of the enzyme to
elucidate the possible interaction between the enzyme and its substrate(43). In our work,
we obtained the substrate bound E.coli BE structures and proposed a hypothesis on how
the enzyme binds with its substrate prior to the catalytic action. The result and discussion

are in the following chapters.

 

Figure 1. 9 Ribbon depiction of the x-ray crystal structure of E. coli BE truncated at
amino acid 113. Residues involved in BE catalysis are shown in green, with atoms
colored by type: red, oxygen; green, carbon; blue, nitrogen. Red indicates the
NHz-terminal domain; orange indicates the central lit/l3 barrel catalytic domain; and blue
indicates the COOH—terminal domain.

1.2 Transcriptional complex TFIIIB/DNA

1.2.1 Transcription

Transcription is the process by which the genetic message in DNA is transcribed into
RNA. DNA is the genetic material of all living organisms except certain viruses(5 6). The
genetic information coded in the double stranded DNA is transferred to RNA in an
accurate multi-step process. This process is accomplished by synthesis of RNA by RNA
polymerases with the help of some cofactors. In eukaryotic cells there are three different
transcription systems. Each consists of one kind of RNA polymerase and a unique set of
transcription factors. RNA polymerase I (pol I ) transcribes large ribosomal RNA ( rRNA)
genes (353-455 depending on species); RNA pol II, located in the nucleoplasm,
synthesizes messenger RNA (mRNA) genes and most of the small nuclear RNA (snRNA)
(e.g. U1,U2) genes while RNA pol III is responsible for transcribing all transfer RNA
(tRNA), 53 rRNA and some other snRNA genes(56-60). The accurate initiation of any
transcription requires RNA polymerase in association with the transcription factors to be
recruited to the specific start sequence of DNA that is called the promoter. In the three
stages of transcription (initiation, elongation and termination), the study of initiation
could reveal how transcription starts and give answers to how the polymerases and
transcription factors interact on the specific promoters.

The genes transcribed by RNA polymerase III (pol III ) encode a variety of small

RNA molecules, many of which have essential functions in cellular metabolism. For

15

example, tRNA and SsRNA are required in the synthesis of protein; 7sLRNA is involved
in intracellular protein transportation while U6 and H1 RNAs are involved in the
post-transcriptional process (5 7, 5 9-60). RNA polymerase III transcribes some genes with
unknown functions as well. The genes transcribed by RNA polymerase III are mainly
shorter than 400 base pairs. This limit is consistent with the elongation properties of RNA
polymerase III, which recognizes a simple run of T residues as a termination signal(5 7).
So far TFIIIB mediated RNA pol III transcription is the only eukaryotic transcription
system that can be reconstituted using only recombinant factors. And this is only
achieved in S. Cerevisiae where TFIIIB binds with the TATA box directly (5 7, 61). So the

TFIIIB mediated pol III transcription becomes a favorite topic.

1.2.2 RNA polymerase III promoters
There are three types of promoters for RNA polymerase III called type 1-3 (Figure
1.8)(57). The SS RNA gene from Xenopus laevis is the only type 1 promoter. It is the
primary binding site of TFIIIA. The structure characteristic of the promoter is the unique
box A sequence, followed by the intermediate element (IE) and a unique conserved
sequence box C. All of these elements constitute the internal control region (ICR)(62-63).
The Adenovirus 2 (Ad2) VAI gene and various tRNA genes from Xenopus laevis and

Drosophila melanogaster are typical type 2 promoters. Type 2 promoters have two highly

16

conserved internal promoter elements, box A and B(64).

The type 3 promoters were found in mammalian U6 snRNA genes, which encode the

snRNA component of the spliceosome(65). Also it is found in the human 7SK gene

which is involved in regulation of the CDK9/cyclin complex(66).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

+1
~—> ICR +120
F—W
type 1:58 [j [1 [j TTTT
A IE C
type 2: tRNA
—> +106
type 3; Hs us E% Tfi'T
DSE PSE TATA
’ +113
Sc U6 {3 fl TT‘l'T fl—
TATA A 8

Figure 1. 10 Different types of RNA pol III promoters: Human U6 (Hs U6) and
S. Cerevisiae U6 (Sc U6) are two examples of type 3 promoters. +1 indicates the starting
point of transcription; TATA is the TATA box region; DSE and PSE stand for distal
sequence element and proximal sequence element which are unique sequence

regions(5 7).

17

1.2.3 TFIIIB transcription factor

In order to initiate transcription accurately and effectively, the RNA polymerases
need the assistance of transcription factors to recruit them to the appropriate start sites of
genes. The transcription factors recognize the specific promoter and bind with DNA, then
load the corresponding polymerase onto the DNA so that the transcription can initiate.
This process involves the interaction between transcription factor and DNA promoter, the
interaction between transcription factor and polymerase and the interaction between DNA
and polymerase. It is obviously a very complicated process. In the study of this process,
different transcription factors were discovered. F or example, the yeast RNA polymerase
III transcription machinery consists of three transcription initiation factors: TFIIIA, which
is a gene-specific factor, TFIIIB and TFIIIC which are general factors for RNA pol III
(6 7-68).

Among all the transcription factors, TFIIIB is the central initiation factor since it
alone can recruit pol III to the transcription starting site and initiate the transcription with
polymerase III (5 7, 67, 69). Besides this function, TFIIIB was also found to participate in
promoter opening steps (70-71). All of these make TFIIIB the research focus recently.

TFIIIC and TFIIIA, the other two components of the core transcription apparatus,
were identified by the separation of cell extract through phosphocellulose
chromatographic separation with TFIIIB together. They bind with DNA and sometimes
serve as assembly factors for TFIIIB in the recognition of specific genes (72- 73).

TFIIIC’S function of loading TFIIIB is already well known. But little is known about how

18

TFIIIA helps to load TFIIIC first(67).
An example of interactions between TFIIIB, TFIIIC and DNA in TFIIIC-dependent

transcription is represented by Figure. 1.9

 

Pol Ill
TFIIIC
_ TFIIIB , .
. L'TA B
tRNA gene

Figure 1. 11 A schematic model of TFIIIC-dependent TFIIIB binding to tRNA
promoter( 74).

It has been demonstrated that TATA-Binding Protein (TBP) can bind DNA segment
with a TATAAA sequence (TATA box) by itself in vitro (71, 75-77). And the recruitment
of pol 111 could happen without the help of TFIIIC and TFIIIA. While TFIIIC and TFIIIA
are unable to recruit pol III to a promoter without TFIIIB in any context, which also
proves that TFIIIB is the core transcription factor for pol III(68, 78-80).

However in most budding yeast pol III transcribed genes do not have strong TATA
boxes near the promoter region. In this case TFIIIC acts as the assembly factor for TFIIIB
to load TFIIIB to the appropriate binding site (75, 81). This is the so-called

TFIIIC-dependent binding. Whether assembled at the promoter by TFIIIC or

independently, TFIIIB alone suffices to recruit pol III for multiple rounds of transcription
and plays an essential role in formation of the open pol III initiation complex. The
structure of the pre-initiation complex of TFIIIB and DNA can reveal how the
transcription factor interacts with the promoter at the very first moment of transcription.
It is the key to understand the initiation of transcription. In our project we try to get the
detailed structure of the TFIIIB/DNA complex at atomic resolution by reconstituting the

pre-initiation complex in vitro.

1.2.4 Architecture of TFIIIB:

TFIIIB transcription factor is a multi-subunit protein. It consists Of three subunit:
TATA-binding protein (TBP), TFIIB-related factor (Brfl, also called TFIIIB70) and B”
(also called de1 or TFIIIB90). Each subunit has been identified and sequenced.

In order to understand how TFIIIB initiates transcription on promoter, the
relationship and individual functions of the TBP, Brf and B” have to be investigated. As
the pol III transcription core apparatus, TFIIIB undergoes both TFIIIC-dependent and
TFIIIC-independent binding to DNA. Figure 1.10 illustrates the binding process in both

cases(5 7).

20

 

type 2

 

   

TATA
box promoter

 

 

 

 

Figure 1. 12 A schematic representation of TFIIIB transcription complexes. There are
two types of complex formation: A) TFIIIC-dependent pathway on a type 2 promoter; B)
TFIIIC-independent pathway on an artificial TATA box containing promoter. Arrows
stand for interaction among different components (57).

Fig 1.10 shows the mechanism of RNA pol III initiation in vitro. Figure A shows that
TFIIIC recruits TFIIIB onto the TATA-less DNA sequence, followed by pol III binding
and initiation. And Figure B depicts that TFIIIB is recruited by the binding of TBP

subunit to TATA box, followed by pol III binding and initiation.

1.2.4.1 TBP protein

It is well known that the TATA-binding protein (TBP) is an essential component of
the transcription machinery of all three RNA polymerase, pol I , pol II and pol III
systems(76-77, 81-82). Thus it is required in the expression of all nuclear genes (75,

83-84). TBP protein binds the TATA box in a TFIIIC-independent binding situation, (85).

21

The C-terminal domain of TBP is highly conserved. This region shows transcriptional
activity in vitro. And the site —directed mutagenesis or internal deletion in this region
results in a dramatic decrease in biological activity (86-87). The N-terminus of TBP is
irrelevant to the transcriptional activity in pol III transcription. In our project we are
going to investigate the specific region that is absolutely required in forming stable
pre-initiation complex and directing subsequent transcriptions(67). The interactions
between TBP and the other two subunits, Brfl and B”, are to be studied as well.

The full length TBP protein from S. cerevisiae has 240 amino acid residues.
Truncated construct (61-240) of this protein was used in the crystallization with TATA
box promoter(88). We also used this protein for our studies. Figure 1.11 shows the

protein sequence of TBP from S. cerevisiae.

l MADEERLKEFKEANKIVFDPNTRQVWENQNRDGTKPATTFQSEEDIKRAA
51 PESEKDTSATSGIVPTLQNIVATVTLGCRLDLKTVALHARNAEYNPKRFA
101 AVIMRIREPKTTALIFASGKMVVTGAKSEDDSKLASRKYARIIQKIGFAA
151 KFTDFKIQNIVGSCDVKFPI RLEGLAFSHGTF SSYEPELFPGLIYRMVKP

201 KIVLLIFVSG KIVLTGAKQR EEIYQAFEAI YPVLSEFRKM

Figure l. 13 Amino acid sequence of S. cerevisiae TBP protein.

22

1.2.4.2 Brf1 protein
Brfl is the TFIIB-related factor of the TFIIIB complex. It is a multi-domain protein
with 596 amino acid residues. The molecular weight of Brfl is about 70 KDa. The

architecture of Brf 1 is represented by Fig 1.12.

 

TFIIB-like domain \ BRF homology regions
Zn finger direct repeats I/ ll III
I l I l m | l l l
4 28 94 164 189 263 286 304 439 515 570 59

 

   

1

Figure l. 14 A schematic graph of S. Cerevisiae Brfl. TFIIB homologous region is
represented by orange color, blue region represents conserved region for Brf(5 7, 74)

596

The Brfl protein has a putative Zn finger domain and two imperfect direct repeat
domains which are homologous to TFIIB. The three conserved domains are only found in
BRF proteins (89-91). The N-terminal half of the protein shows 19% amino acid identity
to TFIIB. It was reported that the N-terminal region of Brf interacts with the C34 subunit
of RNA polymerase III(92). Research shows that the N-terminus retains the ability to
recruit B” to the TBP/DNA complex and to direct efficient transcription on supercoiled
DNA. But the binding with TBP and B” is not mainly stabilized by this half of the

protein(79). Although the N-terminal half can also form TFIIIB-DNA complexes at

23

strong TATA box promoters and recruit pol III and direct multi round transcription as well,
the stability of the complexes is so low that the TFIIIB-DNA complexes assembled with
the N-terminal half of Brfl alone can not be detected by native electrophoresis gel(84).
And in the case of TFIIIC-dependent binding of TFIIIB, the N-terminus of Brfl also has
contact with subunits of TFIIIC (79, 91).

On the other hand, the ~ 30 KD C-terminal half which is only conserved among BRF
is of great importance in the interaction with TBP and DNA. The C-terrninal half of Brfl ,
which is not conserved in TFIIB, but homologous among BRF proteins, shows strong
binding affinity with TBP, B” and DNA(90-91). The c-terminus itself can bind TBP, DNA
and B” in vitro and the complex is stable enough for transcriptional assays(93). Internal
deletion in Brf homology regions 11 or III eliminates the binding ability. But even the
removal of the entire TFIIB homology region and the fungal homology region I of Brfl
doesn’t entirely eliminate the ability to incorporate B” into a TFIIIB/DNA complex(90).

It has been reported that the N-terminus of Brfl is required for directing transcription.
But the C-terminal Brfl itself (especially the region around 435-545) is sufficient for
assembling the TFIIIB/DNA recombinant complex(91). When this domain is
reconstituted with the N-terminal ha1f(1-282) of Brfl, almost full wild-type Brfl activity
is recovered in both TFIIIC-dependent and TFIIIC-independent transcription. The protein
interaction between Brfl and TBP in the TBP/DNA/Brf complex is mainly between the
C-terminus of Brfl and the N-terminal top and stirrup of TBP (90-91 ).

The presence of Brfl is also essential for the loading of B” to the TBP-DNA

24

1“!!— ‘

complex. It has been observed that without Brfl B” can not bind with TBP-DNA stably
(94-95). The C-terrninal mutant 284-596 can hold all of B” tightly to the TBP/DNA
complex. SO wild type Brf (and Brf 435-545 as well) serve as a ‘bridge” between the

TBP/DNA complex and B” in the reconstituted TFIIIB/DNA complex (91, 96).

1.2.4.3 B” protein

B” is the third subunit of TFIIIB. The subunit is also called TFIIIB90 or bdpl. In the
TFIIIB complex, TBP and Brfl by themselves are not competent to direct transcription
initiation by RNA pol III either in vitro or in viva. B” is required for transcription in
duplex DNA or chromatin. But in order to form the pre-initiation complex, Brf has to be
present. B” cross-links very weakly to DNA in TBP/DNA in the absence of Brf or Brf
mutants(96). In other words, B” can bind with TBP/Brf/DNA, but not TBP/DNA. B”
seems to serve as a “scaffold” for holding Brf in the TFIIIB/DNA complex (95-96). The
assembly of the complete pre-initiation complex reconfigures Brf protein inside the
complex. And the recruitment of B” also makes the TFIIIB/DNA complex much more
stable (7], 94).

In yeast S. cerevisiae, B” consists of 594 amino acid residues. It is reported that
two domains, 272-292 and 424-449, are required for TFIIIC-dependent transcription and

at least one of them is required for TFIIIC-independent transcription. This is consistent

25

with the finding that B” without the N-terminal 262 amino acid residues or C-terminal
130 amino acid residues remain competent to assemble into stable TFIIIB-DNA complex
via a TFIIIC-dependent assembly pathway on SUP4 genes(96). It was concluded that the
central ~225 amino acid residues of B” appear to encompass the functional core of the
protein(57). B” has a putative DNA binding SANT (SWI3, ADAZ, fl-COR and IFIIIB)
domain that is homologous among the families. Figure 1.13 illustrates the SANT domain
alignment in human and yeast S. cerevisiae. Figure 1.14 shows the primary sequence of S.

cerevisiae B” protein.

 

 

 

SANT domain
human B" ‘- .’
167 298 355 470 822 1338
-7/////*/122222222221 l1388
S.cerevisiae B" E 21% 143%: 17%:

281 415 472 578

 

Figure l. 15 The similarity in the human and S. cerevisiae B” sequence. The orange
box corresponds to the SANT domain.

26

1. MSSIVNKSGTRFAPKVRQRRAATGGTPTPKPRTPQLFIPESKEIEEDNSD
51 NDKGVDENETAIVEKPSLVGERSLEGFTLTGTNGHDNEIGDEGPIDASTQ
101 NPKADVIEDNVTLKPAPLQTHRDQKVPRSSRLASLSKDNESRPSFKPSFL
151 DSSSNSNGTARRLSTISNKLPKKIRLGSITENDMNLKTFKRHRVLGKPSS
201 AKKPAGAHRISIVSKISPPTAMTDSLDRNEFSSETSTSREADENENYVIS
251 KVKDIPKKVRDGESAKYFIDEENFTMAELCKPNFPIGQISENFEKSKMAK

30]. KAKLEKRRHLRELRMRARQEFKPLHSLTKEEQEEEEEKRKEERDKLLNAD

351 IPESDRKAHTAIQLKLNPDGTMAIDEETMVVDRHKNASIENEYKEKVDEN
401 PFANLYNYGSYGRGRTTDFWTUEEMIKFIKALSHWGTDFNLISQLYPYRS

451 EKQVKAKFVNEEKKPPILIELALRSKLPPNFDEYCCEIKKNIGTVADFNE

501 KLIELQNEHKHHMKEIEEAKNTAKEEDQTAQRLNDANLNKKGSGGIMTND

551 LKVYRKTEVVLGTIDDLKRKKLKERNNDDNEDNEGSEEEPEIDQ

Figure 1. 16 The amino acid sequence of S. cerevisiae B". Region 415-472 is labeled in
red and region 329-357 which is the SANT domain is labeled in blue color.

It has been shown that B” binds predominantly with ~10 base pairs upstream of the
TBP binding site TATA box. In this region, B” and Brf share an extended overlapped
interface with DNA. Upon binding with TBP/Brf/DNA, B” extended its binding ~15 base
pairs upstream of TATA box. An additional 15-20 base pairs upstream will further
stabilize the complex but is not absolutely required( 78, 97). It is also suggested that the

addition of B” in the TFIIIB-DNA complex induces a bend in the DNA conformation

27

between the TATA box and the transcriptional starting site. The bending of DNA is
postulated to contribute to the stabilization of the TFIIIB-DNA complex through

extended interaction between DNA and subunits of TFIIIB (98-100).

1.2.5 Oligonucleotide sequence

From the previous introduction on the three subunits of TFIIIB, it is clear that the
DNA sequences to be investigated should have certain base pairs upstream of the TBP
binding site (TATA box) and a segment downstream of the TATA box.

The TBP protein binds with the TATA box which is numbered around -25 to -30 on
the promoter in most TATA containing DNA sequence(101).

Brf requires 12-15 base pairs downstream of TATA box to form a stable
TBP/Brf/DNA complex (78, 101).

It is reported that two regions of B”, 291-310 and 426-487, contact the DNA
upstream and downstream of the TATA box respectively upon loading by Brf protein (96,
101). And at least 10 base pairs immediately downstream of the TATA box are needed
for B” binding (96).

From the information listed above, we can see that formation of stable TFIIIB/DNA
complexes prefers ~15 bp upstream and ~10 bp downstream of the TBP binding site. This

is a quite long DNA strand which might be detrimental to crystallization of the complex

28

Also from the recent results, TBP/Brf/DNA complex was constructed with truncated
version of TBP and Brf. It has been found that a very short DNA sequence was used to
form the complex stably with the truncated TBP and Brf proteins(102). This gave us a
hint that shorter oligonucleotides might be plausible in forming complexes with truncated
TFIIIB subunits. So a series of oligonucleotides with different lengths will be screened in
the TFIIIB-DNA complex study.

The structures of binary TBP/DNA and tertiary TBP/Brf/DNA complexes have been
solved recently (88, 102). The detailed atomic structures helped us to understand the
basic interaction among TBP, Brf protein and specific promoters. In order to shed light on
the transcription mechanism which will benefit extensive biological and medical research
(103-104), the structure of the complete pre-initiation quaternary complex

TBP/Brf/B”/DNA must be obtained.

29

Reference cited

10.

11.

Beckmann, C. O. (1953) The structure of amylopectin and glycogen, Ann N Y
Acad Sci 57, 384-397.

Lehninger, L. A. N. L. D. C., M. Michael (1993) Principles of Biochemistry,
Second Edition ed., Worth Publishers, Inc., New York.

Ball, S. G, and Morell, M. K. (2003) From bacterial glycogen to starch:
Understanding the biogenesis of the plant starch granule, Annual Review of Plant
Biology 54, 207-233. '

Ozen, H. (2007) Glycogen storage diseases: New perspectives, World J
Gastroentero 13, 2541-2553.

Bell, D. J. (1936) The molecular structure of glycogen from the whole tissues of
Mytilus edulis, Biochem J 30, 2144-2145.

Bell, D. J. (1937) Glycogen: The molecular structure of horse muscle glycogen,
Biochem J31, 1683-1691.

Dumazert, C. (1950) [New model of the structure of starch and glycogen], Bull
Soc Chim Biol (Paris) 32, 998-1008.

Preiss, J., and Sivak, M. N. (1998) Biochemistry, molecular biology and
regulation of starch synthesis, Genet Eng UV Y) 20, 177-223.

Smith, A. M. (1999) Making starch, Current Opinion in Plant Biology 2, 223-229.

van der Maarel, M. J. E. C., van der Veen, B., Uitdehaag, J. C. M., Leemhuis, H.,
and Dijkhuizen, L. (2002) Properties and applications of starch-converting
enzymes of the alpha-amylase family, Journal of Biotechnology 94, 137-155.

Jenkins, P. J., and Donald, A. M. (1995) The influence of amylose on starch

3O

12.

l3.

14.

15.

16.

17.

18.

19.

20.

granule structure, Int J Biol Macromol 1 7, 315-321.

Jenkins, P. J ., Cameron, R. E., Donald, A. M., Bras, W., Derbyshire, G. E., Mant,
G. R., and Ryan, A. J. (1994) In—Situ Simultaneous Small and Wide-Angle

X-Ray-Scattering - a New Technique to Study Starch Gelatinization, J Polym Sci
Pol Phys 32, 1579-1583.

Hinrichs, W., Buttner, G, Steifa, M., Betzel, C., Zabel, V., Pfannemuller, B., and
Saenger, W. (1987) An Amylose Antiparallel Double Helix at Atomic Resolution,
Science 238, 205-208.

Schulz, W., Sklenar, H., Hinrichs, W., and Saenger, W. (1993) The Structure of the
Left-Handed Antiparallel Amylose Double Helix - Theoretical-Studies,
Biopolymers 33, 363-375.

Helbert, W., and Chanzy, H. (1994) Single-Crystals of V-Amylose Complexed
with N-Butanol or N-Pentanol - Structural Features and Properties, Int J Biol
Macromol 16, 207-213.

Yuri, K. (2006) image of wheat starch granules , , (wheat starch granules stained
with iodine, p. f. l. m., Ed.),
http://en.wikimdiaorg/wiki/File:Wheat starch granulesJPG.

 

Abad, M. (2002) the X-ray crystallographic structures of angiostatin and
branching enzyme, In Department of Chemistry, p 127, Michigan State University,
East Lansing.

Tally, S. (2002) Making biofuel from corn stover, Biocycle 43, 62-62.

Sticklen, M., Mei, C. S., Ransom, C., Park, S. H., Sabzikar, R., and Qi, S. (2008)
Modifying the corn genome to improve its biomass biofuel production, In Vitro
Cell Dev-An 44, $29-$29.

Smidansky, E. D., Clancy, M., Meyer, F. D., Lanning, S. E, Blake, N. K., Talbert,
L. E., and Giroux, M. J. (2002) Enhanced ADP-glucose pyrophosphorylase
activity in wheat endosperm increases seed yield, P Natl Acad Sci USA 99,

31

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

1724-1729.

Regierer, B., Fernie, A. R., Springer, F., Perez-Melis, A., Leisse, A., Koehl, K.,
Willmitzer, L., Geigenberger, P., and Kossmann, J. (2002) Starch content and

yield increase as a result of altering adenylate pools in transgenic plants, Nat
Biotechnol 20, 1256-1260.

Crabb, W. D., and Shetty, J. K. (1999) Commodity scale production of sugars
from starches, Curr Opin Microbiol 2, 252-256.

Jobling, S. (2004) Improving starch for food and industrial applications, Current
Opinion in Plant Biology 7, 210-218.

Kossmann, J., and Lloyd, J. (2000) Understanding and influencing starch
biochemistry, Crit Rev Biochem Mol Biol 35, 141-196.

Preiss, J ., Yung, S. (1, and Baecker, P. A. (1983) Regulation of bacterial glycogen
synthesis, Mol Cell Biochem 5 7, 61-80.

Ballicora, M. A., Iglesias, A. A., and Preiss, J. (2004) ADP-Glucose
Pyrophosphorylase: A Regulatory Enzyme for Plant Starch Synthesis, Photosynth
Res 79, 1-24.

Preiss, J. (1996) ADPglucose pyrophosphorylase: basic science and applications
in biotechnology, Biotechnol Annu Rev 2, 259-279.

Ballicora, M. A., Iglesias, A. A., and Preiss, J. (2003) ADP-glucose
pyrophosphorylase, a regulatory enzyme for bacterial glycogen synthesis,
Microbiol Mol Biol Rev 6 7, 213-225, table of contents.

Iglesias, A. A., Barry, G. E, Meyer, C., Bloksberg, L., Nakata, P. A., Greene, T.,
Laughlin, M. J ., Okita, T. W., Kishore, (1 M., and Preiss, J. (1993) Expression of

the potato tuber ADP-glucose pyrophosphorylase in Escherichia coli, J Biol Chem
268, 1081-1086.

Greene, T. W., Chantler, S. E., Kahn, M. L., Barry, G. F., Preiss, J ., and Okita, T.

32

31.

32.

33.

34.

35.

36.

37.

38.

39.

W. (1996) Mutagenesis of the potato ADPglucose pyrophosphorylase and
characterization of an allosteric mutant defective in 3-phosphoglycerate activation,
Proc Natl Acad Sci USA 93, 1509-1513.

Ballicora, M. A., Fu, Y., Nesbitt, N. M., and Preiss, J. (1998) ADP-Glucose
pyrophosphorylase from potato tubers. Site-directed mutagenesis studies of the
regulatory sites, Plant Physiol 118, 265-274.

Fu, Y., Ballicora, M. A., Leykam, J. F., and Preiss, J. (1998) Mechanism of
reductive activation of potato tuber ADP-glucose pyrophosphorylase, J Biol Chem
273, 25045-25052.

Ballicora, M. A., Frueauf, J. B., Fu, Y., Schurmann, P., and Preiss, J. (2000)
Activation of the potato tuber ADP-glucose pyrophosphorylase by thioredoxin, J
Biol Chem 275, 1315-1320.

Bejar, C. M., Ballicora, M. A., Gomez-Casati, D. F., Iglesias, A. A., and Preiss, J.
(2004) The ADP-glucose pyrophosphorylase from Escherichia coli comprises two
tightly bound distinct domains, FEBS Lett 5 73, 99-104.

Jin, X., Ballicora, M. A., Preiss, J ., and Geiger, J. H. (2005) Crystal structure of
potato tuber ADP-glucose pyrophosphorylase, EMBO J 24, 694-704.

Bejar, C. M., Ballicora, M. A., Iglesias, A. A., and Preiss, J. (2006) ADPglucose
pyrophosphorylase's N-terminus: structural role in allosteric regulation, Biochem
Biophys Res Commun 343, 216-221.

Copeland, L., and Preiss, J. (1981) Purification of Spinach Leaf ADPglucose
Pyrophosphorylase, Plant Physiol 68, 996-1001.

Morell, M. K., Bloom, M., Knowles, V., and Preiss, J. (1987) Subunit Structure of
Spinach Leaf ADPglucose Pyrophosphorylase, Plant Physiol 85, 182-187.

Ball, K., and Preiss, J. (1994) Allosteric sites of the large subunit of the spinach
leaf ADPglucose pyrophosphorylase, J Biol Chem 269, 24706-24711.

33

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

Dickinson, D. B., and Preiss, J. (1969) ADP glucose pyrophosphorylase from
maize endosperm, Arch Biochem Biophys 130, 119-128.

Plaxton, W. C., and Preiss, J. (1987) Purification and Properties of Nonproteolytic
Degraded ADPglucose Pyrophosphorylase from Maize Endosperm, Plant Physiol
83, 105-112.

Giroux, M., Smith-White, B., Gilmore, V., Hannah, L. C., and Preiss, J. (1995)
The large subunit of the embryo isoform of ADP glucose pyrophosphorylase from
maize, Plant Physiol 108, 1333-1334.

Abad, M. C., Binderup, K., Rios-Steiner, J ., Ami, R. K., Preiss, J ., and Geiger, J.
H. (2002) The X-ray crystallographic structure of Escherichia coli branching
enzyme, J. Biol. Chem. 277, 42164-42170.

Guan, H. P., Baba, T., and Preiss, J. (1994) Expression of Branching Enzyme-Ii of
Maize Endosperm in Escherichia-Coli, Cell Mol Bio] 40, 981-988.

Guan, H. P., Baba, T., and Preiss, J. (1994) Expression of Branching Enzyme-I of
Maize Endosperm in Escherichia-Coli, Plant Physiology 104, 1449-1453.

Guan, H. P., and Preiss, J. (1993) Differentiation of the Properties of the
Branching Isozymes from Maize (Zea mays), Plant Physiol 102, 1269-1273.

Kuriki, T., Guan, H. P., and Preiss, J. (1994) Structure and Function of Starch
Branching Enzyme, Nippon Nogeik Kaishi 68, 1581-1584.

Hernandez, J. M., Gaborieau, M., Castignolles, P., Gidley, M. J., Myers, A. M.,
and Gilbert, R. G (2008) Mechanistic investigation of a starch-branching enzyme
using hydrodynamic volume SEC analysis, Biomacromolecules 9, 954-965.

Guan, H. R, Li, R, ImparlRadosevich, J., Preiss, J., and Keeling, P. (1997)
Comparing the properties of Escherichia coli branching enzyme and maize
branching enzyme, Archives of Biochemistry and Biophysics 342, 92-98.

Kuriki, T., Stewart, D. C., and Preiss, J. (1997) Construction of chimeric enzymes

34

51.

52.

53.

54.

55.

56.

57.

58.

59.

60.

out of maize endosperm branching enzymes I and II: Activity and properties,
Journal of Biological Chemistry 272, 28999-29004.

Kuriki, T., Stewart, D. C., and Preiss, J. (1997) Construction of chimeric enzymes
out Of maize endosperm branching enzymes I and II: activity and properties, J
Biol Chem 2 72, 28999-29004.

Binderup, K., Mikkelsen, R., and Preiss, J. (2000) Limited proteolysis of
branching enzyme from Escherichia coli, Archives of Biochemistry and
Biophysics 3 77, 366-371.

Abad, M. C., Binderup, K., Preiss, J ., and Geiger, J. H. (2002) Crystallization and
preliminary X-ray diffraction studies of Escherichia coli branching enzyme, Acta
Crystallogr D 58, 359-361.

Binderup, M., Mikkelsen, R., and Preiss, J. (2002) Truncation of the amino
terminus of branching enzyme changes its chain transfer pattern, Archives of
Biochemistry and Biophysics 397, 279-285.

Devillers, C. H., Piper, M. E., Ballicora, M. A., and Preiss, J. (2003)
Characterization of the branching patterns of glycogen branching enzyme
truncated on the N-terminus, Archives of Biochemistry and Biophysics 418, 34-3 8.

lewin, B. (2000) Genes V11, Oxford University Press.

Schramm, L., and Hernandez, N. (2002) Recruitment of RNA polymerase III to
its target promoters, Gene Dev 16, 2593-2620.

Klekamp, M. S., and Weil, P. A. (1987) Properties of Yeast Class-Iii
Gene-Transcription Factor Tfiiib - Implications Regarding Mechanism of Action,
J Biol Chem 262, 7878-7883.

Reznikoff, W. S., Siegele, D. A., Cowing, D. W., and Gross, C. A. (1985) The
Regulation of Transcription Initiation in Bacteria, Annu Rev Genet 19, 355-387.

Willis, 1. M. (1993) Rna Polymerase-Iii - Genes, Factors and Transcriptional

35

61.

62.

63.

64.

65.

66.

67.

68.

69.

Specificity, Eur J Biochem 212, 1-11.

Huang, Y., and Maraia, R. J. (2001) Comparison of the RNA polymerase III
transcription machinery in Schizosaccharomyces pombe, Saccharomyces
cerevisiae and human, Nucleic Acids Res 29, 2675-2690.

Bogenhagen, D. F., Sakonju, S., and Brown, D. D. (1980) Control Region in the
Center of the 53 Rna Gene Directs Specific Initiation of Transcription .2. 3'
Border of the Region, Cell 19, 27-35.

Sakonju, S., Bogenhagen, D. F., and Brown, D. D. (1980) Control Region in the
Center of the SS Rna Gene Directs Specific Initiation of Transcription .1. 5'
Border of the Region, Cell 19, 13-25.

Galli, (1, Hofstetter, H., and Bimstiel, M. L. (1981) 2 Conserved Sequence Blocks
within Eukaryotic Transfer-Rna Genes Are Major Promoter Elements, Nature 294,
626-631.

Krol, A., Carbon, P., Ebel, J. P., and Appel, B. (1987) Xenopus-Tropicalis
U6-Snrna Genes Transcribed by Pol-Iii Contain the Upstream Promoter Elements
Used by Pol-Ii Dependent U-Snma Genes, Nucleic Acids Res 15, 2463-2478.

Nguyen, V. T., Kiss, T. S., Michels, A. A., and Bensaude, O. (2001) 78K small
nuclear RNA binds to and inhibits the activity of CDK9/cyclin T complexes,
Nature 414, 322-325.

Kassavetis, (1 A., Letts, (1 A., and Geiduschek, E. P. (1999) A minimal RNA
polymerase III transcription system, Embo J 18, 5042-5051.

Kassavetis, G. A., and Geiduschek, E. P. (2006) Transcription factor TFIIIB and
transcription by RNA polymerase III, Biochem Soc T 34, 1082-1087.

Joazeiro, C. A. P., Kassavetis, (1 A., and Geiduschek, E. P. (1994) Identical
Components of Yeast Transcription F actor-Iiib Are Required and Sufficient for

Transcription of Tata Box-Containing and Tata-Less Genes, Mol Cell Biol 14,
2798-2808.

36

l finer‘

70.

71.

72.

73.

74.

75.

76.

77.

78.

Kassavetis, (1 A., Kumar, A., Letts, (1 A., and Geiduschek, E. P. (1998) A
post-recruitment fiJnction for the RNA polymerase III transcription-initiation
factor IIIB, P Natl Acad Sci USA 95, 9196-9201.

Kumar, A., Grove, A., Kassavetis, (1 A., and Geiduschek, E. P. (1998)
Transcription factor IIIB: The architecture of its DNA complex, and its roles in

initiation of transcription by RNA polymerase 111, Cold Spring Harb Sym 63,
121-129.

Deprez, E., Arrebola, R., Conesa, W., and Sentenac, A. (1999) A subunit of yeast
TFIIIC participates in the recruitment of TATA-binding protein, Mol Cell Biol 19,
8042-8051.

Rozenfeld, S., and Thuriaux, P. (2001) Genetic interactions within TFIIIC, the
promoter-binding factor of yeast RNA polymerase III, Mo] Genet Genomics 265,
705-710.

Jin, X. (2002) X-ray crystallographic studies of RNA Polymerase III transcription
factor TFIIIB and lL-WO-Inositol l-phosphate synthase, In Department of
Chemistry, Michigan State University, East Lansing.

Huet, J ., and Sentenac, A. (1992) The Tata-Binding Protein Participates in Tfiiib
Assembly on Transfer-Rna Genes, Nucleic Acids Res 20, 6451-6454.

Kassavetis, (1 A., Joazeiro, C. A. P., Pisano, M., Geiduschek, E. P., Colbert, T.,
Hahn, S., and Blanco, J. A. (1992) The Role of the Tata-Binding Protein in the
Assembly and Function of the Multisubunit Yeast Rna Polymerase-Iii
Transcription Factor, Tfiiib, Cell 71, 1055-1064.

Taggart, A. K. P., Fisher, T. S., and Pugh, B. F. (1992) The Tata-Binding Protein
and Associated Factors Are Components of Pol-Iii Transcription Factor Tfiiib,
Cell 71, 1015-1028.

Colbert, T., Lee, S., Schimmack, (1, and Hahn, S. (1998) Architecture of protein
and DNA contacts within the TFIIIB-DNA complex, Mol Cell Biol 18,
1682-1691.

37

79.

80.

81.

82.

83.

84.

85.

86.

87.

88.

Chaussivert, N., Conesa, C., Shaaban, S., and Sentenac, A. (1995) Complex
Interactions between Yeast Tfiiib and Tfiiic, J Biol Chem 270, 15353-15358.

Dieci, (1, Percudani, R., Giuliodori, S., Bottarelli, L., and Ottonello, S. (2000)
TFIIIC-independent in vitro transcription of yeast tRNA genes, J Mol Biol 299,
601-613.

White, R. J., and Jackson, S. P. (1992) Mechanism of Tata-Binding Protein
Recruitment to a Tata-Less Class-Iii Promoter, Cell 71, 1041-1053.

Lobo, S. M., Tanaka, M., Sullivan, M. L., and Hernandez, N. (1992) A pr
Complex Essential for Transcription from Tata-Less but Not Tata-Containing Rna
Polymerase-Iii Promoters Is Part of the Tfiiib Fraction, Cell 71, 1029-1040.

Cormack, B. P., and Struhl, K. (1992) The Tata-Binding Protein Is Required for
Transcription by All 3 Nuclear-Rna Polymerases in Yeast-Cells, Cell 69, 685-696.

Schroder, 0., Bryant, (1 0., Geiduschek, E. P., Berk, A. J ., and Kassavetis, (1 A.
(2003) A common site on TBP for transcription by RNA polymerases II and III,
Embo J22, 5115-5124.

Juo, Z. 8., Chin, T. K., Leiberman, P. M., Baikalov, I., Berk, A. J ., and Dickerson,
R. E. (1996) How proteins recognize the TATA box, J Mol Biol 261 , 239-254.

Shen, Y. H., Kassavetis, G. A., Bryant, (1 0., and Berk, A. J. (1998) Polymerase
(Pol) III TATA box-binding protein (TBP)-associated factor Brf binds to a surface
on TBP also required for activated Pol II transcription, Mol Cell Biol 18,
1692-1700.

Vilalta, A., Trivedi, A., and Johnson, D. L. (1996) Mutation in the tata-binding
protein (TBP) selectively abolishes both RNA polymerase III transcription and the
interaction of TBP with a TFIIIB subunit, Mol Biol Cell 7, 2704-2704.

Kim, Y. C., Geiger, J. H., Hahn, S., and Sigler, P. B. (1993) Crystal-Structure of a
Yeast pr Tata-Box Complex, Nature 365, 512-520.

38

89.

90.

91.

92.

93.

94.

95.

96.

97.

Khoo, B., Brophy, B., and Jackson, S. P. (1994) Conserved Functional Domains
of the Rna-Polymerase-Iii General Transcription Factor Brf, Gene Dev 8,
2879-2890.

Kassavetis, (1 A., Bardeleben, C., Kumar, A., Ramirez, E., and Geiduschek, E. P.
(1997) Domains of the Brf component of RNA polymerase III transcription factor
IIIB (TFIIIB): Functions in assembly of TFIIIB-DNA complexes and recruitment
of RNA polymerase to the promoter, Mol Cell Biol 1 7, 5299-5306.

Kassavetis, G. A., Kumar, A., Ramirez, E., and Geiduschek, E. P. (1998)
Functional and structural organization of Brf, the TFIIB-related component of the

RNA polymerase III transcription initiation complex, Mol Cell Biol 18,
5587-5599.

Ferri, M. L., Peyroche, (1, Siaut, M., Lefebvre, 0., Carles, C., Conesa, C., and
Sentenac, A. (2000) A novel subunit of yeast RNA polymerase III interacts with
the TFIIB-related domain of TFIIIB70, Mol Cell Biol 20, 488-495.

Werner, M., Chaussivert, N., Willis, 1. M., and Sentenac, A. (1993) Interaction
between a Complex of Rna Polymerase-Iii Subunits and the 70-Kda Component
of Transcription Factor-Iiib, J Biol Chem 268, 20721-20724.

Ishiguro, A., Kassavetis, (1 A., and Geiduschek, E. P. (2002) Essential roles of
del, a subunit of RNA polymerase III initiation factor TFIIIB, in transcription
and tRN A processing, Mol Cell Biol 22, 3264-3275.

Kumar, A., Kassavetis, (1 A., Geiduschek, E. P., Hambalko, M., and Brent, C. J.
(1997) Functional dissection of the B" component of RNA polymerase III

transcription factor IIIB: A scaffolding protein with multiple roles in assembly and
initiation of transcription, Mol Cell Biol 1 7, 1868-1880.

Shah, S. M. A., Kumar, A., Geiduschek, E. P., and Kassavetis, (1 A. (1999)
Alignment of the B " subunit of RNA polymerase III transcription factor IIIB in
its promoter complex, J Biol Chem 274, 28736-28744.

Andrau, J. C., and Werner, M. (2001) B "-associated factor(s) involved in RNA
polymerase III preinitiation complex formation and start-site selection, Eur J

39

 

98.

99.

100.

101.

102.

103.

104.

Biochem 268, 5167—5175.

Grove, A., Kassavetis, G. A., Johnson, T. E., and Geiduschek, E. P. (1999) The
RNA polymerase III-recruiting factor TFIIIB induces a DNA bend between the
TATA box and the transcriptional start site, J Mol Biol 285, 1429-1440.

Braun, B. R., Kassavetis, (1 A., and Geiduschek, E. P. (1992) Bending of the
Saccharomyces-Cerevisiae 5s Ribosomal-Rna Gene in Transcription Factor
Complexes, J Biol Chem 267, 22562-22569.

Soragni, E., and Kassavetis, (1 A. (2008) Absolute gene occupancies by RNA
polymerase III, TFIIIB, and TFIIIC in Saccharomyces cerevisiae, J Biol Chem
283, 26568-26576.

Persinger, J., Sengupta, S. M., and Bartholomew, B. (1999) Spatial organization
of the core region of yeast TFIIIB-DNA complexes, Mol Cell Biol 19, 5218-5234.

Juo, Z. S., Kassavetis, (1 A., Wang, J. M., Geiduschek, E. P., and Sigler, P. B.
(2003) Crystal structure of a transcription factor IIIB core interface ternary
complex, Nature 422, 534-539.

White, R. J. (2004) RNA polymerase III transcription and cancer, Oncogene 23,
3208-3216.

White, R. J. (2004) RNA polymerase III transcription - a battleground for tumour
suppressors and oncogenes, Eur J Cancer 40, 21-27.

40

 

CHAPTER 2: X-RAY CRYSTALLOGRAPHIC STUDY OF E.
coli BEIOLIGOSACCHARIDE COMPLEXES

2.1 E.coli Branching enzyme(BE) and polysaccharide

BE cleaves the (I-l,4 linkage of natural linear polysaccharide and transfers the
fragment to form a-1,6 branches. In order to cleave the glycosidic bonds, BE has to bind
with substrate. The natural substrates for BE are long polysaccharide chains (either
glycogen or starch) usually with more than 30 glucose units.

The active site of BE has been identified, it is located inside the (a / B) 3 barrel region.
(Figure 2.1) Several catalytic residues have also been identified by sequence alignment
and site directed mutagenesis: Tyr300, Asp335, His340, Arg403, Glu458, His525 and
Asp526 are involved in the branching function (1-3). This is also consistent with the
active site location of (it-amylase family enzymes. How these residues interact with
substrate, and what is the detailed branching mechanism is still unknown. Although the
investigation of some other (II-amylase family enzymes showed binding between enzymes
and substrate or pseudo-substrate (usually polysaccharide or mimics) (4-6), there is still
no report about BE binding with substrate.

Given the fact that long polysaccharide chains are flexible, it is very likely there are

41

other binding sites outside the (01 / (3)3 barrel region. As mentioned before, so far there is
no data about the binding between branching enzyme and polysaccharide (or natural

substrate) either inside or outside the active site.

active site

i 4’73"
“‘1 1"! y‘l/

0‘

 

Figure 2. l E.coli BE (PDB databank # 1M7X) overall structure (3, 7). First figure
shows the active site is inside the central 01/13 barrel region and the second figure shows

E. coli BE overlaid with isoamylase from Pseudomonas anyloderamosa (PDB databank #
IBF 2) (in magenta color).

42

In order to understand how BE catalyzes the formation of 01-1, 6 branches, first we
want to know how BE interacts with its substrate.

It is known that the E. coli BE lacking its first 113 amino acids (N l 13BE) still retains
the majority of its branching enzyme activity, which means it still has the ability to form
the enzyme-substrate interaction. And also the crystal structure of this construct is known,
so the N113BE was selected to investigate the interaction.

Several kinds of oligosaccharides were selected for the study: maltose, maltotriose,
maltotetraose, maltopentaose, maltohexaose, maltoheptaose, a-cyclodextrin,
B-cyclodextfin and y—cyclodextrin (Figure 2.2). We used homogeneous oligosaccharides

instead of natural starch/glycogen to simplify the problem.

2.2 Why were these oligosaccharides selected?

Cyclodextrins are cyclic oligosaccharides naturally produced from starch by
enzymatic conversion (8-9). There are three major natural cyclodextrins: a-cyclodextrin
(u-CD) (6 glucose units), B-cyclodextrin (B-CD) (7 glucose units) and y-cyclodextrin

(y-CD) (8 glucose units). Figure 2.2 A shows their structures.

43

The natural substrates of BB, glycogen and starch, actually are not straight linear
polymers. Naturally synthesized glucan adopts a helical structure instead of a random
flexible structure in the glycogen /starch granule (10-12). The structure of amylose, the
linear component of natural starch, was determined to be helical (10, 13-16). Amylose
has 6 glucose units in one turn, which is very similar to a-CD. We overlaid the structures
of amylose and a-CD in Figure 2.28. It is clear that amylose has similar size and
curvature as 01-CD; two of the six glucose units of the one turn of the amylose overlay
with a-CD perfectly. This indicates that cyclodextrins have similar structures to natural
helical amylose. So cyclodextrins may serve as substrate mimics in substrate binding

studies.

44

L.-

1m

A)

H
H H O
H 0 OHOHO
HO OH
O HO H
O
HO H
H
OH
0 HO beta CD 0
alpha CD 0.. o 0,, H
"0 OH O
O
O H OH
o OH H00“ 0
0 0H0 OH OH 0 H
0 HO
O
HO OH
OH

OH
O 0 OH
0 OH HO
HO O
H 0
gamma CD

OH
O
OH OH
O OH O
HO OH o o
0 022g
0

H OH

Figure 2. 2 A) Structures of a-cyclodextrin, B-cyclodextrin and y-cyclodextrin. These
three cyclodextrins are cyclic oligosaccharides with 6, 7 and 8 glucose units respectively.

B)

crystal structure
of amylase

    

Figure 2.2 B) The left figure shows the crystal structure of amylose (16); the right figure
shows the overlay of Cl -CD and amylose

There is no research regarding the binding between E. coli BE and oligosaccharides,
but a-glucan binding to potato tuber starch branching enzyme I (SBEI) has been studied
and shows that some cyclodextrins bind with SBEI stronger than linear oligosaccharides
(18). SBEI and E.coli BE share similarities in sequence, fmetion and overall structure.
This gave us a hint that maybe cyclodextrin could bind with E.coli BE as well. Our

research confirmed the binding between E. coli BE and several of these glucans.

 

 

Compound Dissociation constant ( mM)
Maltose >50

Maltotriose l 1.7
Maltotetraose 1 . 1
Maltopentaose 0.75
Maltohexaose 0.25
Maltoheptaose 0.16
a-cyclodextrin 6.0
B-cyclodextrin 0.25
y-cyclodextrin 0.00067

 

Table 2. 1 Dissociation constants for different a-glucans and Potato tuber SBEI (18)

46

2.3 Materials and methods

2.3.1 Experimental design

X-ray crystallography was used to study the interaction between the truncated E. coli
BE protein and selected oligosaccharides. After purification of the truncated E.coli BE
protein, co-crystallization and substrate soaking were both attempted. Many crystals were
used to collect diffraction data at the Advanced Photon Source (APS). Diffraction data
was examined to find out if the substrate was present. Final structures of
substrate-binding E. coli BE protein were refined with the CCP4 and COOT software

packages. The experimental details will be elUcidated.

2.3.2 Protein over-expression and purification

The pET 23d vector based plasmid containing the coding sequence of E.coli BE
lacking the first 113 amino acids(19) at the N-terminus was obtained from our
collaborator Dr. Jack Preiss. It was transformed into E. coli BL21 (DE3) cells and grown
on ampicillin (0.05 mg/mL) resistant agar plate. Colonies appeared after overnight
incubation at 37 °C. Cells were cultivated in 50 ml Luria Broth (L.B) overnight with

ampicillin (0.1mg/mL) and then were transferred into 1 liter fresh L.B media with

47

ampicillin (0.1 mg/mL). Cells were grown at 37 °C until the optical density at 600nm
( O.D600 ) reached 0.5-0.6. Then IPTG (0.5mM) induced production of the protein. After
cultivation at 25 °C for five more hours, the cells were harvested.

The cell pellet was re-suspended in buffer A (50mM Tris-Acetate, pH8.0, lOmM
EDTA, 2.5mM DTT and 10% v/v glycerol) and lysed by sonication. Cell debris was spun
down by centrifugation at 5000 rpm for 30 minutes. The supernatant was precipitated by
adding 40% ammonium sulfate solution. After centrifugation the protein pellet was
re-suspended in buffer A. The solution was dialyzed overnight against buffer A, then
loaded into DEAE fractogel column. Protein fractions were eluted with 0-0.4 M KCl
gradient in buffer A. The protein fractions were centrifuged to about 5 mg / ml. Source Q
anion exchange chromatography was used to purify the protein to almost homogeneity.
Pure protein was eluted with 0-1.0M KCl gradient in buffer A. Size exclusion
chromatography (Superdex200 16/60 column from GE healthcare) was used as the final
purification step. Figure 2.3 shows the size exclusion chromatograph. Comparing the
protein peak position and the protein standard chromatograph (shown in the graph), it is
clear that the protein exists in solution as monomer. SDS-PAGE was used to verify the
homogeneity and purity of the protein. Pure E.coli BE protein fractions were pooled
together and concentrated to about 10 mg/ml for further use. Figure 2.4 shows the

SDS-PAGE gel of final pure truncated E. coli BE protein samples.

48

Elution volume (ml)

 

O 20 40 60 80 1 OO 1 20
' ' l I I i I I ‘5
M.W standard position 158 KD l 44KB 17KD 1.3KD

I‘

I
1

BE protein peak\

illllll l‘» F-TTI'ITN .—L

””II'IIIIIIII‘IIIIIIIII; I’ll’lillil‘“ IIIIHi

..; :1“
:lllllll

Figure 2. 3 Chromatograph of truncated E. coli BE protein afier size exclusion
chromatography purification ( Absorption at 280 nm). Protein was eluted as a single
sharp peak, which indicated the homogeneity of protein particle size.

22OKDa ..
128KDa

82KDa .4. w, ——).Protein

7OKDa

Figure 2. 4 SDS-PAGE gel of several pure truncated E. coli BE protein samples. The
samples were taken from different batches of final protein solution.

49

2.3.3 Protein crystallization

Pure E.coli BE protein was buffer exchanged into Crystallization buffer (25 mM
Na—HEPES, pH7.2). Then protein solution (about 10 mg /ml) was subject to sparse matrix
crystallization screen (under condition: lOOmM Na—HEPES, pH 7.2 at 4 °C) using the
hanging drop vapor diffusion method(3). Protein crystals appeared after 2 weeks, and
reached maximal size (0.3x0.15x0.l mm) in 6 weeks. (See Figure 2.5)

Several batches of protein were used to grow crystals. Single crystals were

reproduced for further substrate soaking experiments.

   

Figure 2. 5 Crystals of E.coli branching enzyme. The first picture shows crystals inside the
crystallization screen drop (crystal size 0.05x0.lx0.2 mm), second picture shows a crystal in
cryo-solution, ready for x-ray diffraction

50

2.3.4 Attempts at co-crystallization of BE with oligosaccharides
Co-crystallization of E.coli BE with various oligosaccharides was attempted.
Oligosaccharides solutions (from 1 mM to 10 mM) were mixed with protein solution
together and incubated at 4 °C for at least 1 hour for binding. Sparse Matrix
crystallization screens were performed at both room temperature and 4 °C. No crystals

were found.

2.3.5 Soaking BE crystals with Oligosaccharide solutions

The complex structures were obtained by substrate soaking. The truncated E. coli BE
protein was crystallized first, then the protein crystals were picked up using a cryo loop
(Hampton Research, Ltd), and dropped into oligosaccharide solutions at 4 °C for various
time. Crystal cracking was carefully monitored. Then crystals were flash frozen with cryo
protection and stored in liquid nitrogen.

Extensive soaking experiments were carried out to search for the optimal soaking
condition. Several factors affect the final result of a soaking experiment: binding strength
between substrate and protein, concentration of the substrate, soaking time and inherent
nature of the protein crystal.

In this research, we managed to overcome the low quality diffraction and soaking
induced crystal cracking problem. Extensive soaking conditions were screened and
soaking induced cracking was carefully monitored.

The substrates were prepared in solutions at the highest concentrations possible.

51

Soaking was carried out for maximal time period before significant cracking was found
on the surface of crystals.

For linear oligosaccharides, the concentration of substrate ranged from 50 mM to 200
mM. A 70 mM solution was used for 01 and y-cyclodextrin. B-cyclodextrin was prepared

at a concentration of 15 mM due to its low solubility.

 

 

Substrate M.W. Cone. Soaking time Electron density of
(Dalton) (mM) before crystal the oligosaccarhde
cracking (hrs)

Maltose 342 200 72 No
Maltotriose 504.4 1 00 72 No
Maltotetraose 666.6 140 1 2 No
Maltopentaose 828.7 1 00 6 Yes
Maltohexaose 990.9 50 3 Yes
Maltoheptaose 1 1 53 9O 3 .5 Yes
a-CD 972.8 70 1 Yes
B-CD 1135 15 12 Yes
Y -CD 1297 70 15 Yes

 

 

Table 2. 2 Summary of substrates used in soaking experiment and soaking conditions.
The concentrations are the highest concentration of oligosaccharides used in the
experiments. The soaking times are the longest possible time before the crystals crack.
All oligosaccharides were from Sigma.

* The x-ray diffraction data were examined to find if there is electron density of the
substrate. But this method can not decide if the substrate binds with E. coli BE protein or
not. There is possibility that the substrate binds with protein, but is flexible to some
extent that its density can not show up in diffraction data.

52

2.3.6 X-ray diffraction data collection and processing

The x-ray diffraction data were collected at beam line #21 (LS-CAT) at Advanced
Photo Source, Argonne National Lab.

Diffraction data were processed using the HKL2000 software package. The
MOLREP program in CCP4 program suite(20)) was used to position the E.coli BE /
oligosaccharide complex in the unit cell using the E. coli BE structure as the search model.
CCP4 and the Coot program suite were used to refine the structures.

Table l and 2 in Appendix list the detailed data collection and refinement statistics
for all the structures obtained. All the ligand-bound structures retain the same cell
dimension as the apo structure (1M7X). All the diffraction data have acceptable
completeness. Ramachandran plots showed that the numbers of disallowed residues were
within the reasonable limit for the resolution of the structure (see Ramachandran plots in

the Appendix).

53

2.4 Three dimensional structure of E.coli BEIcyclOdextrin

complexes

The E. coli BE structure was previously reported by our group(3). The protein was
crystallized in IOOmM Na—HEPES (pH 7.2) solution. It belongs to the P21 space group
with cell dimension of 91, 103 and 185 A. There are 4 monomers in one asymmetric unit
(numbered as chain A, B, C and D respectively). The overall structure is shown in Figure
2.1. We obtained x-ray diffraction data sets of E. coli BE in complex with
(II-cyclodextrin, B-cyclodextrin and y-cyclodextrin.

We found 4 cyclodextrin binding sites in the E. coli BE /cyclodextrin complex. The

binding between 01, B, and y-cyclodextrins and BE are present in many data sets( we
collected multiple data sets for each complex) . Table 2.4 lists the binding sites and the

residues involved in the binding.

 

Binding site Bound CDs BE residues involved in the binding

 

I (on ABCD chains) 01 and B-CDs ARG255, ASN259, ASN260, PHE261 and TRP262
Il(on ABCD chains) (1, B and y-CDs ASP505,PHE508, ILES l l, LEUS 12 and TRP628
III(only on D chain) 01 and y- CDs TRP159, LYSl89, LEU201,GLN211 and GLU215

IV(on1y on C chain) 01, Band y-CDs ASP542, TRP544, GLN545, PRO659 and SER689

 

Table 2. 3 Detailed binding sites in E. coli BE /cyclodextrin complex structures

54

2.4.1 Three dimensional structure of E.coli BEIa- cyclodextrin complex.

a- cyclodextrin (Ix-CD) binds with E.coli BE at all 4 binding sites. 0.- cyclodextrin
binds with binding site I and II on all of the 4 chains in the asymmetric unit. Binding site
111 is only occupied with a- cyclodextrin on Chain C. Also 0.- cyclodextrin only binds
with binding site IV on chain D of the protein.

a— cyclodextrin is a six-glucose cyclic oligosaccharide. The molecule forms a
truncated cone shape. Cyclodextrins are well-known for their ability to carry guest
molecules inside its “cavity” (8, 21). Since the inside of the molecule is hydrophobic
while the outside is hydrophilic, it tends to host small hydrophobic guest molecule inside.
This characteristic makes cyclodextrins suitable for drug delivery. It can carry a small
hydrophobic guest molecule and deliver it into an aqueous environment (22-23). Upon
binding with a protein, it can incorporate a hydrophobic residue side chain inside its
cone-shape cavity, which stabilizes the binding further. Figure 2.6 shows the electron

density map of 01- cyclodextrin from one of our (1- CD/BE structures.

 

Figure 2. 6 1.0 o 2F0-Fc electron density map of a-cyclodextrin binding with E. coli BE
at binding site IV.

55

At binding site I, we found the side chain of PHE261, the benzyl group, extends into
the cavity, forming a hydrophobic interaction with the cyclodextrin. This is a important
stabilizing force in the binding.

TRP262 also plays an important role in stabilizing 01- cyclodextrin binding: Its
aromatic indole side chain is parallel to the +1 glucose moiety. The aromatic stacking
pattern is one of the major types of sugar-protein interaction. Furthermore, the aromatic
side chain of PHE261 is also parallel to the +1 glucose unit. The two aromatic side chains
from the two consecutive residues clamp the glucose unit between them. This interesting
interaction pattern is not very common in sugar-protein binding.

The hydrogen bonds between a hydroxyl group of cyclodextrin and ARG255,
ASN259 and ASN260 also contribute to the binding. Figure2.7 shows the detailed
interaction between 01- cyclodextrin and BE residues at the binding site.

By overlaying the local residues with the un-bound E. coli BE structure (1M7X). We
found there was basically no conformational change upon binding with cyclodextrin
except the side chain of ASN259. It moved up towards the cyclodextrin a little to form a
hydrogen bond (3.48 A). Considering that this hydrogen bond is very weak and the
ASN259 side chain was flexible in the 1M7X structure, we can conclude that there is
little energy needed to distort the conformation of any residue upon binding with
(II-cyclodextrin. This site can only bind with the end of amylose, otherwise the rest of the

amylose chain will clash with the protein.

56

6+ 9 _+ 80¢ @2385:
En 3E: 882m 05 .2me 32> fiesta. a Bow 58885 05 @527. chow—m Ema :1.ch 05 E :32? can muses comet»:
.Amom my? 98 Emmmm .oo~Zm< dwaZm< .mmNOm<v H BE 9:25 5E 9:35 atsxovofimoé £505 £04 5 .N 85E

   

  

NONQWE. mmNom<
ommzm<

a

          

00N2w<

mmNZW<
FonIQ
mmN \ .amed

-- rs I
U .‘A‘ momma...

  

\

      

man s 4
\

‘ .

    

\

    

57

Another binding site is also on the surface of the protein, quite far away from the
active site (about 20 A away from the edge of the beta barrel). Binding site 11 includes
ASP505, THR508, LEU512, and TRP628. The cyclodextrin is bound to the surface of

protein through mainly hydrogen bonds. (see Figure 2.8)

 

ASP505

TH R508

Figure 2. 8 a-cyclodextrin binds with binding site II, residue 505, 508, 512 and 628
have interactions with cyclodextrin. Detailed hydrogen bonds are indicated.

58

The side chain of LEU512 sticks to the cavity of a-cyclodextrin, but can not go
further into the hydrophobic pocket, that is because the steric hindrance of PHE509: the
side chain of PHE509 blocks the way of a glucose unit, this can be seen in Figure 2.9(It
shows the interaction from a difi‘erent angle), otherwise we can imagine that the
hydrophobic side chain of leucine will extend deeply into the cyclodextrin pocket like

binding site I.

 

Figure 2. 9 Binding between a-cyclodextrin and E. coli BE binding site II, from another
angle. It shows that side chain of PHE509 blocks the glucose units from moving closer to
incorporate LEU512 into the cyclodextrin pocket

Binding sites I and II are found on all 4 chains in the asymmetric unit. The binding is
quite strong considering some data sets were obtained by soaking the crystal in CD
solution for only 30 minutes. Also, by overlaying with apo structure 1M7X, we did not
find significant conformational changes around binding sites I and II,

We also found two binding sites (III and IV) which bind with CD on one of the four

59

chains in the asymmetric units. Binding site III is only on the D chain and binding site IV
is found only on the C chain. Binding site 111 consists of TRP159, GLN 211 and GLU

215 on the D chain.

       

PRO214

GLN211

Figure 2. 10 tit-cyclodextrin binds with binding site III. GLN21 1 and GLU215 on D
Chain of the protein have hydrogen bonds with CD.

GLN211 and GLU 215 have hydrogen bonds with CD. Also there is an aromatic
stacking interaction between the indole side chain of tryptophan 159 and CD. As we
mentioned before, this type of interaction between the aromatic side chain of tryptophan

and the glucose plane is very typical of protein-sugar binding (24-29).

60

The apo structure of E. coli BE has some “disordered” loops (the electron density of
the residues are absent because they are flexible inside the crystal). Residue 213-215 is
one of the disordered loops in the E.coli BE apo structure. While upon binding with
ligand, the interaction between the ligand and certain residues of the missing loop helped
to stabilize the loop. As a result, the former disordered loop appeared in the ligand-bound
structure. In our~a-cyclodextrin-bound BE structure, GLU215 not only contributes by
means of hydrogen bonds but also extends its side chain into the cyclodextrin pocket to

form an extra interaction. (see Figure 2.11).

 

Figure 2. 11 2F o-Fc electron density map of loop 213-215 on D chain of E. coli BE
/a-cyclodextrin

61

The last binding site consists of ASP542, TRP544, GLN545, PRO659 and SER689.

Figure 2.12 shows the detailed interaction between the residues and a-CD.

 

ASP542 PR0659

 
  

GLN545

Figure 2. 12 E.coli BE binds with a-CD at binding site IV. a-CD is stabilized by
hydrogen bonds from ASP542, GLN545, PRO659 and SER689. TRP544 contributes to
the binding through aromatic stacking with the glucose unit.

Binding sites 111 and IV only exist in one molecule of the 4 in the asymmetric unit.
Binding site III is on the D chain. Its corresponding position on the A chain is blocked by
both the cyclodextrin bound to binding site II on the D chain and another symmetry mate
due to crystal packing. Although the same site on the B and C chains are not directly

blocked, the pathway to the binding site is not as open as it is in the D chain, so its

62

binding with substrate may be more difficult, which maybe why we did not see the
binding in those molecules.

Binding site IV is only found on the C chain. That is because the same positions on
the A, B and D chains are completely blocked by symmetric molecules due to crystal

packing, making the binding impossible.

2.4.2 Three dimensional structure of E.coli BE I B- cyclodextrin complex

B- cyclodextrin (B-CD) has 7 glucose units. It is bigger than a-cyclodextrin, internal
diameter increases from 57 A to 78 A (9). We expected [3- CD to bind with E. coli BE
more strongly, since the size of inner cavity seems to be an important factor for binding
site I and 11. Due to the unusual low solubility of B- CD (9, 23), we had to increase the
soaking time to get proper binding. the experiment was carried out at low substrate
concentration such as lmM and 5 mM for longer times ( over 10 hrs) , but did not yield
any cyclodextrin bound protein data.

We also tried the soaking experiment with saturated substrate for shorter times (30
minutes, 1 hrs, 2 hrs, and 4 hrs), still no substrate binding was found. The B- CD bound
protein data sets were obtained by soaking experiments with a saturated B- cyclodextrin
solution for a longer time (at least 10 hrs). From the X-ray diffraction data, we found that
[3- CD still binds at binding site II and IV, but the binding with binding sites I and III are

weaker than those seen in the E. coli BE/ a-CD structure.

63

A)

          
 

Qi‘i‘) ‘\: ..
-._., «a. a 1.5 v
“neat”, d‘!- )s
, . .

    
     
   
    

 
 

 

 

we
.

If". —'. .
-1..
I
.u-.
.

 

 

1 r1: ‘.- -

TRP628

THR508
B)

TRP628

PHE509

THR508

Figure 2. 13 B- CD binds with E.coli BE at binding site II. TRP508, PHE509, LEU512
and TRP628 are involved in the interaction by both hydrogen bonds and hydrophobic
interaction. A: the 1.0 o 2Fo-Fc density map of the B- CD; B: The detailed interaction
between the residues and the CD.

Although B-CD is bigger than a-CD, there is basically no difference in terms of the
binding pattern at binding site IV except that ASP542 could donate two hydrogen bonds

instead of one , and tiny changes in hydrogen bond lengths (Figure 2.14).

       
 

SER689

ASP542 PR0659

GLN545

Figure 2. 14 The interaction between E.coli BE binding site IV and B-CD. Like
a-CD binding, ASP542, GLN545, PRO659 and SER689 contribute by hydrogen bonds
and TRP544 through an aromatic stacking force.

65

From Figure 2.14, we can see that B—CD is bigger; the part bound to protein is flatter,
so that part has more contact with the protein. In our structure, it indicates that the protein
stabilizes the B-CD better through one more hydrogen bond coming from ASP542. By
comparing the B-CD bound BE structure with the a-CD bound structure, we noticed that
ASP542 adopted a tiny side chain conformational change to accommodate the hydrogen
bonds.

We reasonably doubted that binding site I binds with B-CD much weaker than it does
with a-CD since in our data sets we never have 4 CD bind to one structure. And usually
the electron density of CD is also much weaker, which indicates the weaker interaction.

(See Figure 2.15 and Figure 2.16 for binding at binding site I )

 
   

viz.
TRP262 34", .—

‘ 'HE261 , 3'06
‘\ .

l... '

‘ .

ASN260
TYR502 ID

a“
'l

l

       
 
   
       
   
 
  

I
‘ 2.61
I

   

ASP258

    

 

ASN259

Figure 2. 15 The detailed interaction in binding site I of the A molecule with B-CD.
ASP258, ASN259, ASN260, TRP261, PHE262 on A molecule and TYR502 on D
molecule are involved in binding.

66

    
  
 

TYR502 / D

ASP258

TRP262

+2 ASN260

Figure 2. 16 The binding between B-CD and E. coli BE binding site I from top.

From Figures 2.15 and 2.16 it is clear that B-CD has more interaction with the
protein than a-CD does. In addition to the a-CD interactions, ASP258, ASN259 and
ASN260 are also involved in the hydrogen bonds. Moreover TYR502 from the D
molecule also binds with B-CD through a hydrogen bond. It might contribute to the

stabilization of the complex in the crystal.

67

There is no electron density for B-CD at binding site 111. After examination of all of
our B-CD bound E. coli BE structures, we only found very weak and incomplete electron

density at this binding site in 2 data sets (out of 21 data sets collected).

2.4.3 Three dimensional structure of E.coli BE] 7- cyclodextrin complex

y-cyclodextrin (7- CD) is the largest natural cyclodextrin; it also has the highest
solubility of the three natural cyclodextrins. So it is easy to prepare a high concentration

substrate solution. Also from the SBEI / cyclodextrin binding assay, y- CD has the lowest

Km value (0.00067 mM)(18). All of this indicated that perhaps 7- CD was a good choice

for soaking experiments. In our soaking experiments, 1 mM, 10 mM and 70 mM y- CD
solutions were used. Soaking time ranged from 30 minutes to 24 hours (beyond 24 hrs,
the crystals cracked completely). The substrate bound data set was obtained by soaking
E. coli BE apo crystals in 70 mM 7- CD for 15 hrs.

y- CD only binds with E.coli BE at binding site II, III and IV. At binding site II, the
binding residues include THRSOS, PHE509, LEU512 and TRP628. These residues have
direct contact with y- CD through hydrogen bonds.

Compared with a-CD and B-CD, 7- CD is larger, the glucose ring is much
looser/flexible, so the electron density of the non-binding glucose units are not as rigid as

in a-CD and B-CD.

68

Figure 2.17, Figure 2.18 and Figure 2.19 show the detailed interactions at binding

site II, III and IV respectively.

   

TRP628

THR508

TRP628

 

Figure 2. l7 7- CD binds with E.coli BE at binding site II. THR508, PHE509, LEU512
and TRP628 are interacting with CD by hydrogen bonds. The top figure shows the 1.0 o

2F o—Fc electron density map of the CD and the bottom figure shows the detailed
interaction.

69

  
   
 

TRP159

LYS189

PR0214
GLN211

Figure 2. 18 y- CD binds with E. coli BE at binding site 111. LYSl89, GLN211,
PR0214, GLU215 and TRP159 interact with 7- CD through hydrogen bonds and
aromatic stacking.

7O

 

2.69

   
  
 

ASP542

SER689

PR0659

GLN545

Figure 2. 19 y- CD binds with E. coli BE at binding site IV. ASP542, GLN545, PRO659,
SER689 and TRP544 bind with 7- CD by hydrogen bonds and aromatic stacking.

71

2.4.5 Overlay of three cyclodextrin bound BE structures

a —CD, [3 —CD and 7 —CD have similar structures and chemical properties. The only
difference is the size of the rings and the curvatures. From the previous discussion, we
already know that they have similar binding patterns with the protein. The different sizes
and geometry of the CD molecules result in slightly different interactions. For example, 7
-CD interacts with more residues than the other 2 CDs at binding site II. These
differences may or may not affect the binding strength.

The overlaying of three CD bound E. coli BE structures did give us some hint about
the substrate preference of E. coli BE. We first overlaid the three structures at binding site

II. (Figure2.20 and Figure 2.21)

  

TRP628

\J/‘g ‘\ \
g k ASP505

Figure 2. 20 Overlay of a—CD, B-CD, y-CD and their bound residues at binding site 11.
Green color represents a—CD and its bound residues; blue represents B-CD and magenta
for y—CD '

72

' . ' ‘ TRP628
PHE509 m

Figure 2. 21 Overlay of a—CD, B-CD, y-CD and their bound residues at binding site 11.
Green color represents a—CD and its bound residues; blue represents B-CD and magenta
stands for y-CD

 

From the figures above, it is clear that the residues bound to the CDs remain in
almost identical conformations. The two glucose units on TRP628 overlay very well
among the three structures. This side is the major hydrogen bond source. As to the
relative position between LEU512 and cyclodextrins, LEU512 itself did not adopt any
conformational change. It remains at the same position in all three structures. But when
the CD rings become bigger (from a—CD to B-CD to y-CD), the CDs no longer sit on top
of LEU512, they tend to include LEUS 12 further inside the hydrophobic pocket. This can

be seen from Figure 2.21 more clearly.

73

          

GLN21 1
TRP159
LYS189

Figure 2. 22 Overlay of a—CD, y-CD and their bound residues at binding site HI. Green
color represents a—CD and its bound residues; and magenta stands for y-CD. There is no
B-CD bound to this binding site. So we use blue color to represent unbound structure.

From Figure 2.22 we can see at binding site 111, a—CD and y-CD binds with the
protein similarly. The side chains of GLU215 participate in the interaction to stabilize
both the CD and the disordered 213—215 loop although their side chains adopt different
conformations. We were barely able to assign the GLU215 residue in the disordered
loop in the unbound structure. But the position of its side chain is different from the

a—CD and y-CD bound structures since there is no interaction to lock its position. As a

74

result, the B factor of GLU215 indicates it is very flexible. Also from the overlaid
structures, we can see that PR0214 blocks y-CD from moving towards GLN211. Figure
2.23 shows the three different CD-bound structures overlaid at binding site IV. We can
see almost all the bound residues remain in the same position. The two glucose units

closest to TRP544 also overlap completely.

 

ASP542 GLN545

Figure 2. 23 Overlay of a—CD, B-CD, y-CD and their bound residues at binding site IV.
Green color represents a—CD and its bound residues; blue represents B-CD and magenta
stands for y-CD and its residues.

75

The last binding site we overlaid is binding site I. This site has strong binding with a
—CD, relatively weak binding with B —CD and no binding with 7 —CD. This trend seems

related to the size of CDs.

TYR502 / D

    

TRP262

Figure 2. 24 Overlay of a—CD, B-CD and their bound residues at binding site I. Green
color represents a—CD and its bound residues; blue represents B-CD

Binding site I binds with a—CD and B-CD. Among the 4 molecules in the asymmetric
unit, the binding site on the B molecule cannot bind with CD because the space is
blocked by a symmetric molecule. The same binding site on the A, C and D molecules is

still open.

76

After examination of all of our data, we found that both a—CD and B-CD bind with
binding site I on molecule A, but 'y-CD does not bind. In order to explain this discrepancy,
we overlaid the a—CD and B—CD bound structures and discovered the reason why 7 —CD
is not found on molecule A. From Figure 2.24 we can see, binding site I holds a —CD
like a tweezers. The side chains of TRP 262 and PHE261 are two tips of the tweezers.
The benzyl group side chain of PHE261 probes into the hydrophobic inner cavity of CD.
Part of the CD lies in the middle of TRP262 and PHE261, with its position quite fixed.
So the bigger CD has to extend away from the “tweezers”. And in that‘direction, we can

see the TYR502 from the D molecule.

  

PHE261

   

TYR502 / D

 

TRP262

Figure 2. 25 Overlay of a—CD, B-CD and their bound residues at binding site I. Also a
free 7 —CD molecule is modeled on binding site. Green color represents a—CD and its
bound residues; blue represents B-CD and its bound residues. Magenta represents a free 7

—CD molecule.

77

a —CD has no contact with the D molecule, since it is the smallest CD. But [3 —CD
already has interaction with TYR502 through a hydrogen bond (2.64 A). Figure 2.25
shows the overlay of a —CD and [3 —CD bound to binding site I on E.coli BE. We also
modeled one 7 —CD onto the binding site, with the right part of the CD (2 glucose units)
positioning between side chain of TRP262 and PHE261. It is clearly seen from the
overlaid structure that there is a clash between 7 -CD and molecule D of E. coli BE (1.44
A). This is probably the reason why 7 —CD can not bind with this binding site by the
soaking method. As we mentioned before, E. coli BE assumes a monomer state in solution
(in viva / in vitro situation), but forms a tetramer in its crystalline state. This indicates that
7 —CD can not bind with this binding site of the BE in this crystal packing pattern. But it
is possible 7 —CD binds with E. coli BE binding site I in the solution since there will no be
clash due to crystal packing.

In our B —CD bound structure; we can see very weak electron density at this binding
site in the C and D molecules, but no density at all in the 7 -CD bound structures. The
reason is still unknown. Because of crystal packing, binding site I on molecule B is
completely blocked by symmetry molecules, so neither [3 —CD nor 7 —CD can bind to this

site on molecule B.

78

2.4.6 Overall binding of cyclodextrins with E.coli BE

In order to show the overall binding site distribution, we overlaid the 4 molecules of
the a-CD bound structure , and superimposed all the 4 bound a-CD onto one molecule
( Figure 2.26) ( in our crystal structures, they never bind to the same molecule: binding
site 111 is only on molecule D and binding site IV only on molecule C).

From the overall composite figure, we can see that all the binding is on the surface of
the protein, the binding sites are distributed around the active site, (a / [3) barrel region. It
is clear the function of these binding sites is to hold the substrate. Whether they are
involved in concerted catalysis action is unknown. But we believe that the natural
substrate of BE will undergo some conformational change before the hydrolysis catalysis
step.

The active site region is filled with the side chains of BE residues. Also the diameter
of the double helix amylose (10.6 A) and a-CD (13.7 A) (9-10) are quite big compared to
the BE active site barrel (diameter of the barrel is about 10-11 A), so we believe that the
natural substrate has no access to the active site in its double helix state. There must be an
unwinding step before the catalysis step. Also acarbose was used to model the
polysaccharide inside the active site. This led us to imagine that the catalysis involves
both double helix state substrate and linear oligosaccharide binding.

So the next part we are going to investigate is the binding between E.coli BE and

linear oligosaccharides.

79

.85: E 83838 5.88835 Ea 83.88 2: 5953 3ng 588% BE. BR 038m 0% SE BEBE
8? E 2322: 38.88 5.. mam—EB =So>o 05 0852: 8 0 25208 Boo womomfitomsm 22> mmoé v =« 25 2385 89$

8:83pm @569 00-5 mo $30208 v =< .mfibxouofizoé 8 canon mm meow mo $5836 3900 05 mo 86388 <

i _ 0% 96:5

 

>_ an 8.9.5

.Eofiz

4 __ 23 96:5
ob

s.
x I. c
I

ogm o>=om 089.com

3 .N 2:5

80

2.5 Three dimensional structures of E.coli BEllinear

oligosaccharides complexes

Linear oligosaccharidess used in the research include: maltose, maltotriose,
maltotetraose, maltopentaose, maltohexaose and maltoheptaose.

Different protein / oligosaccharide ratios were tried in the co-crystallization
experiments. Only apo crystals were grown in the maltose solution. The co-crystallization
experiments with other oligosaccharides failed and no crystals have formed.

In substrate soaking experiments, all oligosaccharides were dissolved into the initial
crystallization solution (100 mM Na-HEPES, pH7.2). Apo E.coli BE protein crystals
were utilized for soaking experiments. It was found that soaking in all oligosaccharides
except maltose caused the protein crystals to crack (visual inspection under microscope).
In order to make sure the observation was accurate and neutral, A control experiment was
performed (soaking was carried out in pure initial crystallization buffer without
oligosaccharides, and even in pure water, absolutely no cracking happened.) This is an
important indication that the oligosaccharides do interact with E. coli BE crystals. Given
the behavior of maltose in the co-crystallization and soaking experiment, we believe that
maltose is not binding with the E. coli BE protein.

At least 20 crystals were soaked with maltotriose solution (from lmM to 45 mM ) for
various times( from 30 minutes to 12 hrs). The diffraction data of the crystals were
examined and no electron density of maltotriose was found. But considering its ability to
“crack” crystals in the soaking, which indicated some kind of interaction between

maltotriose and the protein crystal, it is possible that maltotriose does interact with E. coli

81

BE, but due to binding compromise at other binding sites, the binding of maltotriose is
not sufficient before the crystal lattice was destroyed: so the electron density of
maltotriose is not seen.

We did find some low quality electron density maps of one or two glucose units of
maltotetraose and maltopentaose in the data. But the density is not good enough to
construct the corresponding oligosaccharide molecules. The binding sites they bind with
are also found in the maltohexaose (M6) and maltoheptaose(M7) bound structures, so _
we finally decided to use the maltohexaose and maltoheptaose bound structures to
illustrate the interaction.

Maltohexaose and maltoheptaose bind with E. coli BE at 5 binding sites. Table 2.5

lists the residues involved

 

 

Binding site Bound chain E.coli BE residues involved in the binding
III(on D chain) M6 and M7 TRP159, LYSl89, LEU201,GLN211 and GLU215
lV(on C chain) M6 and M7 ASP542, TRP544, GLN545, PRO659 and SER689
V (on ABCD chains) M6 and M7 ARG255, SER583, ASP585, H18587 and GLU59O
Vl(on B chains) M7 LYSS46, TRP595 and I-IIS596

VII(on A B chain) M7 PRO469, GLY476, TRP478, ASN518 and PHE477

 

Table 2. 4 Binding sites in E. coli BE /linear oligosacchardie complex structures

82

Among the 5 binding sites, 111 and IV are the same as the binding sites in E. coli BE
/CD complex structures. The binding sites can only immobilize two glucose units. The
rest of the linear chain is apparently disordered and can not be seen in the electron density
map. This is a big difference between linear and cyclic oligosaccharides. (CDs are
somewhat rigid. even though only 2 glucose units bind with the protein, the rigidity of the
ring fixes the rest of the ring, So most parts of the cyclic oligosaccharides appear in the
electron density maps.)

Linear oligosaccharides have 3 unique binding sites compared with cyclodextrins.

Those are binding site V, VI and VII. We will discuss these sites in detail.

2.5.1 Three dimensional structure of E.coli BE I maltoheptaose complex
Maltoheptaose (M7) is the longest oligosaccharide we used. It binds with E. coli BE.
And its binding mode on the surface of the protein revealed some very interesting aspects
of the BE mechanism.
We discovered that the non-reducing end of maltoheptaose binds with E. coli BE at

binding site V and the reducing end binds with binding site VI (only on molecule B).

Figure 2.27 shows the F0 —Fc and 2FO —Fc electron density maps calculated at 0.6 and

2.0 0 respectively. The maps depict the electron density of the oligocaccharide at the
binding sites. Figure 2.28 and 2.29 show the detailed interaction between maltoheptaose

and E. coli BE.

83

              

AR6255
2d

       

                 

   

     

     

     

2 V
+ . 1),». V M
\. . .’ A . e s
. annmws. . Hm”. m .m m.
x .111 Xxx. -. .«i ...»..o p... A m , . . r... m
. . . , . . --l I. . ....vn. i i
zany... .. ...... L. n. a... m knit... ._ "an. s . ... ... b
n. . in” o......n;.s,.um... . .7 Flat». ... ....vu. ...“.mwmc
. , I. . . I ‘ V y .r 4.\.¢v'I./.
. '\ thin—v l :l

GLU590

       

  

 

           
   

             

., . .516»,
antl- .\Ar’. 7 ... trig},
....w ...“.- ..Esfl. .1.-5..
7 ,0 ...“..b m0.. ham?!
a LEV , i Q
W. 5.21% a...» .....a
C2. .. u C I ’ I,
.....q seizes,

        

5‘

 

binding site VI

GLUSW

TRPSQS

-Fc electron density map of M7 (calculated at 2.0 c)

Figure 2. 27 The top figure shows the F0

bound at binding site V and VI on molecule B.; The bottom figure shows the 0.6 o 2FO-Fc

electron density map of the M7.

84

 

.5 2m mamas a? WEE e8 was? 2: Ea > as macs; as, mean 3 .8 ED 38828 2: .2322: m mm 2: 2a :2
3953 nowofiowfi wozfiou 05 885mg: Saw—m 2F .H> 98 > moan $553 3 mm tonne HEB £55 3833832 mud 953m

 

83:0
> 25 96:5

mwmdmw

85

.Emno ougooamowzo 95 53> $2285 onwa? :03 on 58 :

Emma 5528 80¢ 530835 2: @505 05mm we .3 can > ozm wficfin “a mm 28M 5? mafia omofiaonozuz aufi 9.sz

     

mnNOm<

hammm<

 

noon—Fr

_> 3m 9.95 . . .. > gm 9.25

86

A linear oligosaccharide has one non-reducing end and a reducing end. Both ends of
maltoheptaose bind with the E. coli BE protein on the surface. From our structure, we can
see that the first two glucose units ( +1, +2) bind with binding site V through extensive
hydrogen bonds with residues ARG255, GLU590, HIS587 and ASP585. The reducing
end interacts with ASP537, LYSS46, TRP595 and HISS96.

The whole linear oligosaccharide forms a “U” shape “jump rope” with binding site V
and V] as two ends, and the middle of the rope was jacked up by the side chain of
ARG576.

ARG576 plays an interesting role in the binding. The side chain of ARG576 makes
hydrogen bonds with hydroxyl groups of the +3 and +4 glucose units. It stabilizes the
middle of the loop so that it is not very flexible. At the same time, it prevents the middle
of the oligosaccharide from contacting the surface of the enzyme. Also because of this
orientation, it makes the whole oligosaccharide extend to its maximal length. When the
maltoheptaose binds with the surface of the protein, several glucose units rotate along the
(it-1,4 link. Also distortion inside the glucose ring was observed: the conformation
deviates from the most energy favorable chair conformation (please see the glucose ring
in previous figures). But the “U” shape jump rope binding is not usual in
polysaccharide-protein binding, twisting the chain along such a short distance ( 7 glucose
units) is likely to cause some units to adopt unusual conformations. Also we believe the
whole linear chain is under extra stress caused by fully extending to its maximal length

(as we mentioned before, the binding on sites V and V1 with interaction of ARGS76 make

87

maltoheptaose fully extend to its maximal extent.)

In our structure, we only found whole maltoheptaose bound to the B molecule. In the
A and D molecules, only the non-reducing ends bound at binding site V were found. We
can see two glucose units at binding site V in the A molecule, and four glucose units in
the D molecule. The rest of the oligosaccharide chain remains flexible, so it is not
detected in X-ray diffraction data. Also there is no binding in the C molecule. Figure

2.30 shows the binding in molecule A in which two glucose units are visible.

H|S587

ASP585

  

235. ," AR<3576

2.46 I : , ' 3.21

 
 

Figure 2. 30 Maltoheptaose binds with E. coli BE at binding site V in the A molecule.
Only two glucose units are visible.

88

Binding site VII only exists in the A and B molecules. Because of the flexibility of

maltoheptaose , only two glucose units are visible in binding(See Figure 2.31).

     
 

TRP478
PHE239

PR0469

PHE477
TRP478

PR0469

PHE239 ARG468

PHE621

Figure 2. 31 Maltoheptaose binds with E. coli BE at binding site VII. Two figures show
the interaction from different angles.

89

The reducing end of M7 binds with this binding site. TRP478 and PRO469 are
parallel to each other, forming an inclusive pocket with PHE477 as well. Glucose +7 was
anchored inside the pocket. The hydrophobic interaction and hydrogen bonds help to
stabilize glucose units. Glucose +6 become flexible although it has a hydrogen bond with
ASN518. The ideal orientation of glucose +7 is to be parallel to the side chains of
TRP478 and PRO469, but because of the steric hindrance from PHE239 and PHE621,
glucose +7’s position was a little distorted. Also glucose +6 turns out of the pocket in a
skewed way, leading the rest of the oligosaccharide chain away from the surface of BE.
+1-+5 glucose units can not be seen because they are flexible. This binding site looks like
a rail docking station: side chain of TRP478 and PRO469 act as the two rails, leading the
glucose inside by aromatic/hydrophobic interactions.

Due to crystal packing, the same space on the C and D molecules was occupied by
side chains of other symmetric molecules. So the binding is only seen at the A and B
molecules.

Similar to CDs, linear oligosaccharides also bind with the E. coli BE at binding sites
111 and IV. Figure 2.32 and Figure 2.33 illustrate the interaction. Binding site III involves
residues TRP159, GLN211, GLU215 and LY8189. This is almost the same as binding
between E.coli BE and CD. Only two glucose units are visible in the electron density
map. Which two glucose units bind with the protein are unknown. The aromatic stacking
force between the indole side chain of TRP1 59 and glucose units, together with hydrogen

bonds between the hydroxyl group of glucose and protein stabilize the oligosaccharide on

90

the surface of protein.

GLU215

   

GLN21 1

LYS189

Figure 2. 32 Maltoheptaose binds with E. coli BE at binding site III.

PR0659

 

2.71

PR0661

Figure 2. 33 Maltoheptaose binds with E. coli BE at binding site IV.

91

 

. Fig

Similarly to the binding at binding site III, the combination of hydrogen bonds and
aromatic stacking immobilizes two glucose units on the surface of the protein. The
interaction at this binding site only anchors two glucose units. Since both sides of the
binding site are quite open, these two glucose units could be any two consecutive units in
the maltoheptaose chain.

From the composite view of the B molecule of the M7-bound structure, we can see
the “jump rope” is pointing toward the active site cavity, and binding site VII is on the

other side of the molecule near the bottom of the (a /B) barrel region (Figure 2.34).

       
   
    

active site binding site V

N-terminal

- A
\ 1
binding site VII ' C—terminal

Figure 2. 34 A composite view of the B molecule of E. coli BE /M7 structure.

92

2.5.2 Three dimensional structure of E.coli BE I maltohexaose complex

Maltohexaose(M6) binds with E. coli BE protein in a pattern similar to that of
maltoheptaose. The binding also happens at five binding sites.

At binding site V, usually only one glucose unit is visible in the A molecule and two
glucose units were indentified in the B, C and D molecules. The binding patterns are
similar to those in the M7-bound structure.

In another M6 bound structure, we found very poor electron density for the entire M6
chain bound at binding sites V and VI, similar to the M7-bound structure. We tried to
build M6 at the binding sites. It indicates the binding at both site V and VI are distorted.
At binding site V only one glucose has interaction with a protein residue; and at binding
site VI, the only bound glucose unit inclines to slide out of the binding site. Since the
density is too poor to construct an acceptable molecule, we only keep the glucose units
bound at the binding sites which have good density in our refined structure. But this still
revealed a quite interesting discovery. Combined with the M7-bound structure, we
believe M7 is the shortest oligosaccharide chain that can simultaneously bind site V and
VI at the same time. M6 is just a little short: but it still shows some binding. But
oligosaccharides shorter than M6 never showed density for the entire chain in the data
sets.

In the M6 bound structure, we did not see binding at binding site VII. The binding at
site 111 and IV are similar to those in the M7-bound structure.

After examining all data sets, we believe that except binding at binding site V and VI

93

in the B molecule, M6 and M7 bind with E. coli BE in exactly the same way: part of the
oligosaccharide chain (usually two glucose units) binds with BB surface residues, and the
rest of the chain remains flexible outside the BE molecule. Thus in this case, the one
glucose difference between M6 and M7 actually made no difference with regard to
binding.

Figure 2.35 — Figure 2.37 shows the detailed interaction at difierent binding sites

H|5587

I

ASP585 {2.4 . 3.02

 

ARG255

Figure 2. 35 Maltohexaose binds with E. coli BE at binding site V in the C molecule.

94

     

TRP159 GLN155

LYS189

GLN211

Figure 2. 36 Maltohexaose binds with E. coli BE at binding site III in the D molecule.

 

SER689

Figure 2. 37 Maltohexaose binds with E.coli BE at binding site IV in the C molecule.

95

At the binding sites, the oligosaccharide was immobilized on the surface of protein
by combination of aromatic stacking and hydrogen bonds. This interaction is very similar

to those in the BE/CD complex.

2.5.3 Overlay of E.coli BEla-CD structure and E.coli BEIM7 structure
We overlaid the binding sites both BE/CD and BE/M7 share: binding site III and IV.
As we indicated in the BE/CD structure section, E.coli BE actually only interacts with

two glucose units of the cyclodextrin.

      
 

GLN21 1

. ’7

  
 
 

. GLU215
\ \/

emrss \.
CO.“ ~
TRP159 \ '

mil

Figure 2. 38 The overlay of BE/u-CD and BE/M7 structures at binding site 111. a-CD
and its bound residue are represented in green , while M7 and its bound residue are in
magenta.

96

Figure 2.38 shows the difference between a-CD and M7 bound structures at binding
site III. In the figure, we can see the difference: first the C6 hydroxy group of M7 glucose
is pointing left, while the same hydroxyl group in a-CD is in the opposite direction. All
the CDs assume the same orientation at this site since they can adopt a more relaxed
conformation at this site with more hydrogen bonds; linear oligosaccharide did not adopt
the same orientation as CDs, but we believe there is no big difference in terms of energy.

Also the bound residues GLNISS, GLU215 and LYSl89 adopt different
conformations when they bind with different substrates. It seems that this is caused by the
opposite orientation of the sugar units. The hydrogen bonds between the hydroxyl groups
of the sugars and protein residues cause the side chains of the residues to move. The
major contributor of aromatic stacking, TRP159, overlaid perfectly. This residue keeps
parallel to the sugar ring plane, interacting with the sugar through aromatic stacking.

Figure 2.39 depicts the overlay of BE/a-CD and BE/M7 structures. From the
overlaid structures we can see the bound residue at the binding site remains in the same
position, and the two glucose units bound to protein overlap perfectly. This indicates that
the interaction between the oligosaccharide and this binding site are mainly on the two
glucose units. From the figure it is clear that the glucose units and bound residues are
almost identical except a minor rotamer change of one C6 hydroxyl group. Also the small
side chain conformational change of ASP542 does not affect the hydrogen bond with the

hydroxyl group of the oligosaccharide.

97

 

Figure 2. 39 The overlay of BE/a-CD and BE/M7 structures at binding site IV. a-CD
and its bound residue are represented in green, while M7 and its bound residue are in
magenta

The overlaid structures showed that binding site III and IV are ready to bind both
cyclic and linear oligosaccharides without need of extra energy. These two sites are on
the surface of protein. It is very likely the natural substrate for BB will keep binding at
this site at different catalytic stages in which the substrate might be either naturally

double helical or unwound to a linear state.

98

2.5.4 Overall composite of E.coli BE I oligosaccharide binding structure
The overall composite binding figure shows the distribution of CD and

oligosaccharide binding sites.

      
   
    
    

binding site VIl

(Linear only)

active site

/
binding site VI
(Linear only)
binding site V ,
(Linear only) binding site i (CD only)

Figure 2. 40 The overall composite binding sites on the E.coli BE surface. All 4
molecules of (ii-CD and 1 molecule of M7 bound structures were overlaid. All 4 a-CDs, 1
M7 and one part of M7 (two glucose units) were superimposed onto one molecule to
illustrate the overall binding. The distance between the CDs and M7 molecules are listed.

99

2.6 Conclusion and hypothesis

In our work, a set of binding sites on the surface of E. coli BE was discovered for the
first time. Also the binding site’s preference for substrate was tested by different substrate
soaking experiments. Four binding sites were found for CD5 (binding site I-IV), and they
in generally bind all CDs similarly. Among the four binding sites, two sites are shared
with linear oligosaccharides (binding site 111 and IV) although binding site 111 bind with
the linear oligosaccharide differently than it does with CDs. Linear oligosaccharides
occupy three unique binding sites (binding site V-VII) in which two of them are found to
bind respectively with the non-reducing and reducing end of maltoheptaose.

All the detailed interactions between substrates and local bound residues of E. coli BB
were depicted. By identifying the residues involved, further site directed mutagenesis

becomes feasible to determine which sites are the most necessary for the activity.

2.6.1 Hypothesis

Afier determining the three dimensional structures of CD and linear oligosaccharide
bound E. coli BE complexes, we realized that none of the binding took place around the
active site. We know the CDs are too big to access the active site cavity, but the fact that
there was no linear oligosaccharide bound to the active site surprised us. Considering that
we did extensive soaking experiments under different conditions, and collected many
diffraction data sets, we can conclude safely that the substrates we used so far (from

Maltose to Maltoheptaose) can not bind tightly to the catalytic residues in the active site.

100

From the literature, we know E.coli BE prefers to transfer chains about 10-12 glucose
units in length (19, 30). So we were interested in how to connect this result to our
discovery of surface binding sites on the E. coli BE protein.

We compared structures of pseudo-maltononaose bound maltohexaose-producing
amylase(5)( PDB database # 1WPC), maltononaose bound cyclodextrin
glycosyltransferase from Bacillus Circulans(6) (PDB data base # 1CXK), acarbose and
oligosaccharide bound alpha-amylase(3I-32) (PDB database # lHXO and 1UA3 ) and
maltohexaose-producing amylase from Bacillus Circulans (33)(PDB database# 2D3N).
These enzymes belong to the glycosyl hydrolase GH13 super family and share similar
overall structures. The structures have oligosaccharide or pseudo-oligosaccharide bound
at or near the active sites that are highly conserved. Among them, there is even a
9-glucose chain bound at the active site. The active sites are very similar within this
family. And from the overlaid structure, we found the substrates also overlaid perfectly.

Their orientations and contour of the chains are similar (see Figure 2.41 top figure).

101

substrate

active site

 

loop of d-amylase

    

active site

Figure 2. 41 Overlay of 1WPC (green), IHXO (blue), 1M7X (magenta) and 1UA3
(yellow). The top figure shows the overlay of all the enzymes and substrates. The bottom
figure shows the close-up of the overlay of 1UA3 (tr-amylase), 1M7X (E. coli BE) and
substrates at the active site.

1WPC: pseudo-maltononaose bound maltohexaose-producing amylase

IHXO: truncated acarbose bound alpha-amylase

1M7X: E. coli BE protein without any substrate

1UA3: oligosaccharide bound alpha-amylase

102

From the overlaid structure we can see the central (ii/B barrel region is conserved
among the proteins. All the substrates follow the same orientation. Since there is no
substrate for E.coli BE, a truncated acarbose was modeled into the E.coli BE active
site(3). We also overlaid the acarbose model with 1M7X and other structures in Figure
2.41, the result proved our hypothesis: the modeled acarbose also superimposed with the
other substrates seamlessly.

From all the information gathered, we believe that inside the active site the substrate
for E.coli BE should also adopt the orientation that other substrates adopted in Figure
2.41. Also from the figure it is clear that the BE does not have the substrate binding loops
of the other enzymes (such as iii-amylase) at the active site (see Figure 2.41 bottom
figure), so the substrate for BB does not need to overcome the loop obstacle to reach into
the active site. This is applied next in the modeling of a glucan chain into the active site.

We have found the binding site of linear oligosaccharide on the surface of E. coli BE,
and we know the orientation of the substrate inside the active site. Based on this
information, we modeled a glucan chain from the binding site into the active site of the
E. coli BE protein. We tried to build the glucan chain over the surface of protein without
any clashes with the surface residues. The glucan chain starts from the non-reducing end
of M7 (binding site I of M7 bound structure), and ends up overlapping with the modeled

acarbose in the active site (See Figure 2.42).

103

I
I
ll

F i

Sll

To

lei;

Siai

Non-reducing
end of M7

 

Figure 2. 42 Oligosaccharide model (blue color) binds to binding site V and the active
site simultaneously. Maltoheptaose (M7) is depicted in yellow and acarbose in magenta.
Top figure shows the overall View of the model glucan over the surface of BB. Bottom
left shows the overlapping part of the glucan with acarbose. Bottom right shows the
starting point of the glucan: the non-reducing end of M7 (at binding site V)

104

In the figure, acarbose (in magenta color) was truncated, only keeping the part
overlapping with the glucan. The modeled glucan chain (in blue) starts from binding site
V, overlapping with the non-reducing end of M7. This glucan chain has 8 glucose units
(to the first overlapping glucose unit with acarbose), and three glucose units are needed
inside the active cavity (the truncated acarbose ) , so at least 11 glucose units are needed
to have a simultaneous binding at both binding site V and the active site. This is
consistent with the data that E. coli BE reacts with amylose with a minimum chain length
of 12 units and mostly transfer chain of 11 glucose units (19).

From all the information above, we proposed our hypothsis: We believe E. coli BE
binds with longer polysaccharide similiarly to that in Figure 2.42. The non-reducing end
of the chain binds with binding site V, and the chain forms a U shape “jump rope”, with
the middle of the rope sinking into the active site; And the reducing end binds loosely
with binding site VI. The cleavage happens inside the active site cavity, actually in the
middle of the chain. This is the reason why the chain has to be at least 12 units long to be
catalyzed. Binding site I fixes the chain on the surface of protein, while site V is only a
“sliding regulator” for the chain. In other words, when the chain is very long, the extra
part of the chain will be floppy outside binding site VI, leaving the appropriate length of
chain between site V and VI to make sure the tip of the jump rope interacts with the
active site properly.

Actually we observed that the binding site V1 is not strictly fixed, by comparing the

M6 and M7 bound structure, we realized that binding site V1 is an area (like a platform,

105

several adjacent residues can provide hydrogen bonds), the glucose anchored on it can
slide a little and still bind. Comparing this behavior with the rigid binding pattern at site V,
we concluded that site VI has two functions: first it regulates the length of the
oligosaccharide between site V and VI to make sure the chain can interact with surface
and active sites of protein comfortably; Secondly its binding with the non-reducing end
(actually maybe not “end”) provides extra stabilization for the whole polysaccharide.

We proposed this hypothesis based on our binding structure, and it did explain the
substrate specificity of E. coli BE: E. coli BE handles the chain transferring specificity by
separating in space the substrate binding site from the active site of the enzyme, with the
substrate making strong interactions with two sugars located 10-11 sugar units away from
the reaction. This is the first hypothesis on a pre-catalysis structure of branching enzyme

based on enzyme-substrate binding information.

2.6.2 Verification of the hypothesis

We know that short oligosaccharides such as acarbose can not bind with the E. coli
BE, while BAYe4609 (a polysaccharide molecule containing 17 or more glucose units) is
an effective inhibitor for the enzyme(34). This could be explained by our hypothesis:
BAYe4609 is long enough to bind with both binding site V and VI simultaneously and
interact with the catalytic residues in the active site while acarbose is too short.

Based on our hypothesis, if we could do substrate soaking experiments with

106

oligosaccharides longer than 10 glucose units (to be safe, we will try substrate with 10, 11,
12 units), very likely we will see glucose binding at binding site V, VI and inside the
active site. We are currently seeking homogeneous polysaccharides that have 10, 11, 12,
13 and 14 glucose units. The polysaccharides mentioned will be synthesized by our
collaborator Professor Xuefei Huang.

Also site directed mutagenesis will be carried out on critical binding residues of the

enzyme. The change in the branching activity will verify the role of each residue in the

branching action.

107

Reference cited

1. Mikkelsen, R., Binderup, K., and Preiss, J. (2001) Tyrosine residue 300 is
important for activity and stability of branching enzyme from Escherichia coli,
Archives of Biochemistry and Biophysics 385, 372-377.

2. Binderup, K., and Preiss, J. (1998) Glutamate-459 is important for Escherichia
coli branching enzyme activity, Biochemistry 3 7, 9033-9037.

3. Abad, M. C., Binderup, K., Rios-Steiner, J ., Ami, R. K., Preiss, J., and Geiger, J.
H. (2002) The X-ray crystallographic structure of Escherichia coli branching
enzyme, Journal of Biological Chemistry 277, 42164-42170.

4. Kagawa, M., Fujimoto, Z., Momma, M., Takase, K., and Mizuno, H. (2003)
Crystal structure of Bacillus subtilis alpha-amylase in complex with acarbose,
Journal of Bacteriology 185, 6981 -6984.

5. Kanai, R., Haga, K., Akiba, T., Yamane, K., and Harata, K. (2004) Biochemical
and crystallographic analyses of maltohexaose-producing amylase from
alkalophilic Bacillus sp 707, Biochemistry 43, 14047-14056.

6. Uitdehaag, J. C. M., Mosi, R., Kalk, K. H., van der Veen, B. A., Dijkhuizen, L.,
Withers, S. G, and Dijkstra, B. W. (1999) X-ray structures along the reaction
pathway of cyclodextrin glycosyltransferase elucidate catalysis in the
alpha-amylase family, Nature Structural Biology 6, 432-436.

7. Katsuya, Y., Mezaki, Y., Kubota, M., and Matsuura, Y. (1998) Three-dimensional
structure of Pseudomonas isoamylase at 2.2 angstrom resolution, Journal of
Molecular Biology 281, 885-897.

8. Sherrod, M. J. (1989) Exploration of Cyclomalto-Oligosaccharide (Cyclodextrin)
Chemistry with Molecular Mechanics - Docking Calculations on the
Complexation of Ferrocenes with Cyclodextrins, Carbohyd Res 192, 17-32.

9. Szejtli, J. (1998) Introduction and. general overview of cyclodextrin chemistry,

108

10.

ll.

12.

13.

14.

15.

16.‘

17.

18.

Chem Rev 98, 1743-1753.

Hinrichs, W., Buttner, G, Steifa, M., Betzel, C., Zabel, V., Pfannemuller, B., and
Saenger, W. (1987) An Amylose Antiparallel Double Helix at Atomic Resolution,
Science 238, 205-208.

Ball, 8., Guan, H. R, James, M., Myers, A., Keeling, P., Mouille, G, Buleon, A.,
Colonna, P., and Preiss, J. (1996) From glycogen to amylopectin: a model for the
biogenesis of the plant starch granule, Cell 86, 349-352.

Schulz, W., Sklenar, H., Hinrichs, W., and Saenger, W. (1993) The Structure of the
Left-Handed Antiparallel Amylose Double Helix - Theoretical-Studies,
Biopolymers 33, 363-375.

Popov, D., Buleon, A., Burghammer, M., Chanzy, H., Montesanti, N., Putaux, J.
L., Potocki-Veronese, G, and Riekel, C. (2009) Crystal Structure of A-amylose: A
Revisit from Synchrotron Microdiffraction Analysis of Single Crystals,
Macromolecules 42, 1167-1174.

Takahashi, Y., Kumano, T., and Nishikawa, S. (2004) Crystal structure of
B-amylose, Macromolecules 3 7, 6827-6832.

Sarko, A., and Marchess.Rh. (1966) Crystal Structure of Amylose Triacetate - a
Nonintegral Helix, Science 154, 1658-&.

Gessler, K., Uson, I., Takaha, T., Krauss, N., Smith, S. M., Okada, S., Sheldrick, G
M., and Saenger, W. (1999) V-amylose at atomic resolution: X-ray structure of a

cycloarnylose with 26 glucose residues (cyclomaltohexaicosaose), P Natl Acad
Sci USA 96, 4246-4251.

Skovvron, S. (2006) chemical structure of the three main types of cyclodextrin,
http://upload. wikimedia. org/wikipea’iw/commons/thumb/5/5 I/Cyclodextrin. 5112/80
Qm-Cycl odextrin. 5112,1921 g.

 

Blennow, A., Vikso-Nielsen, A., and Morell, M. K. (1998) alpha-glucan binding
of potato-tuber starch-branching enzyme I as determined by tryptophan

109

19.

20.

21.

22.

23.

24.

25.

26.

27.

fluorescence quenching, affinity electrophoresis and steady-state kinetics,
European Journal of Biochemistry 252, 331-338.

Guan, H. R, Li, R, ImparlRadosevich, J., Preiss, J., and Keeling, P. (1997)
Comparing the properties of Escherichia coli branching enzyme and maize
branching enzyme, Archives of Biochemistry and Biophysics 3 42, 92-98.

(1994) Collaborative Computational Project, No. 4. The CCP4 suite: programs
for protein crystallography., Acta C rysta. D50, 760-763.

Harata, K. (1979) Structure Chemistry of Cyclodextrin Complexes, J Jpn Soc
Starch Sci 26, 198-209.

Challa, R., Ahuja, A., Ali, J., and Khar, R. K. (2005) Cyclodextrins in drug
delivery: An updated review, Aaps Pharmscitech 6, .

Lofisson, T., and Brewster, M. E. (1996) Pharmaceutical applications of
cyclodextrins .1. Drug solubilization and stabilization, J Pharm Sci 85,
1017-1025.

Sharff, A. J., Rodseth, L. E., and Quiocho, F. A. (1993) Refined 1.8-Angstrom
Structure Reveals the Mode of Binding of Beta-Cyclodextrin to the Maltodextrin
Binding-Protein, Biochemistry 32, 10553-10559.

Knegtel, R. M. K., Strokopytov, B., Penninga, D., Faber, O. G, Rozeboom, H. J .,
Kalk, K. H., Dijkhuizen, L., and Dijkstra, B. W. (1995) Crystallographic Studies
of the Interaction of Cyclodextrin Glycosyltransferase from Bacillus-Circulans
Strain-251 with Natural Substrates and Products, Journal of Biological Chemistry
270, 29256-29264.

Parsiegla, G, Schmidt, A. K., and Schulz, G E. (1998) Substrate binding to a
cyclodextrin glycosyltransferase and mutations increasing the

gamma-cyclodextrin production, European Journal of Biochemistry 255,
710-717.

Uitdehaag, J. C. M., van Alebeek, G. J. W. M., van der Veen, B. A., Dijkhuizen, L.,

110

28.

29.

30.

31.

32.

33.

34.

and Dijkstra, B. W. (2000) Structures of maltohexaose and maltoheptaose bound
at the donor sites of cyclodextrin glycosyltransferase give insight into the

mechanisms of transglycosylation activity and cyclodextrin size specificity,
Biochemistry 39, 7772-7780.

van der Veen, B. A., Uitdehaag, J. C. M., Penninga, D., van Alebeek, G. J. W. M.,
Smith, L. M., Dijkstra, B. W., and Dijkhuizen, L. (2000) Rational design of
cyclodextrin glycosyltransferase from Bacillus circulans strain 251 to increase
alpha-cyclodextrin production, Journal of Molecular Biology 296, 1027-1038.

Boraston, A. B., Healey, M., Klassen, J ., F icko-Blean, E., van Bueren, A. L., and
Law, V. (2006) A structural and functional analysis of alpha-glucan recognition by
family 25 and 26 carbohydrate-binding modules reveals a conserved mode of
starch recognition, Journal of Biological Chemistry 281, 587-598.

Binderup, K., Mikkelsen, R., and Preiss, J. (2002) Truncation of the amino
terminus of branching enzyme changes its chain transfer pattern, Arch Biochem
Biophys 3 97, 279-285.

Qian, M. X., Nahourn, V., Bonicel, J., Bischoff, I-l., Henrissat, B., and Payan, F.
(2001) Enzyme-catalyzed condensation reaction in a mammalian alpha-amylase.
High-resolution structural analysis of an enzyme-inhibitor complex, Biochemistry
40, 7700-7709.

Payan, F., and Qian, M. X. (2003) Crystal structure of the pig pancreatic
alpha-amylase complexed with malto-oligosaccharides, Journal of Protein
Chemistry 22, 275-284.

Kanai, R., Haga, K., Akiba, T., Yamane, K., and Harata, K. (2006) Role of Trp140
at subsite-6 on the maltohexaose production of maltohexaose-producing amylase
from alkalophilic Bacillus sp.707, Protein Science 15, 468-477.

Libessart, N., Binderup, K., and Preiss, J. (1999) Bay e4609, a slow-binding
inhibitor of branching enzyme, Faseb J 13, A1350-A1350.

111

L

CHAPTER 3: OVEREXPERESSION, PURIFICATION AND
LIMIT PROTEOLYSIS 0F BRANCHING ENZYME ll FROM
MAIZE ENDOSPERM

3.1 Introduction to maize starch branching enzyme II

Starch branching enzyme (SBE) plays an important role in the bio-synthesis of
amylopectin (1-2). Using enzymatic methods to modify starch has been emerging for
decades. Research has focused on the bio-synthetic pathway enzymes (2-3). We discussed
the application and need of modified starch in Chapter 1, and discussed the three
dimensional structure of E. coli glycogen BE in complex with cyclic and linear
oligosaccharides. So starch branching enzyme obviously becomes the next research
object.

Starch branching enzyme has more isoforms (3-4) than glycogen branching enzyme.
And the isoforms have different properties such as branching efficiency, substrate
specificity and minimum chain length requirement (3, 5). Starch branching enzyme 11
from maize endosperm showed a different substrate specificity and binding pattern from
E. coli BE (6-7). It is known that the a-amylase family enzymes have similar central a/B
barrel domains. But the different properties mentioned above are obviously related to
structural differences. The subtle difference in their structures must play an important role.

Much biochemical research has been done without real structural information. Several

112

conserved residues have been identified to be necessary for catalysis: ASP386, GLU441,
ASP509(8) and ARG384(9-11) were found necessary for the branching activity. But
without structural information, we still do not have an in-depth understanding about their
detailed role in the branching function. So determination of the three dimensional
structure of SBEII from maize endosperm will be an important achievement in the
structure-mechanism study of starch biosynthetic pathway enzymes.

SBEII from maize has 728 amino acids, with a molecular weight of 84 KDa. The
coding sequence of maize SBEII was obtained(12-1 3) and inserted into T7-based vector
pET-23d (14). The isoforrn was over-expressed in E.coli (15); protein is purified by
DEAE-cellulose chromatography(15-16). It was found that by genetically introducing a
silent mutation in the maize SBEII coding sequence, the expression level increased
significantly(14).

In order to study the structure-function relationship of starch branching enzyme,
SBEII from maize endosperm was selected as the target protein. Our objective in this
research is to determine the three dimensional structure of SBEII by X-ray
crystallography, compare the difference between E. coli glycogen BE and maize Starch

BE to find out the reason why they have different branching specificities.

113

3.2 Materials and method

3.2.1 Protein over-expression and purification

A glycerol stock containing the plasmid of starch branching enzyme 11 from maize
was obtained from our collaborator, Dr. Jack Preiss. The glycerol stock was plated on an
agar plate containing ampicillin (100mg/L). a single colony was picked and incubated in
50ml L.B media with ampicillin (100mg /L) at 37 °C for overnight. Then media was
transferred into 1L fresh L.B media with ampicillin (100mg/L). Growing continued at
37°C until optical density at 600nm ( O.D600 ) reached 0.5. IPTG was induced into
media to final concentration of 0.5 mM. Then the culture was incubated for another 12
hrs at 25 °C before it was harvested (14-15). The protein purification was modified as
below:

The cell pellet from 6 L of media was washed with 50mM Tris-Acetate, pH 7.5,
50mM NaCl solution, and then resuspended in buffer A (50mM Tris-Acetate, pH 7.5,
lOmM EDTA, 5 mM DTT, 10% Glycerol), lysed by sonication and then centrifuged at
5000 rpm for 30 minutes. Then a 3.8M ammonium sulfate solution was added to the
supernatant to final concentration of 0.8 M. The mixture was then centrifuged and
precipitation was discarded. Next the 0.8 M supernatant was brought to 1.6 M by addition
of ammonium sulfate crystals. Lots of protein precipitated, the precipitation was

harvested by centrifugation. The protein pellet was then dissolved in a minimum amount

114

of buffer A, and dialyzed against buffer A at 4 °C overnight. The protein solution was
then applied to DEAE-fractogel column. The column was washed with Buffer A/water
(50:50), and then the protein was eluted with 0-400mM KCl gradient in buffer A. Protein
fractions were pooled together and SDS-PAGE was used to monitor the purity of the
protein. The protein eluted out mostly in the wash buffer. This was concentrated and
applied to the Source Q anion exchange column. Protein was eluted with 0-200 mM KCl
gradient in buffer A. The protein eluted out at about 50 mM KCl. The fractions
containing protein were concentrated to a minimum volume and loaded onto the size
exclusion column (SEC). Protein was eluted with buffer A. Pure protein eluted in a single
sharp peak. From the SEC chromatograph it is clear that the protein is a monomer in
solution. Bradford protein assay was used to measure the concentration of the final
protein solution; SDS-PAGE was used to monitor the purity and Mass Spectrometry was
used to verify the identity of the protein.

Figure 3.1 shows the size exclusion chromatograph and SDS-PAGE of the SBEII protein.

115

Elution volume (ml)
15 3O 45 60 75 90 105 120
l l 1 l I L l I

l I I
M.W standard (KDa) 158 44 17 1.3
27

25129

WW

WWW ” Win

100
75 —» SBEII

50

25 27 28 30

Figure 3. 1 Size exclusion chromatograph and SDS-PAGE of SBEII protein. SBEII
protein eluted out of SEC column in a single peak, and SDS-PAGE showed fractions #27

-# 30 (3 ml per fraction) are pure protein.

 

Figure 3. 2 Two final SBEII protein samples (I and 2) used in crystallization screens.

116

3.2.2 Crystallization screen on maize starch branching enzyme ll (SBEII)

The purified protein (see Figure 3.2) was used to set up a sparse matrix
crystallization screen at both room temperature and 4 °C using the hanging drop vapor
diffusion method. Extensive buffer conditions were attempted; Protein at different
concentrations was also tried, but after 10 months no crystals were found.

We believed that the full length SBEII was not prone to crystallize. A different
strategy is needed to solve this problem.

In the structural study of E. coli BE, a similar situation was encountered (17). The full
length wild type protein did not crystallize, but deleting a significant segment from the
N-terminal of the protein resulted in successful crystallization. Limit proteolysis was used
to identify the new truncated version of the protein. We tried to use limit proteolysis to
investigate the SBEII from maize and found there was a major product after proteolysis
with Trypsin. On SDS-PAGE, a protein band corresponding to about 55 KDa was found

after limit proteolysis (see Figure 3.3).

117

SBEII
75KDa _ / 75KDa

50KDa ' . I ~55KDa 50KDa .... ...- cutting product

M 0.5h 1h 2h 3h 4h M 1 2 3.

...... —. -—-—- cub-i“

Figure 3. 3 Trypsin limit proteolysis test on SBEII at 4 °C.
Lefi gel sample shows incomplete cutting with Protein/ Trypsin = 1 mg / 5 units
Right gel sample shows three complete cutting sample (Protein /Trypsin = 1 mg / 7 units)

From the experiment, we found a major product with molecular weight ~55KDa. By
controlling the Trypsin protein ratio and digestion time, we obtained a small amount of
cut protein. The optimal proteolysis condition is using 7 units Trypsin on 1 mg protein,
cutting at 4 °C for 2 hours.

After Trypsin proteolysis, PMSF was added to stop the enzymatic digestion. And the
protein solution was concentrated and loaded onto the SEC column. Protein was eluted

out with buffer B (50mM Tris-Acetate, pH7.5, 2.5 mM DTT, 10% Glycerol).

118

Elution volume (ml)

 

30 60 90 120
I l L l
l l l I
M.W standard ( KDa) 158 44 17 1.3

 

» i
WWW 1

cut product

75 - /1\

55 ..
M 30313233 34

Figure 3. 4 Size exclusion chromatograph and SDS-PAGE of proteolysis product of
SBEII protein. The proteolysis was eluted out at the tail of the peak. Its molecular weight
is about 55 KDa.

The proteolysis yield was very low: we lost about 80% of the protein in the process.
From many similar tests, the recovery yield of the truncated protein is never higher than
30%. This made the direct cutting-purification method very impractical.

We also found the proteolysis process is progressive for SBEII protein. Another low

molecular weight cutting product was produced when we used more Trypsin in the

119

experiment. In a test 10 units Trypsin was added to a 1 mg protein solution. The
proteolysis result showed another protein band (corresponding to 30KDa) appeared after
0.5hr (see Figure 3.5). This protein is the secondary cutting product of SBEII. From the
gel sample, it is clear that the second band’s intensity increases (~3OKDa protein) and at
the same time the first band (~55KDa) faded accordingly. This indicated that the second
protein is the proteolysis product of the first protein. It is not the byproduct of the first
step cutting. The secondary proteolysis product is about 30 KDa, which indicates it is a
relatively small piece of the wild type protein. It is not as important as the first product
because the second product is too small to contain the active site ( the a/B barrel domain).
Furthermore the existence of this overcutting issue makes controlled-digestion very
difficult. Given this situation, we have to develop another strategy to obtain the 55 KDa

truncated SBEII protein.

 

75 . ...
’ 50 —> ~55
37

_. ~30
25

M 0hr 0.5 1h 2h 3h 4h

Figure 3. 5 SDS-PAGE of a proteolysis experiment. Protein / Trypsin= 1 mg / 10 units.

120

3.3 Further experiment plan

Since the direct cutting-purification method did not work as we expected, we decided
to use protein engineering technique to sub-clone the coding sequence of the truncated
protein into appropriate T7 vector. In order to do identify the Trypsin digesting site,
N-terminal and C-terminal protein sequencing needs to be done. The major proteolysis
product (55KDa) will be sent for protein sequencing analysis to find out the sequence of
the truncated protein. Then we will subclone the coding sequence into several vectors and
over-express them in an E. coli strain. After that the standard purification method will be
also be used to purify the protein. Upon getting pure protein, a crystallization screen will
be pursued. X-ray diffraction data will be collected on good protein crystals. The three
dimensional structure will be determined by molecular replacement.

Although SBEII from maize is only 25% identical to E. coli BE in sequence. But as a
member of the iii-amylase family, it will share a similar overall structure. So the molecular
replacement using E. coli BE as model is feasible. When the three dimensional structure
of the apo enzyme is obtained, further investigation of its substrate specificity and

substrate binding will be planned.

121

Reference cited

1. Hodges, H. F., Creech, R. G, and Loerch, J. D. (1969) Biosynthesis of
Phytoglycogen in Maize Endosperm Branching Enzyme, Biochimica Et
Biophysica Acta 185, 70-&.

2. Fisher, D. K., Boyer, C. D., and Hannah, L. C. (1993) Starch Branching
Enzyme-Ii from Maize Endosperm, Plant Physiology 102, 1045-1046.

3. Guan, H. P., and Preiss, J. (1993) Differentiation of the Properties of the
Branching Isozymes from Maize (Zea mays), Plant Physiol 102, 1269-1273.

4. Boyer, C. D., and Preiss, J. (1978) Multiple Forms of Starch Branching Enzyme
of Maize - Evidence for Independent Genetic-Control, Biochem Bioph Res Co 80,
169-175.

5. Guan, H. R, Li, R, ImparlRadosevich, J., Preiss, J., and Keeling, P. (1997)
Comparing the properties of Escherichia coli branching enzyme and maize
branching enzyme, Archives of Biochemistry and Biophysics 3 42, 92-98.

6. Fisher, D. K., Kim, K. N., Gao, M., Boyer, C. D., and Guiltinan, M. J. (1995) A
Cdna-Encoding Starch Branching Enzyme-I from Maize Endosperm, Plant
Physiology 108, 1313-1314.

7. Guan, H. P., ImparlRadosevich, J., Li, B, Zhang, L., Gao, Z., Sun, J. T., and
Keeling, P. (1997) Understanding the functions and specificities of maize
branching enzyme and starch synthase., Plant Physiology 114, 154-154.

8. Kuriki, T., Guan, H. P., Sivak, M., and Preiss, J. (1996) Analysis of the active
center of branching enzyme II from maize endosperm, Journal of Protein
Chemistry 15, 305-313.

9. Libessart, N., and Preiss, J. (1998) Arginine residue 384 at the catalytic center is

important for branching enzyme 11 from maize endosperm, Arch Biochem Biophys
360,135-141.

122

10.

ll.

12.

13.

14.

15.

16.

17.

Cao, H., and Preiss, J. (1996) Evidence for essential arginine residues at the active
sites of maize branching enzymes, J Protein Chem 15, 291-304.

Cao, H. P., and Preiss, J. (1999) Site-directed mutagenesis evidence for
arginine-384 residue at the active site of maize branching enzyme 11, Journal of
Protein Chemistry 18, 379-386.

Gao, M., Fisher, D. K., Kim, K. N., Shannon, J. C., and Guiltinan, M. J. (1997)
Independent genetic control of maize starch-branching enzymes Ila and 11b -
Isolation and characterization of a Sbe2a cDNA, Plant Physiology 114, 69-78.

Kuriki, T., Stewart, D. C., and Preiss, J. (1997) Construction of chimeric enzymes
out of maize endosperm branching enzymes I and II: activity and properties, J
Biol Chem 272, 28999-29004.

Libessart, N., and Preiss, J. (1998) High-level expression of branching enzyme 11
from maize endosperm in Escherichia coli, Protein Expres Purif14, 1-7.

Guan, H. P., Baba, T., and Preiss, J. (1994) Expression of Branching Enzyme-Ii of
Maize Endosperm in Escherichia-Coli, Cell Mol Biol 40, 981-988.

Boyer, C. D., and Preiss, J. (1981) Evidence for Independent Genetic-Control of
the Multiple Forms of Maize Endosperm Branching Enzymes and Starch
Synthases, Plant Physiology 6 7, 1141-1145.

Binderup, K., Mikkelsen, R., and Preiss, J. (2000) Limited proteolysis of
branching enzyme from Escherichia coli, Archives of Biochemistry and
Biophysics 377, 366-371.

123

CHAPTER 4 X-RAY CRYSTALLOGRAPHIC STUDY OF
RNA POLYMERASE Ill TRANSCRIPTION FACOTOR
TFIIIB COMPLEX

4.1 Our Research Objective

In order to investigate transcription initiation, the first stage of transcription, the
three-dimensional structure of the TFIIIB/DNA complex is essential. Although the binary
TBP/DNA and tertiary TBP/Brf/DNA complexes (1-2) have been crystallized, the
structure of the complete TFIIIB/DNA complex is still not available. Toward this goal,
recombinant mutants of individual subunits of TFIIIB, TBP, Brf and B” containing the
essential functional core domains were designed, expressed in E. coli cells and purified.
Various oligonucleotides were designed based on yeast U6 promoter. TBP/Brf/B”/DNA

quaternary complexes were made and screened for crystallization.

124

4.2 Materials and Methods

Our experimental design is to obtain the recombinant TBP, Brf, B” proteins, purified
oligonucleotides, and use them to make the TFIIIB/DNA quaternary complexes in vitro.

The complexes were concentrated and screened for crystallization.

4.2.1 Over-expression and purification of the S. cerevisiae TBP protein

The S. cerevisiae TBP (61-240) was used in the project. This N-terminal deletion
mutant was used to crystallize the TBP/TATA box complex and the TBP/Brf/TATA box
complex successfully. The contact between TBP, promoter and Brf is mainly at TBP’s
C-terminus. So the N-terminal deletion construct (61-240) of the S. cerevisiae TBP is
used in our crystallization studies. From the previous TBP/TATA, TFIIA/TBP/DNA and
TFIIA/TBP/DNA complexes (1-3) (see Figure 4.1) and recent work on the architecture
of TFIIIB complex (4-5), it is suggested that this construct is sufficient for the
pre-initiation complex assembly. In order to facilitate the purification of the protein, 6x
poly histidine tag with a Thrombin cleavage site at the N-terminus was attached to the

construct.

125

TBP protein

 

Figure 4. 1 Crystal structure of the TBP-TATA box promoter complex (1)

4.2.1.1 Over-expression of the TBP constructs

The pET14 based plasmid encoding the S. cerevisiae TBP was transformed into
BL21 (DE3) stain of E. coli. The transformed E.coli sample was plated on an agar media
plate with ampicillin (50mg /ml) over night at 37°C. Colonies were used to inoculate in
50 ml Terrific Broth (TB) media containing ampicillin (100mg/ml). After overnight
growth at 37°C, the media was transferred into 1L fresh TB media containing ampicillin
(100mg/ml). Growth continued until the optical density (600nm) (O.D600) reaches

0.8-1.0. The media was cooled to 20°C and the inducer IPTG (0.1mM) induced

126

 

producing the protein. The media was grown at 20°C for 10-14 hours before the cells

were harvested (3, 6).

4.2.1.2 Ni-NTA affinity chromatography purification of the TBP constructs
Cells were harvested and re-suspended in lysis buffer A (25 mM Tris-HCl,pH 8.0,
10% glycerol,500mM NaCl, 50mM Ammonium Acetate) with the protease inhibitor
cocktail ( EDTA free) from Roche, and sonicated for 5x1 minuts on an ice bath. Cell
debris was removed by centrifugation at 6000 rpm for 30 minutes. The clear supernatant
containing TBP was loaded onto a Ni-NTA agarose (from Qiagen) column. The
His-tagged TBP was immobilized on the Ni-NTA resin. The column was washed with the
50 ml lysis buffer and 100 ml wash buffer A and buffer B (25 mM Tris-HCl,pH 8.0, 10%
glycerol,500mM NaCl, 50mM Ammonium Acetate, 20mM imidazole ), 30 ml Wash
buffer C (25 mM Tris-HCl,pH 8.0, 10% glycerol,500mM NaCl, 50mM Ammonium
Acetate, IOOmM imidazole ), 30 ml wash buffer D (25 mM Tris-HCl,pH 8.0, 10%
glycerol,500mM NaCl, 50mM Ammonium Acetate, 250mM imidazole ) and 30 ml final
elution buffer E (25 mM Tris-HCl,pH 8.0, 10% glycerol,500mM NaCl, 50mM
Ammonium Acetate, 400mM imidazole ). UV—vis absorption was used to monitor the
protein concentration of each fraction. Fractions containing the TBP protein were
evaluated by the Bradford Assay and SDS-PAGE gel (see Figure 4.2) for concentration

and purity. The purified protein fractions were pooled together for histidine tag removal.

127

 

 

 

31KDa “"-

21 KDa OH- M. ”I u w :23“ wet-5'9- ‘1

“as!

M D1 DZ D3 D4 E1

Ll 1

 

Figure 4. 2 A SDS-PAGE sample of the TBP protein after Ni-NTA column purification.
Buffer D and E fractions were shown.

4.2.1.3 Histidine tag removal

Since the histidine tag may affect crystallization, deleting the poly histidine tag was
tried before making complexes with the TBP protein. In order to remove the 6x histidine
tag (61-240), Thrombin was used to cleave the poly histidine tag from the TBP construct.
After trial experiments, the optimal proteolysis condition was determined to be 15 unit
Thrombin / 1 mg protein for 2 hours at 4 °C. Under this condition, the 6x histidine tag
was removed completely with minimum amount of protein degradation. The result was

reproducible. Figure 4.3 shows a TBP protein proteolysis SDS-PAGE sample.

128

 

 

 

 

TBP

with

(His-tag

E \

‘= 4......

i ' TBP without His-tag
'cut 0.5hr 1hr 2hr

Figure 4. 3 Histidine tag removal in the TBP purification. The first sample was taken
after 0.5 hr of cutting and the second and third samples were taken after 1 and 2 hr
cutting respectively.

4.2.1.4 Chromatographic purification by heparin column

Heparine resin was used in the last purification step. The TBP protein after the
proteolysis step was loaded onto a heparin sepharose (Pharrnacia) column. The
positive-charged TBP protein bound with the polyanionic heparin sepharose resin. And
the protein was eluted against a NaCI gradient (250 mM-lOOOmM) in the buffer A (20
mM Tris-HCI, pH 8.0, 10% glycerol, 250mM NaCl, 50mM Ammonium Acetate). The
protein was eluted at 600-800mM NaCl. The protein fractions were pooled together and

concentrated to about 1.3 mg /ml for making complexes.

129

 

. N’s

‘M.w wash E1 E2 53 E4

N '

Figure 4. 4 SDS-PAGE samples of TBP protein (without Hisidine tag) after the Heparin
column purification. E1-E4 are elution fractions containing most of the pure protein.

 

4.2.2 Over-expression and purification of the Brf protein

From previous unpublished result (6), It is difficult to obtain crystals of the Brf
protein with the Zinc ribbon domain. No full length Brf protein from any source has been
reported to yield good crystals. Also the N-terminus of the Brf protein has only weak
interaction with TBP and DNA, so we decided to exclude it from the protein constructs.
Only the C-terminal domain of Brf was used in our investigation (please see Figure 4.5).
All of our Brf constructs have 6x poly histidine tag on either the C or N terminus. The S.

cerevisiae Brf constructs had Thrombin cleavable 6x histidine tag while the 6x histidine

130

tags on K .lactis constructs were non-cleavable. Our collaborator Dr. Steven Hahn kindly

provided all the plasmids.

 

 

 

 

 

 

 

S.cerevisiae Brf Construct K. Iactis Brf Construct
Mutant Construct Mutant Construct
number ( cleavable His-tag) number (non-cleavable His-tag)

PSH . PSH 552 ,
582 420-551-(HIS)6 302-501-(HIS)6
PSH , PSH 553 ,

5 83 420-531-(HIS)6 395-501-(HIS)6
PSH , PSH 554 ,
584 43 5-53 1 -(HlS)6 395-556-(H1S)6
PSH , PSH 555 ,
585 (HIS)6-435-53l 302-552-(H1s)6
PSH .

586 435-551-(H18)6

 

 

 

 

 

 

Table 4. 1 List of Brf protein constructs

131

TFIIB-like domain BRF homology regions

Zn linger direct repeats I/

 

  

 

 

 

 

 

 

r ‘ I'“l
‘ ~ _ ‘38 94 ‘ 164 189 263 286 304 439 515 570 595
. I - J
1 596
S. Cerevisiae 1 p j . 56
Full LCngth \ '~ rm.
43 5-55 1 :
435 531
435-531: 420
420—531: (“M 551
420-551
K. Lactzs l 5 5 6
(Full Length): ~ « .
302— 501
302-5011
302-556: 395
395-501:
395-556:

Figure 4. 5 A schematic representation of the S. Cerevisiae and K. lactis Brf constructs
used in our project(6). The top figure shows the overall architecture of Brfl protein from
S.Cerevisiae

132

4.2.2.1 Over-expression of the Brf constructs

The Brf protein plasmid was transformed into E. coli BL21 (DE3) cells and plated on
an Argar plate containing 50 mg/L ampicillin. The plate was incubated overnight at 37°C.
A single colony was transferred into 50ml fresh LB media containing 100 mg /ml
ampicillin. Cells were grown overnight at 37°C before being transferred into 1L fresh LB
media containing 100 mg/ml ampicillin. Growth continued until O.D.600 reach 0.6-0.8.
The inducer IPTG (0.4 mM) was added and shaking was maintained for three hours

before the cells were harvested(6).

4.2.2.2 Ni-NTA affinity chromatography purification of the Brf proteins

Cells were re-suspended in the lysis buffer (6 M Guanidine-HCI, 0.1 M NaHzPO4 ,
0.01 M Tris-HCI, PH 8.0) and sonicated on an ice bath. The lysate was centrifuged at
5000 rpm to separate clear supernatant solution. The supernatant was loaded onto a
Ni-NTA agarose (Qiagen) column

Ni-NTA agarose (Qiagen) resin was used for 6x Histidine tagged protein purification.
The histidine-tagged protein was immobilized onto the Ni-NTA resin. Then the Brf
protein was washed with buffer C and D. Buffer E was used to elute the protein fractions.
The fractions were monitored by UV-VIS absorption. The protein fractions in buffer D
and E were pooled together. The protein was purified under denaturing condition.
SDS-PAGE was used to verify the purity of the protein.

The buffers used in the purification include:

133

 

lysis buffer :
Buffer A: 6 M Guanidine-HCI, 0.1 M NaHzPO4 , 0.01 M Tris-HCI, PH 8.0
Buffer B: 6 M urea, 0.1 M NaH2P04, 0.01 M Tris-HCI, PH 8.0
Wash buffer;
Buffer C: 6 M urea, 0.1 M NaHzPO4, 0.01 M Tris-HCI, PH 6.8
Buffer D: 6 M urea, 0.1 M NaHzPO4, 0.01 M Tris-HCl, PH 4.5
Elution buffer:

Buffer E: 6 M Guanidine-HCI, 0.2 M Acetic acid

m

31KB!

21KDI 83"“ w a v H “l Bffpf'oteln
*c*,‘

Figure 4. 6 A SDS-PAGE of the K. lactis Brf (302-501) after Ni-NTA purification.

134

4.2.2.3 Protein refolding

The Brf protein was insoluble, so it was purified in denatured state(6) and then
refolded to its native state. This method was shown to work previously (7). Fractions
after the Ni-NTA column purification were denatured (8).

The protein fractions from the Ni-NTA column were diluted in dilution buffer (6 M
Urea, 10% glycerol, 20 mM Tris-HCI, 500 mM KCl, 5 mM DTT) to about 0.3-0.5 mg/ml.
Then they were dialyzed against the refolding buffer (10% Glycerol, 20 MM Tris-HCI,
500 MM KCl, 2 mM EDTA, 5 mM DTT, 1 mM PMSF) for 4 x 6 hours. Then the protein
was dialyzed against the refolding buffer without PMSF for 2 x 3 hours. The refolding
was carried out at 4 °C. Then the protein was concentrated to ~1.5 mg/ml. The S.
cerevisiae Brf protein was prepared for removal of histidine tag (This step is skipped for

K. lactis constructs since they had non-cleavable histidine tags.)

4.2.2.4 Removal of cleavable histidine tag from the S. cerevisiae Brf protein
Thrombin was used to cut off the poly histidine tag from the S. cerevisiae Brf
constructs. 2.5 unit Thrombin /mg protein was the best ratio at 4 °C. No degradation was
found until 18 hours. Figure 4.7 shows the cleavage test result of one Brf construct.
After the cleavage of the histidine tag, PMSF was added to the protein solution to a
final concentration of lmM to inhibit the Thrombin to prevent further degradation of the

protein.

135

ii

lit ii .
31KDa .. 8” With “‘5 t5'9 31KDa L4 Brf without His-tag
" ' - t = H JJ 5
21KDa “ a :- ..., V f" "' 21KDaH l ...—.degradation
Q Brf without His tag 3” “m ”'5 ‘39
M 15' 30' 45' 1hr 1.5m 2hr M 0'" 1” 2'" ‘8"

Figure 4. 7 The cleavage of His-tag on Brf (420-531): left figure shows the cleavage
test, samples were taken at different time. And at 2 hr the cutting is complete. The right
figure shows the overcutting product appearing after 18 hrs. The test was carried out at 4
°C, Thrombin / protein ratio is 2.5 unit Thrombin/ 1mg protein.

4.2.2.5 FPLC ion exchange chromatography purification

A Pharmacia F PLC system was used to purify the refolded protein without histidine
tag at 4°C. The Source 15 Q matrixes (anion exchange resin with an average particle size
of 15pm) were used in the column. The protein to be purified was diluted with FPLC
dilution buffer (20mM Tris-HCl, 5mM DTT, 10% Glycerol, PH 8.5), then loaded onto the
Source-Q column. Buffer A (20mM Tris-HCl, 50mM KCl, 5mM DTT, 10% Glycerol, PH
8.5) and buffer B (20mM Tris-HCl, 1 M KCl, 5mM DTT, 10% Glycerol, PH 8.5) were
used to elute the protein. The gradient was from 50 mM KCl to 1 M KCl. The Brf
proteins eluted out at the salt concentration of ~200mM KCl. The fractions were
evaluated by Bradford Assay and SDS-PAGE gel for their concentration and purity. After
the ion exchange purification, all the Brf proteins were pure enough for making

complexes. Figure 4.8 shows the SDS-PAGE sample of final K. lactis Brf proteins.

136

 

Figure 4. 8 A SDS-PAGE sample of purified K. lactis Brf proteins. 1: M.W marker; 2:
K. lactis Brf (302-501); 3: K. lactis Brf (395-501); 4: K. lactis Brf (302-556); 5: K. lactis
Brf (395-556).

 

 

 

Linda's
mini 8“ 8“. heavy band B“. Source 0
I J I I I 1
n9 8": -- -- -- 0.5 2 o 5 2 a 16 0.5 2 8
n9 Brf(583): -- -- o 45 >
ng TBPc: -- 2 ,

 

Figure 4. 9 A Gel mobility shift assay using U6 promoter probe, and TBPC and
indicated amount of Brf and B”(9).

The S. Cerevisiae Brf (420-531) constructs were sent to Dr. Steven Hahn for binding
assays. Figure 4.9 shows the Gel mobility shift assay of PSH583 (420-581) on U6
promoter probe. The assay confirmed binding of our Brf, TBP and B” constructs on the

U6 promoter.

137

4.2.2.6 New S.cerevisiae Brf construct design and purification attempt

Besides the current Brf constructs, we also designed a longer Brf construct (69-531)
and five internal deletion mutants of the S. cerevisiae Brf construct. Based on the reported
TBP-Brf-DNA structures and secondary structure predictions on the Brf protein, we
believe that internal deletion constructs (69-531A266-407), (69-531A280-420),
(69-531A323-407), (69-531A335-420) and (69-531A266-430) will bind with the TBP, B”
proteins and DNA, and do not have the existing floppy loop that might be a detrimental
factor in the complex crystallization.

The constructs were cloned and sequences were verified. These constructs of the Brf
protein were also insoluble. After purification in denatured condition, protein refolding
was carried out. Unfortunately, all of the constructs precipitated during the refolding step
in all attempts. After extensive trials at different conditions, we believed that those
constructs can not be refolded correctly to their natural state. Thus those constructs are

not suitable for the crystallization trials.

4.2.3 Over-expression and purification of the B” protein

Different B” constructs were tested extensively for their abilities of binding the

TBP, Brf protein and TATA containing promoters(9). It was determined that the construct

138

-w -'ILJ"’L ”q

A

Jam‘s..- "

 

containing the central SANT domain can bind with the other components of TFIIIB-DNA
complex. We obtained the plasmid of B” (240-520) with uncleavable poly Histidine tag
on its C-terrninus from our collaborator, Dr. Steven Hahn. The construct encompasses the
SANT domain and has been tested to have full activity(9). We sub-cloned the coding
sequence of B” (240-540) and B” (265-540) into a pET vector with a SUMO domain and
a cleavable poly His-tag (made by Geiger lab). The construct design was based on
secondary structure predictions: complete secondary structure elements (alpha helix or
beta sheet) were kept intact while at the same time the whole SANT domain was included

in the construct. The B” constructs were also shown to be active(9).

4.2.3.1 Over-expression of the B” (240-520)

The plasmid of the B” mutant was transformed into E. coli BL21 (DE3) cells. Cells
were plated on an agar plate containing 50mg/L ampicillin and 30mg/L kanamycin. After
inoculated overnight at 37°C, a single colony was used to inoculate 50 mL of LB media
containing lOOmg/ml ampicillin and 60 mg/ml kanamycin. After overnight growth at
37°C, the media was transferred into 1L fresh LB containing ampicillin and kanamycin.

Growth continued at 37°C until O.D.600 of the media reach 0.5-0.6. The media was

cooled to 30°C and IPTG (0.5mM) was added. After 3 hours, cells were harvested by

centrifugation.

139

 

4.2.3.2 Ni-NTA purification of the B” protein

Ni-NTA agarose resin (from Qiagen) was used to purify the His-tagged B” mutants.
The purification was under native condition. The buffers used included:

Lysis buffer: 20mM HEPES, 300mM NaCl, 10% Glycerol, PH 8.0, 5 mM
2-mercaptoethanol (BME) and protease inhibitor added before use)

Wash buffer: lysis buffer plus 20mM imidazole, PH8.0, BME and PMSF added
before use.

Elution bufferzlysis buffer plus IOOmM imidazole, PH8.0, BME and PMSF added
before use.

The cells were re-suspended with a protease inhibitor tablet in the lysis buffer and
sonicated on an ice bath. The supernatant after centrifugation was loaded onto a Ni-NTA
column. The column was washed with the lysis buffer, wash buffer and elution buffer.
Fractions containing the B” protein were evaluated by Bradford Assay and SDS-PAGE

gel. The protein fractions were pooled together for further purification.

4.2.3.3 FPLC ion exchange chromatography purification

A Pharmacia FPLC system was used to purify the the B” mutants at 4°C. The

Source 15 Q matrix was used. The protein solution was loaded onto the Source-Q column.

Buffer A (20mM HEPES, 75mM NaCl, 5mM BME,1mM PMSF, 10% Glycerol, PH 8.0)

and buffer B (20mM HEPES, SOOmM NaCl, 5mM BME,1mM PMSF, 10% Glycerol, PH

8.0 ) were used to elute the protein. The gradient was from 75mM NaCl to 500 mM NaCl.

140

 

The B” protein fractions eluted over a wide salt concentration (100 ~ 200 mM NaCl).

The fractions were evaluated by Bradford Assay and SDS-PAGE gel.

4.2.3.4 Cleavage of poly His tag from the B” (240-540) and B” (265-540)

The B” (240-540) and B” (265-540) have cleavable poly Hisidine tags. SUMO
protease was used to cut off the poly hitidine tags. The purified protein with Histidine tag
was mixed with SUMO protease (prepared by our lab) at a ratio 1mg protein / 0.001 mg
SUMO protease. The cutting was complete at 4 °C after 2 hours. After the cutting, the
mixture was loaded onto a Ni-NTA column containing 1 ml Ni-NTA resin. The solution
containing the cleaved B” was collected. The protein was concentrated to about 2 mg /ml

for making complexes (Figure 4.10).

q ’- -. .- , m. - -. ~

‘5'!

H

-———-- B'withho
1‘

.. J... ...—.... B'withoutlao

B'withouttag
g
I.
-~ .- ——>- Tag
. ‘ —>Tag
r. ’ . I .
‘. MunctflJlI-Zh. p611 . - . M 1 2 3 4

 

Figure 4. 10 A SDS-PAGE sample of B”(240-540) His-tag cleavage. Left shows time
course reaction of SUMO protease cleavage. It can be seen that cutting is complete after
1hr. Right figure: 1 is cutting mixture containing B” protein and the cut tag; 2, 3 and 4 are
flow through of Ni-NTA column showing SUMO His-tag removal

14]

 

4.2.4 Purification of oligonucleotides

In the RNA pol III pre-initiation complex, the promoter anchors all the three subunits
of TFIIIB. The interaction between the proteins and DNA requires certain base pairs
available both upstream and downstream of the TATA box. It was determined that full
length proteins have interaction with DNA from 15 b.p upstream to 10 b.p downstream
(10-11). Considering we were using the truncated Brf, B” and TBP proteins, which might
have less interaction with the promoter, the formation of the complex might require less

nucleotide, so we decided to use a series of DNAs with different lengths for the complex

 

 

 

 

 

making.
DNA Sequence of DNA
iiib44l CGTCCACTATTTTCGGCTACTATAAAAGAATGTT'ITTTTCGCAA
iiib44ll CACTATTTI‘CGGCTACTATA A AAGAATGTTIT'ITTCGCAACTAT
iiib4511 TCGTCCACTAT’I‘I'I‘CGGCTACTATAA AAGAATGTTITTITCGCAA
iiib46I TCCACTATTITCGGCTACTATAAAAGAATG'I'I'ITITTCGCAACTAT
iiib47ll TCGTCCACTATTTTCGGCTACTA’I‘A A AAGAATGTITTTTTCGCAAACT
iiib47lll CGTCCACTAT’I‘TTCGGCTACTATAAAAGAATGTT‘ITTTTCGCAACTA
DNAl CTATTI‘TCGGCTACTATAAAAGAATGTT’I‘TI‘T
DNA2 CTATTI‘TCGGCTACTATAAAAGAATGTT’ITI‘TTC
DNA3 CACTATFTTCGGCTACT AT AAA AGAATGTTITIT
DNA4 CACTA'ITTTCGGCTACTAT A AA AGAATGT’ITTTI‘TC
DNAS TCCACTATTI‘TCGGCTACTA’I‘A AAAGAATGTTTTTI‘
DNA6 T CCACTAT’I’I‘TCGGCT AC’I‘ATAAA AGAATGTT'I‘TITTC
DNA 19 CTATA AAAAAATGTTTTIT
DNA25 CGGCTACTA’I‘AAATAAATGTT'ITI‘T
DNA26 TCGGCTACTA’I‘AAATAAATGT’ITTTT
DNA27 T’I‘CGGCTACTATA A ATAAATGTTTTTT
DNA29 TFTTCGGCTACTATA A ATAAATGTI'ITTT
DNA3O ATTTTCGGCTACTATA A ATAAATGT’ITI‘TT

 

Table 4. 2 DNA used in our project

142

 

The designed oligonucleotides (single strands of DNA) were ordered from Keck
facility, Yale University. The purification and annealing were done in our lab.

Oligonucleotides ordered were purified with a Perkin Elmer HPLC system equipped
_ with a Source-Q column. The single strands of DNA were loaded onto the column. Buffer
A (lOmM NaOH, 0.2M NaCl) and buffer B (lOmM NaOH, 1M NaCl) were used in
gradient. The Oligonucleotide fractions collected were neutralized by Tris buffer and
diluted to pH 7.5. Then they were loaded onto a DEAE column. DNAs eluted out with
buffer C (lOmM Tris, 1M NaCl, PH 7.5). Then they were concentrated and annealed in

equivalent amount(6).

4.2.4 Making quaternary complexes and crystallization screening

One TBP construct, 8 Brf constructs, 3 B” constructs and 18 DNAs were used to
form complexes (in our project, about 130 complexes were made). Quaternary complexes
were formed by mixing TBP, Brf, B” and DNA at a ratio of l:1:1:1.2 on an ice bath for
about 30 minutes. Then the complexes were concentrated to about 200uL. The final
concentrations of complexes for crystallization are ~6-10 mg/ mL. SDS-PAGE was used
to verify complex formation. Figure 4.11 shows the Comassie Blue stained gel sample of
one quaternary complex. Figure 4.12 is one sample UV scan of the TFIIIB only,

Oligonucleotide and the final pure quaternary complex.

143

 

31240540)
/
v u, 81240-520)
.- .. .. TBP(61-240)
* -‘ Brf(420—531)
w
M 1 2

Figure 4. 11 SDS-PAGE of two quaternary complexes. Protein bands were shown on
Commasie blue gel. DNA is not stained in Commasie blue dye.

 

Complex UV scan

 

0' 8 ‘—-—Complex
0. 6 DNA

+ TFDIB mixure

 

 

 

 

230 240 250 260 270 280 290 300

Figure 4. 12 UV spectrum of the TFIIIB protein, DNA and pure quaternary complex

(after size exclusion purification). TFIIIB has Kmax at 280 nm while DNA at 260 nm. The

complex has maximal absorption at about 260 nm, which confirmed the binding between

TFIIIB proteins and the oligonucleotide

144

Sparse Matrix Crystallization screening yielded some tiny crystals, but could not be
improved. The sizes of crystals were smaller than 0.005 mm x 0.005mm x 0.005 mm.

Figure 4.13 shows four pictures of putative quaternary complex crystals.

Figure 4. 13 Photos of putative quaternary complex crystals. The four photos were taken from
different complex crystal screening boxes. All the crystals were too small, and showed no
diffraction. Crystals in the top two photos are from (3.4 M 1,6 hexanediol, 0.1 M Tris, pH 8.5, 0.2
M MgClz ); crystals from bottom left photo are from ( 30% MPD, 0.1 M Sodium Cacodylate, 0.2

M NaOAc, pH 6.5); crystals from bottom right photo are from (30%PEG4000, 0.1M Tris, pH 8.5,
0.2 M NaOAc ).

 

145

4.2.5 SEC chromatographic purification of quaternary complexes

We used proteins and oligonucleotides to make complexes, and then used the
complexes for crystallization screening. After extensive screening failed to yield good
crystals, we realized that forming complex might not be perfect. If the four components
of the pre-initiation complex were not mixed in a perfect ratio, there could be a mixture
of one or two (TBP/DNA), or even three (TBP/Brf/DNA) components. In this case,
homogeneity of the solution will be compromised and crystallization will be difficult.
Since it is difficult to get a perfect ratio because there is always some error in determining
the concentration of the protein and DNA, we sought another way to guarantee the
correct stoichiometry.

Size exclusion chromatography (SEC) was used to further purify the complexes.
SEC separates different proteins or complexes based on their sizes. Bigger complexes or
proteins elute out first since they can not be retained inside the porous stationary matrix,
while the smaller components will be retained inside the matrix much longer. In our
complex forming process, the excess protein, binary or tertiary complex will be eluted
out later then the correct quaternary complex since the quaternary complex is the biggest
in size or molecular weight.

We mixed the proteins and oligonucleotide together to form a complex, then purified
the complex by size exclusion chromatography (SEC). Figure 4.14 shows the SEC
chromatograph of a complex TBP(61-240)/ Brf(435-551) /B”(240-520)/DNA1. #30 and

31 fractions (2 ml each fraction) contain the quaternary complex. In some case, we found

146

we still had binary and tertiary complexes which resulted in the second and third peaks
on the SEC chromatograph. For example, Figure 4.15 shows the SDS-PAGE of the
complex fractions after the SEC purification. From the result, it confirmed that the
quaternary complex formed in vitro, and was stable enough to survive the mild SEC
separation. But it was also clear that other complexes formed when we mixed the proteins
and oligonucleotide. Without SEC purification, there will be several kinds of complexes

in the crystallization solution. The homogeneity of the solution will be significantly

 

impaired, which is fatal to the crystallization screening.
We used this method to purified 10 complexes, and collected the right fractions
containing only the quaternary complex. The fractions were concentrated and used for

crystallization screening. But still no good crystals were found.

Elution volume (ml)

 

   

 

 

 

 

 

 

 

 

 

30 50 80 100
I l _
I I I 7
Molecular standard (KDa) 158 44 17
'.l'.: , 'l' I l ' l l
-i‘ . . l __ ... _ ,, _ ..
i‘ ““i"" r ‘ #30 ‘7‘“ i , l
«l . . é . i: ) l . ;
U i, i ; 3);}:17, + -' , i
. . v; - , ‘1‘ 'A‘ ...t . ..
.-....) -.., i ___,_. .._ -..... ...... m .... h.
. . i I i I ",ri i." . I ._ =1, '.
iiiilil i ‘ 1:1. " 5* .t
7‘)‘. T!’ 3 _. , l . l a " E
[f‘ l. I l ‘ i ..-..i. .. ‘. .2.- ’ -1... j.-. .-.l._
-‘ i l :T . -' ii i I 3i.
i" t; E ji, 1. 'g . i: 43.;
l 4 lfi ; i I; i . ' V ; i ..-. l . ' .
i. 331:3. l) .. ,
I ..., . ... u: .-.. . . -2.... ...—_._' - _._‘, . i . .' h; .
‘ ' - L L . i. 1' I . t 3..
'. Q .‘i 4 i1 ,. I"
11 mm). i .
'i i! r ., .i.. ,‘ .. . .: I .
1.7.1: ‘1" 2"?”"7‘1'" "“’""'§".“7"7’i . ' - ' ,

 

 

 

 

t
r ; I
l i i ' I
f . ' .__ 4 I _ ...
. 4 . j' i H l l I

Figure 4. 14 SEC chromatograph of quaternary complex TBP(61—240)/ Brf(435-551)
B”(240—520)/DNA 1. The complex is in #30, #31 fractions.

147

Elution volume (ml)

30 50 70 100 120
l l I l l _

I fi I
Molecular standard (KDa) 158 44 17,, , 1-3

i l

    

.......

........

i .
’ .-‘. 1" t .I '. . -.'.
.. ~.t>—~ .- ~

Tu W “‘“afi-"C‘flr‘. DNA2
M 28 29 30 31 33 34 37 DNA

Figure 4. 15 The SEC chromatograph and SDS-PAGE of a quaternary complex
TBP(61-240) [Brf (407-531) /B”(265-540)/DNA1 (see Table 4.2). In top figure, peak 1 is
the TBP-Brf-B” fDNA quaternary complex (#28, 29 and 30 in the gel sample); peak 2 is
the TBP-Brf-DNA tertiary complex (#33 and 34 in gel sample); peak 3 represent the
TBP/DNA binary complex (#37 in gel sample). The lower figure is the silver stained
SDS-PAGE sample of the fractions purified by SEC.

148

4.3 Conclusion and future plan on this project

In the TFIIIB-DNA pre—initiation complex project, we purified three subunits of the
TFIIIB factor, and formed the complete pre-initiation complexes in vitro. Extensive
crystallization screenings were attempted on about 130 complexes. Despite all the effort,
unfortunately we only got some tiny crystals that do not diffract. SEC purification was
adopted for 10 complexes, which theoretically should improve the complex homogeneity
and increase the chance of crystallization, but crystals were still not yielded.

We purified the truncated versions of the three subunits of the TFIIIB and made the
complexes in vitro. We should have done more extensive assays on the activity of each
protein, especially the Brf and B” protein. The Brf protein was denatured and refolded in
the purification, which might lead to mis-folding. Whether the protein was still active in
vitro should be investigated after the refolding step. It might give us more accurate
information about the interaction among the three subunits of TFIIIB.

More SEC purification on the complexes should be performed. But considering the
low yield of SEC purification (only, about 30% of the protein can be recovered) we need
much more protein and oligonucleotide to finish the project.

We believe an alternative route for making the pre-initiation complex could be the
co-expression of the three subunits and purification of the TFIIIB together. This should
avoid the risky refolding step for the Brf protein. Also the binding between the three
subunits should be close to the real natural binding in vivo, which is definitely preferred.

This should give us a higher chance of forming good crystals.

149

Reference Cited

1.

Kim, Y. C., Geiger, J. H., Hahn, S., and Sigler, P. B. (1993) Crystal-Structure of a
Yeast pr Tata-Box Complex, Nature 3 65, 512-520.

Juo, Z. S., Kassavetis, G A., Wang, J. M., Geiduschek, E. P., and Sigler, P. B.
(2003) Crystal structure of a transcription factor IIIB core interface ternary
complex, Nature 422, 534-539.

Geiger, J. H., Hahn, S., Lee, S., and Sigler, P. B. (1996) Crystal structure of the
yeast TFIIA/TBP/DNA complex, Science 272, 830-836.

Vilalta, A., Trivedi, A., and Johnson, D. L. (1996) Mutation in the tata-binding
protein (TBP) selectively abolishes both RNA polymerase III transcription and the
interaction of TBP with a TFIIIB subunit, Mol Biol Cell 7, 2704-2704.

Colbert, T., Lee, S., Schimmack, G., and Hahn, S. (1998) Architecture of protein
and DNA contacts within the TFIIIB-DNA complex, Mol Cell Biol 18,
1682-1691.

Jin, X. (2002) X-ray crystallographic studies of RNA Polymerase III transcription
factor TFIIIB and lL-MYO-Inositol l-phosphate synthase, In Department of
Chemistry, Michigan State University, East Lansing.

Librizzi, M. D., Moir, R. D., Brenowitz, M., and Willis, 1. M. (1996) Expression
and purification of the RNA polymerase III transcription specificity factor IIIB70

from Saccharomyces cerevisiae and its cooperative binding with TATA-binding
protein, J Biol Chem 2 71 , 32695-32701.

Sadana, A. (1995) Protein Refolding and Inactivation during Bioseparation -
Bioprocessing Implications, Biotechnol Bioeng 48, 481-489.

Hahn, S. Gel mobility shift assay on Brf and B" protein, Fred Hutchinson Cancer
Research Center, seatle.

150

 

10.

11.

Kassavetis, G. A., and Geiduschek, E. P. (2006) Transcription factor TFIIIB and
transcription by RNA polymerase III, Biochem Soc T 34, 1082-1087.

Huang, Y., and Maraia, R. J. (2001) Comparison of the RNA polymerase III
transcription machinery in Schizosaccharomyces pombe, Saccharomyces
cerevisiae and human, Nucleic Acids Res 29, 2675-2690.

151

 

APPENDIXES

152

:2? scream“... 82 05 8 5%.. 8855qu E 822$

 

 

6.85....” 8.83.” 3.8:.” 6.88.8 8.3 3 3.5 5.2.2
8.5 new 2.5 m8 €8v N8 38% 88 88v ...8 545.2855
882 86: 48:: 5:8 3:2: 888:2 2.3.5
83.88 8848 888m 888 88.8 82. 22.8%.. .38
89.2 as 4.8 8.: Em A _.8 2: 3.8 «.2 o: 8222
ad SH 393 :62 8.33 353 8:85.5—
82-86 8.35.3 $2-85 82-8.: 83-85
8.2... 8.3..“ 88-8 8.78. 8.2% Q: 8...... 8:...er
888 888 888 888 888 QC 5822;
Q8 Q8 Q8 Q8 98 C .2
8.8 ”.8 38 8.8 8.8 Cu
08 Q8 Q8 Q8 28 C .5
SM: SM: EM: SM: EM: 9: o
3.2 :2: 3.2 2.2 2.2 A6 a
58 8.8 58 e8 58 g a
253—555 :09 3::
e a. e e 4 ES. 555855 N
_N a _N a _N n. _N a _N a 9.288%
:2 e: 8... 8.2. no -.. 8.2.35

 

5895 mm 35% 65:56.8»: 2: .«o 8333.». 533:8 Sac 52-x A 5.52:3.

153

 

 

E S 2. 2. o. 3... 832.35
2 .... o. no _.m 8.... 832... 88.2.8
...; ...: ....N. m. : S. 3... 825......
8.8 98 8.8 EM. 2... A8. 8.2... .82

8..— :Eccasuafié

82 :3 m8... 82 E... C 28... 25..

m _ C... so... so... :3. to... 3.. 582 ......m

5:8... ....5 32:
SN 38 EN n8 «.8 3... .2.:
8.: 2: .3. 0.8. ...: A8. is:
13 ...? m3 «.8 ...? Q. .33....— 08.53..
E. e2 8.» no... no -... 3.2.35

 

.539:— Hm 36H 65:56:53. 2: .3 85:3.“ 32:2an ousaoeumfi 5655.3.

154

Appendix 3. Ramachandran plot of the BE/alphaCD structure

PROCHECK

Ramachandran Plot
..aflngLS

    
  
   

  

180

Lyceum.
‘3 (D)

‘ _ - .) Aush‘fi WA) ‘ a .
‘ A) It. ‘
Act:

  

«4b- .'

 
  

 

135

   
 

90 " 4(C')

-.
' 71A)

Psi (degrees)
C

   
 

3(C) A A

ye .ISMWNWMWB. [551312.53 (D,

Hlsmzmu V ‘
I <.

 

 

 

Phi (degrees)

155

Appendix 4. Ramachandran plot of the BE/betaCD structure

PROCHECK

Ramachandran Plot
TLSoutput

l'a

180

      
    
   
  
 

  
 

W5) “ ~b

  

 

 

 

 

A b
ALA i741 "
’ GLUEYRimuOLYS __(> (BI a
135 F AE<’+'f-4><(A> I ___ Ii
' ~!
90- -
45-
E7
g 0 _ .....
3
.a Eh
Q1 ARGJH‘MHDL ,, .
-45 — 'Efikfifigflfih‘) ASN 2.50 1D)
-90~— —
— 7 ASP 1.33.:m
4353*) Wgw ‘
‘Ah'llu, pl g‘: .A. ‘A
I. (I I -M’)“ :5! I:
-180 -135 —9'0 415' 6 4'5 9'0 135 1:

Phi (degrees)

156

Appendix 5. Ramachandran plot of the BE/gamma CD structure

PROCHEC K

Ramachandran Plot
79-20K2_refmac1
WW ~b i

. A§P 414F81

 

   
    
  

' 5M“

R(.l.<’.l)('\ »
u \R(; ”015)

'l'HR |l7 1))
I

‘ I‘.: 17
i ‘ \
. HitFaumg
- ‘ IEL 4.73 C)

‘_ #

[__l . .

~b — 51 mm ~p ‘.

-l35 j LEUl-(i (, » 7;. L

-l80 -135 -90 -45 0 45 90 1 5 180
Phi (degrees)

Psi (degrees)

 

 

157

 

Appendix 6. Ramachandran plot of the BE/M7 structure

PROCHEC K

Ramachandran Plot
M7initialring5_refmacl

Pl 4(0) ARG Hui-433T? rt -
'r ,1 , .4,

, ‘33 C) A

ARG Immumm- " ::

    
  
 
          
 
      

    
   
 

180

     
 

135 " j" (9);

I
i

 

    

90 (A)
45

73‘

e S)

no

0

E

32'
-45 ,‘fi ‘mm “

Aflh‘fim‘m’) A351?“

-90 t?“

ME] 4 f
.

-135 ' VA“: ‘

   

. - “IA v A
‘ .’ A 1 A ‘ m (117-OB)
. . , ‘ _‘ f, A AA A“ - .
g i [(i 252. ‘C ‘ IA
— 5 0 45 90 l 5 10

Phi (degrees)

158

Appendix 7. Ramachandran plot of the BE/M6 structure

 

   
    
  
 
    

PROCHECK
Ramachandran Plot
180 finallrrrefmaCI (il ll 17( ('\,
-- I l : will?) ”b ‘

‘ 1
A MAC | 745‘?)
A I ‘ 7

ALLA}: Vseygmsi‘s 3.3“ (IS
CLlriFFEIlilfik’Rr H ' —

 

 

 

 

 

90 -
ALA 11MB)
PHEJBF D)
I
45-
A
- VJ
8
‘55 m ‘
o ‘ A
'U
V
.5 ‘
an T E” .
‘l“ 7%) A ‘
'45 [EL {fl “$95; :50 (D)
ASP 133L813 475m; I
I I A
—90- _—
A
:h (I. PROM-hm ~p
-135 a! ”I I A A A
P I I I? ‘ A A 5‘
3 ' L i'r'vwm 0‘ U. .‘ ‘ A~D‘
’ ' ‘ ‘ ' .- HLUJS A
I ' .v\sT)I_04:m; fit? J ‘A A IA

 

 

-180 -135 -90 45 6 4'5 9'0 155 10
Phi (degrees)

159

Appendix 8. Protein-ligand close contacts between BE and alpha CD. The interactions

listed are within a 3.8 A cutoff.

 

 

At binding site I

# Chain Residue Atom # Chain Ligand Atom Distance(A)
255 D ARG NH] ] G ACX 02A 3.4]
255 D ARG NH] 1 G ACX 038 3.78
259 D ASN OD] ] G ACX 03D 3.67
259 D ASN OD] l G ACX 02C 3.47
259 D ASN C ] G ACX 03C 3.74
259 D ASN O l G ACX C3C 3.60
259 D ASN O l G ACX 03C 2.83
259 D ASN O l G ACX 028 3.13
260 D ASN CB ] G ACX 038 3.49
260 D ASN CB 1 G ACX 023 3.28
260 D ASN CG ] G ACX 033 3.3]
260 D ASN ND2 ] G ACX C33 3.65
260 D ASN ND2 ] G ACX 038 2.53
260 D ASN ND2 ] G ACX C2B 3.79
260 D ASN ND2 ] G ACX 028 3.55
260 D ASN O l G ACX 02A 3.33
260 D ASN O l G ACX 038 3.28
261 D PHE CA 1 G ACX 03A 3.30
26] D PHE CA ] G ACX 02A 3.77
26] D PHE CB ] G ACX 03A 3.43
261 D PHE CB ] G ACX 02F 3.72
26] D PHE CG ] G ACX C3A 3.75
261 D PHE CG ] G ACX 03A 3.72
26] D PHE CE2 ] G ACX O] B 3.71
26] D PHE CE2 ] G ACX 01A 3.53
26] D PHE CD2 ] G ACX C3A 3.25
26] D PHE CD2 ] G ACX 03A 3.68
26] D PHE CD2 ] G ACX 01A 3.70
26 "l D PHE C l G ACX 03A 3.53
262 D TRP N l G ACX 03A 2.8]
262 D TRP CA I G ACX 03A 3.74
262 D TRP CB ] G ACX 03A 3.50
262 D TRP CG ] G ACX 03A 3.6]
262 D TRP CD2 ] G ACX 03A 3.53
262 D TRP CE3 ] G ACX 03A 3 .4]
262 D TRP O ] G ACX 02F 3.76

 

160

 

 

 

At binding site II

# Chain Residue Atom # Chain Ligand Atom Distance(A)
505 D ASP CA I J ACX 06D 3.47
505 D ASP CB I J ACX 06D 2.96
505 D ASP C I J ACX 06D 3.72
505 D ASP 0 I J ACX 06D 3. I 5
508 D THR CB I J ACX 0613 3.75
508 D THR CB I J ACX 05D 3.35
508 D THR CB I J ACX 06D 3.72
508 D THR 00 I I J ACX 05D 3. I 8
508 D THR 0G I I J ACX 06D 2.79
508 D THR C I J ACX 065 3.6]
508 D THR C I J ACX 06D 3.68
508 D THR O l J ACX C6E 3.06
508 D THR O I J ACX 06E 2.71
509 D PHE N I J ACX 06D 3 .21
509 D PHE CA I J ACX C6D 3.59
509 D PHE CA I J ACX 06D 3.42
509 D PHE CB I J ACX 06D 3.76
509 D PHE CE2 I J ACX 06C 3.37
509 D PHE CD2 I J ACX C6D 3.7I
5 I I D [LE CB I J ACX 0613 3.70
51] D ILE C02 I J ACX 06F 3.55
512 D LEU CD] I J ACX C6C 3.73
512 D LEU CD] I J ACX 06B 3.7]
5 I 2 D LEU CD2 I J ACX 06F 3 .72
5 I 2 D LEU CD2 I J ACX C5E 3.40
5I2 D LEU CD2 I J ACX C6E 3.26
5 I2 D LEU O I J ACX 06A 3 .62
628 D TRP CDI I J ACX CIE 3.74
628 D TRP CD] I J ACX 0513 3.48
628 D TRP CDI I J ACX 06E 3.79
628 D TRP CDI I J ACX C25 3.57
628 D TRP ' NE] I J ACX 05E 3.32
628 D TRP NE] I J ACX C65 3.74
628 D TRP NE] I J ACX 06E 2.66
628 D TRP CE2 I J ACX 065 3.32
628 D TRP C22 I J ACX 0613 3.43

 

161

 

 

 

At binding site 111

# Chain Residue Atom # Chain Ligand Atom Distance(A)
155 D GLN NE2 1 E ACX 023 3.80
189 D LYS CD 1 E ACX 068 3.58
189 D LYS NZ 1 E ACX C6A 3.73
189 D LYS NZ 1 E ACX 068 3.55
201 D LEU CD1 1 E ACX C6A 3.67
201 D LEU CD] 1 E ACX 06A 3.25
211 D GLN CD 1 E ACX C 1 B 3.74
211 D GLN CD 1 E ACX OSB 3.23
21] D GLN OE] 1 E ACX C13 3.69
211 D GLN OE] 1 E ACX OSB 2.74
211 D GLN OE] 1 E ACX CSB 3.63
2] 1 D GLN OE] ] E ACX C68 3.44
211 D GLN NE2 1 E ACX CSC 3.04
21] D GLN NE2 1 E ACX C6C 3.23
2] ] D GLN NE2 1 E ACX 06C 3.70
2]] D GLN NE2 ] E ACX C4C 3.3]
211 D GLN NE2 I E ACX O] B 3.28
21 ] D GLN NE2 1 E ACX C 1 B 2.98
21 1 D GLN NE2 1 E ACX OSB 3.00
212 D MET CB 1 E ACX 06C 3.72
2 I 5 D GLU CB 1 E ACX 05C 3.53
215 D GLU CB 1 E ACX CSC 3.66
215 D GLU CB 1 E ACX 013 3.52
215 D GLU CG 1 E ACX 01C 3.56
215 D GLU CG 1 E ACX 05C 3.40
215 D GLU OE] 1 E ACX 013 3.74
215 D GLU O 1 E ACX CSB 3.70
215 D GLU O l E ACX C68 3 .09
215 D GLU O 1 E ACX 068 2.92
217 D ALA CB 1 E ACX C63 3.49

 

162

At binding site IV

 

 

 

# Chain Residue Atom # Chain Ligand Atom Distance(A)
542 C ASP CG 1 F ACX 02E 3.54
542 C ASP OD] 1 F ACX 03F 3.49
542 C ASP OD] 1 F ACX C2E 3.32
542 C ASP OD] 1 F ACX 02E 2.44
544 C TRP CG 1 F ACX 05E 3.57
544 C TRP NE] 1 F ACX C6E 3.75
544 C TRP CE2 1 F ACX C6E 3.62
544 C TRP CD2 1 F ACX 05E 3.69
545 C GLN CG 1 F ACX 03E 3.7]
545 C GLN CD 1 F ACX 03E 3.54
545 C GLN OE] 1 F ACX 03E 3.38
545 C GLN NE2 1 F ACX OZE 3.63
659 C PRO CB 1 F ACX 02D 3.63
659 C PRO C 1 F ACX 02D 3.58
659 C PRO 0 1 F ACX 03D 3.49
659 C PRO 0 1 F ACX C2D 3.12
659 C PRO 0 1 F ACX 02D 2.62
661 C PRO CG. 1 F ACX 03D 3.51
661 C PRO CD 1 F ACX 03D 3.22
689 C SER CB 1 F ACX 05D 3.72
689 C SER CB 1 F ACX C6D 3.78
689 C SER CB 1 IF ACX 06D 3.24
689 C SER 0G 1 F ACX 06D 2.93
717 C PRO CB 1 F ACX C2D 3.79
717 C PRO CG 1 F ACX C2D 3.69

 

 

163

Appendix 9. Protein-ligand close contacts between BE and beta CD. The interactions

listed are within a 3.8 A cutoff.

 

 

At binding site I

# Chain Residue Atom # Chain Ligand Atom Distance(A)
258 A ASP C 1 I BCD 033 3.77
258 A ASP 0 1 1 BCD 024 3.30
258 A ASP 0 1 ] BCD C33 3.29
258 A ASP 0 1 ] BCD 033 2.6]
258 A ASP 0 1 1 BCD 023 3.44
259 A ASN OD] 1 ] BCD C32 3.44
259 A ASN OD] 1 I BCD 032 3.06
259 A ASN OD] 1 1 BCD 022 3.64
259 A ASN C 1 1 BCD 034 3.79
259 A ASN O 1 1 BCD C35 3.74
259 A ASN O 1 1 BCD 044 3.62
259 A ASN O 1 I BCD C34 3.38
259 A ASN O 1 ] BCD 034 3.29
260 A ASN CB 1 1 BCD 025 3.42
260 A ASN OD] 1 1 BCD 035 3.76
260 A ASN OD] 1 ] BCD 025 3.40
260 A ASN C 1 ] BCD 035 3.80
260 A ASN O 1 1 BCD 026 3.69
260 A ASN O 1 ] BCD 035 3.16
261 A PHE CA 1 1 BCD C36 3.78
261 A PHE CA 1 1 BCD 036 3.48
261 A PHE CA 1 I BCD 026 3.73
261 A PHE CB 1 1 BCD 036 3.49
261 A PHE CE] 1 1 BCD 047 3.73
261 A PHE CE2 1 1 BCD 045 3.55
261 A PHE CD2 1 1 BCD C36 3.40
261 A PHE CD2 1 1 BCD 045 3.39
261 A PHE C 1 I BCD 036 3.66
262 A TRP N 1 1 BCD 036 2.91
262 A TRP N 1 I BCD 026 3.54
262 A TRP CB 1 I BCD 036 3.58
262 A TRP CD2 1 1 BCD 026 3.79
262 A TRP CE3 1 1 BCD - 036 3.43
262 A TRP CE3 1 1 BCD C26 3.67
502 D TYR CA 1 1 BCD 031 3.58

 

164

 

 

 

 

502 D TYR CB 1 1 BCD 031 3.43
502 D TYR C 1 ] BCD 031 3.48
502 D TYR O 1 1 BCD C41 3.64
502 D TYR O 1 1 BCD C31 3.50
502 D TYR O 1 ] BCD 031 2.64
503 D HIS CE] I 1 BCD C22 3.74
503 D HIS CE] 1 1 BCD C12 3.73
At binding site H

# Chain Residue Atom # Chain Ligand Atom Distance(A)
505 B ASP 0 1 F BCD 062 3.61
508 B THR CB 1 F BCD 062 3.65
508 B THR CB 1 F BCD 052 3.39
508 B THR CB 1 F BCD 061 3.69
508 B THR CG] 1 F BCD 062 2.61
508 B THR 001 1 F BCD 052 3.00
508 B THR 001 1 F BCD C12 3.75
508 B THR C 1 F BCD 062 3.52
508 B THR C 1 F BCD 061 3.66
508 B THR O 1 F BCD C61 3.35
508 B THR O 1 F BCD 061 2.83
509 B PHE N 1 F BCD 062 3.12
509 B PHE CA 1 F BCD C62 3.78
509 B PHE CA 1 F BCD 062 3.31
509 B PHE CB 1 F BCD 062 3.72
509 B PHE CE2 1 F BCD C63 3.72
511 B ILE CG] 1 F BCD 067 3.41
511 B ILE C02 1 F BCD C61 2.89
511 B ILE C62 1 F BCD 061 2.54
512 B LEU CD1 1 F BCD C52 3.67
512 B LEU CD] 1 F BCD C62 3.44
512 B LEU CD1 1 F BCD C51 3.67
512 B LEU CD] 1 F BCD C61 2.67
512 B LEU CD1 1 F BCD 061 3.76
512 B LEU CD2 1 F BCD C63 3.71
512 B LEU CD2 1 F BCD C62 3.61
628 B TRP CD] 1 F BCD 051 3.48
628 B TRP CD1 1 F BCD C21 3.72

 

165

 

 

 

 

 

628 B TRP NE] 1 F BCD 061 2.79
628 B TRP NE] 1 F BCD 051 3.36
628 B TRP C E2 1 F BCD 06] 3.48
628 B TRP CZ2 1 F BCD O6] 3 .64
63] B VAL CG] ] F BCD 031 3.77
At binding site IV

# Chain Residue Atom # Chain Ligand Atom Distance(A)
542 C ASP CG 1 J BCD 027 3.3]
542 C ASP OD] 1 J BCD C27 3.50
542 C ASP OD] 1 J BCD 027 2.62
542 C ASP OD] 1 .1 BCD C17 3.67
542 C ASP OD] 1 J BCD 036 3.15
542 C ASP OD2 1 J BCD C27 3.72
542 C ASP OD2 1 J BCD 027 3.27
544 C TRP CB 1 J BCD 057 3.73
544 C TRP CG 1 J BCD 057 3.73
544 C TRP CE3 1 J BCD 037 3.75
545 C GLN CG 1 J BCD 037 3.60
545 C GLN CD 1 J BCD 037 3.69
545 C GLN OE] 1 J BCD 037 2.98
545 C GLN OE] 1 J BCD 027 3.45
659 C PRO CB ] J BCD 02] 3.50
659 C PRO C ] J BCD 02] 3 .67
659 C PRO 0 1 J BCD 031 3.27
659 C PRO 0 1 J BCD C21 3.18
659 C PRO 0 1 J BCD 021 2.71
661 C PRO CG 1 J BCD 022 3.67
661 C PRO CG 1 J BCD C41 3.72
661 C PRO CG 1 J BCD 031 3.67
661 C PRO CD 1 J BCD 022 3.60
66] C PRO CD 1 J BCD 031 3.23
689 C SER CB 1 J BCD C61 3.64
689 C SER CB 1 J BCD 061 3.38
689 C SER CB 1 J BCD 051 3.77
689 C SER 0G 1 J BCD C61 3.13
689 C SER 0G 1 J BCD 061 2.96
717 C PRO CG 1 J BCD C61 3.59

 

166

Appendix 10. Protein-ligand close contacts between BE and gamma CD. The
interactions listed are within a 3.8 A cutoff.

 

 

At binding site II

# Chain Residue Atom # Chain Ligand Atom Distance(A)
505 A ASP OD] B 1 E RCD C6D 3.75
505 A ASP OD2B I E RCD C I C 3.70
508 A THR CB I E RCD 05D 3.45
508 A THR 001 I E RCD C 1 D 3.79
508 A THR 001 I E RCD 05D 2.95
508 A THR 001 I E RCD C5D 3.55
508 A THR CG] 1 E RCD C6D 2.85
508 A THR 00 I I E RCD 06D 3.73
508 A THR C I E RCD 06E 3.66
508 A THR C I E RCD 06D 3.75
508 A THR O I E RCD C6E 3.54
508 A THR O I E RCD 06E 2.72
509 A PHE N I E RCD C6D 3.72
509 A PHE N 1 E RCD 06D 3.23
509 A PHE CA I E RCD 06D 2.98
509 A PHE CB I E RCD 06D 3.26
509 A PHE CG 1 E RCD 06D 3.79
509 A PHE CD2 1 E RCD 06D 3 .41
511 A ILE CB I E RCD 06E 3.65
5 I I A ILE CG2 I E RCD 06F 3.71
512 A LEU CD1 1 E RCD C6E 3.72
512 A LEU CD2 I E RCD C6E 2.81
512 A LEU CD2 1 E RCD 06E 3.62
512 A LEU CD2 1 E RCD 06D 3 .76
628 A TRP CD1 1 E RCD 05E 3.49
628 A TRP CD1 1 E RCD C2E 3.78
628 A TRP NE] 1 E RCD 05E 3.42
628 A TRP NE 1 1 E RCD 06E 3 .08

 

167

 

At binding site 111

 

 

# Chain Residue Atom # Chain Ligand Atom Distance(A)
159 D TRP C0 1 H RCD C6H 3.50
159 D TRP CD] I H RCD C6H 3 .66
159 D TRP CD2 1 H RCD C6H 3.65
159 D TRP CZ2 I H RCD C 10 3.68
189 D LYS CD I H RCD 021'] 3.58
189 D LYS CE I H RCD 03H 3.21
189 D LYS CE 1 H RCD 02H 3.35
I 89 D LYS NZ 1 H RCD 03H 3.28
189 D LYS NZ I H RCD C2H 3 .72
189 D LYS NZ 1 H RCD 021'] 2.67
20] D LEU CD] 1 H RCD C4A 3.79
201 D LEU CD] 1 H RCD C6A 3.59
201 D LEU CD 1 I H RCD 02H 3 .80
21 I D GLN CD 1 H RCD 030 3.67
2] 1 D GLN CD I H RCD 020 3.32
2] I D GLN CE] I H RCD 030 2.79
2]] D GLN OE] I H RCD C20 3.74
21 1 D GLN OE] 1 H RCD 020 3.48
2]] D GLN NE2 I H RCD C20 3.35
2]] D GLN NE2 1 H RCD 020 2.38
215 D GLU CB 1 H RCD 030 3.62
215 D GLU C0 I H RCD 03H 3.64
215 D GLU C0 1 H RCD C30 3.58
215 D GLU C0 1 H RCD 030 3.50
215 D GLU C0 1 H RCD C20 3.61
215 D GLU C0 1 H RCD 020 2.6 I
215 D GLU CD I H RCD C3H 3.73
215 D GLU CD I H RCD 03H 3.23
215 D GLU CD I H RCD 020 3.43
215 D GLU CE] I H RCD C3H 3.55
215 D GLU CE] I H RCD 03H 2.83
217 D ALA CB 1 H RCD 020 3.73

 

168

At binding site IV

 

 

# Chain Residue Atom # Chain Ligand Atom Distance(A)
542 C ASP CG ] 1 RCD 02F 3.51
542 C ASP OD] 1 1 RCD 02F 2.80
542 C ASP OD2 1 I RCD 02F 3.48
544 C TRP CG 1 ] RCD 05F 3.59
544 C TRP C E3 1 1 RCD C4F 3.79
545 C GLN OE] I 1 RCD 03F 3.44
659 C PRO CB 1 1 RCD 02E 3.62
659 C PRO C 1 l RCD 02E 3 .61
659 C PRO 0 1 1 RCD 03E 3.42
659 C PRO 0 1 1 RCD C2E 2.99
659 C PRO 0 1 1 RCD 02E 2.59
661 C PRO CG 1 1 RCD 03E 3.77
66] C PRO CD 1 1 RCD 03E 3.39
661 C PRO CD 1 1 RCD 02D 3.77
689 C SER CB 1 1 RCD C6E 3.73
689 C SER CB 1 I RCD 06E 3.48
689 C SER 0G ] 1 RCD C6E 3.56
689 C SER CG 1 1 RCD 06E 2.77
717 C PRO CG 1 1 RCD 05E 3.58

 

169

Appendix 1 l.

Protein-ligand close contacts between BE and maltoheptaose (M7). The
interactions listed are within a 3.8 A cutoff.

 

 

At binding site 111

# Chain Residue Atom # Chain Ligand Atom Distance(A)
159 D TRP NE] 1 K MAL 062 3.52
159 D TRP NE] 1 l K MAL 052 3.71
159 D TRP CE2 1 K MAL 052 3.71
159 D TRP C22 1 K MAL 052 3 .45
159 D TRP C22 1 K MAL C12 3.56
189 D LYS CD 1 K MAL 021 3.68
189 D LYS CE I K MAL 021 3.68
189 D LYS NZ 1 K MAL 031 3.21
189 D LYS NZ 1 K MAL C21 3.61
189 D LYS NZ 1 K MAL 021 2.60
211 D GLN CD 1 K MAL 032 3.79
211 D GLN CD 1 K MAL C22 3.63
211 D GLN OE] 1 K MAL 032 3.72
211 D GLN OE] l K MAL C22 3.02
21 1 D GLN OEI 1 K MAL 022 2.86
211 D GLN NE2 I K MAL C42 3.78
211 D GLN NE2 1 K MAL C32 3.58
21 I D GLN NE2 1 K MAL 032 3.06
21 I D GLN NE2 1 K MAL C22 3.48
215 D GLU CB 1 K MAL 032 3.22
215 D GLU CB 1 K MAL 022 3.67
215 D GLU CG 1 K MAL 032 3.30
215 D GLU CD 1 K MAL 032 3.11
215 D GLU OE] 1 K MAL C32 3.52
215 D GLU OE] 1 K MAL 032 2.24
215 D GLU O 1 K MAL 022 3.57
216 D THR O 1 K MAL 022 3.64

 

170

At binding site IV

 

 

# Chain Residue Atom # Chain Ligand Atom Distance(A)
542 C ASP CG 1 L MAL 021 3.56
542 C ASP OD] l L MAL 021 3.44
542 C ASP OD2 1 L MAL C21 3.73
542 C ASP OD2 1 L MAL 021 3.03
544 C TRP CB 1 L MAL 051 3.78
544 C TRP CG 1 L MAL 051 3.57
545 C GLN CG 1 L MAL 031 3.66
659 C PRO CB 1 L MAL 022 3.58
659 C PRO C 1 L MAL 022 3.75
659 C PRO 0 1 L MAL 032 3.35
659 C PRO 0 1 L MAL C22 3.10
659 C PRO 0 1 L MAL 022 2.68
661 C PRO CG 1 L MAL 032 3.39
661 C PRO CD 1 L MAL 032 3.00
689 C SER CB 1 L MAL 062 3.63
717 C PRO CG 1 L MAL 052 3 .67

 

171

At binding site V and VI

 

 

# Chain Residue Atom # Chain Ligand Atom Distance(A)
674 A LYS NZ 1 0 M7 037 3.17
254 B ARC! CA I 0 M7 067 3.72
254 B ARG C l 0 M7 067 3.70
255 B ARG N 1 0 M7 067 2.77
255 B ARG CA ] 0 M7 067 3.56
255 B ARO CB ] 0 M7 067 3.77
255 B ARO C I 0 M7 067 3.64
255 B ARG O I 0 M7 067 2.92
255 B ARG O l 0 M7 C67 3.48
257 B THR 00] I 0 M7 047 3.59
260 B ASN CD] 1 0 M7 066 3.57
537 B ASP C0 1 0 M7 022 3.3]
537 B ASP CD] I 0 M7 C32 3.69
537 B ASP CD] I 0 M7 032 3.00
537 B ASP CD] I 0 M7 022 3.53
537 B ASP OD2 l 0 M7 C22 3.60
537 B ASP OD2 I 0 M7 022 2.33
543 B ALA CB ] 0 M7 05] 3.68
546 B LYS CD I 0 M7 06] 3.58
546 B LYS CE I 0 M7 06] 3.67
546 B LYS NZ 1 0 M7 022 2.99
546 B LYS NZ 1 0 M7 06] 2.79
576 B ARG CZ ] 0 M7 025 3.13
576 B ARG NH] 1 0 M7 C25 3.73
576 B ARG NH] 1 0 M7 025 2.39
576 B ARG NH] 1 0 M7 C34 3.19
576 B ARG NH] I 0 M7 034 2.87
576 B ARG NH2 ] 0 M7 026 3.05
576 B ARG NH2 1 0 M7 C35 3.57
576 B ARG NH2 ] 0 M7 025 3.08
583 B SER CB 1 0 M7 067 3.28
583 B SER CB 1 0 M7 C66 3.78
583 B SER 00 l 0 M7 067 2.38
583 B SER 00 l 0 M7 C67 3.25
583 B SER 00 ] 0 M7 057 3.44
584 B LEU O ] 0 M7 036 3.62
585 B ASP CA I 0 M7 036 3.29
585 B ASP C0 I G M7 026 3.39

 

172

 

 

 

 

 

585 B ASP OD2 1 G M7 C26 3.44
585 B ASP OD2 1 G M7 026 2.49
585 B ASP C 1 G M7 036 3.39
586 B TRP N 1 G M7 036 3.28
586 B TRP CB 1 G M7 057 3.59
586 B TRP CB 1 G M7 C17 3.72
586 B TRP CD] 1 G M7 057 3.52
587 B HIS N 1 G M7 C27 3.77
587 B HIS N 1 G M7 027 3.46
587 B HIS N 1 G M7 C17 3.61
587 B HIS N ] G M7 036 3.22
587 B HIS CA 1 G M7 027 3.39
587 B HIS CB 1 G M7 027 3.47
587 B HIS CB 1 G M7 036 3.45
587 B HIS CG 1 G M7 036 3.23
587 B HIS ND] 1 G M7 C36 3.13
587 B HIS ND] 1 G M7 036 2.30
587 B HIS ND] 1 G M7 C26 3.77
587 B HIS ND] 1 G M7 026 3.20
587 B HIS ND] 1 G M7 C55 3.70
587 B HIS CE] 1 G M7 036 3.3]
587 B HIS CE] 1 G M7 026 3.16
587 B HIS CE] 1 G M7 C55 3.65
587 B HIS CE] ] G M7 055 3.77
590 B GLU OE] 1 G M7 027 3.00
595 B TRP CD1 1 G M7 011 3.64
595 B TRP NE] 1 G M7 011 3.32
596 B HIS CE] 1 G M7 031 3.62
596 B HIS NE2 1 G M7 031 3.10
At binding site VII

# Chain Residue Atom # Chain Ligand Atom Distance(A)
467 A SER CA I 1 MAL 06] 3.61
467 A SER C 1 1 MAL C61 3.40
467 A SER C 1 1 MAL 061 2.95
467 A SER O ] 1 MAL C61 3.15
467 A SER O 1 1 MAL 061 3.26

 

 

468
468
468
468
468
468
468
469
469
469
470
470
470
476
476
476
476
476
477
477
477
477
477
477
477
478
478
478
518
518
518

>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>

ARG
ARG
ARG
ARG
ARG
ARG
ARG
PRO
PRO
PRO
GLN
GLN
GLN
GLY
GLY
GLY
GLY
GLY
PHE
PHE
PHE
PHE
PH E
PHE
PH E
TRP
TRP
TRP
ASN
ASN
ASN

00000020000

0

E2
CZ2
C22
0D]
0D]
ND2

———i—_lfl——I————__‘—i—l~H—I———A—d—u—i——————

___—___—_———————_—_——_____—__——

MAL
MAL
MAL
MAL
MAL
MAL
MAL
MAL
MAL
MAL
MAL
MAL
MAL
MAL
MAL
MAL
MAL
MAL
MAL
MAL
MAL
MAL
MAL
MAL

MAL
MAL
MAL
MAL
MAL
MAL

C6]
06]
C61
06]
C61
06]
061
C6]
061
C61
0]]
011
011
06]
O6]
05]
051
C11
06]
06]
051
C5]
C6]
06]
05]
C2]
C2]
021
062
C62
C62

3.55
2.90
3.5]
3.25
3.43
3.02
3.09
3.5]
3.54
3.77
3.67
3.09
3.77
3.18
3.13
3.4]
3.23
3.4]
3.12
3.27
3.69
3.77
2.93
2.33
3.36
3.58
3.5]
3.60
3.07
3.50
3.33

 

174

Appendix 12. Protein-ligand close contacts between BE and maltohexaose (M6). The
interactions listed are within a 3.8 A cutoff.

 

 

At binding site 111

# Chain Residue Atom # Chain Ligand Atom Distance(A)
155 D GLN CG 1 F MAL O62 3.6]
155 D GLN NE2 I F MAL O62 3.46
155 D GLN NE2 ] F MAL C62 3.30
159 D TRP CO I F MAL O6] 3.72
159 D TRP CD] ] F MAL O6] 3.36
159 D TRP NE] 1 F MAL O62 3.55
159 D TRP NE] ] F MAL 052 3.50
159 D TRP NE] ] F MAL O6] 3.57
159 D TRP CZ2 1 F MAL C12 3.63
189 D LYS CD 1 F MAL O2] 3.47
189 D LYS CE ] F MAL O3] 3.66
189 D LYS CE ] F MAL O2] 3.17
189 D LYS NZ 1 F MAL O2] 2.92
20] D LEU CD] I F MAL 02] 3.39
20] D LEU CD] I F MAL C] I 3.5]
2]] D GLN CD I F MAL O32 3.72
2]] D GLN CD I F MAL O22 3.65
2] 1 D GLN OE] I F MAL C42 3.60
2]] D GLN OE] I F MAL C32 3.28
2]] D GLN OE] I F MAL O32 2.63
2]] D GLN OE] I F MAL C22 3.28
2]] D GLN OE] I F MAL O22 3.34
2]] D GLN NE2 I F MAL C22 3.79
2] I D GLN NE2 ] F MAL O22 3.14
215 D GLU C0 I F MAL O22 3.26
215 D GLU CG 1 F MAL O3] 2.93

 

175

At binding site IV

 

 

 

 

 

# Chain Residue Atom # Chain Ligand Atom Distance(A)
542 C ASP CG 1 E MAL 02] 3.42
542 C ASP GD] 1 E MAL C21 3.28
542 C ASP GD] 1 E MAL 021 2.56
542 C ASP GD] 1 E MAL C11 3.52
542 C ASP OD2 1 E MAL 021 3.66
544 C TRP CG 1 E MAL 051 3.65
544 C TRP NE] 1 E MAL C61 3.79
544 C TRP CE2 1 E MAL C6 '1 3.72
545 C GLN CG 1 E MAL 031 3.75
545 C GLN CD I E MAL O3 1 3.59
545 C GLN NE2 1 E MAL 031 3.60
545 C GLN NE2 I E MAL 021 3.12
659 C PRO 0 1 E MAL C22 3.71
659 C PRO 0 1 E MAL 022 2.66
661 C PRO CG 1 E MAL 032 2.97
661 C PRO CD 1 E MAL 032 3.03
689 C SER CB 1 E MAL 062 3.41
689 C SER CB 1 E MAL C62 3.52
689 C SER CB 1 E MAL 052 3.77
689 C SER CG 1 E MAL 062 3.09
689 C SER CG 1 E MAL C62 3.73
At binding site V

# Chain Residue Atom # Chain Ligand Atom Distance(A)
674 C LYS NZ 1 G MAL 032 2.78
254 D ARG CA I G MAL 062 3.21
254 D ARG CB 1 G MAL 062 3.61
254 D ARG CB 1 G MAL C62 3.76
254 D ARG CG 1 G MAL C62 3.76
254 D ARG C I G MAL 062 3.30
255 D ARG N 1 G MAL 062 2.57
255 D ARG N 1 G MAL C62 3.64

 

176

 

 

255
255
255
255
257
257
260
576
583
583
583
583
583
584
584
585
585
585
585
586
586
586
587
587
587
587
587
587
587
590
590
590
590

000000000000000000000000000000000

ARG
ARG
ARG
ARG
THR
THR
ASN
ARG
SER
SER
SER
SER
SER
LEU
LEU
ASP
ASP
ASP
ASP
TRP
TRP
TRP
HIS
HIS
HIS
HIS
HIS
HIS
HIS
GLU
GLU
GLU
GLU

CA

CG2
CG2
OD]
NH2
CB
CB
0G
0G
0G

CA
0D2
OD2

CB
CD]

CB

CG

ND]
ND]
ND]
CE]
CD

0E]
0E1
0E1

~flfl_fl—l—II——l—Afl—l—i~—_fl—fl~—l—i—n__‘—fi—i—nflflflflfl

000000000000000000000000000000000

MAL
MAL
MAL
MAL
MAL
MAL
MAL
MAL
MAL
MAL
MAL
MAL
MAL
MAL
MAL
MAL
MAL
MAL
MAL
MAL
MAL
MAL
MAL
MAL
MAL
MAL
MAL
MAL
MAL
MAL
MAL
MAL
MAL

O62
O62
062
C62
C 32
C61
06]
021
062
051
062
C62
052
03]
C2]
031
C2]
02]
03]
031
052

052 '

03]
03]
03]
C3]
03]
02]
021
022
032
C22
022

3.6]
3.76
3.06
3.38
3.80
3.76
3.60
3.59
3.17
3.50
2.54
3.57
3.58
3.48
3.38
3.49
3.62
2.9]
3.50
3.2]
3.68
3.73
3.06
3.6]
3.62
3.53
2.82
3.29
3.44
3.65
3.76
3.22
2.64

 

177

    

  

211111111

11]

  

1

  

111

111

  

11] 11111111111111]!