Download Mechanistic Studies of Two Selected Flavin

Georgia State University
ScholarWorks @ Georgia State University
Chemistry Dissertations
Department of Chemistry
8-11-2011
Mechanistic Studies of Two Selected FlavinDependent Enzymes: Choline Oxidase and DArginine Dehydrogenase
Hongling Yuan
Georgia State University
Follow this and additional works at: http://scholarworks.gsu.edu/chemistry_diss
Recommended Citation
Yuan, Hongling, "Mechanistic Studies of Two Selected Flavin-Dependent Enzymes: Choline Oxidase and D-Arginine
Dehydrogenase." Dissertation, Georgia State University, 2011.
http://scholarworks.gsu.edu/chemistry_diss/56
This Dissertation is brought to you for free and open access by the Department of Chemistry at ScholarWorks @ Georgia State University. It has been
accepted for inclusion in Chemistry Dissertations by an authorized administrator of ScholarWorks @ Georgia State University. For more information,
please contact [email protected].
MECHANISTIC STUDIES OF TWO SELECTED FLAVIN-DEPENDENT ENZYMES:
CHOLINE OXIDASE AND D-ARGININE DEHYDROGENASE
by
HONGLING YUAN
Under the Direction of Giovanni Gadda
ABSTRACT
Choline oxidase catalyzes the flavin-dependent, two-step oxidation of choline to glycine
betaine via the formation of an aldehyde intermediate. The oxidation of choline includes two
reductive half-reactions followed by oxidative half-reactions. In the first oxidation reaction, the
alcohol substrate is activated to its alkoxide via proton abstraction and oxidized via transfer of a
hydride from the alkoxide α-carbon to the N(5) atom of the enzyme-bound flavin. In the wildtype enzyme, proton and hydride transfers are mechanistically and kinetically uncoupled.
The role of Ser101 was investigated in this dissertation. Replacement of Ser101 with
threonine, alanine, cysteine, or valine demonstrated the importance of the hydroxyl group of
Ser101 in proton abstraction and in hydride transfer. Moreover, the kinetic studies on the
Ser101Ala variant have revealed the importance of a specific residue for the optimization of the
overall turnover of choline oxidase. The UV-visbible absorbance of Ser101Cys suggests Cys101
can form an adduct with the C4a atom of the flavin. The mechanism of formation of the C4acysteinyl adduct has been elucidated.
D-arginine dehydrogenase (DADH) catalyzes the oxidation of D-amino acids to the
corresponding imino acids, which are non-enzymatically hydrolyzed to α-keto acids and
ammonia. The enzyme is strick dehrogenase and deoesnot react with molecular oxygen. Steady
state kinetic studies wirh D-arginine and D-histidine as a substrate and PMS as the electron
acceptor has been investigated. The enzyme has broad substrate specificity for D-amino acids
except aspartate, glutamate and glycine, with preference for arginine and lysine. Leucine is the
slowest substrate in which steady state kinetic parameters can be obtained. The chemical
mechanism of leucine dehydrogenation catalyzed by DADH was explored with a combination of
pH, substrate and solvent kinetic isotope effects (KIE) and proton inventories by using rapid
kinetics in a stopped-flow spectrophotometer. The data are discussed in the context of the
crystallographic structures at high resolutions (<1.3 Å) of the enzyme in complex with
iminoarginine or iminohistidine.
INDEX WORDS: Choline oxidase, Hydroxyl group, C4a-cysteinyl adduct, D-arginine
dehydrogenase, Conformational change, Hydride transfer
MECHANISTIC STUDIES OF TWO SELECTED FLAVIN-DEPENDENT ENZYMES:
CHOLINE OXIDASE AND D-ARGININE DEHYDROGENASE
by
HONGLING YUAN
A Dissertation Submitted in Partial Fulfillment of the Requirements for the Degree of
Doctor of Philosophy
in the College of Arts and Sciences
Georgia State University
2011
Copyright by
Hongling Yuan
2011
MECHANISTIC STUDIES OF TWO SELECTED FLAVIN-DEPENDENT ENZYMES:
CHOLINE OXIDASE AND D-ARGININE DEHYDROGENASE
by
HONGLING YUAN
Committee Chair:
Committee:
Dr. Giovanni Gadda
Dr. Al Baumstark
Dr. Jenny J. Yang
Electronic Version Approved by:
Office of Graduate Studies
College of Arts and Sciences
Georgia State University
August 2011
iv
ACKNOWLEDGMENTS
Five years of Ph.D. life will end with writing of the dissertation, but in my life it is just a
comma. I will face the beginning of another journey. This dissertation would not have been
possible without the help, support, guidance and efforts of a lot of people because of whom my
graduate experience has been one that I will cherish forever. I would like to take this opportunity
to express my gratitude to them for completing the dissertation successfully.
My deepest gratitude is to my advisor, Dr. Giovanni Gadda, for his excellent guidance,
encouragement, caring, supervising and providing me with an excellent atmosphere for doing
research. His encouragement made me feel confident to fulfill my desire and to overcome every
difficulty I encountered. During my studies, he taught me not only the enzymology knowledge
but also his stringent academic attitude to be a good scientist. Until now I just realize without his
“push”, I can never improve myself. I cannot say thank you enough for his tremendous support
and help.
I am grateful to my committee members, Dr. Baumstark and Dr. Yang, for their
encouragement, practical advice and challenge throughout my academic program. Dr.
Baumstark’s encouraging words inspire me to approach my work as a social scientist. Dr. Yang,
I would like to thank you for your motherly care in my work and my personal life. Without their
support, I could not have done what I was able to do. I am also grateful to the all the professors
for their advice and teaching during my Ph.D. studies.
I would also like to thank members past and present in Dr. Giovanni Gadda’s group:
Osbourn, Steffan, Kunchala, Slavica, Kevin, Tran Bao, Merid, Andrea, Stephen, Danila,
Sharonda, Lydia, Philip, Swathi, Crystal, Rizvan and Sandra for sharing their enthusiasm and
comments on my work. It would have been a lonely lab without them. Osbourne, you are the
v
first labmate I met, I am so glad to have you and I still remember you kindly teach me how to
pronounce English. To Slavica, thank you for encouraging me to be a strong woman. I really
enjoy the drinking break and conversation about life. I thank Swathi for helping me with the tone
and my writing. I really had fun in the step class we took. I have been especially fortunate to
have Andrea and Kunchala during my PhD studies. I cannot find any words to express my
thanks. They are like my brother and sister. Without Kunchala, I cannot get used to the abroad
life so easily. Every time when I have problem I know you are there and you help me. Andrea,
thank- you for your patience and time to listen to my complaints. You have always listened when
I needed to talk. The care, discipline and organization that you bring to this work have had a
beneficial effect on my work.
I am also thankful to the following former or current staff at GSU: Edna, Sina, Diane, Will,
Chris, Dan, Stephanie, Robert, Don Hardon etc. your generous and timely help and various
forms of support make my graduate studies easy and enjoyable.
I’d like to convey my heartfelt thanks to my forgiving, generous and loving friends. They
have helped me stay sane and stay focused on my graduate study when I am depressed. Their
care and patience on teaching me driving and cooking helped me overcome the difficulties I met
in the US and without them I could not have survived the process. I greatly value their friendship
and deeply appreciate their help in my personal life.
Most importantly, I would also like to thank my mom, who has always supported,
encouraged and believed in me. My brothers and sisters, their understanding and love
encouraged me to work hard and to continue pursuing a Ph.D. project abroad. They always give
their best wishes to me. Finally, I appreciate the financial support from NSF that funded the
research discussed in this dissertation.
vi
TABLE OF CONTENTS
ACKNOWLEDGMENTS ............................................................................................................. iv
LIST OF TABLES ...........................................................................................................................x
LIST OF FIGURES ....................................................................................................................... xi
LIST OF SCHEMES......................................................................................................................xv
CHAPTER 1 ..................................................................................................................................17
Introduction ....................................................................................................................................17
1.1 Flavoproteins ........................................................................................................................ 17
1.2 Mechanisms of Oxidation of Carbon-Heteroatom Bond ...................................................... 19
1.3 Protein Hydroxyl Group Proximal to the C(4a)-N(5) Flavin Atoms.................................... 24
1.3.1
Physical Properties of Hydroxyl Group .............................................................. 24
1.3.2
Conserved Hydroxyl Group in Flavoenzymes .................................................... 25
1.4 Specific Goals ....................................................................................................................... 37
1.5 References ............................................................................................................................ 41
Chapter 2 ........................................................................................................................................58
Structural and Kinetic Studies on the Ser101Ala Variant of Choline Oxidase: Catalysis by
Compromise ...................................................................................................................................58
2.1 Abstract ................................................................................................................................. 58
2.2 Introduction .......................................................................................................................... 59
2.3 Experimental Procedures ...................................................................................................... 61
2.4 Results .................................................................................................................................. 64
2.5 Discussion ............................................................................................................................. 71
CHAPTER 3 ..................................................................................................................................79
vii
Importance of a Serine Proximal to the C(4a)-N(5) Flavin Atoms for Hydride Transfer in
Choline Oxidase .............................................................................................................................79
3.1 Abbreviations........................................................................................................................ 79
3.2 Abstract ................................................................................................................................. 79
3.3 Introduction .......................................................................................................................... 80
3.4 Experimental Procedures ...................................................................................................... 83
3.5 Results .................................................................................................................................. 86
3.6 Discussion ............................................................................................................................. 94
3.7 Appendix ............................................................................................................................ 102
3.8 Support Information ........................................................................................................... 105
3.9 Acknowledgment ................................................................................................................ 107
3.10 References .......................................................................................................................... 107
CHAPTER 4 ................................................................................................................................114
Mechanism of Flavin C4a-Adduct Formation in a Choline Oxidase Variant .............................114
4.1 Abstract ............................................................................................................................... 114
4.2 Introduction ........................................................................................................................ 114
4.3 Experimental Procedures .................................................................................................... 115
4.4 Results ................................................................................................................................ 120
4.5 Discussion ........................................................................................................................... 130
4.6 Support Information ........................................................................................................... 135
4.7 References .......................................................................................................................... 136
CHAPTER 5 ................................................................................................................................140
viii
Conformational Changes and Substrate Recognition in Pseudomonas aeruginosa D-Arginine
Dehydrogenase .............................................................................................................................140
5.1 Abreviations........................................................................................................................ 140
5.2 Abstract ............................................................................................................................... 140
5.3 Introduction ........................................................................................................................ 141
5.4 Experimental Procedures .................................................................................................... 143
5.5 Results ................................................................................................................................ 146
5.6 Discussion ........................................................................................................................... 162
5.7 Acknowledgment ................................................................................................................ 166
5.8 References .......................................................................................................................... 166
CHAPTER 6 ................................................................................................................................173
Steady State Kinetic Mechanism and Redutive Half-Reaction of D-Arginine Dehydrogenase
from Pseudomonas aeruginosa....................................................................................................173
6.1 Abbreviations...................................................................................................................... 173
6.2 Abstract ............................................................................................................................... 173
6.3 Introduction ........................................................................................................................ 174
6.4 Experimental Procedures .................................................................................................... 175
6.5 Results ................................................................................................................................ 180
6.6 Discussion ........................................................................................................................... 186
6.7 Acknowledgment ................................................................................................................ 196
6.8 References .......................................................................................................................... 196
CHAPTER 7 ................................................................................................................................205
On the Catalytic Mechanism of D-Arginine Dehydrogenase ......................................................205
ix
7.1 Abstract ............................................................................................................................... 205
7.2 Introduction ........................................................................................................................ 205
7.3 Experimental Section .......................................................................................................... 210
7.4 Results ................................................................................................................................ 213
7.5 Discussion ........................................................................................................................... 219
7.6 Support Information ........................................................................................................... 223
7.7 References .......................................................................................................................... 224
CHAPTER 8 ................................................................................................................................231
General Discussion and Conclusions ...........................................................................................231
8.1 References .......................................................................................................................... 237
x
LIST OF TABLES
Table 1.1 List of Flavoenzymes with Hydroxyl Group in the Active Site. .................................. 25
Table 2.1 Crystallographic Data Collection and Refinement Statistics. ....................................... 67
Table 2.2 Comparison of the Kinetic Parameters of Ser101Ala and Wild-Type Choline Oxidase
at pH10.0. ...................................................................................................................................... 71
Table 3.1 Comparison of Spectral and Catalytic Properties of Ser101 Mutant Enzymes with
Wild-type Choline Oxidase........................................................................................................... 87
Table 3.2 Limiting Rate Constants for the Fast and Slow Kinetic Phases of Flavin Reduction
with Choline and 1,2-[2H4]-Choline as Substrate for Ser101 Mutant Enzymes ........................... 90
Table 4.1 pH-Dependence of the Kinetic Parameters with Tris Base for Ser101Cys in 50 mM
Buffer, at 25 oC ........................................................................................................................... 124
Table 4.2 Comparison of the pKa Values of Choline Oxidase Mutant Enzymes with Wild-Type
..................................................................................................................................................... 129
Table 5.1 Crystallographic Data Collection and Refinement Statistics ...................................... 147
Table 5.2 Steady State Kinetics of DADH ................................................................................. 158
Table 6.1 Purification of Untagged P. aeruginosa D-Arginine Dehydrogenase Expressed in E.
coli Rosetta(DE3)pLysS ............................................................................................................. 181
Table 6.2 Kinetic Parameters of D-Arginine Dehydrogenase with D-Arginine and D-Histidine
..................................................................................................................................................... 186
Table S7.1 Reductive Half-Reaction Kinetic Parameters for DADH with Leu as Substrate ..... 223
Table S7.2 Steady State Kinetic Parameters for DADH with Leu as Substrate ......................... 223
xi
LIST OF FIGURES
Figure 1.1 The structures of isoalloxazine, FMN and FAD. Modified from ref. (6). ................... 17
Figure 1.2 Redox states of flavin. Modified from ref. (2). ........................................................... 18
Figure 1.3 The active site of pyranose-2 oxidase (PDB code 1TT0)............................................ 27
Figure 1.4 The active site of flavocytochrome b2 from baker’s yeast Saccharomyces cerevisiae
(PDB code 1FCB). ........................................................................................................................ 29
Figure 1.5 The proton relay in class 2 dihydroorotate dehydrogenase (PDB code 1f78). ............ 31
Figure 1.6 The “aromatic sandwich” in human MAO B (PDB code 1GOS). .............................. 37
Figure 1.7 Left panel, glycine betaine as an osmoprotectant for S. meliloti. Right panel, Growth
under 0.4 M NaCl in control cells Synechococcus sp. PCC 7942 (PAM) and transformed cells
(PAMCOD) in which the codA gene (for choline oxidase) was introduced. Modified from ref.
(110), taken without permission. .................................................................................................. 38
Figure 1.8 Active site of choline oxidase (PDB# 2jbv) ................................................................ 39
Figure 2.1 Comparison of crystal structures of the Ser101Ala and wild-type enzymes of choline
oxidase.. ........................................................................................................................................ 68
Figure 2.2 Anaerobic Double reciprocal plots of Ser101Ala catalyzed oxidation of choline and
betaine aldehyde............................................................................................................................ 69
Figure 2.3 Reductive half-reaction of the Ser101Ala enzyme with betaine aldehyde.................. 70
Figure 3.1 Comparison of the active sites of wild-type and Ser101Ala choline oxidase. ............ 83
Figure 3.2 Anaerobic reduction of the Ser101 variants with choline as a substrate in 50 mM
sodium pyrophosphate at pH 10.0 and 25 °C ............................................................................... 89
Figure 3.3 Anaerobic substrate reduction of the Ser101Ala enzyme with choline (black) and 1,2[2H4]-choline (red) as substrates.. ................................................................................................. 91
xii
Figure 3.4 Anaerobic substrate reduction of Ser101Ala with choline as substrates in H2O (black)
and D2O (red). .............................................................................................................................. 92
Figure 3.5 Simulated reaction traces with a two-step model of a→b→c. .................................... 93
Figure 3.6 Dependence of the rate constant for hydride ion transfer on the hydrophobicity of the
amino acid residue at position 101 in choline oxidase. .............................................................. 100
Figure S3.1 Dependence of the rate constant for hydride ion transfer on the hydrophobicity of the
amino acid residue at position 101 in choline oxidase ............................................................... 105
Figure S3.2 Dependence of the rate constant for hydride ion transfer on the hydrophobicity of the
amino acid residue at position 101 in choline oxidase. .............................................................. 105
Figure S3.3 Dependence of the rate constant for hydride ion transfer on the hydrophobicity of the
amino acid residue at position 101 in choline oxidase. .............................................................. 105
Figure S3.4 Dependence of the rate constant for hydride ion transfer on the hydrophobicity of the
amino acid residue at position 178 in class 2 dihydroorotate dehydrogenase, where the wild-type
is Thr178 ..................................................................................................................................... 106
Figure 4.1 UV-visible absorbance spectra of S101X at pH 8.0. ................................................. 120
Figure 4.2 Inhibition of choline oxidase with Tris. Initial rates were measured in air-saturated 50
mM sodium phosphate at pH 6.0 and 25 oC ............................................................................... 121
Figure 4.3 Formation of flavin C4a adduct with Tris in 50 mM sodium phosphate, at pH 7.0 and
25 oC............................................................................................................................................ 123
Figure 4.4 pH-Dependence of kfor (panel A solid circle), krev (panel A empty circle) and appKd
(panel B) values of formation of flavin C4a adduct with Tris base. Reactions were performed in
50 mM buffer at 25 oC. ............................................................................................................... 125
xiii
Figure 4.5 Proton inventories of kfor (solid circle) and krev (empty circle) values of formation of
flavin C4a adduct at pL 8.8. ........................................................................................................ 126
Figure 4.6 Effect of pH on the spectral properties of CHO-WT and S101C in buffer B ........... 128
Figure S4.1 Effect of pH on the spectral properties of CHO-Ser101X and His99Asn in buffer A
..................................................................................................................................................... 135
Figure 5.1 Overall structure of DADH with iminoarginine. The DADH structure is shown in
cartoon representation. ................................................................................................................ 148
Figure 5.2 (A) The 2Fo-Fc electron density map of FAD contoured at 2σ illustrates the high
quality of the 1.06 Å resolution structure. (B) Schematic diagram of the flavin-apoprotein
interactions in the product-bound DADH conformation. ........................................................... 149
Figure 5.3 DADH interactions with iminoarginine (A, B) and iminohistidine (C, D).. ............. 152
Figure 5.4 Conformational flexibility of DADH structure. ........................................................ 155
Figure 5.5 Comparison of DADH structures in complex with iminoarginine and iminohistidine
..................................................................................................................................................... 157
Figure 5.6 Structural comparison of DADH with pDAAO (A) and LAAO (B). ....................... 160
Figure 5.7 Comparison of the active sites of DADH, pDAAO and LAAO. .............................. 161
Figure 6.1 Lack of reactivity of D-arginine dehydrogenase with molecular oxygen. ................ 182
Figure 6.2 Anaerobic reduction of the D-arginine dehydrogenase with D-histidine as a substrate
in 20 mM Tris-Cl, pH 8.7 and 25 oC........................................................................................... 183
Figure 6.3 Steady state kinetics for the oxidation D-arginine (Panel A) or D-histidine (Panel B)
catalyzed by D-arginine dehydrogenase. .................................................................................... 185
Figure 6.4 Effects of solvent viscosity on the steady state kinetic parameters for the D-arginine
dehydrogenase with D-arginine or D-histidine as a substrate.. .................................................. 185
xiv
Figure 6.5 Structural comparison of active sites of D-arginine dehydrogenase/iminoarginine
(yellow) and D-arginine dehydrogenase/iminohistidine (grey) .................................................. 190
Figure 6.6 Active site of D-arginine dehydrogenase (PDB number 3NYE) .............................. 194
Figure 7.1 Active site of the DADH structure with iminoarginine............................................. 209
Figure 7.2 Effects of pH on the kinetic parameters of RHR (solid circle) and SSK (empty circle)
for DADH with leucine as the reducing substrate. ..................................................................... 214
Figure 7.3 Effects of pH on the kinetic parameters of RHR (solid circle) and SSK (empty circle)
for DADH with leucine as the reducing substrate ...................................................................... 215
Figure 7.4 Observed rate constants as a function of substrate concentration ............................. 216
Figure 7.5 Proton inventory for the reduction rate, kred, using leucine at pL 10.3. ..................... 217
xv
LIST OF SCHEMES
Scheme 1.1 Biological Functions that Involve Flavoenzymes. Modified from ref (1). ............... 18
Scheme 1.2 Catalytic Cycle of Flavoenzymes. Modified from ref. (7). ....................................... 19
Scheme 1.3 Oxidation of Carbon-Heteroatom Bonds Catalyzed by Flavoenzymes. ................... 19
Scheme 1.4 Proposed Mechanisms of Flavin Reduction Catalyzed by Flavoenzymes. ............... 20
Scheme 1.5 Proposed Carbanion Mechanism of β-Cl-Alanine with DAAO. .............................. 21
Scheme 1.6 Proposed Hydride Transfer Mechanism of DAAO. .................................................. 22
Scheme 1.7 The Structure of Hydroxyl Amino Acids. ................................................................. 24
Scheme 1.8 Hydride Transfer Mechanism for the Oxidation of Choline to Betaine Aldehyde
Catalyzed by Choline Oxidase ...................................................................................................... 39
Scheme 1.9 Mechanism Proposed for Oxidation of D-Amino Acids by DADH. ........................ 40
Scheme 2.1 The Steady State Kinetic Mechanism of Choline Oxidation Catalyzed by Choline
Oxidase. ........................................................................................................................................ 61
Scheme 3.1 The Two-step Oxidation of Choline Catalyzed by Choline Oxidase. ....................... 82
Scheme 3.2 Proposed Kinetic Mechanism for Reductive Half-Reaction of Ser101 Variants of
Choline Oxidase ............................................................................................................................ 97
Scheme 3.3 Roles of Ser101 in the Cleavages of the OH (a) and CH (b) Bonds of Choline. ...... 97
Scheme A3. 1 .............................................................................................................................. 102
Scheme 4.1 Proposed Formation of Flaivn C4a Adduct Mechanism from the Kinetics of the
Attack Steps ............................................................................................................................... 122
Scheme 4.2 Proposed Fromation of Flavin C4a Adduct Mechanism in 50 mM Sodium
Pyrophosphate, pH 8.8 ................................................................................................................ 130
xvi
Scheme 4.3 Proposed Formation of Flaivn C4α Adduct Mechanism from pH Titration in the
Present and Absent of Tris Base ................................................................................................. 134
Scheme 6.1 Oxidation of D-Arginine by D-Arginine Dehydrogenase. ...................................... 174
Scheme 6.2 Proposed Steady State Kinetic Mechanism for the Oxidation of D-Arginine
Catalyzed by D-Arginine Dehydrogenase .................................................................................. 187
Scheme 6.3 Isomerization of the Michaelis Complex with D-Histidine .................................... 191
Scheme 7.1 Proposed Mechanisms of Amine Oxidation............................................................ 208
Scheme 7.2 Oxidation of D-Arginine Dehydrogenase by DADH .............................................. 209
17
CHAPTER 1
Introduction
1.1
Flavoproteins
Flavoproteins contain either flavin mononucleotide (FMN), flavin adenine dinucleotide
(FAD) or both as a cofactor. The core function group of flavins is the isoalloxazine ring, as
shown in Figure 1.1. Flavoproteins are important in metabolic pathways, such as DNA repair,
halogenations, light emission, protein folding and chromatin remodeling, as shown in Scheme
1.1 (1). Flavoproteins can transfer one or two electrons in the reaction. Therefore, a flavin can
exist in three redox states: oxidized, semiquinone (one-electron reduced) and hydroquinone (twoelectron reduced) (2), as shown in Figure 1.2. Each different state shows spectroscopic
properties which allow kinetic studies of flavoproteins. The UV-visible absorbance spectrum of
the oxidized flavin shows bands at ~ 360 nm and ~ 450 nm. For the neutral semiquinone, a broad
absorbance at long wavelength, 580-620 nm, has been observed whereas the anionic
semiquinone shows a strong absorbance around 380 nm and an extra sharp peak at 490 nm (3, 4).
The neutral hydroquinone has characteristic absorbance at ~ 295 nm and ~ 395 nm whereas the
anionic hydroquinone has absorbance around 285 nm and 340 nm (5).
OH OH
R=
OH
O
O
P
O
O
flavin mononucleotide (FMN)
O
2
N
10a
4
4a
1
10
OH OH
9a
9
8
R=
7
3
HN
R
N
5
5a
6
N
O
7,8-dimethylisoalloxzine
OH
O
O CH2
O P O P
O
OO O
HO
N
OH N
flavin adenine dinucleotide (FAD)
Figure 1.1 The structures of isoalloxazine, FMN and FAD. Modified from ref. (6).
N
NH2
N
18
energy production
reduction
R
N
oxidation
oxygenaton
chromatin remodeling
DNA repair
N
protein folding
NH
N
neural development
O
light emission
apoptosis
O
detoxification
non-redox reactions
biodegradation
biosynthesis
Scheme 1.1 Biological Functions that Involve Flavoenzymes. Modified from ref (1).
R
N
-
1e , H
R
N
N
N
+
O
NH
O
Oxidized (yellow)
1e-
N
O
1e-, H+
H
N
O
NH
N
O
H
Neutral Hydroquinone
(colorless)
NH
N
H
O
Neutral Semiquinone
(blue)
H+ pKa 8.3
R
N
N
O
R
N
-
+
H+ pKa 6.7
R
N
N
O
1e ,H
N
NH
O
Anionic Semiquinone
(red)
NH
N
H
O
Anionic Hydroquinone
(colorless)
Figure 1.2 Redox states of flavin. Modified from ref. (2).
Flavoenzymes catalyze a variety of reactions in which the enzyme is reduced by the substrate
and the turnover is completed with the oxidation of flavin by an electron acceptor, as shown in
Scheme 1.2. Based on the electron acceptor, the enzymes with molecular oxygen as electron
acceptor and hydrogen peroxide as product are named oxidase. Those using other electron
acceptors instead of oxygen are dehydrogenases or transferases. Monooxygenases insert a single
oxygen atom into the substrate and form a transient C4a-hydroperoxide intermediate.
19
EFLox
AH2
BH2
Reductive
half-reaction
Oxidative
half-reaction
EFLH-
A
B
Scheme 1.2 Catalytic Cycle of Flavoenzymes. Modified from ref. (7).
1.2
Mechanisms of Oxidation of Carbon-Heteroatom Bond
Many flavoenzymes catalyze the oxidation of a carbon-heteroatom bond, including oxidation
of alcohols, α-hydroxy acids, α-amino acids and amines (Scheme 1.3). The oxidation of alcohols
has been shown to be consistent with a hydride transfer mechanism. The hydroxyl proton of the
substrate is abstracted by an active site base and a hydride ion is transferred from the substrate αcarbon to the flavin N5 atom (8, 9), as shown in Scheme 1.4a. The reaction can be catalyzed by
the GMC oxidoreductase enzyme superfamily, which includes choline oxidase (10, 11), pyranose
2-oxidase (12), glucose oxidase (13, 14), cholestrerol oxidase (15), methanol oxidase (16) and
cellobiose dehydrogenase (17).
H
R
FAD FADH2
OH
(a)
R
H
O
(b)
FAD
COO
R
H
(c) NH3
(d)
FADH2
O
O
H
O
O
OH
O
R1
H
H R2
C NH
H R3
FAD FADH2
COO
R
+ H2O
COO
+ NH4
R
O
NH2
FAD
FADH2 R
R2
1
C N
R3
H
O
+ H 2O
R1
H
+
R2
H2
N
R3
Scheme 1.3 Oxidation of Carbon-Heteroatom Bonds Catalyzed by Flavoenzymes.
20
(a) Hydride transfer mechanism
R1
N
N
NH
N
H
R2
R1
N
O
O
O
H H
N
NH
N
H
O
O
R2
B
O
HB
H
(b) Carbanion mechanism
R
N
N
O
NH
N
R
N
O
NH
N
N
O
NH
N
O
H O
H3C C NH2
COO
H3C C NH2
COO
B:
O
H
H3C C NH2
COO
N
R
N
N
O
NH
N
O
H
R C NH2
H
R
N
N
NH
N
H
R C NH2
H
R
N
O
N
H
O
R
C
H
N
O
NH
O
NH2
(d) Nucleophilic mechanism
R
N
N
N
H
O
R C NH2
H
O
NH
R
N
N
H
N
R
N
O
NH
O
R C NH2
H
R
C
H
N
N
H
O
NH2
N
H
N
O
NH
O
NH2
H 3C C
COO
(c) Radical mechanism
R
N
R
N
O
NH
Scheme 1.4 Proposed Mechanisms of Flavin Reduction Catalyzed by Flavoenzymes.
21
In contrast, the chemical mechanism of amine oxidation by flavoenzymes, such as D-amino
acid oxidase (DAAO) and monoamine oxidase (MAO), has been debated for years. In addition to
the hydride transfer mechanism, several mechanisms were also proposed for flavoenzyme
reductive half-reactions. As shown in Scheme 1.4b, the carbanion mechanism involves the
formation of a carbanionic intermediate after abstraction of the substrate α-proton (18, 19). In the
radical mechanism, a single-electron is transferred from the amine nitrogen to the flavin and the
substrate α-proton is removed forming a carbanion radical that transfers a second electron to the
flavin (Scheme 1.4c) (20, 21). It has been proposed that in the nucleophilic mechanism amine
attacks the C4a position of the flavin, leading to the formation of a C4a adduct intermediate
(Scheme 1.4d) (22).
The carbanion mechanism is proposed based on the reaction of oxidation of β-Cl-alanine by
DAAO (18, 19). As shown in Scheme 1.5, the observation of the elimination of HCl and the
oxidation of β-Cl-alanine to β-Cl-pyruvate suggests the formation of a carbanion intermediate.
Moreover, the same substrate isotope effects with α-2H-Cl-alanine on the formation of β-Clpyruvate and pyruvate support the conclusion that a C-H bond cleavage occurs in a common
intermediate.
B:
H
H2C C NH3
COO
Cl
FAD
H2C C NH3
Cl
COO
Carbanion intermediate
FADH
H2C C NH2
Cl
COO
H2 O
H2C C NH2
Cl
COO
NH4
Oxidation reaction
ClH2C C NH3
COO
H2 O
H3C C NH2
H3C C NH2
COO
COO
Elimination reaction
NH4
Scheme 1.5 Proposed Carbanion Mechanism of β-Cl-Alanine with DAAO.
However, substrate isotope effect studies of DAAO containing 5-deazaFAD with α-3Halanine showed that the reduced flavin at N5 position was tritium-labeled, which suggested
22
hydride from the substrate α-carbon was transferred to the modified form of the cofactor (23,
24). The crystal structures also support a hydride transfer mechanism when pig kidney DAAO
structures were solved in 1996 (25, 26). Based on crystal structure, there is no active site base
which is required in carbanion mechanism properly placed to abstract the substrate α-carbon.
The isotope effects studies on DAAO carried out by Fitzpatrick’s group also show that the
reaction mechanism is consistent with hydride transfer (27-29). First, the pH dependence of the
15
N isotope effect on the kcat/Km values with D-serine as a substrate suggested that a proton was
released from free D-serine to the enzyme-bound amino acid and it is at least a partially ratelimiting step (27-29). However, lack of solvent isotope effect on the kcat/Km indicated that the
proton released from D-serine does not occur when the C-H bond is cleaved (28). Therefore,
they proposed that there is a proton released in a pre-equilibrium state and a hydride transfer
from the α-carbon atom of the anionic amine to the flavin N(5) atom as shown in Scheme 1.6.
Computational studies on the free energy profile of the reaction and the electronic distribution
have also supported a hydride transfer mechanism from the anionic amino acid to the flavin
rather than a carbanion mechanism (30).
R
N
N
N
O
H
R C NH3
COO
R
N
O
NH
N
N
H
R
N
O
NH
O
H
R C NH2
COO
N
H
R
N
O
NH
O
NH2
C
COO
Scheme 1.6 Proposed Hydride Transfer Mechanism of DAAO.
Two other mechanisms for the reductive half reaction have been proposed based on the
studies of MAO: a radical mechanism and a nucleophilic mechanism. The radical mechanism is
supported by the inhibition studies of MAO with cyclopropyl or cyclobutyl rings from McEwen
23
(31), Paech (32) and Silverman (20, 21) groups. The studies show that the inhibitor bound either
to the protein or the flavin to inactive the MAO. Silverman (20, 21) concluded that the
inactivation mechanism occurred by one electron transfer from the inhibitor to the flavin N5
atom, which indicated the formation of radical intermediate, as shown in Scheme 1.4c. However,
the studies with trans-2-phenyl(aminomethyl)-cyclopropane showed the radical intermediate is
not required for the inactivation of MAO (33, 34). Moreover, the lack of flavin semiquinone
UV-visible spectrum in the stopped-flow experiment (35, 36) and radical spectrum in EPR (37)
suggested that radical mechanism is less tenable. Also, the radical intermediate mechanism
became even weaker when the flavin N5 adduct was observed in the crystal structure of MAO
with cyclopropyl ring (38).
An alternative mechanism, nucleophilic mechanism, was proposed by Hamilton based on
model chemistry (22). According to the crystal structure of MAO with cyclopropyl ring, this
mechanism was supported by Edmondson (38, 39). However, there are also several observations
that are not consistent with the nuclephilic mechanism. First, the large kinetic isotope effect with
deuterium benzylamine and the absence of an C4a adduct intermediate (40) implied the
formation of the adduct is reversible and sufficiently unfavorable. Second, the rate constants for
reduction of MAO A in a series of ring-substituted benzylamine are sensitive to the volume
instead of the electron donating ability of the benzylamine (41). Therefore, based on these
studies, the mechanism for primary amine oxidation catalyzed by flavin-dependent enzymes is
still controversial.
24
1.3
Protein Hydroxyl Group Proximal to the C(4a)-N(5) Flavin Atoms
1.3.1
Physical Properties of Hydroxyl Group
Amino acids with a hydroxyl group, including serine, threonine and tyrosine, are polar,
uncharged at physiological pH, and hydrophilic. The hydroxyl group enables them to form
hydrogen bonds in a protein structure. As shown in Scheme 1.7, serine and threonine have
hydroxyl groups attached to aliphatic side chains, which make them more hydrophilic. The
hydroxyl groups on sp3-hybridized carbons can be a donor and two acceptors in the formation of
hydrogen bond. Compared with serine, threonine is less polar, sterically more hindered and
chemically less reactive since it has one more methyl group than serine. Tyrosine has a hydroxyl
group attached to the benzene ring. The hydroxyl group on sp2-hybridized carbon can only be a
donor and an acceptor in the formation of hydrogen bond since the R-OH bond has partial
double-bonded character (42). The phenolic hydroxyl group makes tyrosine reactive and imparts
a significant degree of polarity. The pKa of the hydroxyl groups of serine and threonine are very
high with values about 15 and not easy to ionize. The phenolic hydroxyl of tyrosine with a pKa of
~10 is significantly more acidic than the aliphatic hydroxyls of either serine or threonine.
serine
pK1=2.2
O
H3N CH C O
CH2
OH
pK2=9.2
threonine
pK1=2.1
O
H3N CH C O
CH OH
CH3
pK2=9.6
tyrosine
pK1=2.2
O
H3N CH C O
CH2
pK2=9.1
OH
pK3=10.0
Scheme 1.7 The Structure of Hydroxyl Amino Acids.
25
1.3.2
Conserved Hydroxyl Group in Flavoenzymes
According to the crystal structure of flavoenzymes that are available in the PDB (protein data
bank) and kinetic studies, amino acid residues with hydroxyl groups are highly conserved in
proximity to the flavin C(4a)-N(5) atoms and play an important role in the flavin reduction and
oxidation reactions (Table 1.1).
Table 1.1 List of Flavoenzymes with Hydroxyl Group in the Active Site.
enzymes
residues
PDB
choline oxidase
Ser101
2jbv
D-arginine dehydrogenase
Tyr249/53
3nye
Pyranose 2-oxidase
Thr169
1tt0
NADPH-cytochrome P450 oxidoreductase
Ser457
1amo
ferredoxin-NADP+ reductase
Ser96
1frn
old yellow enzyme
Thr37
1oyb
alditol oxidase
Ser106
2vfs
glucose oxidase
Thr110
1cf3
pentaerythritol tetranitrate reductase
Thr26
1h51
morphinone reductase
Thr32
1gwj
S-mandelate dehydrogenase
Ser108/Tyr131
1huv
human glycolate oxidase
Tyr132/26
2rdu
flavocytochrome b2
Tyr254/143
1fcb
porcine D-amino acid oxidase
Tyr228/224/Thr317 1kif
L-lactate oxidase
Tyr146/40
2du2
chorismate synthase
Ser127/16
1qxo
yeast D-amino acid oxidase
Ser335/Tyr338
1cop
class 2 dihydroorotate dehydrogenase
Ser175/Thr178
1f78
monomeric sacrosine oxidase
Thr48/Tyr317
1el5
heterotetrameric sacrosine oxidase
Thr66
2gag
aryl-alcohol oxidase
Tyr92
3fim
nitroalkane oxidase
Ser267/ser171
3d9g
aklavinone-11-hydroxylase
Tyr224
3ihg
2-nitropropane dioxygenase
Ser288
2gjl
p-hydroxybenzoate hydroxylase
Ser212/Tyr201
1dod
glycine oxidase
Tyr246
1ryi
chromate reductase
Tyr177
3hf3
glucooligosaccharide oxidase
Tyr429
1zr6
medium chain acyl-CoA dehydrogenase
Thr136/168
1udy
porcine cytochrome b5 reductase
Ser99/tyr65
1ndh
trimethylamine dehydrogenase
Tyr169
1djn
human monoamine oxidase B
Tyr435
1gos
26
1.3.2.1 Stabilization of the Transition State or Reaction Intermediate
The transition state is the most unstable state in the reaction pathway as the chemical bonds
are in the process of being made and broken. Transition state theory for enzyme-catalyzed
reactions proposes that catalysis needs tight-binding at the transition state (43, 44). The binding
forces of the transition state can be affected by hydrogen and ionic bond energy on bond
distance, angle and solvent environment. Dynamic electronic environments and hydrogen
bonding network in enzymatic catalytic site increase the probability of forming the transition
state (45). Changes made at the active site could fail to stabilize the transition state. In the
flavoenzymes, the hydroxyl group in the active site can form a hydrogen bond with flavin or
substrate in the transition state. Removal of the hydroxyl group could destroy the stabilization of
transition state.
In pyranose-2 oxidase, Thr169 sits above the flavin si-side, as shown in Figure 1.3. Steady
state and transient kinetics of wild-type, Thr169Ser, Thr169Asp, Thr169Gly and Thr169Ala
variants with D-glucose or D-galactose as a substrate suggest that Thr169 is important to
stabilize the C4a-hydroperoxy-FAD intermediate through its interaction with the flavin N5/O4
locus (46). The oxidative half-reaction of wild-type carried out in stopped-flow instrument shows
that a transient intermediate is observed (47). However, there is no intermediate detected in any
of the Thr169 variants (46). In addition, the crystal structure of wild-type, Thr169Ser and
Thr169Gly also showed that the side chains of Thr169 form an H-bond with the N5/O4 locus.
27
Figure 1.3 The activ
ve site of pyrranose-2 oxid
dase (PDB ccode 1TT0).
Ser45
57 in NADP
PH-cytochrom
me P450 ox
xidoreductasee was also sstudied with isotope effeect on
the reducctive half-reeaction of th
he alanine, cysteine andd threonine variants (48). The 1000-fold
decrease of the flav
vin reduction
n rate consttant and moore than 2-ffold increasee of the priimary
ggests an increase in thhe rate limitting of hydrride transfer. The
substrate KIE in Serr457Ala sug
steady sttate kinetics and flavin content
c
studiies show thaat the mutatiion has no eeffect on coffactor
binding. Furthermoree, the decreaase of the red
dox potentiaal of FAD seemiquinone in the Ser4557Ala
s
the inability to
o stabilize th
he FAD sem
miquinone. The results suggest thaat the
mutant suggests
Ser457 iss important to
t stabilize th
he transition
n state for hyydride transffer by H-bonnding interacction.
Ser96
6Val variantt of ferredox
xin-NADP+ reductase ffolds properrly based onn the UV-viisible,
fluorenseence, crystal structure an
nd circular dichroism speectra. Howevver, the turnnover numbeer and
second-o
order rate co
onstant (kcat/K
/ m) are sig
gnificantly ddifferent froom wild-typpe enzyme. Also,
there is no charge--transfer speecies accum
mulated in rrapid reactiion of Ser996Val variannt as
d by little sp
pectral chan
nge at 700 nm
n during thhe reductionn of the enzyyme by NAD
DPH.
suggested
Thus, thee removal of
o hydroxyl group resultts the loss oof interaction between nnicotinamidee and
flavin wh
hich disruptt the transitiion state durring the hyddride transfeer between nnicotinamidee and
28
FAD. The absence of neutral semiquinone in Ser96Val variant in the photoreduction also shows
the destabilization of the semiquinone (49).
In the old yellow enzyme family, Thr37 is conserved and proposed to be H-bonded with
C(4)=O based on the crystal structure (50, 51). However, the only evidence is the decreased
redox potential in the Thr37Ala variant compared to wild-type. According to the crystal structure
and sequence alignment, the hydroxyl protein residue is conserved in several flavoenzymes
which are proposed to form a H-bond with either N5 or C(4)=O of flavin, such as Ser106 in
alditol oxidase (52), Thr110 in glucose oxidase (14), Thr26 in pentaerythritol tetranitrate
reductase (53), Thr32 in morphinone reductase (53), Thr24 in xenobiotic reductase B (53), Thr24
in glycerol trinitrate reductase (53), Thr48 in estrogen-binding proton (53), Thr32 in 12oxophytidienoic acid reductase (53), Ser105 in long chain hydroxy acid oxidase (54), Ser108 in
S-mandelate dehydrogenase (55), Ser106 in spinach glycolate oxidase (55).
Also, the hydroxyl protein group can stabilize the transition state by H-bonding with the
substrate, as shown in Figure 1.4. Several enzymes and the corresponding variants have been
studied. The Lederer’s group studied the Tyr254Phe and Tyr254Leu mutants of flavocytochrome
b2 (56, 57). The higher Ki value with propionate in the wild-type compared to the small Km of Llactate suggests the hydroxyl group of the substrate is important for binding via H-bonding with
protein. The transition state destabilization induced by the mutation and binding modeling
suggests that the Tyr254 hydroxyl group stabilizes the Michaelis complex formation and
transition state. Later on, the pH profile of wild-type and Tyr254Phe mutant investigated by
Fitzpatrick’s group also showed the importance of neutral Tyr254 in the catalysis (58).
29
Figure 1.4 The activ
ve site of flav
vocytochrom
me b2 from bbaker’s yeastt Saccharom
myces cerevissiae
de 1FCB).
(PDB cod
The studies of Tyr129
T
in spinach glyccolate oxidaase also suppport the staabilization oof the
transition
n state by H--bonding bettween the su
ubstrate and hydroxyl grroup of Tyr129 (59). Oxxalate,
transition
n state analo
ogue, can bind
b
and inh
hibit glycolaate oxidase wild-type ((60). The looss of
binding and
a inhibitio
on in the Tyrr249Phe enzy
yme suggestt that the rem
moval of hyddroxyl groupp lose
the interaaction with transition
t
state analoguee. Tyr152 inn lactate monooxygenasse was studieed by
replacing
g with pheny
ylalanine (61
1). The correelation of biinding of oxxalate and caatalytic efficiency
of wild-ttype and Ty
yr152Phe su
uggests the Tyr152 is involved inn binding of transition state
analoguee oxalate. Mo
oreover, the dissociation
n constant foor sulfite is 1100-fold highher in Tyr1552Phe
which is also observed in homollogous mutation Tyr1299Phe in glycoolate oxidase and Tyr2554Phe
in flavoccytochrome b2. In porccine D-amin
no acid oxiddase, since the characteeristics of ssulfite
binding and N5 sulfite adduct in tyrosine mutants annd the chem
mical modificcation studiies of
Tyr228Ph
he are similaar to the Tyrr129Phe in glycolate
g
oxiidase and Tyyr254Phe in flavocytochhrome
b2 (62), the
t similar ro
ole of Tyr22
28 was propo
osed.
with modeliing and inhibbition
Tyr20
01Phe in p-h
hydroxybenzzoate hydrox
xylase was innvestigated w
studies (6
63, 64). C(4
4a)-flavinhyd
droperoxide was used ass a model too describe thhe transition state.
Accordin
ng to the caalculation of
o activation
n barrier, Tyyr201 stabillizes both gground statee and
transition
n state via H-bonding
H
with
w substratee. The failurre to stabilizze the substtrate analoguue, 4-
30
fluorobenzoate, and inhibitor, 4-hydroxycinnamate in the Tyr201Phe mutant suggests Tyr201 is
important to activate substrate by forming H-bond with the hydroxyl group of substrate.
Based on sequence alignment and crystal structure studies, Tyr132 in human glycolate
oxidase (65), Tyr146 in L-lactate oxidase (66), Tyr131 in mandelate dehydrogenase (55) are
proposed to stabilize the transition state.
The role of Tyr143 in flavocytochrome b2 has also been probed with crystal structure, sitedirected mutagenesis, stopped-flow, steady state and kinetic isotope effects (67). The substrate
kinetic isotope effect on kred of Tyr143Phe was smaller and the Kd, which has been calculated
based on the Km and susbtrate kinetic isotope effect on kcat and kcat/Km, was 4-5 fold higher than
the wild-type. Based on the calculated free energy, these data suggest that the Tyr143 stabilizes
the Michaelis complexes via H-bonding with the carboxylate group of the substrate.
1.3.2.2 Component of Proton Relay Systems
Hydroxyl protein groups can also be part of a proton relay system in a H-bonding network.
The single and double mutants of Ser127 and Ser16 in chorismate synthase show the importance
of these two serine residues in proton relay among the FMN, His106 and ligand EPSP (68).
These two residues are conserved in bacterial, fungal, plant and protozoan chorismate synthase.
The unchanged binding affinity to the FMN in all these mutants and slightly increased binding
affinity only in the double mutants show the enzymes fold properly. However, the decreased
activity in single mutant and no activity in the double mutants, combined with the crystal
structure (69) shows that the H-bonding network is disrupted in the double mutants.
The roles of Ser175 and Thr178 in class 2 dihydroorotate dehydrogenase were studied by
crystal structure, molecular dynamic simulation, site-directed mutagenesis, redox potential, and
pH profile of flavin reduction (70-73). The data showed that the Ser175Ala, Thr178Ser,
31
Thr178A
Ala and Thr178Val mutan
nt enzymes were
w folded properly annd the active site environnment
was littlee affected baased on the absorbance
a
spectra and ccircular dichhroism spectrra. The decreeased
redox po
otential in th
he Ser175Ala and Thr17
78Ala mutannts, about 40 mV, was due to hydrrogen
bond disrruption. Thee slower hyd
dride transfeer in the muttants togetheer with simuulation show
ws the
proton traansfer from substrate to Ser175, thro
ough water tto Thr178 too deprotonatee the substraate, as
shown in
n Figure 1.5
5. Thr48 in monomeric
m
sacrosine
s
oxxidase (74) aand Thr66 inn heterotetram
meric
sacrosinee oxidase (75
5) were also
o proposed to
o be a part oof a proton rrelay system
m from the crrystal
structure. However, thus
t
far the functions
f
of these residuues have not yet kineticaally studied.
on relay in cllass 2 dihydrroorotate dehhydrogenasee (PDB codee 1f78).
Figure 1.5 The proto
1.3.2.3 Aligning
A
the Substrate into
i
the Rig
ght Position for Catalyssis
The role
r of Tyr22
23 in yeast D-amino
D
acid
d oxidase haas been probed by studyiing the wild--type,
Tyr223Ph
he and Tyr2
223Ser varian
nts (76). The spectra, liggand binding and redoxx potential sttudies
suggest that
t mutagen
nesis has no significant effect
e
on reddox potentiall and the binnding pockett. The
kinetic parameters
p
for
f Tyr223P
Phe are simiilar to wild--type. In coontrast, the Kd in Tyr2223Ser
mutant iss increased 60-fold
6
and product releease is 800- fold slower.. These dataa suggest thee side
chain of Tyr223 is no
ot importantt for catalysiis. Probably the steric efffect of the sside chain ass well
32
as H-bonding is important for aligning the substrate in the right position for catalysis. Tyr238
residue was also investigated by the same group (77). Rapid reaction studies of the enzymes
show the rates of the Tyr238Phe and Tyr238Ser mutants are similar to wild-type, suggesting this
residue is not responsible for acid/base catalysis. The binding of inhibitors and substrate
specificity of the mutants are also not significant different from wild-type, therefore, the residue
238 is not important for the substrate binding. However, the simulation analysis of experiment
traces with program A shows that the substrate binding and product release steps are slower in
the mutant. The effect of pH on the ligand binding and the flapping of Tyr238 side chain in the
crystal structure as well as the kinetic parameters obtained from Tyr238 mutants propose that
Tyr238 acts as a gate to control the substrate and product exchange.
Tyr92 residue in aryl-alcohol oxidase was also proposed to have the similar role of Tyr238
base on the molecular docking and steady state kinetic studies of wild-type and Tyr92Ala/Phe
mutants. The results show that there is no catalytic activity observed in Tyr92Ala mutant,
however, the activity of Tyr92Phe is similar to the wild-type (78). The crystal structure of arylalcohol oxidase also suggests the aromatic side chain of Tyr92 act as a gate which controls the
entry of the substrate to the active site (79).
Tyr317 in monomeric sacrosine oxidase was proposed to bind and activate substrate (80).
The kinetic studies of Tyr317Phe show the kcat/Km and kred values were 15-fold and 20-fold lower
than wild-type, respectively. The pH profile of kred in Tyr317Phe with the similar pKa as in the
wild-type rules out the possibility of Tyr317 as a base. The Kd values of Tyr317Phe are pH
dependent with a pKa value of 9.0 which is not observed in wild-type enzyme. Moreover, the Kd
values of Tyr317Phe with inhibitors are 14- and 21-fold increased compared to the wild-type.
33
Together, the Tyr317 residue plays an important role in binding the substrate for efficient
catalysis.
Ser267 is not in the active site in nitroalkane oxidase. However, steady state studies of
Ser267Ala show that kcat and kcat/Km values are significantly lower than wild-type with
nitroethane, 1-nitrohexane or 1-nitrooctane as substrate. The kred was also 80-fold lower than
wild-type enzyme (81). The crystal structure of Ser267Ala shows that the residue Asp402 which
was proposed to abstract proton of the substrate was not in the right position, compared with
wild-type enzyme (81). Therefore, the hydroxyl group of Ser267 plays a role in aligning the
Asp402 to the right position via H-bonding.
The crystal structure of aklavinone-11-hydroxylase shows Tyr224 forms a H-bond with
aklavinone (82). No catalytic activity was detected in Tyr224Phe variant. These data suggests
that the Tyr224 plays an important role in catalysis and maybe position the substrate for
catalysis. In 2-nitropropane dioxygenase, the role of Ser288 was proposed (83). Ser288Ala of
cannot be expressed in E. coli (83). Based on the crystal structure, the side chain of Ser288 was
proposed to form H-bond with the nitro group of 2-nitropropane and bring it to the active site
(83).
1.3.2.4 Involvement in Binding Substrate
The hydroxyl protein group is also proposed to affect the substrate binding by H-bond with
the substrate carboxylate group, such as Tyr24 in spinach glycolate oxidase (84), Tyr26 in
human glycolate oxidase (65) and Ser212 in p-hydroxybenzoate hydroxylase (85). Steady state
kinetic, rapid reaction and crystallography were used to study the wild-type spinach glycolate
oxidase and Tyr24Phe mutants (84). The results show that the Km and Kd of glycolate in the
mutant are 10- and 17-time increased, respectively. However, the turnover number and flavin
34
reduction rate constant are only slightly affected. Substitution of Ser212 in p-hydroxybenzoate
hydroxylase to alanine results the Kd value 10-fold increase and no change in the catalytic
activity (85). Moreover, there are two binding conformations in the Ser212Ala with alternate
substrate, suggested by the two products formed with HPLC analysis (85). Based on the crystal
structure, Tyr246 in glycine oxidase (86), Tyr40 in L-lactate oxidase (87) (66), Tyr249 in Darginine dehydrogenase (88) and Tyr26 in modelate dehydrogenase (55) are also H-bond with
the substrate carboxylate group and may play a similar role in substrate binding.
1.3.2.5 Acting as a Base or Acid
Ser175 in class 2 dihydroorotate dehydrogenase was proposed to be a weak base which
abstracts the proton from C5 of DHO substrate (73, 89, 90). Substitution of serine to alanine or
cysteine causes the activity to decrease 10,000- and 500-fold, respectively (89). The flavin
reduction in Ser175Ala is increased while increasing in pH with no observable pKa, which is
different from wild-type enzyme with a pKa of 9.5 (69). The crystal structure (71) and studies of
conserved residue Cys130 in class 1 dihydroorotate dehydrogenase (91) also suggests that
Ser175 is a base in this enzyme.
In chromate reductase from Thermus scotoductus SA-01, the FMN is reduced by NADPH
and a hydride from N5 of reduced FMN with a proton transfers to the substrate. Tyr177 forms Hbond with substrate and was proposed to be a proton donor (92). However, the phenylalanine
mutant was studied and only results in 5-fold decreasing of the activity. In the absence of
Tyr177, the other residue or water may act as a proton donor. For the sequence alignment,
Tyr196 in old yellow enzyme was also proposed to be a proton donor.
35
Tyr429 in glucooligosaccharide oxidase acts as a general base to abstract the hydroxyl proton
of substrate based on the crystal structure (93). The maximum activity at pH 10 also supports this
proposal (94).
1.3.2.6 Contributing to the Affinity of the Flavin to Apoprotein
In human medium chain acyl-CoA dehydrogenase, the role of Thr136 was studied by
substitution with serine, aspartate and leucine (95). The analyses of in vitro and prokaryotic
expression products show that the Thr136Ser has the similar biogenesis and activity as the wildtype. By adding riboflavin and FAD to culture broth and buffers, the Thr136Asp can form a
catalytically active tetramer. However, no activity can be rescued in the Thr136Leu. Based on
the crystal structure, the hydroxyl group of the Thr136 forms a H-bond with flavin N1-C2 atoms
(96). Altogether, the hydrophilic group of Thr136 was proposed to bind FAD. The important of
Thr168 to bind FAD in this enzyme was also studied by Andresen’s group (97). The Thr168Ala
was not well folded and it is presumed to have lost the ability to bind FAD.
In pig kidney D-amino acid oxidase, Thr317 in the N1-C2 position plays a role in the affinity
of FAD (98). The alanine mutant has a lower affinity to FAD than in the wild-type. The 2-13C
labeled FAD-NMR spectra of wild-type and Thr317 shows that Thr137 forms a H-bond with
flavin C(2)=O. Based on the sequence alignment, Tyr338 in yeast D-amino acid oxidase (99),
Thr317 in human D-amino acid oxidase (99) are conserved and proposed to have similar roles.
Ser99 porcine cytochrome b5 reductase is proposed to maintain the conformation of FAD by
H-bonding to the phosphate group (100, 101). Absorbance spectra and denaturation studies were
investigated in the Ser99Thr, Ser99Ala and Ser99Val mutants. Except for the Ser99Thr, the other
two mutants result in obvious spectral and stability changes. Therefore, the hydroxyl group of
Ser99 is important to maintain the conformation of FAD. Based on the sequence alignment,
36
Ser492 in NADPH-cytochrome p450 reductase, Thr98 in nitrate reductase, Ser133 in ferredoxinNADP+ reductase, Ser84 in phthalate dioxygenase reductase and Ser76 in flavodoxin reductase
are proposed to have the same role.
1.3.2.7 Maintaining the Protein Structure
Ser171 in nitroalkane oxidase was studied by substitution with alanine, valine and threonine
(102). There is no significant difference in the mutant from the wild-type enzyme with steady
state kinetic and kinetic isotope effect studies. However, the crystal structure shows the hydroxyl
group of Ser171 forms a H-bond with N5 flavin. Therefore, Ser171 is believed to not be involved
in catalysis but may be important for maintaining the protein structure.
Tyr65 in porcine cytochrome b5 reductase is close to 4’-hydroxyl of the ribityl moiety of the
FAD and the hydroxyl group and is important in protein stability based on the studies of
Tyr65Ala and Tyr65Phe mutants (100, 101). The absorbance spectra and CD spectrum of
Tyr65Ala exhibit large changes instead of Tyr65Phe. Also, the fluorescence emission spectra of
both mutants were different from the wild-type.
1.3.2.8 Other Functions
The role of Tyr169 in trimethylamine dehydrogenase has been investigated by site-directed
mutagenesis, stopped-flow and the EPR spectroscope. Based on the crystal structure, Tyr169 is
Van der Waals bonded to the flavin C(2)=O locus and H-bonded to His172 (103). In comparison
to the EPR spectra of the wild-type with the mutant, the spin-interaction between flavin
semiquinone and reduced 4Fe/4S center seen in the wild-type was not observed in the Tyr169Phe
mutant, which suggest that Tyr169 residue plays an important role in mediating the spininteraction (104).
37
The two
t
tyrosinee residues were
w
found at the activve site of huuman monoaamine oxidaase B
which fo
orms an “aro
omatic sandw
wich” with substrate
s
anaalog, as show
wn in Figurre 1.6 (105). The
wild-typee enzyme ass well as Ty
yr435Phe, Tyr435His,
T
T
Tyr435Leu aand Tyr435T
Trp were stuudied
with crysstal structuree and steady state kineticcs (106). Cryystal structur
ures of mutannts shows there is
no signifficant differeence from th
he wild-type,, suggesting that the muttants are covvalently linkked to
FAD and fold welll as the wiild-type. An
nalyzing thee kcat/Km vaalues with tthe dipole-ddipole
interactio
on energy to
ogether propo
osed nucleop
philic mechanism indicaates that thee dipolar effeect of
tyrosine side chains activate
a
the substrate
s
am
mine for catallysis by incrreasing its nuucleophilicitty.
Figure 1.6 The “arom
matic sandw
wich” in hum
man MAO B ((PDB code 11GOS).
1.4
Specific Goa
als
d
aims to stud
dy the role of the serine proximal too the flavin C
C(4a)-N(5) aatoms
The dissertation
in cholin
ne oxidase an
nd the kinetiic mechanism
m of D-arginnine dehydroogenase to ggain insight iin the
mechanissm for alcoh
hol or amine oxidation.
Choliine oxidase (E.C.1.1.3.17) from A.. globiformiis catalyzes the oxidation of choline to
glycine betaine
b
via tw
wo reductivee half-reactio
ons and two oxidative haalf-reactionss. This reactiion is
of consid
derable inteerest for meedical and biotechnolog
b
gical appliccations becaause the prooduct,
38
glycine betaine,
b
is a compatible solute (107--109). Experrimental datta have show
wn that cellss with
the codA
A gene encod
ding choline oxidase can
n grow well iin the presennce of 0.4 M NaCl. How
wever,
cells witthout the codA gene caannot surviv
ve, as shownn in Figure 1.8 (110). In the wildd-type
enzyme, catalysis is triggered by
y the kineticcally fast absstraction of the substratee hydroxyl ggroup
by a cataalytic base, followed
fo
by the
t transfer of
o a hydride ion from the α-carbon oof the substrate to
the N5 atom
a
of flavin, as shown
n in Schem
me 1.8. Severral residues in the activve site have been
studied to
o contribute the catalysiis of oxidatio
on of cholinne (Figure 11.9). His99, ccovalently liinked
to FAD, is importantt to align FA
AD in the op
ptimal positioon to acceptt hydride ionn in the oxiddation
of cholin
ne (111). Glu
u312 is impo
ortant for su
ubstrate bindding via elecctrostatic intteraction witth the
trimethyllammonium group of ch
holine (112, 113). Val4 64 providess a non-polaar site for oxxygen
access in
nto the activee site (114, 115). His35
51 and His4666 are propoosed to stabiilize the alkooxide
species (116, 117). Asn510
A
is im
mportant for the
t flavinylaation reactionn in which F
FAD is covallently
attached to the proteiin moiety in choline oxid
dase (118).
Figure 1.7
1 Left paneel, glycine betaine
b
as an
n osmoprotecctant for S. m
meliloti. Rigght panel, Grrowth
under 0.4
4 M NaCl in
n control ceells Synechococcus sp. P
PCC 7942 (PAM) and ttransformed cells
(PAMCO
OD) in whicch the codA gene (for ch
holine oxidaase) was inttroduced. M
Modified from
m ref.
(110), tak
ken without permission.
39
a) Proton
n abstraction
n:
N
H99
N
N
H99
R
N
N
N
R
O
N
N
O
H
NH
N
E312N
F357
N
H466
N
E312-
H
O
H
W331
O
H
H
N
O
H
N
B
H
H
NH
F3357
W 331
N
H 466
N
H
O
H
N
BH
H
N
V464
V464
N
H351
N
H 351
b) Hydrid
de transfer:
H99
R
N
N
N
E312-
H
N
O
H
W331
V4644
N
R
N
O
H
NH N
N
F357
N
N
N
H99
BH
B
H
N
N
H351
3
d
N
E312F357
W331
V464
H
N
H
O
O N
d H
O
H466
BH
H
N
N
F357
H351
O
H
NH N
O
H
V464
N
N
H
E312W331
N
R
N
N
DMS
SO
H
NH N
N
H466
O N
H
H99
H466
O N
H
H
BH
H
N
N
H351
M
for the Oxiidation of C
Choline to B
Betaine Aldeehyde
Scheme 1.8 Hydridee Transfer Mechanism
Catalyzed by Cholinee Oxidase
Figure 1.8 Active sitte of cholinee oxidase (PD
DB# 2jbv)
Based
d on the cry
ystal structurre of cholinee oxidase wiild-type, a seerine residuee at positionn 101,
which is less than 4 Å from thee flavin N5 atom and w
within hydroogen bondinng distance tto the
40
ligand DMSO, was investigated to probe the role of the hydroxyl group of Ser101 in the reaction
catalyzed by the enzyme. To elucidate this, the variants in which the serine was replaced with
alanine, threonine, cysteine or valine were prepared and purified. The effects of hydroxyl group
at position 101 on the catalytic efficiency of choline oxidase in both the reductive and the
oxidative half-reactions were studied with crystallographic and mechanistic approaches. The
details will be described in Chapters 2, 3 and 4.
D-arginine dehydrogenase (DADH) from Pseudomonas aeruginosa catalyzes the oxidation
of D-amino acids to the corresponding imino acids, which are non-enzymatically hydrolyzed in
solution to α-keto acids and ammonia, as shown in Scheme 1.9. D-amino acids are important for
bacteria as fundamental elements of the bacterial cell wall peptidoglycan layer (119). It was
reported that D-leucine, D-methionine, D-tyrosine and D-tryptophan can regulate disassembly of
bacterial biofilms (120). D-aspartate was found in the brain and other tissues of mammals and
plays a role in regulating the development of these organs (121). D-serine may act as an
endogenous agonist of the N-methyl-D-aspartate receptor in the rat brain (122). In P. aeruginosa,
D-arginine can be the sole source of carbon and nitrogen (123). Dr. Lu’s group, Biology
department in Georgia State University, found that the racemization of D- to L-arginine
conversion is provided by a two-enzyme-coupled system: DADH and L-arginine dehydrogenase.
The products of DADH, keto acid and ammonia, is catalyzed by L-arginine dehydrogenase to Larginine. The discovery of D-arginine dehydrogenase suggests the existence of the fourth
arginine catabolic pathway in P. aeruginosa (124).
unknown e- acceptor
COO
R
H
NH3
FAD
FADH2
COO
R
NH2
+ H2O
COO
+ NH4
R
O
Scheme 1.9 Mechanism Proposed for Oxidation of D-Amino Acids by DADH.
41
The mechanism for primary amine oxidation catalyzed by flavin-dependent enzymes has
been debated for several decades. In addition to the hydride transfer mechanism, several
mechanisms were also proposed for flavoenzyme reductive half-reactions. In order to understand
the mechanism of carbon-nitrogen oxidation, the DADH can be a candidate to study. The studies
will be discussed together with the structural information in chapters 5, 6 and 7.
1.5
References
1.
Joosten, V., and van Berkel, W. J. (2007) Flavoenzymes, Curr Opin Chem Biol 11, 195202.
2.
Ghisla, S., and Massey, V. (1986) New flavins for old: artificial flavins as active site
probes of flavoproteins, Biochem J 239, 1-12.
3.
Muller, F., Brustlein, M., Hemmerich, P., Massey, V., and Walker, W. H. (1972) Lightabsorption studies on neutral flavin radicals, Eur J Biochem 25, 573-580.
4.
Massey, V., and Hemmerich, P. (1980) Active-site probes of flavoproteins, Biochem Soc
Trans 8, 246-257.
5.
Ghisla, S., Massey, V., Lhoste, J. M., and Mayhew, S. G. (1974) Fluorescence and
optical characteristics of reduced flavines and flavoproteins, Biochemistry 13, 589-597.
6.
Miura, R. (2001) Versatility and specificity in flavoenzymes: control mechanisms of
flavin reactivity, Chem Rec 1, 183-194.
7.
Mattevi, A. (2006) To be or not to be an oxidase: challenging the oxygen reactivity of
flavoenzymes, Trends Biochem Sci 31, 276-283.
8.
Fitzpatrick, P. F. (2001) Substrate dehydrogenation by flavoproteins, Acc Chem Res 34,
299-307.
42
9.
Fitzpatrick, P. F. (2004) Carbanion versus hydride transfer mechanisms in flavoproteincatalyzed dehydrogenations, Bioorg Chem 32, 125-139.
10.
Fan, F., and Gadda, G. (2005) On the catalytic mechanism of choline oxidase, J Am
Chem Soc 127, 2067-2074.
11.
Gadda, G. (2008) Hydride transfer made easy in the reaction of alcohol oxidation
catalyzed by flavin-dependent oxidases, Biochemistry 47, 13745-13753.
12.
Sucharitakul, J., Wongnate, T., and Chaiyen, P. (2010) Kinetic isotope effects on the
noncovalent flavin mutant protein of pyranose 2-oxidase reveal insights into the flavin
reduction mechanism, Biochemistry 49, 3753-3765.
13.
Gibson, Q. H., Swoboda, B. E., and Massey, V. (1964) Kinetics and Mechanism of
Action of Glucose Oxidase, J Biol Chem 239, 3927-3934.
14.
Wohlfahrt, G., Witt, S., Hendle, J., Schomburg, D., Kalisz, H. M., and Hecht, H. J.
(1999) 1.8 and 1.9 A resolution structures of the Penicillium amagasakiense and
Aspergillus niger glucose oxidases as a basis for modelling substrate complexes, Acta
Crystallogr D Biol Crystallogr 55, 969-977.
15.
Vrielink, A., and Ghisla, S. (2009) Cholesterol oxidase: biochemistry and structural
features, FEBS J 276, 6826-6843.
16.
Sherry, B., and Abeles, R. H. (1985) Mechanism of action of methanol oxidase,
reconstitution of methanol oxidase with 5-deazaflavin, and inactivation of methanol
oxidase by cyclopropanol, Biochemistry 24, 2594-2605.
17.
Hallberg, B. M., Henriksson, G., Pettersson, G., Vasella, A., and Divne, C. (2003)
Mechanism of the reductive half-reaction in cellobiose dehydrogenase, J Biol Chem 278,
7160-7166.
43
18.
Walsh, C. T., Schonbrunn, A., and Abeles, R. H. (1971) Studies on the mechanism of
action of D-amino acid oxidase. Evidence for removal of substrate -hydrogen as a proton,
J Biol Chem 246, 6855-6866.
19.
Walsh, C. T., Krodel, E., Massey, V., and Abeles, R. H. (1973) Studies on the elimination
reaction of D-amino acid oxidase with -amino- -chlorobutyrate. Further evidence for
abstraction of substrate -hydrogen as a proton, J Biol Chem 248, 1946-1955.
20.
Silverman, R. B. (1983) Mechanism of inactivation of monoamine oxidase by trans-2phenylcyclopropylamine and the structure of the enzyme-inactivator adduct, J Biol Chem
258, 14766-14769.
21.
Silverman, R. B. (1995) Radical ideas about monoamine oxidase, Acc. Chem. Res. 28,
335-342.
22.
Brown, L. E., and Hamilton, G. A. (1970) Some model reactions and a general
mechanism for flavoenzyme-catalyzed dehydrogenations, J Am Chem Soc 92, 7225-7227.
23.
Hersh, L. B., and Jorns, M. S. (1975) Use of 5-deazaFAD to study hydrogen transfer in
the D-amino acid oxidase reaction, J Biol Chem 250, 8728-8734.
24.
Edmondson, D., and Ghisla, S. (1999) Flavoenzyme structure and function. Approaches
using flavin analogues, Methods Mol Biol 131, 157-179.
25.
Mattevi, A., Vanoni, M. A., Todone, F., Rizzi, M., Teplyakov, A., Coda, A., Bolognesi,
M., and Curti, B. (1996) Crystal structure of D-amino acid oxidase: a case of active site
mirror-image convergent evolution with flavocytochrome b2, Proc Natl Acad Sci U S A
93, 7496-7501.
44
26.
Mizutani, H., Miyahara, I., Hirotsu, K., Nishina, Y., Shiga, K., Setoyama, C., and Miura,
R. (1996) Three-dimensional structure of porcine kidney D-amino acid oxidase at 3.0 A
resolution, J Biochem 120, 14-17.
27.
Denu, J. M., and Fitzpatrick, P. F. (1994) Intrinsic primary, secondary, and solvent
kinetic isotope effects on the reductive half-reaction of D-amino acid oxidase: evidence
against a concerted mechanism, Biochemistry 33, 4001-4007.
28.
Kurtz, K. A., Rishavy, M. A., Cleland, W. W., and Fitzpatrick, P. F. (2000) Nitrogen
isotope effects as probes of the mechanism of D-amino acid oxidase, J Am Chem Soc
122, 12896-12897.
29.
Fitzpatrick, P. F., and Massey, V. (1982) Proton release during the reductive half-reaction
of D-amino acid oxidase, J Biol Chem 257, 9958-9962.
30.
Tilocca, A., Gamba, A., Vanoni, M. A., and Fois, E. (2002) First-principles molecular
dynamics investigation of the D-amino acid oxidative half-reaction catalyzed by the
flavoenzyme D-amino acid oxidase, Biochemistry 41, 14111-14121.
31.
McEwen, C. M., Jr., Sasaki, G., and Jones, D. C. (1969) Human liver mitochondrial
monoamine oxidase. II. Determinants of substrate and inhibitor specificities,
Biochemistry 8, 3952-3962.
32.
Paech, C., Salach, J. I., and Singer, T. P. (1980) Suicide inactivation of monoamine
oxidase by trans-phenylcyclopropylamine, J Biol Chem 255, 2700-2704.
33.
Silverman, R. B., Zhou, J. P., and Eaton, P. E. (1993) Inactivation of monoamine oxidase
by (aminomethyl)cubane. First evidence for an α-amino radical during enzyme catalysis,
J Am Chem Soc 115, 8841-8842.
45
34.
Choi, S. Y., Eaton, P. E., Newcomb, M., and Yip, Y. C. (1992) Picosecond radical
kinetics. Bond cleavage of the cubylcarbinyl radical, J Am Chem Soc 114, 6326-6329.
35.
Husain, M., Edmondson, D. E., and Singer, T. P. (1982) Kinetic studies on the catalytic
mechanism of liver monoamine oxidase, Biochemistry 21, 595-600.
36.
Miller, J. R., Edmondson, D. E., and Grissom, C. B. (1995) Mechanistic probes of
monoamine oxidase B catalysis: rapid-scan stopped flow and magnetic field
independence of the reductive half-reaction, J Am Chem Soc 117, 7830-7831.
37.
Tan, A., Glantz, M. D., Piette, L. H., and Yasunobu, K. T. (1983) ESR analysis of the
FAD in bovine liver monoamine oxidase, Biochem Biophys Res Commun 117, 517-523.
38.
Binda, C., Li, M., Hubalek, F., Restelli, N., Edmondson, D. E., and Mattevi, A. (2003)
Insights into the mode of inhibition of human mitochondrial monoamine oxidase B from
high-resolution crystal structures, Proc Natl Acad Sci U S A 100, 9750-9755.
39.
Edmondson, D. E., Binda, C., and Mattevi, A. (2007) Structural insights into the
mechanism of amine oxidation by monoamine oxidases A and B, Arch Biochem Biophys
464, 269-276.
40.
Miller, J. R., and Edmondson, D. E. (1999) Structure-activity relationships in the
oxidation of para-substituted benzylamine analogues by recombinant human liver
monoamine oxidase A, Biochemistry 38, 13670-13683.
41.
Walker, M. C., and Edmondson, D. E. (1994) Structure-activity relationships in the
oxidation of benzylamine analogues by bovine liver mitochondrial monoamine oxidase
B, Biochemistry 33, 7088-7098.
42.
Whitford, D. (2005) Proteins:structure and function, John Wiley & Sons, Ltd.
46
43.
Wolfenden, R. (1969) Transition state analogues for enzyme catalysis, Nature 223, 704705.
44.
Schramm, V. L. (1998) Enzymatic transition states and transition state analog design,
Annu Rev Biochem 67, 693-720.
45.
Schramm, V. L. (2005) Enzymatic transition states and transition state analogues, Curr
Opin Struct Biol 15, 604-613.
46.
Pitsawong, W., Sucharitakul, J., Prongjit, M., Tan, T. C., Spadiut, O., Haltrich, D., Divne,
C., and Chaiyen, P. (2010) A conserved active-site threonine is important for both sugar
and flavin oxidations of pyranose 2-oxidase, J Biol Chem 285, 9697-9705.
47.
Prongjit, M., Sucharitakul, J., Wongnate, T., Haltrich, D., and Chaiyen, P. (2009) Kinetic
mechanism of pyranose 2-oxidase from trametes multicolor, Biochemistry 48, 4170-4180.
48.
Shen, A. L., and Kasper, C. B. (1996) Role of Ser457 of NADPH-cytochrome P450
oxidoreductase in catalysis and control of FAD oxidation-reduction potential,
Biochemistry 35, 9451-9459.
49.
Aliverti, A., Bruns, C. M., Pandini, V. E., Karplus, P. A., Vanoni, M. A., Curti, B., and
Zanetti, G. (1995) Involvement of serine 96 in the catalytic mechanism of ferredoxinNADP+ reductase: structure--function relationship as studied by site-directed
mutagenesis and X-ray crystallography, Biochemistry 34, 8371-8379.
50.
Xu, D., Kohli, R. M., and Massey, V. (1999) The role of threonine 37 in flavin reactivity
of the old yellow enzyme, Proc Natl Acad Sci U S A 96, 3556-3561.
51.
Fox, K. M., and Karplus, P. A. (1994) Old yellow enzyme at 2 Å resolution: overall
structure, ligand binding, and comparison with related flavoproteins, Structure 2, 10891105.
47
52.
Forneris, F., Heuts, D. P., Delvecchio, M., Rovida, S., Fraaije, M. W., and Mattevi, A.
(2008) Structural analysis of the catalytic mechanism and stereoselectivity in
Streptomyces coelicolor alditol oxidase, Biochemistry 47, 978-985.
53.
Messiha, H. L., Bruce, N. C., Sattelle, B. M., Sutcliffe, M. J., Munro, A. W., and
Scrutton, N. S. (2005) Role of active site residues and solvent in proton transfer and the
modulation of flavin reduction potential in bacterial morphinone reductase, J Biol Chem
280, 27103-27110.
54.
Cunane, L. M., Barton, J. D., Chen, Z. W., Le, K. H., Amar, D., Lederer, F., and
Mathews, F. S. (2005) Crystal structure analysis of recombinant rat kidney long chain
hydroxy acid oxidase, Biochemistry 44, 1521-1531.
55.
Sukumar, N., Xu, Y., Gatti, D. L., Mitra, B., and Mathews, F. S. (2001) Structure of an
active soluble mutant of the membrane-associated (S)-mandelate dehydrogenase,
Biochemistry 40, 9870-9878.
56.
Dubois, J., Chapman, S. K., Mathews, F. S., Reid, G. A., and Lederer, F. (1990)
Substitution of Tyr254 with Phe at the active site of flavocytochrome b2: consequences
on catalysis of lactate dehydrogenation, Biochemistry 29, 6393-6400.
57.
Gondry, M., Dubois, J., Terrier, M., and Lederer, F. (2001) The catalytic role of tyrosine
254 in flavocytochrome b2 (L-lactate dehydrogenase from baker's yeast). Comparison
between the Y254F and Y254L mutant proteins, Eur J Biochem 268, 4918-4927.
58.
Sobrado, P., and Fitzpatrick, P. F. (2003) Solvent and primary deuterium isotope effects
show that lactate CH and OH bond cleavages are concerted in Y254F flavocytochrome
b2, consistent with a hydride transfer mechanism, Biochemistry 42, 15208-15214.
48
59.
Macheroux, P., Kieweg, V., Massey, V., Soderlind, E., Stenberg, K., and Lindqvist, Y.
(1993) Role of tyrosine 129 in the active site of spinach glycolate oxidase, Eur J Biochem
213, 1047-1054.
60.
Macheroux, P., Massey, V., Thiele, D. J., and Volokita, M. (1991) Expression of spinach
glycolate oxidase in Saccharomyces cerevisiae: purification and characterization,
Biochemistry 30, 4612-4619.
61.
Muh, U., Williams, C. H., Jr., and Massey, V. (1994) Lactate monooxygenase. III.
Additive contributions of active site residues to catalytic efficiency and stabilization of an
anionic transition state, J Biol Chem 269, 7994-8000.
62.
Pollegioni, L., Fukui, K., and Massey, V. (1994) Studies on the kinetic mechanism of pig
kidney D-amino acid oxidase by site-directed mutagenesis of tyrosine 224 and tyrosine
228, J Biol Chem 269, 31666-31673.
63.
Bach, R. D., and Dmitrenko, O. (2004) Model studies on p-hydroxybenzoate
hydroxylase. The catalytic role of Arg-214 and Tyr-201 in the hydroxylation step, J Am
Chem Soc 126, 127-142.
64.
Eschrich, K., van der Bolt, F. J., de Kok, A., and van Berkel, W. J. (1993) Role of Tyr201
and Tyr385 in substrate activation by p-hydroxybenzoate hydroxylase from Pseudomonas
fluorescens, Eur J Biochem 216, 137-146.
65.
Murray, M. S., Holmes, R. P., and Lowther, W. T. (2008) Active site and loop 4
movements within human glycolate oxidase: implications for substrate specificity and
drug design, Biochemistry 47, 2439-2449.
66.
Umena, Y., Yorita, K., Matsuoka, T., Kita, A., Fukui, K., and Morimoto, Y. (2006) The
crystal structure of L-lactate oxidase from Aerococcus viridans at 2.1A resolution reveals
49
the mechanism of strict substrate recognition, Biochem Biophys Res Commun 350, 249256.
67.
Rouviere-Fourmy, N., Capeillere-Blandin, C., and Lederer, F. (1994) Role of tyrosine
143 in lactate dehydrogenation by flavocytochrome b2. Primary kinetic isotope effect
studies with a phenylalanine mutant, Biochemistry 33, 798-806.
68.
Rauch, G., Ehammer, H., Bornemann, S., and Macheroux, P. (2008) Replacement of two
invariant serine residues in chorismate synthase provides evidence that a proton relay
system is essential for intermediate formation and catalytic activity, FEBS J 275, 14641473.
69.
Maclean, J., and Ali, S. (2003) The structure of chorismate synthase reveals a novel
flavin binding site fundamental to a unique chemical reaction, Structure 11, 1499-1511.
70.
Norager, S., Jensen, K. F., Bjornberg, O., and Larsen, S. (2002) E. coli dihydroorotate
dehydrogenase reveals structural and functional distinctions between different classes of
dihydroorotate dehydrogenases, Structure 10, 1211-1223.
71.
Small, Y. A., Guallar, V., Soudackov, A. V., and Hammes-Schiffer, S. (2006) Hydrogen
bonding pathways in human dihydroorotate dehydrogenase, J Phys Chem B 110, 1970419710.
72.
Kow, R. L., Whicher, J. R., McDonald, C. A., Palfey, B. A., and Fagan, R. L. (2009)
Disruption of the proton relay network in the class 2 dihydroorotate dehydrogenase from
Escherichia coli, Biochemistry 48, 9801-9809.
73.
Fagan, R. L., Nelson, M. N., Pagano, P. M., and Palfey, B. A. (2006) Mechanism of
flavin reduction in class 2 dihydroorotate dehydrogenases, Biochemistry 45, 1492614932.
50
74.
Jorns, M. S., Chen, Z. W., and Mathews, F. S. (2010) Structural characterization of
mutations at the oxygen activation site in monomeric sarcosine oxidase, Biochemistry 49,
3631-3639.
75.
Chen, Z. W., Hassan-Abdulah, A., Zhao, G., Jorns, M. S., and Mathews, F. S. (2006)
Heterotetrameric sarcosine oxidase: structure of a diflavin metalloenzyme at 1.85 A
resolution, J Mol Biol 360, 1000-1018.
76.
Harris, C. M., Molla, G., Pilone, M. S., and Pollegioni, L. (1999) Studies on the reaction
mechanism of Rhodotorula gracilis D-amino-acid oxidase. Role of the highly conserved
Tyr-223 on substrate binding and catalysis, J Biol Chem 274, 36233-36240.
77.
Boselli, A., Sacchi, S., Job, V., Pilone, M. S., and Pollegioni, L. (2002) Role of tyrosine
238 in the active site of Rhodotorula gracilis D-amino acid oxidase. A site-directed
mutagenesis study, Eur J Biochem 269, 4762-4771.
78.
Ferreira, P., Ruiz-Duenas, F. J., Martinez, M. J., van Berkel, W. J., and Martinez, A. T.
(2006) Site-directed mutagenesis of selected residues at the active site of aryl-alcohol
oxidase, an H2O2-producing ligninolytic enzyme, FEBS J 273, 4878-4888.
79.
Fernandez, I. S., Ruiz-Duenas, F. J., Santillana, E., Ferreira, P., Martinez, M. J.,
Martinez, A. T., and Romero, A. (2009) Novel structural features in the GMC family of
oxidoreductases revealed by the crystal structure of fungal aryl-alcohol oxidase, Acta
Crystallogr D Biol Crystallogr 65, 1196-1205.
80.
Zhao, G., and Jorns, M. S. (2005) Ionization of zwitterionic amine substrates bound to
monomeric sarcosine oxidase, Biochemistry 44, 16866-16874.
51
81.
Heroux, A., Bozinovski, D. M., Valley, M. P., Fitzpatrick, P. F., and Orville, A. M.
(2009) Crystal structures of intermediates in the nitroalkane oxidase reaction,
Biochemistry 48, 3407-3416.
82.
Lindqvist, Y., Koskiniemi, H., Jansson, A., Sandalova, T., Schnell, R., Liu, Z., Mantsala,
P., Niemi, J., and Schneider, G. (2009) Structural basis for substrate recognition and
specificity in aklavinone-11-hydroxylase from rhodomycin biosynthesis, J Mol Biol 393,
966-977.
83.
Ha, J. Y., Min, J. Y., Lee, S. K., Kim, H. S., Kim do, J., Kim, K. H., Lee, H. H., Kim, H.
K., Yoon, H. J., and Suh, S. W. (2006) Crystal structure of 2-nitropropane dioxygenase
complexed with FMN and substrate. Identification of the catalytic base, J Biol Chem 281,
18660-18667.
84.
Stenberg, K., Clausen, T., Lindqvist, Y., and Macheroux, P. (1995) Involvement of Tyr24
and Trp108 in substrate binding and substrate specificity of glycolate oxidase, Eur J
Biochem 228, 408-416.
85.
Moran, G. R., Entsch, B., Palfey, B. A., and Ballou, D. P. (1999) Mechanistic insights
into p-hydroxybenzoate hydroxylase from studies of the mutant Ser212Ala, Biochemistry
38, 6292-6299.
86.
Mortl, M., Diederichs, K., Welte, W., Molla, G., Motteran, L., Andriolo, G., Pilone, M.
S., and Pollegioni, L. (2004) Structure-function correlation in glycine oxidase from
Bacillus subtilis, J Biol Chem 279, 29718-29727.
87.
Leiros, I., Wang, E., Rasmussen, T., Oksanen, E., Repo, H., Petersen, S. B.,
Heikinheimo, P., and Hough, E. (2006) The 2.1 A structure of Aerococcus viridans L-
52
lactate oxidase (LOX), Acta Crystallogr Sect F Struct Biol Cryst Commun 62, 11851190.
88.
Fu, G., Yuan, H., Li, C., Lu, C. D., Gadda, G., and Weber, I. T. (2010) Conformational
changes and substrate recognition in Pseudomonas aeruginosa D-arginine dehydrogenase,
Biochemistry 49, 8535-8545.
89.
Bjornberg, O., Gruner, A. C., Roepstorff, P., and Jensen, K. F. (1999) The activity of
Escherichia coli dihydroorotate dehydrogenase is dependent on a conserved loop
identified by sequence homology, mutagenesis, and limited proteolysis, Biochemistry 38,
2899-2908.
90.
Palfey, B. A., Bjornberg, O., and Jensen, K. F. (2001) Insight into the chemistry of flavin
reduction and oxidation in Escherichia coli dihydroorotate dehydrogenase obtained by
rapid reaction studies, Biochemistry 40, 4381-4390.
91.
Bjornberg, O., Rowland, P., Larsen, S., and Jensen, K. F. (1997) Active site of
dihydroorotate dehydrogenase A from Lactococcus lactis investigated by chemical
modification and mutagenesis, Biochemistry 36, 16197-16205.
92.
Opperman, D. J., Sewell, B. T., Litthauer, D., Isupov, M. N., Littlechild, J. A., and van
Heerden, E. (2010) Crystal structure of a thermostable old yellow enzyme from Thermus
scotoductus SA-01, Biochem Biophys Res Commun 393, 426-431.
93.
Huang, C. H., Lai, W. L., Lee, M. H., Chen, C. J., Vasella, A., Tsai, Y. C., and Liaw, S.
H. (2005) Crystal structure of glucooligosaccharide oxidase from Acremonium strictum:
a novel flavinylation of 6-S-cysteinyl, 8alpha-N1-histidyl FAD, J Biol Chem 280, 3883138838.
53
94.
Lin, S. F., Yang, T. Y., Inukai, T., Yamasaki, M., and Tsai, Y. C. (1991) Purification and
characterization of a novel glucooligosaccharide oxidase from Acremonium strictum T1,
Biochim Biophys Acta 1118, 41-47.
95.
Saijo, T., Kim, J. J., Kuroda, Y., and Tanaka, K. (1998) The roles of threonine-136 and
glutamate-137 of human medium chain acyl-CoA dehydrogenase in FAD binding and
peptide folding using site-directed mutagenesis: creation of an FAD-dependent mutant,
T136D, Arch Biochem Biophys 358, 49-57.
96.
Kim, J. J., Wang, M., and Paschke, R. (1993) Crystal structures of medium-chain acylCoA dehydrogenase from pig liver mitochondria with and without substrate, Proc Natl
Acad Sci U S A 90, 7523-7527.
97.
Andresen, B. S., Bross, P., Udvari, S., Kirk, J., Gray, G., Kmoch, S., Chamoles, N.,
Knudsen, I., Winter, V., Wilcken, B., Yokota, I., Hart, K., Packman, S., Harpey, J. P.,
Saudubray, J. M., Hale, D. E., Bolund, L., Kolvraa, S., and Gregersen, N. (1997) The
molecular basis of medium-chain acyl-CoA dehydrogenase (MCAD) deficiency in
compound heterozygous patients: is there correlation between genotype and phenotype?,
Hum Mol Genet 6, 695-707.
98.
Setoyama, C., Nishina, Y., Tamaoki, H., Mizutani, H., Miyahara, I., Hirotsu, K., Shiga,
K., and Miura, R. (2002) Effects of hydrogen bonds in association with flavin and
substrate in flavoenzyme d-amino acid oxidase. The catalytic and structural roles of
Gly313 and Thr317, J Biochem 131, 59-69.
99.
Kawazoe, T., Tsuge, H., Pilone, M. S., and Fukui, K. (2006) Crystal structure of human
D-amino acid oxidase: context-dependent variability of the backbone conformation of the
54
VAAGL hydrophobic stretch located at the si-face of the flavin ring, Protein Sci 15,
2708-2717.
100.
Kimura, S., Nishida, H., and Iyanagi, T. (2001) Effects of flavin-binding motif amino
acid mutations in the NADH-cytochrome b5 reductase catalytic domain on protein
stability and catalysis, J Biochem 130, 481-490.
101.
Nishida, H., Inaka, K., Yamanaka, M., Kaida, S., Kobayashi, K., and Miki, K. (1995)
Crystal structure of NADH-cytochrome b5 reductase from pig liver at 2.4 A resolution,
Biochemistry 34, 2763-2767.
102.
Valley, M. P., Fenny, N. S., Ali, S. R., and Fitzpatrick, P. F. (2010) Characterization of
active site residues of nitroalkane oxidase, Bioorg Chem 38, 115-119.
103.
Trickey, P., Basran, J., Lian, L. Y., Chen, Z., Barton, J. D., Sutcliffe, M. J., Scrutton, N.
S., and Mathews, F. S. (2000) Structural and biochemical characterization of recombinant
wild type and a C30A mutant of trimethylamine dehydrogenase from methylophilus
methylotrophus (sp. W(3)A(1)), Biochemistry 39, 7678-7688.
104.
Basran, J., Jang, M. H., Sutcliffe, M. J., Hille, R., and Scrutton, N. S. (1999) The role of
Tyr-169 of trimethylamine dehydrogenase in substrate oxidation and magnetic interaction
between FMN cofactor and the 4Fe/4S center, J Biol Chem 274, 13155-13161.
105.
Binda, C., Newton-Vinson, P., Hubalek, F., Edmondson, D. E., and Mattevi, A. (2002)
Structure of human monoamine oxidase B, a drug target for the treatment of neurological
disorders, Nat Struct Biol 9, 22-26.
106.
Li, M., Binda, C., Mattevi, A., and Edmondson, D. E. (2006) Functional role of the
"aromatic cage" in human monoamine oxidase B: structures and catalytic properties of
Tyr435 mutant proteins, Biochemistry 45, 4775-4784.
55
107.
Graham, J. E., and Wilkinson, B. J. (1992) Staphylococcus aureus osmoregulation: roles
for choline, glycine betaine, proline, and taurine, J Bacteriol 174, 2711-2716.
108.
Peddie, B. A., Chambers, S. T., and Lever, M. (1996) Is the ability of urinary tract
pathogens to accumulate glycine betaine a factor in the virulence of pathogenic strains?, J
Lab Clin Med 128, 417-422.
109.
Sakamoto, A., Valverde, R., Alia, Chen, T. H., and Murata, N. (2000) Transformation of
Arabidopsis with the codA gene for choline oxidase enhances freezing tolerance of
plants, Plant J 22, 449-453.
110.
Chowdhury, S. P., Nagarajan, T., Tripathi, R., Mishra, M. N., Le Rudulier, D., and
Tripathi, A. K. (2007) Strain-specific salt tolerance and osmoregulatory mechanisms in
Azospirillum brasilense, FEMS Microbiol Lett 267, 72-79.
111.
Quaye, O., Cowins, S., and Gadda, G. (2009) Contribution of flavin covalent linkage
with histidine 99 to the reaction catalyzed by choline oxidase, J Biol Chem 284, 1699016997.
112.
Quaye, O., Lountos, G. T., Fan, F., Orville, A. M., and Gadda, G. (2008) Role of Glu312
in binding and positioning of the substrate for the hydride transfer reaction in choline
oxidase, Biochemistry 47, 243-256.
113.
Quaye, O., and Gadda, G. (2009) Effect of a conservative mutation of an active site
residue involved in substrate binding on the hydride tunneling reaction catalyzed by
choline oxidase, Arch Biochem Biophys 489, 10-14.
114.
Finnegan, S., and Gadda, G. (2008) Substitution of an active site valine uncovers a
kinetically slow equilibrium between competent and incompetent forms of choline
oxidase, Biochemistry 47, 13850-13861.
56
115.
Finnegan, S., Agniswamy, J., Weber, I. T., and Gadda, G. Role of valine 464 in the flavin
oxidation reaction catalyzed by choline oxidase, Biochemistry 49, 2952-2961.
116.
Rungsrisuriyachai, K., and Gadda, G. (2008) On the role of histidine 351 in the reaction
of alcohol oxidation catalyzed by choline oxidase, Biochemistry 47, 6762-6769.
117.
Ghanem, M., and Gadda, G. (2005) On the catalytic role of the conserved active site
residue His466 of choline oxidase, Biochemistry 44, 893-904.
118.
Rungsrisuriyachai, K., and Gadda, G. (2010) Role of asparagine 510 in the relative
timing of substrate bond cleavages in the reaction catalyzed by choline oxidase,
Biochemistry 49, 2483-2490.
119.
Pollegioni, L., Piubelli, L., Sacchi, S., Pilone, M. S., and Molla, G. (2007) Physiological
functions of D-amino acid oxidases: from yeast to humans, Cell Mol Life Sci 64, 13731394.
120.
Kolodkin-Gal, I., Romero, D., Cao, S., Clardy, J., Kolter, R., and Losick, R. D-amino
acids trigger biofilm disassembly, Science 328, 627-629.
121.
Yoshimura, T., and Esak, N. (2003) Amino acid racemases: functions and mechanisms, J
Biosci Bioeng 96, 103-109.
122.
Hashimoto, A., Nishikawa, T., Oka, T., and Takahashi, K. (1993) Endogenous D-serine
in rat brain: N-methyl-D-aspartate receptor-related distribution and aging, J Neurochem
60, 783-786.
123.
Haas, D., Matsumoto, H., Moretti, P., Stalon, V., and Mercenier, A. (1984) Arginine
degradation in Pseudomonas aeruginosa mutants blocked in two arginine catabolic
pathways, Mol Gen Genet 193, 437-444.
57
124.
Jann, A., Matsumoto, H., and Haas, D. (1988) The fourth arginine catabolic pathway of
Pseudomonas aeruginosa, J Gen Microbiol 134, 1043-1053.
58
Chapter 2
Structural and Kinetic Studies on the Ser101Ala Variant of Choline Oxidase: Catalysis by
Compromise
(This chapter has been published verbatim in Finnegan, S., Yuan, H., Wang, Y., Orville, A.M.,
and Gadda, G., (2010), Archives of Biochemistry and Biophysics 501, 207-213. The author’s
contribution to this chapter involved the preparation and purification of the mutant enzyme, the
determination of the steady state kinetic parameters and the reductive half-reaction kinetic
parameters.)
2.1
Abstract
The oxidation of choline catalyzed by choline oxidase includes two reductive half-reactions
where FAD is reduced by the alcohol substrate and by an aldehyde intermediate transiently
formed in the reaction. Each reductive half-reaction is followed by an oxidative half-reaction
where the reduced flavin is oxidized by molecular oxygen. In the present study, we have used
site-directed mutagenesis to prepare the Ser101Ala mutant of choline oxidase and have
investigated the impact of this mutation on the structural and kinetic properties of the enzyme.
The X-ray crystallographic structure of the Ser101Ala enzyme resolved to 2.2 Å indicates that
the only differences between the mutant and wild-type enzymes are the lack of a hydroxyl group
on residue 101 and a more planar configuration of the flavin isoalloxazine ring with respect to
that of the wild-type enzyme. Steady state and rapid reaction kinetics established that
replacement of Ser101 with alanine yields a mutant enzyme with increased efficiencies in the
oxidative half-reactions and decreased efficiencies in the reductive half-reactions. This is
accompanied by a significant decrease in the overall rate of turnover with choline. Thus, this
mutation has revealed the importance of a specific residue for the optimization of the overall
59
turnover of choline oxidase, which requires fine-tuning of four consecutive half-reactions for the
conversion of an alcohol to a carboxylic acid.
2.2
Introduction
The reaction of choline oxidation catalyzed by choline oxidase (E.C. 1.1.3.17; choline-
oxygen 1-oxidoreductase) has been extensively characterized (see (1) for a recent review) due to
its relevance for genetically engineering crops with increased tolerance to environmental stress
and the potential for designing therapeutic agents against pathogenic bacteria (2-5). In brief, the
reaction includes two reductive half-reactions where the FAD cofactor is reduced by the alcohol
substrate and by an enzyme-associated aldehyde intermediate (Scheme 2.1) (6). Each reductive
half-reaction is followed by an oxidative half-reaction where the reduced FAD cofactor is
oxidized by molecular oxygen with formation of hydrogen peroxide (6). In the wild-type
enzyme, the first reductive half-reaction is initiated by a kinetically fast abstraction of the
hydroxyl proton of choline, which results in the formation of a transient alkoxide intermediate
(6). This is followed by a rate-limiting hydride ion transfer from the α-carbon of the alkoxide
intermediate to the N(5) atom of the flavin resulting in the oxidation of choline to betaine
aldehyde and reduction of the flavin (6). Betaine aldehyde is subsequently hydrated in the active
site of the enzyme to form gem-diol choline (7, 8). In the second reductive half-reaction, the
gem-diol choline is oxidized to the product, glycine betaine. In the oxidative half-reactions, the
reduced flavin reacts with oxygen by transferring two electrons to oxygen to form oxidized
flavin and hydrogen peroxide (9). These reactions occur when the organic product of the
enzymatic reaction is bound at the active site, rather than after its release to the solvent (6, 9).
Maximal overall enzymatic turnover is therefore attained through a fine balancing of the
60
requirements that are necessary for each of the half-reactions to go forward efficiently without
negatively impacting any of the other half-reactions.
By using site-directed mutagenesis the mechanistic roles of several functional groups in the
active site of the choline oxidase from Arthrobacter globiformis have been elucidated. His99,
which is the site of covalent attachment of the flavin to the protein moiety, is important for the
optimal positioning of FAD in the enzyme-alkoxide complex, thereby facilitating the
environmentally assisted transfer of the hydride ion in the oxidation of choline (10). Glu312 is
the
main
site
of
substrate
anchoring
through
electrostatic
interaction
with
the
trimethylammonium group of choline, playing a role for the optimal positioning of the substrate
for efficient catalysis (11, 12). His351 is important for the binding and positioning of the
substrate for efficient hydride ion transfer, for stabilization of the transition state developed
during choline oxidation, and for fine-tuning the polarity of the active site (13). Val464
participates primarily in the oxidative half-reaction where the reduced flavin is oxidized by
oxygen, with minimal effects on the reaction of choline oxidation (14, 15). Finally, His466
modulates the electrophilicity of the enzyme-bound flavin, the polarity of the active site and
contributes to the stabilization of the transition state for the oxidation of choline to betaine
aldehyde (16, 17). Interestingly, none of the three active site histidines of choline oxidase (i.e.,
His99, His351, and His466) provides electrostatic stabilization of either the superoxide anion
intermediate or the transition state that are formed in the reaction of the reduced flavin with
oxygen (10, 13, 16). In contrast, such an electrostatic stabilization is exerted by the positive
charge of the enzyme-bound organic molecule undergoing oxidation in the reaction catalyzed by
choline oxidase, as suggested by mechanistic studies with a substrate analog devoid of positive
charge (18, 19).
61
In the crystal structure of wild-type choline oxidase, the side chain of Ser101 is less than 4 Å
from the N(5) atom of FAD and within hydrogen bonding distance of the oxygen atom of
DMSO, a ligand that was used in the crystallization of the enzyme (11). This suggests that
Ser101 may be actively involved in the oxidation of choline catalyzed by the enzyme. Here, we
report the expression, purification, as well as the crystallographic and kinetic characterizations of
the Ser101Ala variant of the enzyme. The results showed that replacement of Ser101 with
alanine increases the apparent rates for the oxidative half-reactions by three-fold, while
decreasing those for the reductive half-reactions and the overall turnover of the enzyme by tenfold. Thus, while the hydroxyl side chain of Ser101 is not essential for catalysis, it is important
for the optimization of the enzymatic turnover of choline oxidase.
E-FADox-CH
CH, k1
k3
k2
k4
E-FADox
GB
E-FADred-BA
O2, H2O
k11
k5
H2O2
E-FADox-BA
E-FADox-GB
k9
H2O2
O2
k8
k7
E-FADred-GB
Scheme 2.1 The Steady State Kinetic Mechanism of Choline Oxidation Catalyzed by Choline
Oxidasea. aE, enzyme; FADox, oxidized flavin; CH, choline; FADred, reduced flavin; BA, betaine
aldehyde; GB, glycine betaine.
2.3
Experimental Procedures
The mutant gene for the Ser101Ala enzyme was prepared using the QuikChangeTM Site-
Directed Mutagenesis kit following the manufacturer’s instructions in the presence of 2%
62
DMSO, as previously described (17, 20). The pET/codAmg plasmid harboring the wild-type
gene was used as template for mutagenesis (20). Upon mutagenesis, the entire mutant gene
(pET/codAmg-Ser101Ala) was sequenced at the DNA Core Facility of Georgia State University
to confirm the presence of the desired mutation. As an expression host, competent Escherichia
coli Rosetta(DE3)pLysS cells were transformed with the mutant plasmid by electroporation. The
mutant enzyme was expressed and the resulting enzyme was purified to homogeneity as
previously described for wild-type choline oxidase (17, 18, 20).
Crystals of the Ser101Ala enzyme were grown by the hanging-drop vapor-diffusion method
at room temperature. Purified Ser101Ala (2 μL) at a concentration of 5 mg mL-1 was mixed with
2 μL from a 500 μL reservoir solution consisting of 80 mM sodium cacodylate, 20% v/v
PEG6000, 20% v/v glycerol, 150 mM Mg-acetate at pH 6.0. Single crystals were transferred into
a cryoprotectant consisting of reservoir solution containing 25% (v/v) glycerol and allowed to
soak for approximately 2 min prior to flash-freezing in liquid nitrogen for data collection at
Beamline 12B of the National Synchrotron Light Source at Brookhaven National Laboratory,
NY. The data were integrated, scaled, and merged using the HKL2000 package (21). The
structures were solved by molecular replacement with PHASER (22) using the structure of wildtype choline oxidase (2JBV from the Protein Data Bank) as the starting model (23). Refinement
was carried out using Refmac5 (24) in CCP4 (25) and manual adjustment used the molecular
graphics program COOT (26). Structural figures were made with PyMol software (27).
Steady state kinetic parameters were measured with the method of the initial rates (28) at
varying concentrations of both choline, or betaine aldehyde, and oxygen in 50 mM sodium
pyrophosphate, pH 10.0, at 25 oC. Kinetic assays were performed at pH 10.0 because at this pH
value the kinetic parameters kcat and kcat/Km of choline oxidase are maximal and independent of
63
pH (9). Initial rates were determined by monitoring the rate of oxygen consumption with a
computer-interfaced Oxy-32 oxygen monitoring system (Hansatech Instrument Ltd.) at 25 °C.
The assay reaction mixture was equilibrated at the desired concentration of oxygen by sparging
the appropriate O2/N2 gas mixture for 10 min before the reaction was started with the addition of
the enzyme. The initial rates measured with choline were fit to eq 1, which describes a sequential
steady state kinetic mechanism in which Kcholine and Koxygen are the Michaelis constants for
choline and oxygen and kcat is the overall turnover number of the enzyme (e) when saturated with
both substrates. The initial rates measured with betaine aldehyde were fit to eq 2, which
describes a sequential steady state kinetic mechanism of the type described by eq 1 where
Kaldehyde « KoxygenKia.
k cat [choline][oxygen]
v
=
e K choline [oxygen] + K oxygen [choline] + [choline][oxygen] + K ia K oxygen
(1)
kcat [betaine − aldehyde][oxygen]
v
=
e K ia K oxygen + K oxygen[betaine − aldehyde] + [betaine − aldehyde][oxygen]
(2)
The reductive half-reaction with betaine aldehyde was carried out by using a Hi-Tech SF-61
stopped-flow spectrophotometer at 25 °C and pH 10.0. The rate constants for flavin reduction
were measured by monitoring the decrease in absorbance at 452 nm that results from the
anaerobic mixing of the enzyme and betaine aldehyde, as previously described for the wild-type
enzyme (6). Glucose (5 mM) and glucose oxidase (0.5 µM) were added to the substrate and
enzyme solutions to scavenge possible trace amounts of oxygen. The Ser101Ala enzyme was
mixed anaerobically with an equal volume of betaine aldehyde, obtaining reaction mixtures with
10 µM enzyme and 0.2-5 mM betaine aldehyde. At each concentration of the substrate, the rate
constants for flavin reduction were recorded in triplicate, with measurements usually differing by
≤5 %. Stopped-flow traces were fit to eq 3, which describes a single exponential process where
64
kobs is the observed first-order rate constant for flavin reduction, A is the value of absorbance at
452 nm at time t, B is the amplitude of the absorbance change, and C is an offset value that
accounts for the non-zero absorbance value at infinite time. Kinetic parameters for the reductive
half-reaction were determined by using eq 4, where kobs is the observed first-order rate constant
for the reduction of the enzyme-bound flavin at any given concentration of substrate, kred is the
limiting first-order rate constant for flavin reduction at saturated substrate concentration, and
app
Kd is the macroscopic dissociation constant for binding of the substrate to the enzyme.
2.4
A = B exp(− kobst ) + C
(3)
k A
kobs = app red
Kd + A
(4)
Results
The Ser101Ala enzyme was expressed and purified to high yields by using the same protocol
previously used for the wild-type enzyme (11, 20). As in the case of wild-type choline oxidase,
the enzyme-bound flavin was present as a mixture of oxidized and anionic flavosemiquinone
throughout the purification procedure, with the fully oxidized enzyme obtained after extensive
dialysis at pH 6.0 (9, 20). The purified Ser101Ala enzyme showed absorbance bands centered at
373 nm (ε = 8.9 mM-1cm-1) and 452 nm (ε = 9.5 mM-1cm-1) at pH 8.0, compared to 367 nm (ε =
10.0 mM-1cm-1) and 455 nm (ε = 11.4 mM-1cm-1) for the wild-type enzyme (9), consistent with
an altered flavin microenvironment in the mutant enzyme. The enzymatic activity of the
Ser101Ala enzyme was 3.3 s-1 with saturated choline (i.e., 10 mM) at pH 7.0 and 25 oC,
compared to 15 s-1 for the wild-type (11), suggesting that Ser101 is important, but not essential,
for catalysis.
65
The Ser101Ala enzyme crystallized in the primitive monoclinic space group P21 with eight
subunits assembled as four homodimers in the asymmetric unit. X-ray data to 2.2 Å were used to
refine the structure of the Ser101Ala enzyme to a final R-factor of 22.8% (Table 2.1). The final
model consisted of residues 1 to 5301, FAD, and acetate from the crystallization solution in the
active site of the enzyme. The acetate ligand in the active site of the Ser101Ala enzyme had the
same orientation and positioning in the eight subunits that crystallized; its oxygen atoms are 3.8
Å from both the flavin N(5) atom and the N(3) atom of His351 (Figure 2.1C). When the
homodimers of the Ser101Ala enzyme were superimposed on those of the wild-type enzyme the
average rmsd values were 0.41 Å for 527 equivalent Cα atoms in each chain, indicating no
significant structural differences between the backbone atoms of the two enzymes. The active
sites of the two structures are compared in Figure 2.1, showing essentially no differences in the
relative location of the flavin and the active site residues in the two enzymes, although there are
differences in the conformation of the flavin. Interestingly, although the replacement of the side
chain on residue 101 from serine to alanine resulted in loss of a hydrogen bond with His351, all
of the active site residues are located in the same positions and orientations in the mutant and
wild-type enzymes. Moreover, the electron density maps clearly indicate that FAD in the
Ser101Ala enzyme is covalently linked to the His99 Nє2 atom as seen in the wild-type (11).
A notable difference between the wild-type and the Ser101Ala enzymes is the isoalloxazine
moiety of the FAD cofactor in the variant enzyme being more planar than that of the wild-type.
In the latter enzyme, the C(4a) atom was shown to be in an sp3 hybridization due to the presence
of an oxygen adduct that is not observed in the Ser101Ala variant enzyme resulting in the C(4a)
1
In the 3D structure of the Ser101Ala enzyme reported here (3NNE), the last 16-18 amino acids at the C-terminus
of different subunits are not observable. Similarly, the wild-type enzyme was reported with coordinates for the first
528 residues and no ordered density for the last 18 amino acid residues at the C-terminus (2JBV). A third structure
of a choline oxidase variant recently reported also has coordinates for the first 530 residues, but not for the last 16
residues at the C-terminus (3LJP). This is most likely due to the C-terminal end of the protein being highly flexible.
66
atom being sp2 hybridized (11, 29). In the Ser101Ala enzyme the isoalloxazine ring system
showed a slight bend between the two planes containing the benzene and pyrimidine moieties
defining an 8o angle along the N(5)-N(10) axis (Figure 2.1B). A similar bend was recently
observed in the structure of another active site mutant form of choline oxidase where Val464 is
replaced with alanine, for which the C(4a) atom of FAD was also shown to be sp2 hybridized
(14). In summary, the major differences observable in the structures of the Ser101Ala and wildtype enzymes are the hybridization of the flavin C(4a) atom and the loss of the hydroxyl group of
Ser101.
The steady state kinetic parameters with choline or betaine aldehyde as substrate for the
Ser101Ala enzyme were determined at varying concentrations of oxygen, by measuring the
initial rates of oxygen consumption with a Clark-type oxygen electrode at pH 10.0 and 25 oC. As
for the case previously reported for the wild-type enzyme (8, 9), the best fits of the data were
obtained with eq 1 for choline and eq 2 for betaine aldehyde (Figure 2.2). As summarized in
Table 2.2, the Ser101Ala enzyme showed between three-fold and four-fold increases in the
second-order rate constants for oxygen (kcat/Koxygen) with choline and betaine aldehyde. In
contrast, a ten-fold decrease in both the second-order rate constant for the capture of choline
(kcat/Kcholine) and the overall turnover number of the enzyme with choline as substrate (kcat) were
observed. Since the kcat/Kbetaine
aldehyde
value could not be determined by using a steady state
kinetic approach due to Kaldehyde « KoxygenKia, the reductive half-reaction with betaine aldehyde
was investigated by anaerobically mixing the enzyme and the substrate in a stopped-flow
spectrophotometer at pH 10.0 and 25 oC. As previously reported for the wild-type enzyme (6),
the absorbance at 452 nm of the Ser101Ala enzyme decreased in a single exponential manner
upon mixing with betaine aldehyde (Figure 2.3). A plot of the kobs value for flavin reduction as a
67
function of the concentration of betaine aldehyde followed saturation kinetics, allowing the
determination of the limiting rate constant for flavin reduction at saturated substrate (kred) and the
macroscopic equilibrium constant for the formation of the enzyme-betaine aldehyde complex
(appKd). A second-order rate constant for the capture of betaine aldehyde to yield species
committed to catalysis (kred/Kd) of 52,000 M-1s-1 was calculated from the stopped-flow kinetic
data (Table 2.2), which was six-fold lower than the value of 320,000 M-1s-1 previously
determined for the wild-type enzyme (6). Thus, replacement of Ser101 with alanine results in
lowered second-order rate constants for both the reductive half-reactions, but increased secondorder rate constants for both the oxidative half-reactions catalyzed by the enzyme.
Table 2.1 Crystallographic Data Collection and Refinement Statistics.
Space group
P21
Unit cell dimensions: (Å)
A
69.26
B
346.03
C
105.92
β
94.33
Unique reflections
150,964
Rmerge (%) overall (final shell)
11.1(25.8)a
I/σ(I) overall (final shell)
14 (3.2)
Completeness (%) overall (final shell)
90.4 (50.3)
Data range for refinement (Å)
10-2.47
Rcryst (%)b
22.8
29.2
Rfree (%)c
No. of solvent atoms
333
(total occupancies)
(207.3)
RMS deviation from ideality
Bonds (Å)
0.014
Angle distance (Å)
0.021
Average B-factors (Å2)
Main-chain atoms
26.4
Side-chain atoms
27.2
FAD
18.5
Solvent
20.3
a
Values in parentheses are given for the highest resolution shell
b
Rcryst = Σ|Fobs-Fcal|/ΣFobs.
c
Rfree = Σtest(|Fobs|-|Fcal|)2/Σtest|Fobs|2.
68
Figure 2.1
2 Comparisson of crystaal structures of the Ser1001Ala and w
wild-type enzymes of chooline
oxidase.T
The residuess with carbon
ns in green represent
r
thee Ser101Ala structure, w
whereas the
carbon attoms for the wild-type en
nzyme are in
n gray. Paneel A shows thhe interactioons at the
Ser101A
Ala mutation site. Hydrog
gen bonds on
nly present inn the wild-tyype enzyme are colored red,
those onlly in the Ser101Ala enzy
yme are colo
ored green, aand those obsserved in booth enzymes are
colored cyan.
c
Panel B illustrates the superposition of the FAD isoallooxazine ringg from the w
wildtype (greey) with the C(4a)-oxyge
C
en adduct (reed) and the S
Ser101Ala ennzymes (colored by elem
ment
type) of choline
c
oxid
dase. Note th
hat the isoallo
oxazine ringg of the Ser101Ala enzym
me is more
planar than that of th
he wild-type structure. Paanel C show
ws the omit-F
Fourier densiity map withh
contour level
l
at 2 forr the active site
s of the Seer101Ala struucture. Paneel D shows a comparisonn of
the activee sites of willd-type and Ser101Ala
S
choline
c
oxidaase. The willd-type enzym
me structuree is
from pdb
b file 2jbv (1
11).
69
Figure 2.2
2 Anaerobic Double recciprocal plotts of Ser101A
Ala catalyzeed oxidation of choline aand
betaine aldehyde.
a
Ch
holine oxidasse activity was
w measuredd at varying concentratioons of both
choline and
a oxygen or
o betaine ald
dehyde and oxygen in 1 00 mM sodiium pyrophoosphate, pH 10.0,
at 25 oC.. Panel A, e//vo as a funcction of the in
nverse choliine concentraation determ
mined at seveeral
fixed con
ncentrations oxygen: (●)) 0.033 mM; (○) 0.068 m
mM; (■) 0.1772 mM; (□) 00.344 mM. P
Panel
B, e/vo as
a a function of the inverrse oxygen concentrationn determinedd at several ffixed
concentraations cholin
ne: (●) 0.1 mM;
m (○) 0.2 mM;
m (■) 0.55 mM; (□) 2 mM; (▲) 5 mM. Panel C,
e/vo as a function of the inverse betaine
b
aldeh
hyde concenntration deteermined at seeveral fixed
23 mM; (■) 00.36 mM; (□
□) 0.47 mM. Panel D, e/vvo as
concentraations oxygeen: (●) 0.16 mM; (○) 0.2
a function of the inveerse oxygen concentratio
on determineed at severall fixed conceentrations beetaine
aldehydee: (●) 0.5 mM
M; (○) 2 mM
M; (■) 5 mM;; (□) 10 mM
M; (▲) 40 mM
M.
70
Figure 2.3
2 Reductivee half-reactio
on of the Serr101Ala enzzyme with beetaine aldehyyde. Panel A
shows the reduction traces
t
with 0.1
0 mM (black), 0.2 mM
M (blue), 0.3 mM (red), 22.5 mM (greeen)
and 5 mM
M (fuscia) beetaine aldehy
yde. All tracces were fit tto eq 3. Timee indicated iis after the ennd of
flow, i.e., 2.2 ms. Forr clarity, onee experimenttal point eveery 5 is show
wn (vertical llines). Panell B
shows the observed rate
r constantts for flavin reduction
r
ass a function oof the concenntration of
betaine aldehyde.
a
Daata were fit to
o eq 4. Buffe
fer used was 50 mM sodiium pyrophoosphate, pH 10.0.
71
Table 2.2 Comparison of the Kinetic Parameters of Ser101Ala and Wild-Type Choline Oxidase
at pH10.0.
substrate
kinetic parameters
choline
kcat, s-1
kcat/Km, M-1s-1
kcat/Koxygen, M-1s-1
Km, mM
Koxygen, mM
Kia, mM
betaine aldehyde
k3 k 7 d
)
k3 + k 7
k1 k 3
k 2 + k3
k5k9
k5 + k9
k 7 (k 2 + k3 )
k1 ( k 3 + k 7 )
k 3 k 7 (k 5 + k 9 )
k 5 k 9 (k 3 + k 7 )
k 2 k 4 k9
k1k 3 ( k 5 + k 9 )
Ser101Ala
wild-type
6.7 ± 0.1
60 ± 1
25,600 ± 2000
237,000 ± 9000
261,000 ± 6500
86,400 ± 3600
0.26 ± 0.02
0.25 ± 0.01
0.026 ± 0.001
0.69 ± 0.03
3.6 ± 0.1
0.14 ±0.01
(
kcat, s-1
k7
43 ± 3
133 ± 4
kcat/Koxygen, M-1s-1
k9
203,400 ± 33,400
53,400 ± 1600
k7/k9
0.2 ± 0.03
2.5 ± 0.01
kred, s-1
k7
47 ± 2
135 ± 4
Kd, mM
-e
Koxygen, mM
a
definitionc
0.9 ± 0.1
o
b
0.45 ± 0.1
c
Conditions: 50 mM sodium pyrophosphate, pH 10.0, at 25 C. From ref (6). Kinetic data are
readily accounted for with the kinetic mechanism of Scheme 2.1 [6]. d While k7 >> k3, kcat is
equal to k3, Km is equal to (k2+ k3)/k1 and Koxygen is equal to k3(k5+ k9)/(k5 k9), as for the case of
the Ser101Ala enzyme. eNot applicable to the mechanism of Scheme 2.1.
2.5
Discussion
In choline oxidase, each of the four half-reactions in which the flavin is reduced by the
organic molecules and oxidized by molecular oxygen is adjusted to achieve the most efficient
overall turnover of the enzyme rather than maximal efficiencies. This conclusion stems from the
comparison of the steady state and rapid reaction kinetic data of the Ser101Ala and the wild-type
enzymes. Indeed, in spite of increased efficiencies in the oxidative half-reactions catalyzed by
the enzyme lacking the hydroxyl group on residue 101, there is a ten-fold decrease in the overall
72
turnover of the enzyme with choline that is imputable to decreased efficiencies in the reductive
half-reactions catalyzed by the enzyme. In the wild-type enzyme, Ser101 is on the si-face of the
isoalloxazine moiety of the flavin with its hydroxyl group being 4 Å away from the flavin N(5)
atom. Such a geometric arrangement of residue 101, as well as of all of the other residues in the
active site of choline oxidase, is maintained in the crystallographic structure of the Ser101Ala
enzyme, which clearly lacks the hydroxyl group. Thus, the differences seen in the kinetic
properties of the mutant and wild-type enzymes are solely due to the lack of the hydroxyl group
on residue 101, rather than originating from other geometric or steric factors.
Replacement of Ser101 with alanine results in decreased efficiencies in the reductive halfreactions with betaine aldehyde and choline. Evidence for this conclusion comes from the
comparison of the steady state and rapid reaction kinetic data determined with the Ser101Ala and
the wild-type enzymes presented here. Indeed, the second-order rate constants for the capture of
choline (kcat/Kcholine) and betaine aldehyde (kred/Kd) to yield enzyme-substrate complexes that are
committed to flavin reduction were ten- and six-fold lower in the enzyme lacking the hydroxyl
group on residue 101. In the wild-type enzyme, the chemical steps of flavin reduction with
choline and the aldehyde intermediate are fully rate-limiting in both reductive half-reactions and
are the sole contributors to the overall rate of turnover of the enzyme (6). Moreover, the rate for
the overall turnover with betaine aldehyde as substrate is limited solely by the chemical step of
flavin reduction, with a rate constant of 135 s-1 (6). The latter conclusion also applies to the
Ser101Ala enzyme, since the rate constant for flavin reduction with betaine aldehyde is very
similar to the turnover number of the enzyme (i.e., 47 s-1 and 43 s-1 for kred and kcat, respectively).
In contrast, the overall turnover of the Ser101Ala enzyme with choline is primarily limited by the
chemical step of flavin reduction with choline (e.g., with a kcat value of 6.7 s-1 and a kred value of
73
6.8 s-1 with choline as substrate at pH 10.0, see chapter 3) without contribution of the subsequent
chemical step of flavin reduction involving the aldehyde intermediate. Since the substrate for the
second reaction of flavin reduction in choline oxidase is the gem-diol species rather than the
aldehyde, it is possible that the extra hydroxyl group of the gem-diol substrate compensates for
the lack of the hydroxyl group on residue 101 in the mutant enzyme (7), thereby making the
chemical step of flavin reduction significantly faster with betaine aldehyde than choline. The
alternative possibility of betaine aldehyde being released from the active site of the Ser101Ala
enzyme turning over with choline rather than proceeding through the second step of flavin
reduction seems less likely based on the similarity of the three dimensional structures of the
mutant and wild-type enzymes and the fact that the active site is completely secluded from the
bulk solvent in choline oxidase (11, 14, 29)
Lack of a hydroxyl group on residue 101 in the active site of choline oxidase results in
increased efficiencies in the oxidative half-reactions with either choline or betaine aldehyde as
substrate. Evidence for this conclusion comes from the comparison of the steady state kinetic
data, showing that the second-order rate constants for reaction of the reduced Ser101Ala enzyme
with oxygen (kcat/Koxygen) were three- to four-fold larger than those observed with the wild-type
enzyme. The microenvironment in proximity of the reactive C(4a) atom of the reduced flavin is
considerably less hydrophilic in the Ser101Ala enzyme than in the wild-type enzyme, suggesting
that this may be the cause of the increased reactivity with oxygen of the reduced flavin in the
mutant enzyme. The importance of a hydrophobic microenvironment in proximity of the C(4a)
atom of the flavin was recently demonstrated in choline oxidase with variant enzymes where
Val464 was replaced with either alanine or threonine (14). In that case, the less hydrophobic
character of the mutant enzymes at position 464, which is proximal to the C(4a) atom of the
74
flavin on the opposite face of the flavin than Ser101, resulted in fifty-fold decreases efficiencies
in the kcat/Koxygen value with choline as substrate for the mutant enzyme (14). Alternatively, the
increased efficiencies in the oxidative half-reactions may be due to the absence of the hydrogenbonding interaction between the O(4) atom of the flavin and the side chain of the residue 101.
Interestingly, the Ser101Ala variant is the first case of increased efficiency in the oxidative halfreaction among many mutant variants of choline oxidase where no changes were observed with
respect to the wild-type enzyme (His99Asn, Glu312Asp, His351Ala, and His466Ala) or where
decreased efficiencies were observed (Val464Ala and Val464Thr) (10, 11, 13-15). To our
knowledge, the choline oxidase Ser101Ala mutant represents the first instance in which the
efficiency of a flavoprotein oxidase for the reaction of the reduced flavin with oxygen has been
increased by at least three-fold with respect to the wild-type enzyme.
The flavin adduct between the C(4a) atom of FAD and oxygen that was previously observed
in the crystal structure of the wild-type form of choline oxidase (11) was not observed in the
structure of the Ser101Ala enzyme reported in this study. In the case of the wild-type enzyme
many interactions were shown to stabilize the flavin-adduct within the active site of the enzyme,
among which a hydrogen bond involving the O(4) atom of the flavin and the hydroxyl group of
Ser101 (11, 29). Thus, it is possible that the absence of the flavin adduct in the Ser101Ala
enzyme is due to the inability to form a hydrogen bond between the O(4) atom of the
isoalloxazine ring and the side chain of residue 101. Alternatively, the lack of a flavin adduct in
the structure of the Ser101Ala enzyme may simply be due to the FAD not being reduced during
the data collection at the synchrotron, thereby preventing the formation of the C(4a) adduct
between the reduced enzyme-bound flavin and oxygen. A more planar flavin with respect to the
wild-type enzyme was recently reported for another mutant form of choline oxidase, where
75
Val464 was replaced with alanine (14). In that case, a significantly lowered apparent affinity of
the reduced enzyme for oxygen, which could not be saturated at concentrations of oxygen as
high as 1 mM, was proposed to explain the lack of a flavin adduct in the crystal structure of the
mutant enzyme (14).
In conclusion, replacement of Ser101 with alanine in the active site of choline oxidase
produced a variant enzyme with increased efficiencies in the oxidative half-reactions where the
reduced flavin reacts with oxygen and decreased efficiencies in the reductive half-reactions with
choline and the aldehyde intermediate of reaction. Interestingly, this mutation has disclosed the
importance of a protein hydroxyl group in close proximity of the flavin C(4a) atom in the
optimization of the overall turnover of choline oxidase, which requires fine tuning of four
consecutive half-reactions for the conversion of an alcohol to a carboxylic acid.
2.6
Acknowledgment
Use of the National Synchrotron Light Source, Brookhaven National Laboratory, was supported
by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences, under
Contract No. DE-AC02-98CH10886.
2.7
1.
References
Gadda, G. (2008) Hydride transfer made easy in the reaction of alcohol oxidation
catalyzed by flavin-dependent oxidases, Biochemistry 47, 13745-13753.
2.
Waditee, R., Bhuiyan, N. H., Hirata, E., Hibino, T., Tanaka, Y., Shikata, M., and Takabe,
T. (2007) Metabolic engineering for betaine accumulation in microbes and plants, J Biol
Chem 282, 34185-34193.
76
3.
Quan, R., Shang, M., Zhang, H., Zhao, Y., and Zhang, J. (2004) Engineering of enhanced
glycine betaine synthesis improves drought tolerance in maize, Plant Biotechnol J 2, 477486.
4.
Sakamoto, A., Valverde, R., Alia, Chen, T. H., and Murata, N. (2000) Transformation of
Arabidopsis with the codA gene for choline oxidase enhances freezing tolerance of
plants, Plant J 22, 449-453.
5.
Holmstrom, K. O., Somersalo, S., Mandal, A., Palva, T. E., and Welin, B. (2000)
Improved tolerance to salinity and low temperature in transgenic tobacco producing
glycine betaine, J Exp Bot 51, 177-185.
6.
Fan, F., and Gadda, G. (2005) On the catalytic mechanism of choline oxidase, J Am
Chem Soc 127, 2067-2074.
7.
Fan, F., Germann, M. W., and Gadda, G. (2006) Mechanistic studies of choline oxidase
with betaine aldehyde and its isosteric analogue 3,3-dimethylbutyraldehyde, Biochemistry
45, 1979-1986.
8.
Gadda, G. (2003) Kinetic mechanism of choline oxidase from Arthrobacter globiformis,
Biochim Biophys Acta 1646, 112-118.
9.
Ghanem, M., Fan, F., Francis, K., and Gadda, G. (2003) Spectroscopic and kinetic
properties of recombinant choline oxidase from Arthrobacter globiformis, Biochemistry
42, 15179-15188.
10.
Quaye, O., Cowins, S., and Gadda, G. (2009) Contribution of flavin covalent linkage
with histidine 99 to the reaction catalyzed by choline oxidase, J Biol Chem 284, 1699016997.
77
11.
Quaye, O., Lountos, G. T., Fan, F., Orville, A. M., and Gadda, G. (2008) Role of Glu312
in binding and positioning of the substrate for the hydride transfer reaction in choline
oxidase, Biochemistry 47, 243-256.
12.
Quaye, O., and Gadda, G. (2009) Effect of a conservative mutation of an active site
residue involved in substrate binding on the hydride tunneling reaction catalyzed by
choline oxidase, Arch Biochem Biophys 489, 10-14.
13.
Rungsrisuriyachai, K., and Gadda, G. (2008) On the role of histidine 351 in the reaction
of alcohol oxidation catalyzed by choline oxidase, Biochemistry 47, 6762-6769.
14.
Finnegan, S., Agniswamy, J., Weber, I. T., and Gadda, G. Role of valine 464 in the flavin
oxidation reaction catalyzed by choline oxidase, Biochemistry 49, 2952-2961.
15.
Finnegan, S., and Gadda, G. (2008) Substitution of an active site valine uncovers a
kinetically slow equilibrium between competent and incompetent forms of choline
oxidase, Biochemistry 47, 13850-13861.
16.
Ghanem, M., and Gadda, G. (2005) On the catalytic role of the conserved active site
residue His466 of choline oxidase, Biochemistry 44, 893-904.
17.
Ghanem, M., and Gadda, G. (2006) Effects of reversing the protein positive charge in the
proximity of the flavin N(1) locus of choline oxidase, Biochemistry 45, 3437-3447.
18.
Gadda, G., Powell, N. L., and Menon, P. (2004) The trimethylammonium headgroup of
choline is a major determinant for substrate binding and specificity in choline oxidase,
Arch Biochem Biophys 430, 264-273.
19.
Gadda, G., Fan, F., and Hoang, J. V. (2006) On the contribution of the positively charged
headgroup of choline to substrate binding and catalysis in the reaction catalyzed by
choline oxidase, Arch Biochem Biophys 451, 182-187.
78
20.
Fan, F., Ghanem, M., and Gadda, G. (2004) Cloning, sequence analysis, and purification
of choline oxidase from Arthrobacter globiformis: a bacterial enzyme involved in
osmotic stress tolerance, Arch Biochem Biophys 421, 149-158.
21.
Otwinowski, Z., and Minor, W. (1997) Processing of X-ray diffraction data collected in
oscillation mode, Methods Enzymol 276, 307-326.
22.
McCoy, A. J., Grosse-Kunstleve, R. W., Adams, P. D., Winn, M. D., Storoni, L. C., and
Read, R. J. (2007) Phaser crystallographic software, J Appl Crystallogr 40, 658-674.
23.
Berman, H. M., Westbrook, J., Feng, Z., Gilliland, G., Bhat, T. N., Weissig, H.,
Shindyalov, I. N., and Bourne, P. E. (2000) The Protein Data Bank, Nucleic Acids Res 28,
235-242.
24.
Murshudov, G. N., Vagin, A. A., and Dodson, E. J. (1997) Refinement of
macromolecular structures by the maximum-likelihood method, Acta Crystallogr D Biol
Crystallogr 53, 240-255.
25.
Potterton, E., Briggs, P., Turkenburg, M., and Dodson, E. (2003) A graphical user
interface to the CCP4 program suite, Acta Crystallogr D Biol Crystallogr 59, 1131-1137.
26.
Emsley, P., and Cowtan, K. (2004) Coot: model-building tools for molecular graphics,
Acta Crystallogr D Biol Crystallogr 60, 2126-2132.
27.
DeLano, W. L. (2002) The pymol molecular graphics system, San Carlos, CA.
28.
Allison, R. D., and Purich, D. L. (1979) Practical considerations in the design of initial
velocity enzyme rate assays, Methods Enzymol 63, 3-22.
29.
Orville, A. M., Lountos, G. T., Finnegan, S., Gadda, G., and Prabhakar, R. (2009)
Crystallographic, spectroscopic, and computational analysis of a flavin C4a-oxygen
adduct in choline oxidase, Biochemistry 48, 720-728.
79
CHAPTER 3
Importance of a Serine Proximal to the C(4a)-N(5) Flavin Atoms for Hydride Transfer in
Choline Oxidase
(This chapter has been published verbatim in Yuan, H. and Gadda, G., (2010), Biochemistry 50:
770-779)
3.1
Abbreviations
Ser101Ala enzyme, choline oxidase with Ser101 substituted with alanine; Ser101Thr enzyme,
choline oxidase with Ser101 substituted with threonine; Ser101Val enzyme, choline oxidase with
Ser101 substituted with valine; Ser101Cys enzyme, choline oxidase with Ser101 substituted with
cysteine; Asn510His enzyme choline oxidase with Asn510 substituted with histidine; DMSO,
dimethyl sulfoxide; KIE, kinetic isotope effect.
3.2
Abstract
Choline oxidase catalyzes the flavin-dependent, two-step oxidation of choline to glycine
betaine with the formation of an aldehyde intermediate. In the first oxidation reaction the alcohol
substrate is initially activated to its alkoxide via proton abstraction. Substrate oxidation occurs
via hydride transfer from the alkoxide α-carbon to the N(5) atom of the enzyme-bound flavin. In
the wild-type enzyme, proton and hydride transfers are mechanistically and kinetically
uncoupled. In this study, we have mutagenized an active site serine proximal to the C(4a)-N(5)
atoms of the flavin and investigated the reactions of proton and hydride transfers by using
substrate and solvent kinetic isotope effects. Replacement of Ser101 with threonine, alanine,
cysteine or valine resulted in biphasic traces in anaerobic reductions of the flavin with choline
investigated in a stopped-flow spectrophotometer. Kinetic isotope effects established that the
80
kinetic phases correspond to the proton and hydride transfer reactions catalyzed by the enzyme.
Upon removal of Ser101 there is at least a 15-fold decrease in the rate constants for proton
abstraction, irrespective of whether threonine, alanine, valine or cysteine is present in the mutant
enzymes. A logarithmic decrease spanning 4 orders of magnitude is seen in the rate constants for
hydride transfer with increasing hydrophobicity of the side chain at position 101. This study
shows that the hydrophilic character of a serine residue proximal to the C(4a)-N(5) flavin atoms
is important for efficient hydride transfer.
3.3
Introduction
Choline oxidase (E.C.1.1.3.17; choline-oxygen 1-oxidoreductase) catalyzes the flavin-
mediated oxidation of choline to glycine betaine (1). This reaction is of considerable interest for
medical and biotechnological applications because the intracellular accumulation of glycine
betaine, a compatible solute that does not interfere with cytoplasmic function at high
concentration, is important for normal cell function under conditions of osmotic and temperature
stress in bacteria and plants (30-33). Furthermore, glycine betaine is much more effective than
other compatible solutes in the stabilization of the structure and function of macromolecules and
the highly ordered state of membranes (34). In biotechnology, the codA gene encoding for
choline oxidase is used to improve the tolerance of economically important crops like tomatoes,
potatoes and rice towards environmental stresses due to hypersalinity of the soil or high and low
temperatures (35-37). A membrane associated choline dehydrogenase that is homologous to the
cytosolic oxidase is found in many pathogenic bacteria, where it was recently shown that the
betA gene encoding for the dehydrogenase is likely associated with virulence (38). Thus, the
study of the mechanism of choline oxidase offers an opportunity to control glycine betaine
81
biosynthesis in pathogenic bacteria and to provide osmotic balance with the environment in
plants.
Choline oxidase from Arthrobacter globiformis has been investigated using biophysical (9,
17, 39, 40), structural (11, 29), mechanistic (7, 9, 41-46), computational (47), and site-directed
mutagenic (11, 13-17, 48-52) approaches. The reaction of choline oxidation catalyzed by choline
oxidase includes two reductive half-reactions where the FAD cofactor is reduced in subsequent
steps by the alcohol substrate and an aldehyde intermediate of the reaction. Each reductive halfreaction is followed by an oxidative half-reaction where the reduced FAD cofactor is oxidized by
molecular oxygen with formation of hydrogen peroxide (Scheme 3.1) (for a recent review, see
(1)). In the wild-type enzyme, the first oxidation is triggered by the kinetically fast abstraction of
the hydroxyl proton of choline by a catalytic base with pKa of ~7.5, which has not been identified
yet (13, 16, 49), resulting in the formation of a transient alkoxide (9, 42, 43, 46). The side chains
of His351 and His466 contribute to the stabilization of the alkoxide species through hydrogen
bonding and electrostatic interactions, respectively (13, 16). A rate-limiting transfer of a hydride
ion from the α-carbon of the alkoxide species to the N(5) atom of the flavin results in the
oxidation of choline and the reduction of the enzyme-bound flavin. This was established with
solvent, substrate, and multiple kinetic isotope effects on the steady state kinetic parameter
kcat/Km with choline as a substrate for the enzyme (1, 43). The abstraction of the hydroxyl proton
of the substrate is associated with an isomerization of the enzyme-substrate complex that is
kinetically independent from the subsequent reaction of hydride ion transfer. This is suggested in
the wild-type enzyme by the comparison of the kinetic isotope effects on the flavin reduction
step under reversible and irreversible catalytic regimes (46), and in active site mutant enzymes
with replacements of residues that are not directly involved in catalysis, namely Glu312 to
82
aspartate or Val464 to either alanine or threonine, by substrate or solvent kinetic isotope effects
(11, 15).
H2O2
FAD
N
OH
O2
H2O2
H2O
FADHH
N
O
H
N
OH
FAD
O2
FADH-
OH
O
N
O
Scheme 3.1 The Two-step Oxidation of Choline Catalyzed by Choline Oxidase.
In the X-ray structure of choline oxidase resolved to 1.86 Å, the Oγ atom of Ser101 is less
than 4 Å from the N(5) atom of FAD and within hydrogen bonding distance (i.e., <3 Å) of the
oxygen atom of DMSO (Figure 3.1), an additive that was used in the crystallization of the
enzyme (11). This suggests that Ser101 may interact with the substrate and be actively involved
in the oxidation of choline catalyzed by the enzyme. Moreover, the recent determination of the
X-ray structure of the Ser101Ala enzyme to a resolution of 2.5 Å showed lack of structural
changes in the active site and the overall structure of the enzyme upon replacing the side chain
on residue101 (Figure 3.1) (52). This prompted us to investigate the role of Ser101 in the
reductive half-reaction catalyzed by choline oxidase by using site-directed mutagenesis and rapid
kinetics.
In the present study, the effects of replacing Ser101 with alanine, threonine, cysteine or
valine on the reductive half-reaction catalyzed by choline oxidase have been investigated using
both solvent and substrate kinetic isotope effects in a stopped-flow spectrophotometer. The
mutated enzymes were slower than the wild-type enzyme in their ability to activate choline to the
alkoxide species, resulting in biphasic reductions of the flavin in anaerobic stopped-flow
reactions with choline. Kinetic isotope effects allowed us to assign the two kinetic phases of
flavin reduction to the stepwise cleavages of the substrate OH and CH bonds. Finally, it was
83
shown th
hat the rate of hydridee ion transfeer from the alkoxide too the flavinn decreased with
decreasin
ng hydrophillicity of the side
s chain on
n residue 1001.
Figure 3.1
3 Compariison of the active sites of wild-typpe and Ser1101Ala choline oxidase. The
residues with carbon
ns in green represent
r
thee mutant struucture, wherreas the carbbon atoms foor the
wild-typee enzyme arre in gray. The
T C(4a)-ox
xygen adducct observed w
with the wildd-type enzym
me is
shown in
n red. The structures
s
off wild-type and
a the S1001A mutant enzymes arre from pdbb files
2JBV (11
1) and 3NNE
E (52), respeectively.
3.4
Experimenttal Procedu
ures
Mateerials. Escheerichia coli strain Rosetta(DE3)pLyysS was froom Novagenn (Madison, WI).
QIAprep Spin Minip
prep kit wass from Qiag
gen (Valenciia, CA). The QuickChaange site-dirrected
mutageneesis kit waas from Stratagene (L
La Jolla, C
CA). Oligonnucleotides for site-dirrected
mutageneesis and seq
quencing of the mutant genes were from Sigma Genosys ((The Woodllands,
TX). 1,2-[2H4]-Choliine and sodiium deuteriu
um oxide (999%) were frrom Isotec Innc. (Miamissburg,
OH). Pollyethylene glycol 6000 was
w from Flu
uka (St. Louuis, MO). Chholine chloride was from
m ICN
84
Pharmaceutical Inc. (Irvine, CA). All other reagents used were of the highest purity
commercially available.
Instruments. UV-visible absorbance spectra were recorded using an Agilent Technologies
model HP 8453 diode array spectrophotometer equipped with a thermostated water bath. The
enzymatic activity of choline oxidase was measured polarographically using a computerinterfaced Oxy-32 oxygen monitoring system (Hansatech Instrument Ltd.). Stopped-flow
experiments were carried out using a Hi-Tech SF-61 Double Mixing Stopped Flow
spectrophotometer.
Site-directed Mutagenesis and Protein Purification. The mutated genes for the choline
oxidase variants containing alanine, threonine, cysteine or valine instead of Ser101 were
prepared as previously described (39). The resulting mutant genes were sequenced at the DNA
Core Facility of Georgia State University using an Applied Biosystems Big Dye Kit on an
Applied Biosystems model ABI 377 DNA Sequencer to confirm the presence of the mutation.
The mutant enzymes were expressed and purified to homogeneity as previously described for
wild-type choline oxidase (17, 39, 42). The enzymes were stored at -20 oC in 20 mM Tris-Cl, pH
8.0, and were found to be stable for at least six months.
Kinetic Assays. The apparent steady state kinetic parameters at atmospheric oxygen of the
mutant enzymes were measured with the method of the initial rates by monitoring oxygen
consumption as described for wild-type choline oxidase (43).
Observed rate constants for flavin reduction of the enzymes were determined anaerobically at
varying concentrations of substrate (choline or 1,2-[2H4]-choline) in 50 mM sodium
pyrophosphate, pH 10.0, using a stopped-flow spectrophotometer thermostated at 25 oC, as
previously described (11). Equal volumes of the enzyme and substrate were mixed anaerobically
85
in the stopped-flow spectrophotometer in the presence of a mixture of glucose/glucose oxidase (5
mM/0.5 μM) yielding a final enzyme concentration of ~10 µM. Substrate concentrations were
≥50 μM, ensuring pseudo-first order conditions. Data were collected in the dual beam mode
using photomultiplier detection set at 452 nm.
For the determination of solvent kinetic isotope effects on the observed first-order rate
constants for flavin reduction, all of the reagents were prepared using 99.9% deuterium oxide by
adjusting the pD value with NaOD. The pD values were determined by adding 0.4 to the pH
electrode readings (53). Solvent viscosity effects were measured in the presence of 0.0211g/mL
PEG-6000 as viscosigen (which is equivalent to relative viscosity of 1.26) (54), in both the
tonometer containing the enzyme and the syringes containing the organic substrates.
Data Analysis. Kinetic data were fit with KaleidaGraph software (Synergy Software,
Reading, PA) and the Kinetic Studio Software Suite (Hi-TgK Scientific, Bradford on Avon, U.
K.). The apparent steady state kinetic parameters at atmospheric oxygen were determined by
fitting the data to the Michaelis-Menten equation for one substrate.
Stopped-flow traces were fit to eq. 1, which describes a double-exponential process; kobs1 and
kobs2 represent the observed first-order rate constants associated with the absorbance changes of
the fast and slow phases, t is time, A is the absorbance at 452 nm at any given time, B and C are
the amplitudes of the absorbance changes for the fast and slow phases, and D is the absorbance at
infinite time. The kinetic parameters associated with the fast and slow phases of flavin reduction
seen in the reductive half-reaction where determined by using eqs 2 and 3, respectively (see
Appendix for derivation). Here, kobs1 and kobs2 represent the observed first-order rate constants
associated with the absorbance changes of the fast and slow phases at any given concentration of
substrate (S); klim1 and klim2 are the limiting rate constants at saturated substrate concentration of
86
the fast and slow phases, respectively; appKfast and appKslow are the apparent dissociation constants
defining equilibria between free enzyme and free substrate and enzyme-substrate complexes; and
c is a constant.
A = B exp(−k obs1t) + C exp(−k obs2 t) + D
k obs1 =
kobs2 =
3.5
k
(1)
S+c
K +S
(2)
klim 2S
Kslow + S
(3)
lim 1
app
fast
app
Results
Expression and Purification of Ser101 Mutant Enzymes. The choline oxidase variants in
which Ser101 is replaced with alanine, threonine, cysteine or valine were expressed and purified
by using the same protocol previously used for the wild-type enzyme (39). A mixture of airstable anionic semiquinone and oxidized flavin species was observed throughout the purification
procedure of all the mutant enzymes, as previously reported for the wild-type enzyme (9, 39).
The fully oxidized and active mutant enzymes were obtained by dialysis at pH 6.0, following the
procedure described for the wild-type enzyme (39). In all cases, the flavin was covalently linked
to the protein as established by heat denaturation of the mutant enzymes followed by
spectroscopic analysis of the samples after removal of the denatured proteins by centrifugation.
The spectral properties of the oxidized Ser101 mutant enzymes are summarized in Table 3.1,
along with the apparent steady state kinetic parameters determined with choline at atmospheric
oxygen concentration and pH 7.0. Both the
app
(kcat/Km) and
app
kcat values of the mutant enzymes
were between 1.5- and 1,000-fold lower than in the wild-type enzyme, indicating that Ser101 is
an important residue for catalysis in choline oxidase.
87
Table 3.1 Comparison of Spectral and Catalytic Properties of Ser101 Mutant Enzymes with
Wild-type Choline Oxidase.
app
app
app
ε
λmax
(kcat/Km)
kcat
Km
(mM-1cm-1)
(s-1)
(nm)b
(M-1s-1)
(mM)
a
Wild-typea
367, 455
10, 11.4
25000
15
0.6
Ser101Thr
371, 450
8.7, 9.3
18000 ± 2000
3.0 ± 0.1
0.17 ± 0.02
Ser101Ala
373, 452
8.9, 9.5
4600 ± 400
3.3 ± 0.1
0.73 ± 0.07
Ser101Cys
374, 450
9.2, 9.9
1100 ± 70
0.076 ± 0.001
0.07 ± 0.01
Ser101Val
371, 452
9.2, 10.9
21 ± 1
0.017 ± 0.001
0.82 ± 0.03
b
o
From ref. (11). Absorbance spectra were recorded at pH 8.0, 25 C. Enzymatic activities were
measured in 50 mM potassium phosphate, pH 7.0 and 25 oC, using fully oxidized enzymes.
Reductive Half-Reaction. The reductive half-reactions of the Ser101Ala, Ser101Thr,
Ser101Cys and Ser101Val enzymes were studied in a stopped-flow spectrophotometer by mixing
anaerobic solutions of the enzyme with different concentrations of anaerobic choline at pH 10.0
and 25 oC under pseudo-first-order conditions (i.e., 0.01 mM enzyme and 0.05 mM to 5 mM
choline). Alkaline pH was chosen because previous results established that catalysis is pH
independent in the wild-type and a number of active site mutant enzymes of choline oxidase at
pH 10.0 (7, 9, 11, 13, 16, 43, 44). The reduction of enzyme-bound oxidized FAD was monitored
by the changes in absorbance at 452 nm. With all enzymes two kinetic phases were observed
(Figure 3.2A). Full reduction of the enzyme-bound flavin to the anionic hydroquinone species
was observed at the end of the slow kinetic phase with all the mutant enzymes and at all substrate
concentrations tested, as illustrated in the example of Figure 3.2B for the Ser101Ala enzyme. In
agreement with a biphasic process for the decrease in absorbance at 452 nm the stopped-flow
traces with all of the mutant enzymes were fit best to eq. 1, which describes a double exponential
process. When the observed first-order rate constants for the fast (kobs1) and slow (kobs2) phases
were plotted as a function of choline concentrations they both showed saturation kinetics (Figure
88
3.2C shows the example of the Ser101Ala enzyme). Accordingly, the associated kinetic
parameters for the reductive half-reactions of the mutant enzymes were determined by using eqs
2 and 3 (see Appendix for derivations). With all of the mutant enzymes the fits of the rapid
kinetic data yielded
app
Kfast and
app
Kslow values which were below the lowest concentration of
choline that could be used in order to maintain pseudo first-order conditions (data not shown),
thereby preventing accurate determinations of these kinetic parameters. In contrast, accurate
determinations of the limiting rate constants at saturated substrate concentration for the two
phases seen in the stopped-flow traces (klim1 and klim2) could be attained. As summarized in Table
3.2, the klim1 values for all of the Ser101 mutant enzymes were between 40 s-1 and 125 s-1. In
contrast, the klim2 values spanned over two orders of magnitude between 0.016 s-1 and 6.6 s-1,
with Thr101≈Ala101 > Cys101 > Val101 (Table 3.2). For comparison, the reductive halfreaction with the wild-type enzyme was previously shown to be monophasic with a limiting rate
constant for flavin reduction of 93 s-1 (43).
Substrate Deuterium KIE. Substrate KIEs were employed to probe whether the cleavage of
the CH bond of choline is associated with either the fast or the slow kinetic phase observed in the
stopped-flow spectrophotometer with the Ser101 mutant enzymes. The Ser101Val enzyme was
too slow for an accurate determination of the KIE and was not investigated further. As
exemplified in Figure 3.3 for the Ser101Ala enzyme, substitution of choline with 1,2-[2H4]choline further slowed down the slow kinetic phase observed in the stopped-flow
spectrophotometer with all of the Ser101 mutant enzymes. In contrast, there were no significant
changes in the fast phases for all the mutant enzymes (Figure 3.3A shows the Ser101Ala enzyme
as example). Consequently, significant decreases in the klim2 values were observed, whereas the
klim1 values were minimally affected by the substitution of choline with 1,2-[2H4]-choline (Table
89
3.2). Acccordingly, sm
mall substraate KIEs ≤1.1 were assoociated withh the klim1 values determ
mined
with the Ser101Thr/A
Ala/Cys enzzymes, wherreas large suubstrate KIE
Es between 5.0 and 6.5 were
T
resultts establish thhe slow kineetic phase off flavin reduuction
associateed with the klim2 values. These
as being primarily du
ue to the cleaavage of the substrate CH
H bond.
3 Anaerob
bic reduction
n of the Ser101 variantss with choliine as a subbstrate in 500 mM
Figure 3.2
sodium pyrophospha
p
ate at pH 10..0 and 25 °C
C. Panel A shhows the redduction tracees with saturrating
choline for
f the Ser10
01Ala (5 mM
M choline; black
b
curve),, Ser101Thr (5 mM choline; blue cuurve),
Ser101Cy
ys (10 mM choline; red
d curve) and
d Ser101Vaal (10 mM ccholine; greeen) enzymess. All
traces weere fit with eq. 1. Timee indicated is
i after the eend of the fflow, i.e. 2.22 ms. In ordder to
improve clarity, one experimenttal point eveery 5 is show
wn (verticall lines). Pannel B shows UVvisible absorbance
a
spectra of the oxidized enzyme (blue) and reduced (bblack) speciees of
Ser101A
Ala enzyme variant
v
obtaiined by mix
xing anaerobbically the ooxidized enzzyme with bbuffer
and 5 mM
M choline in
n 50 mM sod
dium pyroph
hosphate bufffer at pH 100.0 and 25 °C
C. Panel C sshows
the obseerved rate constants
c
forr the fast and
a
slow kiinetic phasees as a funnction of chholine
concentraation for Seer101Ala enzzyme varian
nt. Data werre fit to eq. 2 and eq. 3 for the fasst and
slow phaases, respectiively.
90
Table 3.2 Limiting Rate Constants for the Fast and Slow Kinetic Phases of Flavin Reduction
with Choline and 1,2-[2H4]-Choline as Substrate for Ser101 Mutant Enzymes a.
parameters
klim1, s-1
klim2, s-1
definition
k3
k5
choline
Ser101Thr
41 ± 1
3.07 ± 0.05
Ser101Ala
125 ± 2
6.6 ± 0.1
Ser101Cys
47 ± 1
0.10 ± 0.01
Ser101Val
49 ± 1
0.016 ± 0.001
Wild-type
no b
93 ± 1
1,2-[2H4]-choline
a
Ser101Thr
36 ± 1
0.48 ± 0.02
Ser101Ala
125 ± 6
1.02 ± 0.02
Ser101Cys
44 ± 1
0.020 ± 0.001
Wild-type
no b
10.3 ± 0.3
Determined in 50 mM sodium pyrophosphate, pH 10.0 and 25 oC.
b
Not observed
Solvent KIE. Solvent KIE were used with the Ser101Ala enzyme to probe whether the
cleavage of the OH bond of choline is associated with the fast phase observed in the stoppedflow spectrophotometer. When water was substituted with D2O there was a significant decrease
in the observed rate constants for the fast kinetic phase at 452 nm, with minimal effects on the
slow phase (Figure 3.4). The klim1 and klim2 values determined in D2O for the Ser101Ala enzyme
were 32.8 ± 0.5 s-1 and 4.0 ± 0.1 s-1, respectively, yielding solvent KIEs of 3.8 ± 0.1 and 1.65 ±
0.05 for the fast and slow phases observed in the stopped-flow spectrophotometer. The effect of
increased solvent viscosity on the reductive half-reaction was investigated as a control for the
solvent KIE, since D2O has a relative viscosity of 1.25 at 25 oC. PEG-6000, at a concentration of
0.0211 g/mL that is equivalent to a 100% solution of D2O, was used in the stopped-flow
91
spectroph
hotometer. The
T limiting
g rate constaants at saturaating concenntrations off choline forr both
the fast and
a slow phaases were sim
milar to thosse determineed in aqueouus solution, yielding ratiios of
(klim1)H2OO/(klim1)PEG an
nd (klim2)H2O/(klim2)PEG vaalues of 1.055 ± 0.01 andd 0.98 ± 0.022. Lack of soolvent
viscosity
y effects estab
blishes the solvent
s
KIEss on the klim1 and klim2 vaalues as beinng associatedd with
transition
n states with
h exchangeab
ble protons being
b
in flighht rather thaan solvent-seensitive equiilibria
of enzym
me-substrate complexes. These data,, in turn, aree consistent with the fasst phase obseerved
in the sto
opped-flow spectrophoto
s
ometer as beiing due to thhe cleavage oof the substrrate OH bondd.
Figure 3.3
3 Anaerobiic substrate reduction
r
off the Ser101A
Ala enzyme with cholinee (black) andd 1,2[2H4]-cho
oline (red) as substrattes. The ex
xperiments were carrieed out in 50 mM soodium
pyrophossphate, pH 10.0, at 25 °C. Panel A shows stoopped-flow ttraces at satturating substrate
concentraations. All trraces were fit
f with eq. 1.
1 Time indiccated is afteer the end off the flow, i.ee. 2.2
ms. In orrder to impro
ove clarity, one
o experim
mental point eevery 5 is shhown (verticcal lines). Paanel B
shows th
he observed rate
r constan
nts for the faast and slow kinetic phasses as a funcction of substrate
concentraation. Data were
w fit to eq
q. 2 and eq. 3 for the fastt phase and slow phasess, respectivelly.
92
Figure 3.4
3 Anaerobiic substrate reduction
r
of Ser101Ala w
with cholinee as substratees in H2O (bblack)
and D2O
O (red). The experiments
e
were carried out in 50 m
mM sodium
m pyrophosphhate, pH 10, at 25
°C. Panel A shows sttopped-flow traces at satturating subsstrate concenntrations. Thhe curves weere fit
to eq. 1. Time indicaated is after the end of th
he flow, i.e. 2.2 ms. In order to impprove clarityy, one
experimeental point every
e
5 is sh
hown (vertical lines). Paanel B showss the observved rate consstants
for the faast and slow kinetic phasses as a funcction of substtrate concenntration. Dataa were fit to eq. 2
and eq. 3 for the fast and slow ph
hases, respecctively.
Kinettic Data Sim
mulation. Th
he two kinettic phases obbserved in thhe reductive half-reactioons of
the Ser101 mutant enzymes
e
aree consistent with the m
minimal kineetic model oof eq 4, witth the
oxidized enzyme (a)) giving rise to a transieent species ((b) that evenntually decayys to the redduced
enzyme (c).
( In eq 4, a and b aree reversibly connected too account foor the saturaation kineticss as a
function of substrate concentratio
on that is observed for thhe slow kineetic phase in the stopped-flow
hotometer (5
55). Figure 3.5A showss the progresss curves foor the relativve amounts oof the
spectroph
three speecies a, b an
nd c that haave been geenerated for the Ser101Ala enzymee upon usinng the
experimeental data with
w
cholinee of Table 3.2. Similaar time-courrses that show the trannsient
accumulaation of thee intermediaate species b could be generated ffor the otheer Ser101 m
mutant
enzymes (data not shown).
s
Sim
mulation of the kinetic data acquireed with 1,22-[2H4]-choline in
umulation off the transiennt species b with respecct to the casee with
water cleearly shows a larger accu
when
choline, as expected due to the similar
s
ratess of formatioon and a sloower rate of decay of b w
choline is substituted
d with 1,2-[2H4]-choline (cfr. black aand blue currves in Figure 3.5B). Siimilar
93
accumulaation of the transient speecies b, but at a delayedd time, is eviident in the simulation oof the
kinetic data
d
with 1,,2-[2H4]-choline in D2O with respect to thosee with cholline in wateer, as
expected
d due to the slowing dow
wn of the clleavage of bboth OH andd CH bondss of the substrate
(cfr. greeen and black curves in Figure 3.5
5B). Finally, a lower acccumulationn of the trannsient
species b is seen in the
t simulatio
on of the kin
netic data wiith choline iin D2O with respect to w
water,
as expectted due to a slower rate of formatio
on and similaar rates of ddecay of b (ccfr. red and bblack
curves in
n Figure 3.5B
B).
a ⇔b⇒c
(4)
3 Simulateed reaction traces with a two-stepp model of a→b→c wiith the equaations
Figure 3.5
A0 k lim1
B=
((exp(−k lim1t ) − exp(−k limm 2 t ))
A = A0 exp(
e −k lim1t )
,
and
k −k
lim 2
lim1
k lim 2 exp((−k lim1t ) − k liim1 exp(−k lim 2 t )
) . Kineetic parameeters are ttaken from
m the
k lim1 − k liim 2
experimeentally deterrmined valuees of Table 3.2.
3 Panel A shows the progress currves for the three
enzyme-ssubstrate species a (oxidized enzym
me-substratee complex),, b (transiennt species) aand c
(reduced enzyme-prroduct comp
plex) of th
he Ser101Alla enzyme during anaaerobic substrate
reduction
n with 10 mM
M choline att pH 10.0 and 25 oC. Pannel B shows the time couurses of speccies b
during reeduction in 50
5 mM sodiium pyropho
osphate, pH 10.0, with ccholine in H2O (black), D2O
2
(red), 1,2
2-[ H4]-choliine in H2O (b
blue) and D2O (green) uunder anaerobic conditions at 25 oC.
C = A0 (1 +
94
3.6
Discussion
The reductive half-reactions of the active site variants of choline oxidase containing alanine,
threonine, cysteine or valine instead of serine at position 101 were investigated to establish the
role of Ser101 in the reaction of choline oxidation catalyzed by the enzyme. In the case of the
Ser101Ala enzyme, the three dimensional structure of the mutant enzyme was recently shown to
be practically identical to that of the wild-type enzyme, with an average rmsd value of 0.41 Å for
527 equivalent Cα atoms when the structure of the mutant enzyme is overlaid to that of the wildtype enzyme (Figure 3.1) (11, 52). In the present study we have shown that all of the mutant
enzymes shared a number of kinetic, mechanistic and biochemical properties with the wild-type
enzyme: 1) the UV-visible absorbance spectrum had maxima at 367-374 nm and 450-455 nm in
the oxidized state and at 350-360 nm in the hydroquinone state; 2) the flavin was covalently
linked to the protein; 3) an anionic flavosemiquinone was aerobically stabilized at pH 8.0 during
purification; and 4) the substrate KIEs on the limiting rate constant for anaerobic flavin reduction
were ≥5.0. This allowed us to compare and contrast the mechanistic properties of the enzymes
and to draw conclusions on the role that is played by Ser101 in the activation of the alcohol
substrate and the subsequent transfer of a hydride ion from the activated alkoxide to the flavin.
Timing of OH and CH Bond Cleavages. The abstraction of the hydroxyl proton of the
substrate precedes and is mechanistically independent of the transfer of the hydride ion from the
activated substrate to the flavin in the Ser101 mutant enzymes (Scheme 3.2). Evidence for this
conclusion comes from the anaerobic reductive half-reactions with choline of the four variants of
choline oxidase substituted on residue 101 and the associated solvent and substrate KIEs. Two
kinetic phases, which are well separated from one another, were observed in the stopped-flow
traces (Figure 3.2). A large solvent KIE with a value of ~4 was determined on the klim1 value for
95
the Ser101Ala enzyme, consistent with the fast kinetic phase reflecting the cleavage of the
substrate OH bond. Negligible substrate KIEs of ≤1.1 on the klim1 values and large substrate KIEs
of ≥5.0 on the klim2 values were seen with the Ser101Ala, Ser101Thr, and Ser101Cys enzymes,
consistent with the slow kinetic phase reflecting the cleavage of the substrate CH bond.
The analytical equations that define the limiting rate constants at saturating concentration of
choline for the two kinetic phases observed in the stopped-flow traces are given by eqs 5 and 6,
with numbering based on Scheme 3.2 (see Appendix for derivation). These eqs simplify further
to eqs 7 and 8 due to the cleavage of the substrate OH bond being significantly faster than both
its reverse reaction (i.e., k3>k4) and the cleavage of the CH bond (i.e., k3>k5). The first condition
is stipulated by the extrapolation to the origin of the hyperbola fitting the observed rate constants
for the fast kinetic phase in the stopped-flow traces as a function of the concentration of choline
(Figure 3.2C). The second condition is indirectly validated by the results of the KIEs since
otherwise both klim1 and klim2 would incorporate both kinetic steps and be sensitive to both
substrate and solvent KIEs, as illustrated by eqs 5 and 6. Thus, the timing for the cleavages of the
OH and CH bonds in the Ser101 mutant enzymes is similar to that of the wild-type choline
oxidase (43). What is significantly different between the two enzymes is how transition states
and reaction intermediates are stabilized in the active site when the serine residue at position 101
is absent or present in the active site (vide infra).
k lim1 = k 3 + k 4 + k 5
(5)
k3 ⋅ k5
k3 + k 4 + k5
(6)
k lim 2 =
klim1 = k3
(7)
klim2 = k5
(8)
96
Hydroxyl Proton Abstraction. The hydroxyl group of Ser101 stabilizes the transition state
for the abstraction of the hydroxyl proton of choline. Evidence for this conclusion comes from
the observation that the substitution of Ser101 with other amino acid residues results in at least a
15-fold decrease in the rate constant for the cleavage of the substrate OH bond. The fast kinetic
phases seen in the stopped-flow traces of the Ser101 mutant enzymes, which report on the
cleavage of the substrate OH bond (vide ante), have limiting rate constants at saturating choline
(klim1) comprised between 40 s-1 and 125-1. Instead, cleavage of the substrate OH bond in the
wild-type enzyme has been shown to occur within the dead time of the stopped-flow
spectrophotometer (43), which corresponds to 2.2 ms. If one considers that 6 half-lives are
required to complete 98.5% of a pseudo-first order process of the type considered here2, a lower
limiting value of 1900 s-1 can be estimated for the cleavage of the substrate OH bond in the wildtype choline oxidase. In agreement with substrate OH bond cleavage occurring in the dead time
of the stopped-flow spectrophotometer, previous mechanistic investigations showed that the
solvent KIE associated with the anaerobic reduction of the wild-type choline oxidase has a value
of 0.99 (43). It is likely that in the wild-type enzyme the transition state for the OH bond
cleavage is stabilized by a hydrogen bond involving the side chain of Ser101 and the oxygen
atom of the alkoxide species, which is necessarily more electronegative than in the enzymesubstrate complex due to the development of the charged alkoxide intermediate (Scheme 3.3). In
the mutant enzymes containing alanine, valine or cysteine in place of Ser101, such a transition
state stabilization cannot be achieved, thereby making the abstraction of the substrate hydroxyl
proton slow. In the case of the Ser101Thr enzyme, a sub-optimal geometry of the hydrogen bond
2
In the four enzymes mutated on Ser101 cleavage of the substrate OH bond is a first-order process (Figure 3.2A). It
is therefore reasonable to assume that cleavage of the substrate OH bond by the wild-type enzyme is also a firstorder process.
97
due to the presence of the extra methyl may be responsible for the slow rate of proton
abstraction.
k1
CH + E-FADox
k2
k3
E-FADox-CH
substrate binding
k5
E-FADox-CHintermediate
k4
hydroxyl proton abstraction
E-FADred-BA
k6
hydride transfer
Scheme 3.2 Proposed Kinetic Mechanism for Reductive Half-Reaction of Ser101 Variants of
Choline Oxidase. E, enzyme; FADox, oxidized flavin; CH, choline; FADred, reduced Flavin; BA,
betaine aldehyde.
a) OH bond cleavage:
H99
FAD
E312−
O CH2
S101
H
H
N
F357
H99
H99
O
H
B
H
W331
H351
V464
E312−
H
N
F357
O CH2
S 101
H
δ
O
δ
H
−
O
BH+
H
H351
V464
H466
O CH2
S 101
H
H
N
W331
H351
V464
E312
F357
B
H
W331
H466
FAD
FAD
H466
b) CH bond cleavage:
H99
H99
FAD
E312−
F357
O CH2
S 101
H
H
N
O
H
W331
V464
H351
BH+
H466
H99
δ
FAD
E312−
F357
H
N
O CH2
S 101
H
Oδ
H
W331
V464
BH+
FADH-
E312−
F357
N
H466
O
H
W331
H351
O CH2
S 101
H
V464
H351
BH+
H466
Scheme 3.3 Roles of Ser101 in the Cleavages of the OH (a) and CH (b) Bonds of Choline. Other
residues roles are: E312, substrate binding (15); H351, hydrogen bonding O atom of the substrate
(29); and H466, electrostatic stabilization of alkoxide species (34).
Hydride Ion Transfer. The hydroxyl group of Ser101 provides a hydrophilic
microenvironment with hydrogen bonding capability that facilitates the transfer of the hydride
98
ion from the activated alkoxide to the flavin (Scheme 3.3). Evidence for this conclusion comes
from a plot of the log (k5) as a function of the hydrophobicity of residue 101, showing a linear
decrease over 4 orders of magnitude of the rate constant for hydride transfer with decreasing
hydrophilicity of residue 101 (Figure 3.6). Possible steric effects can be immediately ruled out
as being responsible for the decrease in the k5 values because no correlation was found when the
log (k5) or k5 itself was plotted as a function of the volume of the side chain on residue 101 (data
not shown). Among the various hydrophobicity scales that are available the best correlation was
found by using the GES scale (R2 = 0.91), where the hydrophobicity of the amino acids side
chains is defined by the free energy for transfer of the amino acid from water to an organic
solvent while taking into account the hydrogen bonding ability of the amino acid side chains
(56). Interestingly, a significantly improved fit of the data was obtained when the Ser101Ala
enzyme was excluded (R2 = 0.99), with the latter enzyme displaying a k5 value 8-times larger
than the value expected from the fit of the data obtained with the other four enzymes (Figure
S3.1 in SI). For comparison, R2 values of 0.75 (with data for Ala101) and 0.87 (without data for
Ala101) were obtained upon using the more popular Kyte hydrophobicity scale (Figure S3.2
and S3.3 in SI), where hydrophobicity is defined by the free energy for the transfer from water
to vapor phase (57). These observations collectively suggest that besides its hydrophilic
character, and irrespective of whether the redox potential of the flavin is affected by the
substitution or not, the hydrogen bonding capability of the side chain on residue 101 is also
important for the hydride transfer reaction catalyzed by choline oxidase. The presence of a
somewhat mobile water molecule in place of the less mobile side chain on the Ser101Ala
enzyme would readily explain why the hydride transfer reaction is significantly faster in the
Ser101Ala.
99
Alkoxide Intermediate vs. Transition State. In the Ser101 mutant enzymes the alkoxide
intermediate promotes detectable perturbations in the absorbance spectrum of the enzyme-bound
flavin, as indicated by the decreased intensity at 450 nm ensuing in the fast kinetic phase of the
stopped-flow traces3. A previous study showed that another variant of choline oxidase with the
active site Asn510 substituted with histidine also catalyzes a stepwise oxidation of choline with
cleavage of the OH and CH bonds occurring in the time frame of the stopped-flow
spectrophotometer (51). However, in that case anaerobic reduction of the flavin was monophasic
and the stepwise timing for the cleavages of the OH and CH bonds of choline could be
established only by using substrate and solvent KIEs (51). A readily explanation for this apparent
incongruence is that in the Ser101 mutant enzymes the alkoxide formed in catalysis is a stable
reaction intermediate. In contrast, in the reaction of choline oxidation catalyzed by the
Asn510His enzyme the alkoxide is a transition state. The former, being long-lived and stable,
would be detected spectrophotometrically within the time frame of the stopped-flow
spectrophotometer, whereas the latter would be undetectable with probes other than KIEs due to
its lifetime in the 10-12 s.
3
The decrease in absorbance at 450 nm that is associated with the cleavage of the substrate OH bond (Figure 3.2A)
is too large to be ascribed to the mere presence of the negatively charged alkoxide intermediate close to the flavin in
the active site of the enzymes mutated on Ser101. Indeed, cleavage of the OH bond of choline did not result in
absorbance changes of the type reported here in both the wild-type and a mutant form of choline oxidase where
Asn510 is replaced with His (20, 30). A spectroscopic characterization of the enzyme-alkoxide intermediate in the
reaction catalyzed by the Ser101 mutant enzymes is currently underway to elucidate the biophysical rationale for
such a significant change in the absorbance of the oxidized enzyme-alkoxide intermediate complex. This will be
presented in a separate study.
100
Figure 3.6
3 Dependen
nce of the raate constant for hydride ion transfer on the hydrrophobicity oof the
amino accid residue at
a position 10
01 in cholinee oxidase. T
The GES scaale for hydroophobicity iss used
(56).
Impliications forr Hydride Transfer
T
Reactions Cattalyzed by F
Flavin-Depeendent Enzyymes.
Several flavin-depen
f
ndent enzymees have in th
heir active ssites either a serine, threeonine or tyrrosine
in a posiition that is similar, if not
n equivaleent, to that oof Ser101 of choline oxxidase. Exam
mples
include, but are not limited to, class 2 dihy
ydroorotate ddehydrogenaase (58), pyyranose 2-oxxidase
ucose oxidasse (60), nitriic oxide syn
nthase (61), the old yelllow enzymee (62), NAD
DPH(59), glu
cytochrom
me P450 ox
xidoreductasse (63), aldiitol oxidase (64), UDP--galactose 44-epimerase (65),
cholesterrol oxidase (66),
(
nitronaate monooxy
ygenase (67) , heterotetraameric sacossine oxidase (68).
In many instances crrystallograph
hic or site-d
directed mutaagenesis stuudies have suuggested thaat the
t
orr tyrosine residues
r
aree important for catalyssis because they
active siite serine, threonine
hydrogen
n bond to eitther the O(4)) or N(5) ato
oms of the fl
flavin (60-622, 64, 66, 68)), the substraate in
the enzym
me-substratee complex (5
59, 65) or th
he transition state (63) foor the reactioon catalyzedd. The
only case in the literature for which multtiple substituutions of ann active sitee threonine (i.e.,
were engineeered and thhe effects of the
Thr178) equivalent to Ser101 of choline oxidase w
r constantt for the hyd
dride transferr reaction weere investigaated in a stopppedsubstitutiions on the rate
flow specctrophotomeeter is class 2 dihydrooro
otate dehydrrogenase (588). Interestinngly, a plot oof the
101
log (kred) as a function of the hydrophobicity of residue 178 shows that the rate constant for the
hydride transfer reaction catalyzed by class 2 dihydroorotate dehydrogenase increases with
increasing hydrophilicity of the residue 178 (Figure S3.4 in SI), as for the case presented here
for choline oxidase. With dihydroorotate dehydrogenase, however, the slope of the line that fits
the data to a GES hydrophobicity scale is shallower than that with choline oxidase, e.g., -0.36 ±
0.05 as compared to -1.9 ± 0.3. This suggests that the conclusions that are drawn in our study of
choline oxidase, most likely in combination with other features such as the ability of the
hydroxyl side chain to hydrogen bond to the 7,8,-dimethyl-isoalloxazine of the flavin, may have
general relevance for those flavin-dependent enzymes that utilize a hydride transfer mechanism
for the oxidation of their organic substrate.
Conclusions. In summary, the results presented in this study on the choline oxidase variants
in which serine at position 101 was replaced by alanine, cysteine, threonine or valine allow to
support the conclusions that Ser101 is important for both the activation of the alcohol substrate
and the subsequent hydride transfer reaction catalyzed by the enzyme. This is likely exerted by
the side chain of residue 101 acting as a hydrogen bond donor to the oxygen atom of the alcohol
substrate and of the alkoxide intermediate, respectively. This study represents the first instance in
which the hydrophilic character of the serine residue proximal to the C(4a)-N(5) flavin atoms in
a flavin-dependent enzyme, besides its hydrogen bonding capability, has been shown to be
important for efficient hydride transfer from the substrate to the enzyme-bound flavin. Analysis
of published data on class 2 dihydroorotate dehydrogenase indicates that the hydrophilic
character of Thr178 is also important for the hydride transfer reaction catalyzed by that enzyme.
In the future it would be interesting to complement these experimental findings with
computational approaches to provide a physical rationale for these conclusions.
102
3.7
Appendix
The proposed mechanism for the reaction catalyzed by the choline oxidase Ser101 variant is
described. The derivation of the equation that describes the rate of the OH bond and CH bond
cleavage of the substrate follows the logic described by Bernasconi (69) and the method by
Chaiyen (70).
Scheme A3. 1
CH + E-FADox
k1
E-FADox-CH
k2
substrate binding
k3
k4
E-FADox-CHintermediate
hydroxyl proton abstraction
k5
k6
E-FADred-BA
hydride transfer
In Scheme A3.1, E-FADOX represents oxidized choline oxidase. CH is the substrate choline.
BA represents bataine aldehyde. E-FADred is reduced choline oxidase. k1, k2, k3, k4, k5 and k6 are
the rate constants. Binding step is in rapid equilibrium, therefore, kobs0 is very large and is in the
dead time of the system and we cannot measure.
dΔC int ermediate
= k 3 ⋅ ΔC E − FADox−CH − (k 4 + k 5 )ΔC int ermediate + k 6 ⋅ C E − FADred − BA ⋅ ΔC E − FADred − BA (1)
dt
Since ΔC E − FADox−CH =
k1 ⋅ CCH ⋅ ΔCCH
k2
ΔCCH + ΔC E − FADox−CH + ΔCint ermediate + ΔC E − FADred− BA = 0
Therefore ΔC E − FADox−CH =
(2)
(3)
k1 ⋅ CCH ⋅ (−ΔC E − FADox−CH − ΔCint ermediate − ΔC E − FADred− BA )
k2
ΔC E − FADox−CH =
k1 ⋅ CCH ⋅ (ΔCint ermediate + ΔC E − FADred− BA )
k 2 + k1 ⋅ CCH
Replace the ΔCE − FADox−CH into eq. 1
dΔCint ermediate
k ⋅k ⋅C
k ⋅k ⋅C
= −( 1 3 CH + k 4 + k 5 )ΔCint ermediate − ( 1 3 CH − k 6 ⋅ C E − FADred− BA ) ⋅ ΔC E − FADred− BA
dt
k 2 + k1CCH
k 2 + k1CCH
103
dΔC E − FAD− BA
= k 5 ⋅ ΔCint ermediate − k 6 ⋅ C E − FADred − BA ⋅ ΔC E − FADred − BA
dt
Based on the logic described by Bernasconi (69), let a11 =
a12 =
k1 ⋅ k 3 ⋅ CCH
− k 6 ⋅ C E − FADred− BA ,
k 2 + k1CCH
k1 ⋅ k 3 ⋅ CCH
+ k 4 + k5 ,
k 2 + k1CCH
a 21 = k 5 , a 22 = k 6 ⋅ C E − FADred− BA ,
x1 = ΔCint ermediate ,
x2 = ΔC E − FADred− BA
ฬ
ܽଵଵ െ ݇௢௕௦
ܽଶଵ
ܽଵଶ
ฬ
ܽଶଶ െ ݇௢௕௦
So,
k obs1 =
1
1
(a11 + a 22 ) + ( (a11 + a 22 )) 2 + a12 a 21 − a11 a 22
2
2
k obs 2 =
1
1
(a11 + a 22 ) − ( (a11 + a 22 )) 2 + a12 a 21 − a11 a 22
2
2
By taking the sum and the product
A = k obs1 + k obs 2 = a11 + a 22 =
k1 ⋅ k 3 ⋅ S
+ k 4 + k 5 + k 6 ⋅ C E − FADred − BA
k 2 + k1 ⋅ S
B = k obs1 ⋅ k obs 2 = a11 a 22 − a12 a 21 =
k1 ⋅ k 3 ⋅ S
(k 5 + k 6 ⋅ C E − FADred − BA ) + k 4 ⋅k 6 ⋅ C E − FADred − BA
k 2 + k1 ⋅ S
The extrapolation to the origin of the curve fitting the kinetic data on the dependence of the
kobs2 value as a function of the substrate concentration in Figures 3.2B, 3.3B and 3.4B
establishes that the kinetic step k6 is close to zero (71).
A = k obs1 + k obs 2 =
So
B = k obs1 ⋅ k obs 2 =
k1 ⋅ k 3 ⋅ S
+ k4 + k5
k 2 + k1 ⋅ S
k1 ⋅ k 3 ⋅ k 5 ⋅ S
k 2 + k1 ⋅ S
Substitute k obs1 = A − k obs 2 into B to obtain
104
2
− k obs 2 + k obs 2 ⋅ A − B = 0
k obs 2
4B
A ± A2 − 4B A A
=
= ±
1− 2 ;
2
2 2
A
While A 2 ≥ 4 B
1−
4B
2B
≈
−
1
A2
A2
Therefore, k obs2 = A − B
B
k = B or A − B
A or A and obs1
A
A
Since kobs1 > kobs2
k obs 2 = B
A
=
k1 ⋅ k 3 ⋅ k 5 ⋅ S
k1 ⋅ k 3 ⋅ S + k1 ⋅ k 4 ⋅ S + k1 ⋅ k 5 ⋅ S + k 2 (k 4 + k 5 )
k3 ⋅ k5
⋅S
k3 + k 4 + k5
=
k 2 (k 4 + k5 )
S+
k1 (k 3 + k 4 + k 5 )
So k lim 2 =
k obs1 =
k3 ⋅ k5
k 2 (k 4 + k 5 )
, K slow =
k3 + k 4 + k5
k1 (k 3 + k 4 + k 5 )
B
k obs 2
=
=
k1 ⋅ k 3 ⋅ k 5 ⋅ S k1 (k 3 + k 4 + k 5 ) ⋅ S + k 2 (k 4 + k 5 )
⋅
k 2 + k1 S
k1 ⋅ k 3 ⋅ k 5 ⋅ S
k1 (k 3 + k 4 + k 5 ) ⋅ S + k 2 (k 4 + k 5 )
k1 ⋅ S + k 2
(k3 + k 4 + k5 ) ⋅ S + k 2 (k 4 + k5 )
=
k1
S+
k2
So k lim1 = k 3 + k 4 + k 5 , K fast =
k1
k2
k1
105
3.8
Support Infformation
Figu
ure S3.1 Deppendence off the rate connstant
for hydride ion trannsfer on the
hydrrophobicity of the aminno acid residdue at
posiition 101 in choline oxiidase. Data were
fit w
without conssidering the Ala101 enzzyme.
The GES scalee for hydropphobicity is used
(72)).
Figu
ure S3.2 Deppendence off the rate connstant
for hydride ion trannsfer on the
hydrrophobicity of the aminno acid residdue at
posiition 101 inn choline oxxidase. The Kyte
scalee for hydropphobicity is uused (72).
Figu
ure S3.3 Deppendence off the rate connstant
for hydride ion trannsfer on the
hydrrophobicity of the aminno acid residdue at
posiition 101 in choline oxiidase. Data were
fit w
without conssidering the Ala101 enzzyme.
The Kyte scalee for hydropphobicity is used
(72)).
106
Figu
ure S3.4 Deppendence off the rate connstant
for hydride ion trannsfer on the
hydrrophobicity of the aminno acid residdue at
posiition 178 in class 2 dihydroorrotate
dehyydrogenase, where thhe wild-typpe is
Thr1178 (58). The GE
ES scale for
hydrrophobicity is used (72)..
107
3.9
Acknowledgment
The authors thank Prof. Dabney W. Dixon for insightful discussions on hydrophobicity scales
and Prof. Donald Hamelberg for insightful discussions.
3.10
1.
References
Gadda, G. (2008) Hydride transfer made easy in the reaction of alcohol oxidation
catalyzed by flavin-dependent oxidases, Biochemistry 47, 13745-13753.
2.
Vilhelmsson, O., and Miller, K. J. (2002) Humectant permeability influences growth and
compatible solute uptake by Staphylococcus aureus subjected to osmotic stress, J. Food
Prot. 65, 1008-1015.
3.
Peddie, B. A., Wong-She, J., Randall, K., Lever, M., and Chambers, S. T. (1998)
Osmoprotective properties and accumulation of betaine analogues by Staphylococcus
aureus, FEMS Microbiol. Lett. 160, 25-30.
4.
Peddie, B. A., Chambers, S. T., and Lever, M. (1996) Is the ability of urinary tract
pathogens to accumulate glycine betaine a factor in the virulence of pathogenic strains?,
J. Lab Clin. Med. 128, 417-422.
5.
Sakamoto, A., Alia, and Murata, N. (1998) Metabolic engineering of rice leading to
biosynthesis of glycinebetaine and tolerance to salt and cold, Plant Mol. Biol. 38, 10111019.
6.
Papageorgiou GC, M. N. (1995) The unusually strong stabilizing effects of glycine
betaine on the structure and function of the oxygen-evolving photosystem II complex.,
Photosynth Res. 44, 243-252.
108
7.
Alia, Hayashi, H., Sakamoto, A., and Murata, N. (1998) Enhancement of the tolerance of
Arabidopsis to high temperatures by genetic engineering of the synthesis of
glycinebetaine, Plant J. 16, 155-161.
8.
Alia, Kondo, Y., Sakamoto, A., Nonaka, H., Hayashi, H., Saradhi, P. P., Chen, T. H., and
Murata, N. (1999) Enhanced tolerance to light stress of transgenic Arabidopsis plants that
express the codA gene for a bacterial choline oxidase, Plant Mol. Biol. 40, 279-288.
9.
Sakamoto, A., Valverde, R., Alia, Chen, T. H., and Murata, N. (2000) Transformation of
Arabidopsis with the codA gene for choline oxidase enhances freezing tolerance of
plants, Plant J. 22, 449-453.
10.
Andresen, P. A., Kaasen, I., Styrvold, O. B., Boulnois, G., and Strom, A. R. (1988)
Molecular cloning, physical mapping and expression of the bet genes governing the
osmoregulatory choline-glycine betaine pathway of Escherichia coli, J. Gen. Microbiol.
134, 1737-1746.
11.
Ghanem, M., Fan, F., Francis, K., and Gadda, G. (2003) Spectroscopic and kinetic
properties of recombinant choline oxidase from Arthrobacter globiformis, Biochemistry
42, 15179-15188.
12.
Fan, F., Ghanem, M., and Gadda, G. (2004) Cloning, sequence analysis, and purification
of choline oxidase from Arthrobacter globiformis: a bacterial enzyme involved in
osmotic stress tolerance, Arch. Biochem. Biophys. 421, 149-158.
13.
Ghanem, M., and Gadda, G. (2006) Effects of reversing the protein positive charge in the
proximity of the flavin N(1) locus of choline oxidase, Biochemistry 45, 3437-3447.
14.
Hoang, J. V., and Gadda, G. (2007) Trapping choline oxidase in a nonfunctional
conformation by freezing at low pH, Proteins 66, 611-620.
109
15.
Quaye, O., Lountos, G. T., Fan, F., Orville, A. M., and Gadda, G. (2008) Role of Glu312
in binding and positioning of the substrate for the hydride transfer reaction in choline
oxidase, Biochemistry 47, 243-256.
16.
Orville, A. M., Lountos, G. T., Finnegan, S., Gadda, G., and Prabhakar, R. (2009)
Crystallographic, spectroscopic, and computational analysis of a flavin C4a-oxygen
adduct in choline oxidase, Biochemistry 48, 720-728.
17.
Gadda, G. (2003) pH and deuterium kinetic isotope effects studies on the oxidation of
choline to betaine-aldehyde catalyzed by choline oxidase, Biochim. Biophys. Acta. 1650,
4-9.
18.
Gadda, G. (2003) Kinetic mechanism of choline oxidase from Arthrobacter globiformis,
Biochim. Biophys. Acta. 1646, 112-118.
19.
Gadda, G., Powell, N. L., and Menon, P. (2004) The trimethylammonium headgroup of
choline is a major determinant for substrate binding and specificity in choline oxidase,
Arch. Biochem. Biophys. 430, 264-273.
20.
Fan, F., and Gadda, G. (2005) On the catalytic mechanism of choline oxidase, J. Am.
Chem. Soc. 127, 2067-2074.
21.
Fan, F., and Gadda, G. (2005) Oxygen- and temperature-dependent kinetic isotope effects
in choline oxidase: correlating reversible hydride transfer with environmentally enhanced
tunneling, J. Am. Chem. Soc. 127, 17954-17961.
22.
Fan, F., Germann, M. W., and Gadda, G. (2006) Mechanistic studies of choline oxidase
with betaine aldehyde and its isosteric analogue 3,3-dimethylbutyraldehyde, Biochemistry
45, 1979-1986.
110
23.
Gadda, G., Fan, F., and Hoang, J. V. (2006) On the contribution of the positively charged
headgroup of choline to substrate binding and catalysis in the reaction catalyzed by
choline oxidase, Arch. Biochem. Biophys. 451, 182-187.
24.
Fan, F., and Gadda, G. (2007) An internal equilibrium preorganizes the enzyme-substrate
complex for hydride tunneling in choline oxidase, Biochemistry 46, 6402-6408.
25.
Xin, Y., Gadda, G., and Hamelberg, D. (2009) The cluster of hydrophobic residues
controls the entrance to the active site of choline oxidase, Biochemistry 48, 9599-9605.
26.
Quaye, O., and Gadda, G. (2009) Effect of a conservative mutation of an active site
residue involved in substrate binding on the hydride tunneling reaction catalyzed by
choline oxidase, Arch. Biochem. Biophys. 489, 10-14.
27.
Quaye, O., Cowins, S., and Gadda, G. (2009) Contribution of flavin covalent linkage
with histidine 99 to the reaction catalyzed by choline oxidase, J. Biol. Chem. 284, 1699016997.
28.
Quaye, O., Nguyen, T., Gannavaram, S., Pennati, A., and Gadda, G. (2010) Rescuing of
the hydride transfer reaction in the Glu312Asp variant of choline oxidase by a substrate
analogue, Arch. Biochem. Biophys. 499, 1-5.
29.
Rungsrisuriyachai, K., and Gadda, G. (2008) On the role of histidine 351 in the reaction
of alcohol oxidation catalyzed by choline oxidase, Biochemistry 47, 6762-6769.
30.
Rungsrisuriyachai, K., and Gadda, G. (2010) Role of asparagine 510 in the relative
timing of substrate bond cleavages in the reaction catalyzed by choline oxidase,
Biochemistry 49, 2483-2490.
111
31.
Finnegan, S., and Gadda, G. (2008) Substitution of an active site valine uncovers a
kinetically slow equilibrium between competent and incompetent forms of choline
oxidase, Biochemistry 47, 13850-13861.
32.
Finnegan, S., Agniswamy, J., Weber, I. T., and Gadda, G. Role of valine 464 in the flavin
oxidation reaction catalyzed by choline oxidase, Biochemistry 49, 2952-2961.
33.
Finnegan, S., Yuan, H., Wang, Y. F., Orville, A. M., Weber, I. T., and Gadda, G. (2010)
Structural and kinetic studies on the Ser101Ala variant of choline oxidase: Catalysis by
compromise, Arch. Biochem. Biophys.
34.
Ghanem, M., and Gadda, G. (2005) On the catalytic role of the conserved active site
residue His466 of choline oxidase, Biochemistry 44, 893-904.
35.
Schowen, K. B., and Schowen, R. L. (1982) Solvent isotope effects of enzyme systems,
Methods Enzymol. 87, 551-606.
36.
Kirinčič, S., and Klofutar, C. (1999) Viscosity of aqueous solutions of poly(ethylene
glycol)s at 298.15 K, Fluid Phase Equilibria 115, 311-325.
37.
Cook, P. F., and Cleland, W. W. (2007) Enzyme Kinetics and Mechanism.
38.
Engelman, D. M., Steitz, T. A., and Goldman, A. (1986) Identifying nonpolar
transbilayer helices in amino acid sequences of membrane proteins, Annu. Rev. Biophys.
Biophys. Chem. 15, 321-353.
39.
Kyte, J., and Doolittle, R. F. (1982) A simple method for displaying the hydropathic
character of a protein, J. Mol. Biol. 157, 105-132.
40.
Kow, R. L., Whicher, J. R., McDonald, C. A., Palfey, B. A., and Fagan, R. L. (2009)
Disruption of the proton relay network in the class 2 dihydroorotate dehydrogenase from
Escherichia coli, Biochemistry 48, 9801-9809.
112
41.
Pitsawong, W., Sucharitakul, J., Prongjit, M., Tan, T. C., Spadiut, O., Haltrich, D., Divne,
C., and Chaiyen, P. A conserved active-site threonine is important for both sugar and
flavin oxidations of pyranose 2-oxidase, J. Biol. Chem. 285, 9697-9705.
42.
Wohlfahrt, G., Witt, S., Hendle, J., Schomburg, D., Kalisz, H. M., and Hecht, H. J.
(1999) 1.8 and 1.9 Å resolution structures of the Penicillium amagasakiense and
Aspergillus niger glucose oxidases as a basis for modelling substrate complexes, Acta.
Crystallogr. D Biol. Crystallogr. 55, 969-977.
43.
Panda, S. P., Gao, Y. T., Roman, L. J., Martasek, P., Salerno, J. C., and Masters, B. S.
(2006) The role of a conserved serine residue within hydrogen bonding distance of FAD
in redox properties and the modulation of catalysis by Ca2+/calmodulin of constitutive
nitric-oxide synthases, J. Biol. Chem. 281, 34246-34257.
44.
Xu, D., Kohli, R. M., and Massey, V. (1999) The role of threonine 37 in flavin reactivity
of the old yellow enzyme, Proc. Natl. Acad. Sci. U. S. A. 96, 3556-3561.
45.
Shen, A. L., and Kasper, C. B. (1996) Role of Ser457 of NADPH-cytochrome P450
oxidoreductase in catalysis and control of FAD oxidation-reduction potential,
Biochemistry 35, 9451-9459.
46.
Forneris, F., Heuts, D. P., Delvecchio, M., Rovida, S., Fraaije, M. W., and Mattevi, A.
(2008) Structural analysis of the catalytic mechanism and stereoselectivity in
Streptomyces coelicolor alditol oxidase, Biochemistry 47, 978-985.
47.
Liu, Y., Thoden, J. B., Kim, J., Berger, E., Gulick, A. M., Ruzicka, F. J., Holden, H. M.,
and Frey, P. A. (1997) Mechanistic roles of tyrosine 149 and serine 124 in UDPgalactose 4-epimerase from Escherichia coli, Biochemistry 36, 10675-10684.
113
48.
Lyubimov, A. Y., Lario, P. I., Moustafa, I., and Vrielink, A. (2006) Atomic resolution
crystallography reveals how changes in pH shape the protein microenvironment, Nat.
Chem. Biol. 2, 259-264.
49.
Ha, J. Y., Min, J. Y., Lee, S. K., Kim, H. S., Kim do, J., Kim, K. H., Lee, H. H., Kim, H.
K., Yoon, H. J., and Suh, S. W. (2006) Crystal structure of 2-nitropropane dioxygenase
complexed with FMN and substrate. Identification of the catalytic base, J. Biol. Chem.
281, 18660-18667.
50.
Chen, Z. W., Hassan-Abdulah, A., Zhao, G., Jorns, M. S., and Mathews, F. S. (2006)
Heterotetrameric sarcosine oxidase: structure of a diflavin metalloenzyme at 1.85 Å
resolution, J. Mol. Biol. 360, 1000-1018.
51.
Bernasconi, C. (1976) Relaxation kinetics.
52.
Sucharitakul, J., Prongjit, M., Haltrich, D., and Chaiyen, P. (2008) Detection of a C4ahydroperoxyflavin intermediate in the reaction of a flavoprotein oxidase, Biochemistry
47, 8485-8490.
53.
Strickland, S., Palmer, G., and Massey, V. (1975) Determination of dissociation constants
and specific rate constants of enzyme-substrate (or protein-ligand) interactions from rapid
reaction kinetic data, J. Biol. Chem. 250, 4048-4052.
114
CHAPTER 4
Mechanism of Flavin C4a-Adduct Formation in a Choline Oxidase Variant
4.1
Abstract
In the active site of choline oxidase, serine 101 is proximal to the C4a and N5 flaivn atom. In
Ser101Cys mutant protonated Tris base induce a cysteinyl adduct at the flavin C4a position,
which can convert to oxidized form reversibly while protonated Tris is removed for the active
site. The activities for these two forms enzyme are not significantly different. We measured pH
and isotope effects on the formation of adduct. Between pH 6.7 and 9.0, adduct formation was
observed. Ser101Cys showed a nearly 7-fold slowing of adduct formation or breakage in D2O
relative to H2O, indication that rate-limiting step involves one proton transfer to flavin N5 atom.
4.2
Introduction
Flavoenzymes contain either FMN or FAD at the active site as a cofactor which can transfer
one or two electrons. The isoalloxazine ring of the flavin offers several possibilities for
interaction with proteins. Most of them are noncovalently linked to the apoprotein. The first
reported flavin covalently linked enzyme was mammalian succinate dehydrogenase (1). Most of
the covalently attachment occurs at either 8α-methyl or 6-position of the isoalloxazine (2).
Recently, a bicovalently linkage to both positions was found, like glucooligosaccharide oxidase
and berberine bridge ezymes (3, 4). There are several residues have been observed to be bound to
the 8α position, such as histidine, cysteine or tyrosine. For the 6-position so far only cysteine
residue was found in the linkage. Mutangenesis, spectrometry and kinetic studies in some
enzymes have been shown that covalently linkage will affect redox poteintial, stability and
electron transfer or protect flavin from deteterious modifications (5-9).
115
There are also some enzymes found that can be covalently linked to C4a or N5 flavin and
form intermediate adducts during reaction. Such as flavin C4a-thiol adduct in pig heart lipoamide
dehydrogenase (10), mercuric ion reductase (11), LOV domains (12, 13), C4a-hydroperoxy-FAD
in pyranose 2-oxidase (14), C4a oxygen adduct in choline oxidase (15) and flavin-N5 adduct in
nitroalkane oxidase (16). Studies have shown that these kinds of adduct can be reversibly
reverted to each other. Such as LOV domain can form FMN-cysteine adduct induced by blue
light that slowly reverts to oxidized form in the dark (13).
In the crystal structure of wild-type choline oxidase and Ser101Ala enzyme shows that
Ser101 is proximal to the C4a and N5 flavin atoms (17, 18). In this study, the generated
Ser101Cys mutant shows that flavin C4a-cysteinyl adduct can be formed in a certain condition.
The adduct enzyme and oxidized enzyme can convert to each other reversibly and the enzyme
activity remains the same. In order to study this adduct, the mechanism of adduct formation was
investigated using pH profile, solvent isotope effects, solvent viscosity effects proton inventories,
inhibition and spectroscopic studies. The results shows that the formation of adduct was induced
by the protonated Tris base at the active site of the enzyme. Moreover, the formation of flavin
adduct is pH-dependent and when the pH lower than 6.5 or higher than 9, the flavin adduct will
convert reversibly to oxidized form.
D2O
(k3) and
D2O
(k4) with a value around 7 shows that the
rate-limiting step involves a proton flight.
4.3
Experimental Procedures
Materials. The Ser101 variants of choline oxidase were purified according to Yuan et al..(19)
Escherichia coli strain Rosetta (DE3)pLysS was from Novagen (Madison, WI). EDTA was from
Fisher. Choline chloride and amplicillin were from ICN Pharmaceutical Inc. All other reagents
were of the highest purity commercially available.
116
Instruments. UV-visible absorbance spectra were recorded using an Agilent Technologies
model HP 8453 diode array spectrophotometer equipped with a thermostated water bath. The
enzymatic activity of choline oxidase was measured polarographically using a computerinterfaced Oxy-32 oxygen monitoring system (Hansatech Instrument Ltd.). Stopped-flow
experiments were carried out using a Hi-Tech SF-61 Double Mixing Stopped Flow system.
Expression and Purification of CHO mutant enzymes. Permanent stocks of E. coli Rosetta
(DE3) pLysS cell harboring plasmid pET/codAmg1-Ser101X or His99Asn were used to
inoculate 7.5 L of Luria-Bertani broth medium containing 50 µg/mL ampicillin and 34 µg/mL
chloramphenicol. Ser101X and His99Asn mutant enzymes were purified as described previously
(19, 20).
Enzyme Assays. The apparent steady state kinetic parameters at atmospheric oxygen of the
Ser101Cys enzyme were measured with the method of the initial rates by monitoring oxygen
consumption as described for wild-type choline oxidase (21). Inhibition studies were performed
by varying the concentrations of both choline and Tris base in air-saturated 50 mM sodium
phosphate, pH 6.0, by monitoring the rate of oxygen consumption with a computer-interfaced
Oxy-32 oxygen-monitoring system (Hansatech Instrument) thermostated at 25 oC.
The formation of S101C flavin adduct experiments were carried out using an SF-61DX2 HITECH KinetAayst high performance stooped-flow spectrophotometer thermostatted at 25 oC.
Rates of flavin adduct formation were determined at varying concentrations of Tris base (2.5-100
mM) in 50 mM sodium phosphate and pH 7.0. Equal volumes of the enzyme and Tris base were
mixed aerobically in the stopped-flow spectrophotometer yielding a final enzyme concentration
of ~10 µM. The pH dependence of the adduct formation was studied in the stopped-flow as
described above in the pH range from 6.7 to 9.0.
117
For the determination of the solvent kinetic isotope effects on the formation of flavin adduct,
all of the reagents were prepared as described above expect that D2O was used to dissolve both
the enzyme and substrate, and the pH of the buffered solution containing the enzyme was
determined by adding 0.4 to the pH electrode reading, account for the isotope effect on the
ionization of sodium pyrophosphate (22). Solvent viscosity effects were measured in the
presence of 8% (mass) glucose as viscosogen, in both the tonometer containing the enzyme and
the syringes containing Tris base. The resulting relative viscosity at 25 oC was 1.25, which is
slightly above the value of 1.23 representing a 100% solution of D2O (23). For the proton
inventories in solvents containing varying mole fractions of D2O, the pD values were adjusted
using DCl and NaOD based on the empirical relationship (eq 1) that exists between the pH-meter
reading and the pD value at varying mole fractions of D2O (n) (22).
(ΔpH ) n = 0.076n 2 + 0.3314n
(1)
Spectral Studies. The pH dependence of the UV-visible absorbance spectra of choline
oxidase was investigated as previously described (24) to determine the effect of pH on the
spectral properties of the enzymes. The UV-visible absorbance of the flavin-bound variant
enzyme was initially recorded after gel-filtration with 20 mM NaH2PO4 and 20mM Na4P2O7
(buffer A) or 20 mM NaH2PO4, 20mM Na4P2O7 and 20mM Tris (buffer B), pH 6.0 at 15 °C to a
volume of ~2 mL and a final concentration of ~30 μM. The absorbance spectrum and pH of the
enzyme solution were recorded upon each addition of 1-10 μL of 1 M potassium hydroxide and
stirring until the pH was changed gradually and successively to ~12.
Data Analysis. Data were fit with KaleidaGraph software (Synergy Software, Reading, PA)
and the Hi-Kinetic Studio Software Suite (Hi-Tech Scientific, Bradford on Avon, U. K.). The
kinetic data with inhibitor were fit to eq. 2, which describe competitive inhibition pattern of the
118
inhibitor versus choline. I is the concentration of inhibitor and Kis is the inhibition constant for
the slope term.
v
=
e
k cat A
⎛
I ⎞
⎟⎟ + A
K a ⎜⎜1 +
⎝ K is ⎠
(2)
Stopped-flow traces were fit to eqs. 3 and 4, which describe single- and double- exponential
processes, respectively. kobs, kobs1 and kobs2 represent the observed first-order rate constant
associated with the absorbance change at 455 nm, t is time, At is the absorbance at 455 nm at any
given time, A, A1, and A2 are the amplitudes of the absorbance changes, and A∞ is the absorbance
at infinite time. Formation of flavin adduct parameters were determined by using eq. 5, where
kobs is the observed first-order rate constant for the formation of flavin adduct at any given
concentration of Tris base, k3 is the limiting first-order rate constant for flavin flavin adduct at
saturating concentrations of Tris base, k4 is the rate of the reverse step and
app
Kd is the
dissociation constant for binding of Tris base to the enzyme.
At = A exp(−k obs t ) + A∞
(3)
At = A1 exp(−k obs1t ) + A2 exp(−k obs 2 t ) + A∞
(4)
kobs =
k3 A
+ k4
app
Kd + A
(5)
The pH dependence of k3 was determined by fitting initial rate data to eq. 6, which describes
a curve with a slope of +1 and a plateau region at high pH. The pH profile of the
app
Kd values
was fit to eq. 7, which describes a curve with a slope of +1 and plateau regions at low pH. C is
the pH-independent value of the kinetic parameter of interest. The pH dependences of the k4
were determined by fitting initial rate data with eq. 8 where YL and YH are the limiting values at
119
low and high pH, respectively, and Ka is the dissociation constants for the ionization of groups
which are relevant to catalysis.
log(k 3 ) = log
C
10 − pH
1 + − pK a
10
(6)
log(app K d ) = log(C(1 + 10 pH − pKa ))
log( k 4 ) = log
YL + YH ×
10 − pK a
10 − pH
10 − pK a
1 + − pH
10
(7)
(8)
The effect of pH on the UV-visible absorbance spectrum was demonstrated in a plot of Δ
absorbance versus wavelength, nm, and the presence of ionizable groups within the
microenvironment of the flavin cofactor was determine in a plot of absorbance versus pH. The
data of pH dependencies of the Ser101X absorbance spectra with buffer A was fitted to eq. 9, the
data with buffer B was fitted to eq. 10 and the data for His99Asn was fitted to eq. 11, which
describes a curve with a slope of 1 and -1 and plateau regions at low and high pH where A and B
represent the absorbance values at the selected wavelength at low and high pH, respectively.
Y=
( A × 10 − pK a 1 + B × 10 − pH ) ( B × 10 − pK a 2 + A × 10 − pH )
+
(10 − pK a 1 + 10 − pH )
(10 − pK a 2 + 10 − pH )
Y=
( A × 10 − pK a 1 + B × 10 − pH ) ( B × 10 − pK a 2 + A × 10 − pH )
+
(10 − pK a 1 + 10 − pH )
(10 − pK a 2 + 10 − pH )
( B × 10 − pK a 2 + A × 10 − pH )
+
(10 − pK a 2 + 10 − pH )
Y=
(YL × 10 − pKa1 + YH × 10 − pH )
(10 − pK a1 + 10 − pH )
(9)
(10)
(11)
120
4.4
Results
Specttroscopic An
nalysis of CH
HO-Ser101C
Cys. The UV
V-visible abssorbance speectra of the C
CHO-
Ser101Cy
ys were reco
orded in 20 mM
m buffer, pH 8.0, as shhown in Figgure 4.1A. T
The Ser101C
Cys in
20 mM Tris
T displayeed a significaantly differeence in the vvisible regionn with the w
wild-type enzzyme,
indicating subtle alteerations of the
t FAD env
vironment. T
The spectrum
m of this sppecies exhibiits an
absorban
nce band at 378
3 nm with
h ε of 8.1 cm
m-1 mM-1. Coomparison thhe spectrum
m of Ser101C
Cys in
some bufffers suggessts that the requirement
r
of Tris basee for the forrmation of tthe flavin addduct.
Howeverr, the spectraa of Ser101A
Ala, Ser101T
Thr and Serr101Val are similar to w
wild-type enzzyme
and displlayed typical flavin max
xima in the near-UV
n
andd visible regions centereed at 370 andd 450
nm, as ex
xpected for an
a enzyme with
w bound flavin
fl
in the ooxidized statte, as shownn in Figure 44.1B.
Figure 4.1
4 UV-visib
ble absorban
nce spectra of S101X aat pH 8.0. ((A) Spectra of Ser101C
Cys at
different buffer pH 8.0.
8 Black cu
urve represnets 20 mM T
Tris buffer, R
Red curve reepresents 200 mM
TES bufffer, blue cu
urve represen
nts 20 mM HEPES bufffer and greeen curve reepresents 200 mM
potassium
m phosphatee buffer. (B
B) Spectra recorded in 20 mM Triis and pH 88.0. Black ccurve
representts CHO-wt, yellow cu
urve represeent CHO-Seer101Ala, R
Red curve rrepresents C
CHOSer101Th
hr, blue curv
ve representss CHO-Ser10
01Cys, and ggreen curve represents C
CHO-Ser1011Val.
Appa
arent Steadyy State Kin
netic Studiees. The appparent steaddy state kinnetic param
meters
determin
ned with cho
oline at atmo
ospheric oxy
ygen concenntration in 500 mM Tris bbuffer, at pH
H 8.0
with fullly oxidized Ser101Cys and adduct Ser101Cys , respectively. Both thee
app
app
(kcat/Km) and
kcat vallues of the tw
wo species enzymes
e
werre not signifi
ficantly diffeerent, with vaalues of 5600 ± 30
121
M-1s-1, 0..094 ± 0.001
1 s-1 for fully
y oxidized Ser101Cys
S
annd 580 ± 400 M-1s-1, 0.0886 ± 0.002 s-1 for
adduct Ser101Cys,
S
respectively
y, indicating
g that the ooxidized form
m and adduuct form caan be
reversibly
y converted to each other without lo
oss of enzym
me activity.
Tris Base
B
Inhibittion Studiess. The inhibittion pattern of Tris versuus choline w
was determinned in
50 mM sodium pho
osphate, pH 6.0, using the
t steady sstate kinetics approach with atmospphere
oxygen. Assuming that
t
the possitive charged choline is importannt for the suubstrate binnding,
protonateed Tris base is expected to compete with the subbstrate cholinne. As show
wn in Figuree 4.2,
patterenss with lines intersecting
i
fferent inhibiitor concentrration
on the y-axiis were obersseved at diff
(between
n 0 and 10 mM)
m in doub
ble reciprocal plots of thee initial ratees of reactionn as a functiion of
the substtrate concenttration. In acccordance with
w the obseerved kineticc patterns, thhe best fits oof the
kinetic data
d is compeetitive inhib
bition. The reelevant kineetic parameteers determinned were:
appp
Kis =
18 ± 2 mM.
m
Figure 4.2
4 Inhibition
n of choline oxidase with
h Tris. Initiaal rates weree measured inn air-saturateed 50
mM sodiium phosphaate at pH 6.0
0 and 25 oC . Tris conceentrations w
were 0 (●), 2 (■), 5(○) annd 10
mM (□).
122
Kinetics of Formation of the Flavin Adduct at pH 7.0 Induced by Binding of Tris. The
formation of flavin adduct was investigated in a stopped-flow spectrophotometer at 25 °C. The
CHO-Ser101Cys exhibit dramatic changes in its absorbance spectrum on titration with Tris, as
shown in Figure 4.3A. Difference spectrums show the loss of some of the oxidized flavin at the
457 nm and the formation of a new component. The reactions exhibit first-order kinetics over a
range of Tris concentrations (1-100 mM). At pH 7, the decrease in absorbance was fit best to a
single exponential process, as illustrated in Figure 4.3B. The observed rate of the flavin adduct
(kobs) were dependent on the substrate concentration with a finite intercept on the y-axis,
allowing the determination of limiting rate constant for the formation and breakage of flavin
adduct (kfor and krev) and the macroscopic equilibrium constant for Tris base binding at the active
site of the enzyme
app
Kd (Figure 4.3C). The value of krev is determined as 2.0 ± 0.3 s-1 from the
y-intercept of the plot. The value of kfor is determined as 8.2 ± 0.3 s-1 from the limiting value of
kobs. A generalized reaction scheme consistent with these data is shown in Scheme 4.1. Tris base
binds with Ser101Cys. And after the formation of Ser101Cys-Tris complex, Tris base promotes
the formation of a reversible flavin adduct.
S101Cox
Tris
S101Cox-Tris
Keq
k3
k4
S101Cadduct-Tris
k5
k6
S101Cadduct-Tris *
Scheme 4.1 Proposed Formation of Flaivn C4a Adduct Mechanism from the Kinetics of the
Attack Steps a
a
S101Cox, oxidized CHO-S101C; S101Cadduct, S101C flavin C4a adduct; k5, k6 steps only present
at the high pH reactions (pH not lower than 8.0).
123
Figure 4.3
4 Formation of flavin C4a
C adduct with
w Tris in 50 mM sodiium phosphaate, at pH 7.00 and
25 oC. Panel
P
A show
ws spectral changes ind
duced by addditions of T
Tris base. Paanel B show
ws the
reduction
n traces with
h 1 mM (b
black), 2.5 mM
m (red), 5 mM (bluee), 10 mM ((purple), 25 mM
(green), 50 mM (cy
yan) and 100
0 mM (brow
wn) Tris basse. All tracees were fit with eq 3. Time
indicated
d is after the end of the flow,
f
i.e. 2.2
2 ms. For claarity, one exxperimental point every 20 is
shown (v
vertical liness). Panel C shows
s
the ob
bserved rate of flavin reeduction as a function off Tris
concentraation. Data were
w fit to eq
q 5.
Effecct of pH on
n the Forma
ation of Ad
dduct Inducced by Triss. The pH deependence oof the
kinetic parameters
p
formation
f
off adduct ind
duced by T
Tris were deetermined w
with stopped-flow
spectroph
hotometer at
a 25 °C in the pH rang
ge from 6.77 to 9.0. Abbove pH 8.00, the reactiion is
distinctly
y biphasic, with
w a rapid decrease fo
ollowed by a slower deccrease at 4557 nm (tracees not
shown heere). The ob
bserved rate for the fastt phase show
wed saturatioon kinetic w
while plottedd as a
function of Tris conccentration. A plot of the observed ratte for the sloow phase waas independeent on
124
the Tris concentration with an average value of 0.7 s-1 (Table 4.1). The rates for the second phase
are independent on the Tris base concentration (data not shown here); probably due to the
conformational change as proposed in the blue part of Scheme 4.1. At all pH values where flavin
adduct can be formed, the observed rate constant for flavin adduct formation increased with Tris
concentration. As shown in Figure 4.4A, the observed values for and kfor and krev were pHdependent. Within this pH range, the rates increased with increasing pH, allowing an
approximate pKa = 6.7 ± 0.1 and 8.2 ± 0.4 to be determined for a group that must be
deprotonated for kfor and krev, respectively. The kfor is small at low pH, it increases with pH and
reached a limiting value of 12.3 ± 0.8 s-1 at pH 9.0. The
app
Kd are strongly pH-dependent, as
shown in Figure 4.4B. The data are fit well to eq. 7 and give a pKa value of 8.5 ± 0.1, suggesting
that there is a protonated group involved in binding.
Table 4.1 pH-Dependence of the Kinetic Parameters with Tris Base for Ser101Cys in 50 mM
Buffer, at 25 oC
app
Kd, mM
krev, s-1
pH
kfor, s-1
kobs2 , s-1
total
protonated
deprotonate
6.7
5.3 ± 0.2
10.7 ± 2.1
10.5
0.17
2.3 ± 0.2
7.0
8.2 ± 0.3
10.8 ± 2.2
10.5
0.33
2.0 ± 0.3
-
7.5
8.0
8.5
9.0
8.1 ± 0.3
10.2 ± 0.5
10.4 ± 0.5
12.3 ± 0.8
10.7 ± 2.1
13.6 ± 3.8
20.5 ± 5
42 ± 9
9.7
10.3
10.3
10.1
0.97
3.27
10.25
31.9
2.0 ± 0.3
4.7 ± 0.7
5.2 ± 0.5
5.5 ± 0.3
0.67
0.66
0.71
125
Figure 4.4
4 pH-Depeendence of kfor (panel A solid circlee), krev (paneel A empty circle) and appKd
(panel B)) values of formation
f
off flavin C4a adduct withh Tris base. R
Reactions w
were perform
med in
50 mM buffer
b
at 25 oC. Data werre fit to eqs. 6, 7 and 8.
Solveent KIEs. The
T effects of deuteratted solvent were determ
mined to prrobe whetheer the
kinetic stteps involvin
ng solvent exchangeable
e
e protons arre rate-limitiing or not. T
The flavin addduct
was inveestigated in H2O and D2O, and thee relevant
D22O
kfor, D2Okrevv and
D2O appp
(
Kd) values were
determin
ned at pL 8.8
8 and 25 oC with
w Tris baase. Substituttion of H2O with D2O, llarge solventt KIE
values off D2Okfor and
D2O
krev with
h 6.1 ± 0.3 and
a 7.8 ± 0.66 were obserrved respecttively, suggeesting
that the kinetic
k
steps of solvent exchangeable
e
e protons aree also at leasst partially raate limiting iin the
formation
n and break
kage of addu
uct. There is a solvent KIE with
D
D2O app
( Kd) vaalue of 1.6 ± 0.3
observed
d, probably due
d to pH efffect because Kd is not pH
H-independeent at pL 8.8.
Solveent KIEs weere also deteermined at pL
p 8.5 to avvoid artifactuual contributtion arising from
pH effectt. At pL 8.5,, a similar so
olvent KIE of
o 6.3 ± 0.2 aand 7.3 ± 0.4 were founnd for kfor andd krev,
respectiv
vely, suggestting that the observed vaalues at pL 88.5 are in the pH-indepeendent region. To
establish whether thee observed solvent
s
KIEss originated from the chhemical kineetics effect rrather
than bein
ng due to an
a increased viscosity of
o D2O withh respect to H2O, the eeffects of soolvent
viscosity
y on the form
mation of flav
vin adduct were
w
investiggated. The eexperiment w
was carried oout at
pH 8.8 an
nd 25 oC in solutions co
ontaining 8%
% glucose, wh
which providees a relative solvent visccosity
126
equivalen
nt to a 100%
% solution off D2O. Simillar kfor and k rev values w
were observeed in the pressence
and abseence of visccosigen (ratios are 1.00 ± 0.03 andd 1.09 ± 0.005 respectivvely). Thesee data
indicate that
t
the observed solven
nt kinetic KIEs
K
at pL 8 .8 were werre not contriibuted by soolvent
viscosity
y.
Proto
on Inventorries. To gain
n further und
derstanding of the obserrved solvent KIEs on kfoor and
krev, the proton
p
inven
ntory techniq
que was useed to gain innsights into the number of exchanggeable
protons that
t produces the solven
nt KIEs. Protton inventoriies studies w
were investiggated in soluutions
with vary
ying mole frractions of deuterium
d
ox
xide during tthe formationn of flavin aadduct. As shhown
in Figuree 4.5, linear relationship
ps were seen
n upon plottiing the kfor annd krev values as a functiion of
the mole fraction of deuterated solvent,
s
sugg
gesting that tthe solvent K
KIE is due tto a single pproton
in flight in
i the transittion state forr the reaction
n of flavin reeduction.
Figure 4.5
4 Proton in
nventories off kfor (solid circle)
c
and k rev (empty ccircle) values of formation of
flavin C4
4a adduct at pL 8.8.
Effeccts of pH on
n the UV-Viisible Absorrbance Specctra of Free Oxidized V
Variant Enzyymes.
The effect of pH on
n the UV-viisible absorb
bance spectrrum of the Ser101X vaariant enzym
me in
comparisson with thee wild-type enzyme
e
wass investigateed respectiveely to determ
mine whetheer the
pKa values were affeected by repllacing the seerine at posiition 101. Thhe effects w
were monitorred in
127
two buffer systems due to the observed flavin adduct of Ser101Cys induced by Tris base. In the
buffer A system which is in the absence of Tris base, the UV-visible absorbance spectrum of the
serine variants enzyme showed significant changes in absorbance around 390 nm and 500 nm
with increasing pH which are similar to the wild-type enzyme around the same wavelength
suggesting no differences in the flavin microenvironment of the enzymes (Figure S4.1 for wildtype, S101Ala, Ser101Cys, Ser101Thr and Ser101Val). A plot of the absorbance against pH for
CHO-Ser101X (Figure S4.1, inset) showed two pKa values of ~8 and ~12, suggesting the
presence of two ionizable groups. The UV-visible absorbance spectrum of the His99Asn variant
enzyme showed significant changes in absorbance around 400 nm with increasing pH suggesting
differences in the flavin microenvironment from the Ser101X enzymes (Figure S4.1). A plot of
the absorbance against pH for CHO-His99Asn (Figure S4.1 insert) showed one pKa with a value
of ~10 and consistent with the presence of one ionizable group. In the case of buffer B system
which is in the presence of Tris base, the spectrums of wild-type still shows oxidized state at all
the pH (Figure 4.6A). The observed spectral changes, along with the fitting of the data with
eq.10, are consistent with the presence of three pKa values associated three ionizable groups
(Figure 4.6B). As shown in Figure 4.6C, Ser101Cys is in oxidized state, when the pH is
increased from 6.0 to 6.5. Further increasing of the pH to 9.5 yielded significantly spectral
changes. Further increasing of pH above 9.5, the spectrum shows flavin goes back to the
oxidized state. Three pKa values of of ~7, ~9.5 and ~12 were determined by fitting of the data
with eq. 10 (Figure 4.6D), which is similar to the values obtained with wild-type under the same
condition. The relevant pKa values for variants enzyme are summarized in Table 4.2.
128
0.3
Absorbance
0.36
ΔAbsorbance
0.2
0.1
0.3
0.24
6
8 pH 10
12
0
-0.1
B
--0.2
300
400
50
00
600
0
W
Wavelength,
n
nm
Abs
ΔAbsorbance
460
0.4
0.6
0.5
0.4
0.2
0.3
6
8
10
12
pH
0.0
D
-0.2
300
400
50
00
600
W
Wavelength,
nm
m
Figure 4.6
4 Effect off pH on thee spectral prroperties of CHO-WT aand S101C iin buffer B. UVvisible ab
bsorbance sp
pectra were recorded
r
at different pH
H for CHO-W
WT (A) and S101C (C) aabout
o
21 μM, at
a 15 C. On
nly selected difference absorbance
a
sspectra for C
CHO-WT (B
B) and S101C
C (D)
in the pH
H range betw
ween 6 and 12
1 are shown
n. Inset: UV
V-visible abssorbance vallues as a funnction
of pH, th
he curves aree fits of the data
d to eq.10.
129
Table 4.2 Comparison of the pKa Values of Choline Oxidase Mutant Enzymes with Wild-Type.
kinetic
WTa
Ser101Ala
Ser101Cys
Ser101Thr
Ser101Val
His99Asn
WTb
Ser101Cys
parameters
pKa1
8.2 ± 0.1
8.6 ± 0.1
8.0 ± 0.1
8.0 ± 0.1
8.2 ± 0.1
7.9 ± 0.1
6.9 ± 0.1
pKa2
>12
>12
>12
>12
>12
9.8 ± 0.1
>12.0
>12
pKa3
10.3 ± 0.2
9.4 ± 0.1
a
Spectrophotometric titration of CHO-Ser101X in the 20 mM sodium phosphate, sodium pyrophosphate buffer by KOH, 15 oC.
b
Spectrophotometric titration of CHO-Ser101X in the 20 mM sodium phosphate, sodium pyrophosphate and 20 mM Tris buffer by
KOH, 15 oC.
130
4.5
Discussion
For the previous studies, the serine residue is located in proximity to the C4a-N5 atom of the
flavin cofactor of choline oxidase and plays an important role in the flavin reduction reactions. In
this paper, the Ser101Cys mutant was generated and shows the cysteine can form flavin adduct
with C4a atom in a certain condition. The formation of flavin adduct was studied with stoppedflow spectrometry. From pH profile, inhibition and solvent kinetic isotope effects studies, the
formation of flavin adduct was proposed as in Scheme 4.2. First, Tris base gets to the active site
and lower the pKa of cysteine. Deprotonated cysteine will nucleophilically attack on the C4a of
flavin isoalloxazine ring to create the cysteinyl-flavin adduct. The evidences for this mechanism
are discussed in the following.
R
R
N CH2
+
N
N
O
HN
NH
N
H+
-
S
N CH2
+
N
Cys
O
NH
N
H
O
N
HN
S
O
Cys
Scheme 4.2 Proposed Fromation of Flavin C4a Adduct Mechanism in 50 mM Sodium
Pyrophosphate, pH 8.8
Absorbance Spectrum of Ser101Cys Suggests the Formation of a Flavin C(4a)- Cysteinyl
Adduct. The absorbance of Ser101Cys in the Tris buffer exhibits maximum peak at 378 nm,
suggesting the formation of a flavin C4a-cysteinyl adduct. Adduct formation is accompanied by
changes in the visible absorption spectrum. It is identified as adduct formed between the sulfur of
a cysteine residue and the C4a position of the flavin, based on the similarity of its spectral
properties with covalent thiolate adducts observed with LOV, lipoamide dehydrogenase and
mercuric ion reductase. The spectrum of sulfide-adduct close to a spectrum for endogenous
heterotetrameric sarcosine oxidase adduct (λmax=383 nm, ε=7.3 mM-1cm-1) (25) and spectra
131
reported for 4a-thiolate adducts in lipoamide dehydrogenase (λmax=384 nm, ε=8.7 mM-1cm-1)
(10), LOV domain with the λmax=378 nm, ε=8.7 mM-1cm-1 (26) and mercuric reductase
(λmax=382 nm, ε=7.5 mM-1cm-1) (11) where the complexes are formed by reaction of the flavin
with an active site cysteine. The spectra that exhibit spectral characteristics similar to this are
those measured for the formation of flavin C(4a)-cysteinyl adduct which typically have a
maximum absorbance at ~380 nm.
Protonated Tris Is Important for the Formation of Flavin Adduct. Data supporting this
conclusion comes from the pH profile of the
app
app
Kd and krev. From the pH profile study on the
Kd, the substrate affinity decreased while increasing the pH and a pKa value of 8.5 ± 0.1 was
determined which implicates that there is a protonated group involved in substrate binding.
Based on the pKa and
app
Kd values, the concentration of different Tris base species were
calculated. As summarized in Table 4.1,
app
Kd for the protonated form is around 10 mM for all
the pH, which indicated that the pKa value was assigned to Tris base; otherwise, different
concentration for the protonated form will be obtained for all the pH. The inhibition studies
showed that Tris base is a competitive inhibitor of the enzyme; therefore, Tris base is at the
active site of the enzyme. These results clearly indicate that the protonated form of Tris will get
into the active site and is the relevant species responsible for inducing the formation of flavin
adduct. And also the pKa value about 8.2 ± 0.4 on the breakage of the adduct which is similar to
the pKa value on the appKd could be Tris base. At high pH, Tris base will be deprotonated and the
rate of adduct breakage increases. The requirement of protonated group for the formation of
flavin C4a adduct are also seen in other enzymes with flavin C4a-cysteinyl adduct, such as the
adduct in lipoamide dehydrogenase only observes upon binding of NAD+ (10) and the adduct
formation in the mercuric ion reductase only occurs in the presence of NADP+ (11). The possible
132
explanation was that the electrophilicity of the oxidized flavin increases by the closeby positive
charge which promote the accumulation of the flavin adduct (11).
Both Flavin C(4a)-Cysteinyl Adduct Formation and Breakage Involving Proton Transfer
Are Rate-Limiting Steps. Evidence supporting this conclusion comes from the large solvent
kinetic isotope effects and proton inventories on the observed reaction rate determined with
stopped-flow spectrometer. The solvent KIE on the forward and reverse reactions with values of
about ~7 indicates that solvent-derived protons are in flight during the formation and breakage of
flavin adduct. In order to ascertain the number of protons in flight, a proton inventory was
conducted. The proton inventory on kfor and krev is linear, indicating that a single proton transfer
that participates in the flavin adduct formation and breakage. From the pH profile on kfor, there is
a group need to be deprotonated with a pKa value of 6.7 for the formation of the Flaivn C(4a)Cysteinyl adduct. This group is proposed to be cysteine 101 at the active site. Therefore, the
solvent isotope effects observed at pL 8.8 suggest that there is one proton in flight from the
solvent or a solvent exchangeable site to the flavin N5 atom.
Ser101 Does Not Change the Polarity of the Active Site of Choline Oxidase. Evidence
supporting this conclusion comes from the UV-visible absorbance pH titration in buffer A of the
Ser101 variants of choline oxidase in comparison with that of the wild-type and His99Asn
enzyme from pH 6 to 12. The UV-visible absorbance pH titration for choline oxidase wild-type
and Ser101X variants enzyme from this study showed two pKa values of ~8 and ~11. In
His99Asn mutant, the absence of the species with the pKa value of ~8 is therefore consistent with
the His99 being responsible for the pKa. A low pKa value of ~6 in a UV-visible absorbance pH
titration between 5.5 and 8.5 has been shown in cholesterol oxidase from two different sources
and assigned to the ionization of the histidine residues involved in the covalent attachment of the
133
flavin cofactor to the protein moiety in that enzyme (27). In wild-type choline oxidase, pKa value
of ~8 was obtained in a similar pH titration between pH 6 and 11 and was thought to be the
ionization of the N(3) position of the isoalloxazine ring (24). A pH ranges used for the pH
titrations for cholesterol oxidase and choline oxidase could only determine the low pKa. The
second pKa could be assigned to the ionization of the N3 locus of the oxidized enzyme-bound
FAD (24). The one pKa (~10) observed in the asparagine variant enzyme is suitably attributable
to the ionization of the N(3) position of the isoalloxazine ring. As summarized in Table 4.2, the
pKa values for wild-type enzyme and Ser101X variants are similar in buffer A, suggesting that
the residue at position 101 did not change the polarity of the active site of choline oxidase. The
UV-visible absorbance pH titration of wild-type and CHO-Ser101Cys variant in the buffer B
suggests that a thermodynamic pKa value of 9.4 was determined for Tris base. As shown in
Table 4.2, pKa values of 6.9 ± 0.1, 9.4 ± 0.1 and ~ 12 for S101C and pKa values of 7.7 ± 0.1, 9.5
± 0.2 and ~ 12 for the wild-type, both of them showed three ionizations. Comparing the pKa
values determined in buffer A, the three pKa values could be assigned to ionization of His99, Tris
base and the N3 locus of the oxidized FAD, respectively. The cysteine residue at the active site
of the Ser101Cys mutant may affect the pKa of His99 and cause the difference compared to the
wild-type.
There Is a Commitment to Catalysis in the Formation of a Flavin C(4a)- Cysteinyl Adduct.
Evidence for this conclusion comes from the different pKa values for Tris base obtained in the
pH-profile of the
app
Kd and pH titration flavin spectrum of Ser101Cys. The thermodynamic pKa
value of 9.4 observed is an intrinsic value. The different pKa value with 8.5 is indicative of the
presence of a forward commitment to form the flavin C4a adduct. Based on the equations 11 and
12, if k3<<k2, the appKd = trueKd, while k3 is not significantly smaller than k2, the pKa value of appKd
134
will be shift. The pH profile studies show the kfor will increase with increasing of pH. Therefore,
at high pH, k3 cannot be ignored, the pKa values for appKd shifted to the left.
app
Kd =
k 2 + k3
k1
(11)
true
Kd =
k2
k1
(12)
Proposed Model for the Formation of a Flavin C(4a)- Cysteinyl Adduct. Ser101Cys
exhibits spectral properties significantly different from those of other enzymes. This difference is
due to the fact that the flavin in the enzyme forms a reversible covalent 4a-adduct with a cysteine
residue. As indicated in Scheme 4.3, in the presence of Tris base, the His99 and Cys101 will get
deprotonated with a pKa of 6.9 and deprotonated cysteine residue nucleophilically attack to the
flavin C4a atom. Further increasing pH, Tris base will get deprotonated which cause the
breakage of flavin adduct. While the pH goes above 11, flavin N3 atom will get deprotonated. In
the absence of Tris base, no flavin adduct was observed. The ionization of His99 and flavin N3
give two pKa, as shown in Scheme 4.3 red color.
R
R
N CH2
+
N
N
O
HN
N CH2
pKa=6.9
NH
N
H
S
N
O
NH
N
H
O
S
O
Cys
Cys
pK
a =8
pKa of Tris 9.4
.0 2
R
R
N CH2
N
N
N
N
O
N
N
N
-
S
O
Cys
pKa>11
N CH2
N
N
O
N
NH
N
-
O
S
Cys
Scheme 4.3 Proposed Formation of Flaivn C4α Adduct Mechanism from pH Titration in the
Present and Absent of Tris Base
135
4.6
Support Infformation
Figure S4.1
S Effect of pH on the spectral prop
perties of CH
HO-Ser101X
X and His99Asn in buffeer A.
Absorban
nce spectra were
w recordeed at an enzy
yme concenttration aboutt 21 μM, at 115 oC. Only
selected difference
d
ab
bsorbance sp
pectra in the pH range beetween 6 and 12 are shoown. Inset: U
UVvisible ab
bsorbance vaalues of CHO
O-Ser101X as
a a functionn of pH; the curves are ffits of the datta to
eqs. 9 and 11.
136
4.7
1.
References
Walker, W. H., and Singer, T. P. (1970) Identification of the covalently bound flavin of
succinate dehydrogenase as 8-alpha-(histidyl) flavin adenine dinucleotide, J Biol Chem
245, 4224-4225.
2.
Mewies, M., McIntire, W. S., and Scrutton, N. S. (1998) Covalent attachment of flavin
adenine dinucleotide (FAD) and flavin mononucleotide (FMN) to enzymes: the current
state of affairs, Protein Sci 7, 7-20.
3.
Winkler, A., Hartner, F., Kutchan, T. M., Glieder, A., and Macheroux, P. (2006)
Biochemical evidence that berberine bridge enzyme belongs to a novel family of
flavoproteins containing a bi-covalently attached FAD cofactor, J Biol Chem 281, 2127621285.
4.
Heuts, D. P., Winter, R. T., Damsma, G. E., Janssen, D. B., and Fraaije, M. W. (2008)
The role of double covalent flavin binding in chito-oligosaccharide oxidase from
Fusarium graminearum, Biochem J 413, 175-183.
5.
Fraaije, M. W., van den Heuvel, R. H., van Berkel, W. J., and Mattevi, A. (1999)
Covalent flavinylation is essential for efficient redox catalysis in vanillyl-alcohol oxidase,
J Biol Chem 274, 35514-35520.
6.
Mewies, M., Basran, J., Packman, L. C., Hille, R., and Scrutton, N. S. (1997)
Involvement of a flavin iminoquinone methide in the formation of 6-hydroxyflavin
mononucleotide in trimethylamine dehydrogenase: a rationale for the existence of 8alphamethyl and C6-linked covalent flavoproteins, Biochemistry 36, 7162-7168.
7.
Lu, X., Nikolic, D., Mitchell, D. J., van Breemen, R. B., Mersfelder, J. A., Hille, R., and
Silverman, R. B. (2003) A mechanism for substrate-Induced formation of 6-
137
hydroxyflavin mononucleotide catalyzed by C30A trimethylamine dehydrogenase,
Bioorg Med Chem Lett 13, 4129-4132.
8.
Caldinelli, L., Iametti, S., Barbiroli, A., Bonomi, F., Fessas, D., Molla, G., Pilone, M. S.,
and Pollegioni, L. (2005) Dissecting the structural determinants of the stability of
cholesterol oxidase containing covalently bound flavin, J Biol Chem 280, 22572-22581.
9.
Kim, J., Fuller, J. H., Kuusk, V., Cunane, L., Chen, Z. W., Mathews, F. S., and McIntire,
W. S. (1995) The cytochrome subunit is necessary for covalent FAD attachment to the
flavoprotein subunit of p-cresol methylhydroxylase, J Biol Chem 270, 31202-31209.
10.
Thorpe, C., and Williams, C. H., Jr. (1976) Spectral evidence for a flavin adduct in a
monoalkylated derivative of pig heart lipoamide dehydrogenase, J Biol Chem 251, 77267728.
11.
Miller, S. M., Massey, V., Ballou, D., Williams, C. H., Jr., Distefano, M. D., Moore, M.
J., and Walsh, C. T. (1990) Use of a site-directed triple mutant to trap intermediates:
demonstration that the flavin C(4a)-thiol adduct and reduced flavin are kinetically
competent intermediates in mercuric ion reductase, Biochemistry 29, 2831-2841.
12.
Salomon, M., Eisenreich, W., Durr, H., Schleicher, E., Knieb, E., Massey, V., Rudiger,
W., Muller, F., Bacher, A., and Richter, G. (2001) An optomechanical transducer in the
blue light receptor phototropin from Avena sativa, Proc Natl Acad Sci U S A 98, 1235712361.
13.
Crosson, S., and Moffat, K. (2001) Structure of a flavin-binding plant photoreceptor
domain: insights into light-mediated signal transduction, Proc Natl Acad Sci U S A 98,
2995-3000.
138
14.
Prongjit, M., Sucharitakul, J., Wongnate, T., Haltrich, D., and Chaiyen, P. (2009) Kinetic
mechanism of pyranose 2-oxidase from trametes multicolor, Biochemistry 48, 4170-4180.
15.
Orville, A. M., Lountos, G. T., Finnegan, S., Gadda, G., and Prabhakar, R. (2009)
Crystallographic, spectroscopic, and computational analysis of a flavin C4a-oxygen
adduct in choline oxidase, Biochemistry 48, 720-728.
16.
Gadda, G., Edmondson, R. D., Russell, D. H., and Fitzpatrick, P. F. (1997) Identification
of the naturally occurring flavin of nitroalkane oxidase from fusarium oxysporum as a 5nitrobutyl-FAD and conversion of the enzyme to the active FAD-containing form, J Biol
Chem 272, 5563-5570.
17.
Quaye, O., Lountos, G. T., Fan, F., Orville, A. M., and Gadda, G. (2008) Role of Glu312
in binding and positioning of the substrate for the hydride transfer reaction in choline
oxidase, Biochemistry 47, 243-256.
18.
Finnegan, S., Yuan, H., Wang, Y. F., Orville, A. M., Weber, I. T., and Gadda, G. (2010)
Structural and kinetic studies on the Ser101Ala variant of choline oxidase: catalysis by
compromise, Arch Biochem Biophys 501, 207-213.
19.
Yuan, H., and Gadda, G. (2011) Importance of a serine proximal to the c(4a) and n(5)
flavin atoms for hydride transfer in choline oxidase, Biochemistry 50, 770-779.
20.
Quaye, O., Cowins, S., and Gadda, G. (2009) Contribution of flavin covalent linkage
with histidine 99 to the reaction catalyzed by choline oxidase, J Biol Chem 284, 1699016997.
21.
Fan, F., and Gadda, G. (2005) On the catalytic mechanism of choline oxidase, J Am
Chem Soc 127, 2067-2074.
139
22.
Schowen, K. B., and Schowen, R. L. (1982) Solvent isotope effects of enzyme systems,
Methods Enzymol 87, 551-606.
23.
Lide, D. R. (2000) Handbook of Chemistry and Physics, CRC Press, Boca Raton, FL.
24.
Ghanem, M., and Gadda, G. (2005) On the catalytic role of the conserved active site
residue His466 of choline oxidase, Biochemistry 44, 893-904.
25.
Hynson, R. M., Mathews, F. S., and Schuman Jorns, M. (2006) Identification of a stable
flavin-thiolate adduct in heterotetrameric sarcosine oxidase, J Mol Biol 362, 656-663.
26.
Salomon, M., Christie, J. M., Knieb, E., Lempert, U., and Briggs, W. R. (2000)
Photochemical and mutational analysis of the FMN-binding domains of the plant blue
light receptor, phototropin, Biochemistry 39, 9401-9410.
27.
De Francesco, R., and Edmondson, D. E. (1988) pKa values of the 8 alpha-imidazole
substituents in selected flavoenzymes containing 8 alpha-histidylflavins, Arch Biochem
Biophys 264, 281-287.
140
CHAPTER 5
Conformational Changes and Substrate Recognition in Pseudomonas aeruginosa DArginine Dehydrogenase
(This chapter has been published verbatim in Fu, G., Yuan, H., Li, C., Lu, C., Gadda, G. and
Weber, I., (2010), Biochemistry 49: 8535-8545. The author’s contribution to this chapter
involved the preparation and purification of the untagged wild-type enzyme and the
determination of the steady state kinetic parameters with a number of D-amino acids as
substrates.)
5.1
Abreviations
DADH, D-arginine dehydrogenase; LADH, L-arginine dehydrogenase; DAAD, D-amino
acid dehydrogenase; pDAAO, porcine D-amino acid oxidase; hDAAO, human D-amino acid
oxidase; LAAO, L-amino acid oxidase; PMS, phenazine methosulfate; RMSD, root-mean-square
deviation.
5.2
Abstract
DADH catalyzes the flavin-dependent oxidative deamination of D-amino acids to the
corresponding α-keto acids and ammonia. Here we report the first X-ray crystal structures of
DADH at 1.06 Å resolution, and its complexes with iminoarginine (DADHred/iminoarginine) and
iminohistidine (DADHred/iminohistidine) at 1.30 Å resolution. The DADH crystal structure
comprises an unliganded conformation and a product-bound conformation, which is almost
identical to the DADHred/iminohistidine crystal structure. The active site of DADH was partially
occupied with iminoarginine product (30% occupancy) that interacts with Tyr53 in the minor
conformation of a surface loop. This flexible loop forms an “active site lid”, similar to those seen
in other enzymes, and may play an essential role in substrate recognition. The guanidinium side
chain of iminoarginine forms a hydrogen bond interaction with the hydroxyl of Thr50 and an
141
ionic interaction with Glu87. In the structure of DADH in complex with iminohistidine, two
alternate conformations were observed for iminohistidine where the imidazole groups formed
hydrogen bond interactions with the side chains of His48 and Thr50 and either Glu87 or Gln336.
The different interactions and very distinct binding modes observed for iminoarginine and
iminohistidine are consistent with the 1000-fold difference in kcat/Km values for D-arginine and
D-histidine. Comparison of the kinetic data for the activity of DADH on different D-amino acids
and the crystal structures in complex with iminoarginine and iminohistidine establishes that this
enzyme is characterized by relatively broad substrate specificity, being able to oxidize positively
charged and large hydrophobic D-amino acids bound within a flask-like cavity.
5.3
Introduction
All the standard amino acids except glycine can exist as either L- or D- optical isomers. Lamino acids are predominant in proteins synthesized by all living organisms. D-amino acids are
important for bacteria as fundamental elements of the bacterial cell wall peptidoglycan layer (1).
Several D-amino acids, D-leucine, D-methionine, D-tyrosine and D-tryptophan, were recently
reported to regulate disassembly of bacterial biofilms (2). Also, specific D-amino acids have
been discovered in many physiological processes in vertebrates, including humans (3). Dunlop et
al. identified D-aspartate in the brain and other tissues of mammals and suggested it may play a
role in regulating the development of these organs (4). D-serine was identified at significant
levels in rat brain, at about a third of the level of L-serine (5). Moreover, D-serine in the rat brain
is distributed in close association with N-methyl-D-aspartate (NMDA) and it may act as an
endogenous agonist of the NMDA receptor in mammalian brains (5).
Conversion of L- and D- amino acids in living organisms is commonly catalyzed by
racemases. Various aminoacid racemases have been identified in bacteria, archaea, and
142
eucaryotes including mammals. These racemases are classified into two groups: pyridoxal 5’phosphate (PLP) -dependent and -independent enzymes (3). In mammals, D-serine racemases
and D-aspartate racemases are the most intensively studied enzymes due to their involvement in
cell aging and neural signaling (6-7). In bacteria, D-amino acids are deaminated by flavindependent dehydrogenases (8), allowing the bacteria to grow using D-amino acids as the sole
carbon, nitrogen and energy source, in a concentration-sensitive manner, since some D-amino
acid analogs have toxic effects on bacterial growth (9-10).
Pseudomonas aeruginosa, an opportunistic human pathogen, can grow with D-arginine as
the sole source of carbon and nitrogen (11). D-to-L arginine racemization by two coupled
dehydrogenases serves as prerequisite of D-arginine utilization through L-arginine catabolic
pathways (12-13). One enzyme is an FAD-dependent catabolic DADH and the other is an
NAD(P)H-dependent anabolic LADH. DADH catalyzes the conversion of D-arginine into 2ketoarginine and ammonia, and LADH converts 2-ketoarginine to L-arginine. In order to
understand the reaction mechanism and substrate specificity of DADH, the crystal structure of
DADH was determined at the atomic resolution of 1.06 Å, while the structures of DADH
crystallized in the presence of D-arginine and D-histidine were both determined at 1.30 Å
resolution. Well-defined electron density for the non-covalently bound FAD and imino
intermediate of the reaction allowed detailed analysis of the enzyme active site. A loop region
with alternative conformations was identified in the DADH structure and is involved in substrate
binding. Very distinct binding modes were observed for iminoarginine and iminohistidine, in
agreement with detailed kinetic analysis on substrate specificity reported previously (14) and in
this paper. The structural characteristics described here provide insights into substrate
recognition and the catalytic reaction mechanism of DADH.
143
5.4
Experimental Procedures
Materials. D-amino acids and L-amino acids were from Alfa Aesar and Sigma-Aldrich (St.
Louis, MO). Phenazine methosulfate (PMS) and phenylmethanesulfonyl fluoride (PMSF) were
obtained from Sigma-Aldrich (St. Louis, MO). All of the other reagents were of the highest
purity commercially available.
Expression and Purification of DADH and SeMet-DADH. Hexa-histidine-tagged DADH
was expressed in Escherichia coli TOP10 and purified as described previously (13). No FAD
was added during expression and purification. The selenomethionine (SeMet) DADH protein
was prepared following a protocol with slight modifications as described in a previous study
(15). DTT (10-20 mM) was incorporated in the buffer throughout the purification steps in order
to avoid oxidation of selenium. Mass spectrometry confirmed that about 88% of the eight
methionines were replaced by Se-Met.
Crystallization and X-ray Data Collection. Purified DADH and SeMet-DADH were
concentrated to 6 mg/mL and 3 mg/mL, respectively, in 50 mM Tris at pH 7.5. Crystals were
grown by the hanging drop vapor diffusion method using a well solution of 0.1M 2-(Nmorpholino)-ethanesulfonic acid (MES) pH 6.5-7.0, 5% glycerol, and 6-10% (w/v) PEG6000.
Crystals can grow to a size of 0.1-0.2 mm3 within two weeks. DADH was co-crystallized with
D-arginine or D-histidine under similar conditions using a 1:10 molar ratio of the enzyme
(~0.15mM) to substrate. The crystals were soaked in the reservoir solution with 20% glycerol as
cryoprotectant for ~1 min and frozen immediately in liquid nitrogen. X-ray data were collected at
100 K on beamline 22-ID of the Southeast Regional Collaborative Access Team (SER-CAT) at
the Advanced Photon Source, Argonne National Laboratory. Single-wavelength anomalous
144
diffraction (SAD) data for SeMet-DADH were collected at the wavelength of 0.97182 Å (high
energy remote) on beamline 22-BM of SER-CAT at Argonne National Laboratory.
Structure Determination and Model Refinement. The X-ray data were integrated and
scaled using HKL2000 (16). Program SGXPRO was used to perform phasing and automatic
tracing with scaled but unmerged SeMet-DADH data (17). With this approach, 331 of the 381
residues (375 amino acids residues of the enzyme and the N-terminal hexa-histidine tag) (~87%)
were successfully built. This model was then used for automated model building by ARP/WARP
(18), in which 96% of the structure was fit. The structure of SeMet-DADH was then employed to
solve the native DADH data set (1.06 Å) by molecular replacement using PHASER (19) in the
CCP4i suite of programs (20). Notably, all the FAD atoms were distinctly visible in the electron
density maps. FAD was refined with 100% occupancy. The structure of DADH was used to
solve the structures of DADH co-crystallized with D-arginine or D-histidine by molecular
replacement. Crystal structures were refined with SHELX97 (21). Manual adjustments and
rebuilding were performed using the molecular graphics program COOT (22). The structures of
DADH co-crystallized with D-arginine and D-histidine showed difference density for
iminoarginine and iminohistidine, respectively, bound at the enzyme active site, as observed for
D-amino acid oxidase crystallized in the presence of D-tryptophan (23). The D-amino acids are
converted enzymatically to imino acids (R-C=NH), which then dissociate from the enzyme and
are hydrolyzed to the final keto acids (R-C=O) in a non-enzymatic reaction. The iminoarginine
was refined with 100% occupancy, while two alternate conformations were visible for the
iminohistidine that refined to relative occupancies of 60% and 40%. Further analysis showed that
the structure of DADH, which was crystallized without added D-amino acids, contained extra
density for a low occupancy iminoarginine adjacent to a disordered loop showing two alternate
145
conformations. This structure was refined with two conformers corresponding to 70% occupancy
of a ligand-free, open conformation DADH, and 30% occupancy of iminoarginine bound to a
closed conformation of DADH. Higher peaks in the 2Fo-Fc electron density (>5σ) were
observed for the two main-chain oxygen atoms of iminoarginine, while smaller density peaks
(~1.3σ) were visible for other atoms. Therefore, in the ligand-free conformation, two water
molecules were refined near the carboxylate oxygen atoms of iminoarginine. Solvent molecules
were inserted at stereochemically reasonable positions based on the peak height of the 2Fo–Fc
and Fo–Fc electron density maps, hydrogen bond interactions and interatomic distances.
Anisotropic B-factors were refined for all the structures. Hydrogen atom positions were included
in the last stage of refinement using all data.
Sequence similarity searches were performed using BLAST (24). Protein structures were
superimposed on Cα atoms by using the secondary structure matching (SSM) module in COOT
(25). Figures of the structures were generated with PYMOL (http://www.pymol.org).
Enzyme Assay. The determination of the apparent steady state kinetic parameters was carried
out by measuring initial rates of reaction with an oxygen electrode at varying concentrations of
the D-amino acid substrate and a fixed concentration of 1 mM for PMS as electron acceptor. The
reaction mixture (1 mL) was first equilibrated with organic substrate and PMS at the desired
concentrations before the reaction was started by the addition of DADH. Enzyme assays were
conducted in 20 mM Tris-Cl, pH 8.7, 25 ̊C. All reaction rates are expressed per molar
concentration of enzyme-bound flavin.
146
5.5
Results
Overall Structure of DADH. The DADH was crystallized without the addition of FAD or
D-amino acid, and the structure was solved in the orthorhombic space group P212121with one
molecule per asymmetric unit using single anomalous dispersion (SAD) phasing and automated
tracing. The structure was refined to an R-factor of 13.2% at 1.06 Å resolution. Structures were
obtained also for DADH crystallized with substrates D-arginine or D-histidine under similar
conditions. The structures were solved in the same space group P212121 by using molecular
replacement with the DADH structure as a template, and were both determined at the resolution
of 1.30 Å and refined to R-factors of 13.4% and 12.8%, respectively. The crystallographic data
and refinement statistics are summarized in Table 5.1. In the three structures, all 375 DADH
residues and N-terminal hexa-histidine tag were defined clearly in the electron density map.
Further analysis of the N-terminus of the structure revealed that the His-tag residues form several
direct or water-mediated polar interactions with residues from symmetry related molecules,
which stabilize the flexible terminus. Overall, the three structures are almost identical with
pairwise RMSD values of 0.10-0.11 Å for 381 Cα atoms. The protein folds into two domains: an
FAD-binding domain and a substrate-binding domain (Figure 5.1). The FAD-binding domain
includes residues 1-82, 147-217 and 309-375. It consists of a central six-stranded β-sheet
surrounded by five α-helices on one side and a three-stranded anti-parallel β-sheet with two αhelices on the other side. The substrate-binding domain is formed by residues 83-142 and 218297, and consists of an eight-stranded β-sheet and two short anti-parallel β-strands forming a
sandwich surrounded by four α-helices. The substrate binding site was identified near the re face
of the FAD 7, 8-dimethyl-isoalloxazine ring at the interface of the two domains.
147
Table 5.1 Crystallographic Data Collection and Refinement Statistics
SeMet
DADH
DADH/
DADH
D-Arg
Data Collection Statistics
Redox state of flavin
Mixeda
Reduced
Wavelength (Å)
0.97182 b
0.99999
1.00000
Space group
P212121
P212121
P212121
a (Å)
62.65
62.18
62.17
b (Å)
78.21
78.08
78.43
c (Å)
89.47
89.72
89.95
Resolution range (Å)
50-1.50
50-1.06
50-1.30
Total observations
1016,247
1258,212
600,949
Unique reflections
70270
190,545
98,021
c
Completeness
99.7 (98.0)
96.2 (69.1)
90.6(53.9)
27.6(16.6)
26.5(2.3)
18.5(2.7)
<I/σ(I)>
Rsym (%)
7.6(18.1)
6.2(41.4)
8.2(37.3)
Figure of merit
0.72
f'
-4.2234
f''
3.7853
Refinement Statistics
Resolution range (Å)
10-1.06
10-1.30
13.2
13.4
Rwork (%)
Rfree (%)
15.9
16.6
Mean B-factor (Å2)
16.0
15.9
Protein
14.4
14.6
FAD
7.3
6.5
Ligand
29.5
37.0
Water
27.9
29.7
Number of atoms
Protein
3102
2961
FAD
53
53
Ligand
12
12
Water
467
358
r.m.s. deviations
Bond length (Å)
0.017
0.013
Angled
0.035
0.031
a
The DADH structure contains both ligand-free and product-bound conformations.
b
The SeMet DADH data were collected at the wavelength of high energy remote.
c
Values in parentheses are given for the highest resolution shell.
d
The angle RMSD in SHELX97 is indicated by distance in Å.
DADH/
D-His
Reduced
1.00000
P212121
62.10
78.15
89.59
50-1.30
627,618
97,110
90.4(54.7)
31.8(3.2)
4.9(30.6)
10-1.30
12.8
16.3
17.5
16.1
8.3
16.3
31.8
2971
53
22
367
0.012
0.030
148
Figure 5.1
5 Overall structure off DADH witth iminoargiinine. The D
DADH struccture is show
wn in
cartoon representatio
on. Iminoarrginine and cofactor F
FAD are shhown as sticks and coolored
magenta and red, resp
pectively. Loops that co
ontribute to thhe active sitte are coloredd green.
FAD
D-Binding Siite. DADH contains
c
a no
oncovalentlyy bound FAD
D cofactor ass identified iin the
X-ray cry
ystal structures. All non-hydrogen atoms
a
of the FAD were cclearly visibble in the eleectron
density map
m (Figuree 5.2A). Thee flavin adop
pts an elonggated conform
mation withh the adeninee ring
buried in
nside the FA
AD-binding domain
d
and the isoalloxxazine ring loocated at thee interface oof the
two dom
mains. The ad
denine portion of the co
ofactor pointts toward thhe flavin binnding domainn and
the 7, 8-d
dimethyl-iso
oalloxazine ring
r
is directted towards the interfacee of the two domains, siimilar
to the FA
ADs observed in other flavin-depend
dent enzymees, such as L
L-proline dehhydrogenase (26),
D-amino acid oxidasse (27) and L-amino
L
acid
d oxidase (228-29). As illlustrated in Figure 5.2B
B, the
ng moiety has
h hydrogeen bond inteeractions wiith the side chain of G
Glu32, two w
water
ribose rin
moleculees and the main
m
chain am
mide of Arg
g33. Numeroous hydrogeen bonds aree present bettween
the phosp
phate groups and the peeptidyl atom
ms of Ala14,, Thr42, Glyy331 and 5 water moleccules.
149
The centtral part of the
t conserveed glycine-riich sequencee GXGXXG
G in the N-teerminal region of
the polyp
peptide is paart of an α-heelix (residuees 13-25) thaat points tow
ward the phosphate groupp and
is assumeed to contrib
bute to the stabilization
s
of the two nnegative chaarges of the diphosphatee(30).
The 2'-O
OH and 4'-OH
H of the ribiityl moiety of
o the cofacctor establishh hydrogen bbond interacctions
with hyd
droxyl groups of Ser41, Thr42
T
and Ser45. The 3''-OH group interacts witth peptidyl aatoms
of Gly331 and Gly33
34.
Figure 5.2
5 (A) The 2Fo-Fc electron densitty map of F
FAD contourred at 2σ illlustrates the high
quality of
o the 1.06 Å resolution structure. FAD
F
(coloreed by elemennt type) adoopts an elongated
conformaation. (B) Scchematic diaagram of thee flavin-apopprotein interractions in thhe product-bbound
DADH conformation
c
n. Hydrogen
n bonds aree indicated bby dashed llines. The fflavin cofacttor is
colored blue.
b
For clarrity, H atom
ms are not sho
own.
hyl-isoalloxaazine is shaarply bent bbetween thee two planes containingg the
The 7, 8-dimeth
benzene and pyrim
midine moieeties, definiing a 15o angle alonng the N5--N10 axis. This
conformaation is in agreement with several other
o
crystalllographic sttructures of flavin-depenndent
enzymes in which th
he flavin is in the redu
uced state (331). The pyrrimidine porrtion of the 7, 8dimethyll-isoalloxazin
ne ring interracts throug
gh several hyydrogen bonnds with baackbone atom
ms of
the proteein (i.e. C2=
=O atom witth N atom of
o Ile335, G
Gln336, N3-H atom witth the O atoom of
150
His48 and a water molecule, and C4=O atom with the N atom of His48). The benzene portion of
the 7, 8-dimethyl-isoalloxazine ring is excluded from solvent and interacts with residues Arg44,
Ser45, Ala46, Arg222, Tyr249, Gly303 and Arg305.
Interactions of DADH with Iminoarginine. In order to investigate the structural basis for the
substrate recognition of DADH, the enzyme was crystallized with D-arginine that was converted
to iminoarginine. Iminoarginine binds with the plane formed by its Cα atom, carboxylate group
and imino group approximately parallel to the re face of the flavin, while the side chain points
away from the FAD (Figure 5.3A). The α-carbon of the iminoarginine is 3.3 Å away from the
FAD N5 atom, which is compatible with direct involvement of these atoms in the arginine
oxidation catalyzed by DADH. The iminoarginine imino group is pointed out of the plane of the
carboxylate atoms and can form a 2.9 Å long hydrogen bond interaction with the FAD O4 atom
(Figure 5.3B). Several DADH residues form extensive polar interactions with the two mainchain carboxyl oxygen atoms of iminoarginine that anchor the ligand in the active site. One of
the iminoarginine carboxyl oxygen forms hydrogen bonds with the side-chain hydroxyl of
Tyr53, the guanidinium side chain of Arg305, and the carbonyl oxygen of Gly332. The other
ligand carboxyl oxygen forms hydrogen bonds with the side-chain nitrogen of Arg222 and sidechain hydroxyl of Tyr249. Two water molecules are located near the ligand imino group, and
form a hydrogen bond network extending to the imidazole side chain of His48. The side chain of
iminoarginine forms hydrophobic interactions with the Val242 side chain, and a hydrogen bond
interaction with the hydroxyl of Thr50. A strong ionic interaction (2.5 Å distance between pairs
of O and N atoms) was observed between the guanidinium group of iminoarginine and the
carboxylate group of Glu87. Hence, Glu87 may play an important role in the high specificity of
151
DADH for D-arginine, in agreement with the kinetic analysis on substrate specificity of DADH
reported previously (14) and performed in this study (vide infra).
Interactions of DADH with Iminohistidine. The structure of DADH was also determined
for crystals grown in the presence of D-histidine, thereby yielding a DADHred/iminohistidine
complex. The iminohistidine binds to the DADH active site in two discrete conformations with
clear density for the overlapping Cα atoms and carboxylate groups and weaker density for the
alternate positions of the side chains (Figure 5.3C). The two iminohistidine conformations are
related by a rotation of approximately 180º. In fact, there might be two additional conformations
that are indistinguishable in this structure due to the potential 180º rotation of the imidazole ring.
The imidazole orientation showing more hydrogen bond interactions with DADH is illustrated in
Figure 5.3D. The main-chain atoms of conformation A (60% occupancy) lie nearly parallel to
the isoalloxazine ring of FAD, whereas its imidazole side chain is rotated by about 30º. The
iminohistidine atoms of conformation B (40% occupancy) all lie on the same plane. DADH
residues Tyr53, Arg222, Tyr249, Arg305, and Gly332 form conserved polar interactions with the
ligand main-chain oxygen atoms in both conformations (Figure 5.3D). The imino group of the
ligand forms a polar interaction with Gly332 in conformation A and with the hydroxyl of Tyr249
in conformation B. In both conformations the side chain of iminohistidine forms hydrogen bonds
with the side chain of His48 and the hydroxyl of Thr50. The imidazole group also interacts with
the side chain of Glu87 and the O4 of FAD in conformation A, and the side chain of Gln336 in
conformation B. When the imidazole is rotated by 180º the interactions with His48, Thr50 and
Gln336 are retained for conformation B. The interactions of conformation A with Glu87 and
FAD are lost, however, suggesting they are not critical since they appear in only one of four
possible conformations of iminohistidine.
152
Figure 5.3
5 DADH in
nteractions with
w iminoarginine (A, B
B) and iminoohistidine (C
C, D). Carbonns are
colored yellow
y
for DADH
D
activee site residuees and greenn for the iminno acids. FA
AD is represented
by its iso
oalloxazine ring
r
in mageenta. The Fo--Fc omit maap of the ligaand is indicaated as blue mesh
and conto
oured at 3σ. Dashed linees represent hydrogen
h
boonds and ionnic interactioons.
Confformationall Flexibility of the Activve Site. No s ubstrate or oother ligand was added tto the
protein solution
s
duriing crystallizzation of thee DADH. H
However, thee solved ligaand-free struucture
contained
d weaker deensity at thee substrate-b
binding site that was fiit by 30% ooccupancy oof the
153
iminoarginine reaction product. It is possible that the ligand was trapped during the bacterial
expression of the protein, as observed for other enzymes, for example, a tetrahedral reaction
intermediate was discovered in the crystal structure of bacterial Est30 (15). The iminoarginine
lies adjacent to residues 50-56 of loop L1 with two alternate conformations that were refined
with 0.7/0.3 relative occupancy (Figure 5.4A). This loop had a single well-defined conformation
in the electron density for structures of DADH with iminoarginine and iminohistidine that
corresponds to the higher occupancy conformation in the ligand-free structure. Therefore, the
crystal structure of DADH contains both ligand-free and product-bound conformations, which
suggests that a conformational change occurs in the adjacent loop regions upon binding of the
substrate. Comparison of the two conformations of DADH revealed distinct changes in two
regions of loop L1: the residues 50-56 adjacent to the imino product and the residues 45-47 near
the flavin ring of FAD.
One of the major conformational changes between the ligand-free and the product-bound
conformations in the DADH structure was observed at the substrate binding site for residues 5056 of loop L1 region (Figure 5.4B). Four loops (L1, residues 33-56; L2, 244-248; L3, 261-276;
L4, 329-336) form a flask-like substrate binding pocket. The binding pocket has a small entrance
but expands at the bottom near FAD. The iminoarginine product binds to DADH in a similar
manner as observed in the structures of pDAAO/iminotryptophan (23) and hDAAO/iminoserine
(32). In the ligand-free conformation, the side chain of Tyr53 points away from the active site
and forms a hydrogen bond with Thr137, whereas in the structure of the product-bound enzyme,
the aromatic ring of Tyr53 moves into the active site and forms hydrogen bonds with the
carboxyl oxygen of the imino acid and also the side chains of Glu246 and Arg305. Other
residues also form different interactions in the two conformations of this loop. In the ligand-free
154
conformation, the side chain of Thr50 forms hydrogen bonds with Asp39 and Ala52, while
Gly54 and Thr55 both form polar interactions with Arg59. In the product-bound conformation,
Thr50 interacts with Glu87 and Gln336, and its main-chain peptidyl oxygen forms a hydrogen
bond with the main-chain amide of Tyr53. The hydroxyl side chain of Thr55 forms a hydrogen
bond with Asp312. The interactions of Tyr53 and Thr50, in particular, stabilize the ligand bound
in the active site, suggesting that this flexible loop L1 region may act as a lid controlling
substrate accessibility to the active site. The shape and flexibility of the active site enables
DADH to control the entrance of substrates and accommodate a variety of substrates.
The second region showing significant differences in the ligand-free and product-bound
conformations comprised residues 45-47 at the N-terminal end of loop L1, which is located at the
si face of the flavin ring (Figure 5.4C). This region possesses two conformations in the DADH
structure and shows a single well-defined conformation in the structures of DADH with high
occupancy products. In the ligand-free conformation, the Ser45 hydroxyl group forms a
hydrogen bond with the flavin N5 atom, and the side chain of Ala46 is pointed away from FAD.
In the product-bound conformation, this region is flipped about 90º with the side chain of Ser45
pointed away from FAD and the Ala46 side chain approaching closer to the FAD. This structural
change causes the loss of the hydrogen bond interaction between Ser45 and FAD and, instead, a
new hydrogen bond is formed between the main-chain amide of Ala46 and the N5 of FAD. A
similar interaction was reported between the Ala49 of pDAAO and its FAD N5 atom (23, 27).
Due to the conformational changes of residues 50-56 and 45-47 in loop L1 region, the active site
of DADH shrinks by about 3.8 Å when the product is bound. The distances between the Cα
atoms of Ala46 and Tyr53 are 17.5 Å and 13.7 Å in the ligand-free conformation and productbound conformation, respectively (Figure 5.4B). Loop L2 also has a slight conformational
155
change and
a moves ab
bout 0.7 Å closer to loop
p L1 based oon the distannces betweenn the Cα atom
ms of
loop L1 Tyr53
T
and lo
oop L2 Asp2
245.
Figure 5.4
5 Conform
mational flexibility of DA
ADH structuure. (A) Thee DADH strructure compprises
two confformers: a 30% occupied population
n with iminooarginine prroduct, and 770% one wiithout
ligand. The
T iminoarrginine and its adjacentt loop regioon with low
w occupancyy (product-bbound
conformaation) are sh
hown as greeen sticks. The Fo-Fc omit map oof this regioon (red messh) is
contoured at 2.5σ.. The corrresponding loop regioon with hiigh occupanncy (ligandd-free
conformaation) is sho
own as blue sticks. (B) Comparison
C
of ligand-frree conformaation (greenn) and
product-b
bound confo
ormation (m
magenta) at loop L1 andd L2 regionss in DADH structure. A
Ala46,
Tyr53 an
nd iminoarg
ginine are reepresented as
a sticks. A hydrogen bond (blackk dotted linne) is
formed between Tyr53
T
and iminoargin
nine in thhe productt-bound connformation. (C)
Conform
mational chan
nge of residu
ues Ser45-A
Ala46-Ala47 at the si facce of flavin rring. Ser45 iin the
ligand-free conformaation (green carbon atom
ms) forms a 3.0 Å hydroogen bond w
with the FAD
D N5
ay
carbon
at
toms).
Ala46
6
forms
a
3.3
3
Å
long
hyd
drogen
bond
d
with
the
N5
5
atom
of
FA
AD
in
atom (gra
the produ
uct-bound co
onformation (yellow carb
bon atoms).
156
Substrate Specificity. The steady state kinetic parameters were determined by measuring
initial rates of reaction with various D-amino acids in 20 mM Tris-Cl, pH 8.7 at 25 °C. With both
D-arginine and D-histidine as substrate for DADH, the Km value for PMS is ~10 μM at pH 8.7;
moreover, with both substrates the enzyme displays a ping-pong bi-bi steady state kinetic
mechanism (H. Yuan and G. Gadda; unpublished results). The concentration of the electron
acceptor PMS was kept fixed at 1 mM PMS, ensuring saturation of the enzyme during steady
state turnover. The kcat values determined in this study were larger than those previously reported
(14), which were expressed per molar concentration of protein rather than enzyme-bound flavin.
As illustrated in Table 5.2 the enzyme showed substrate preferences for amino acids with
positively charged side chains, of which D-arginine appears to be the best substrate. D-lysine, Dtyrosine and D-methionine were also good substrates. In contrast, D-glutamate, D-aspartate, Larginine and glycine were not detectably oxidized by the enzyme. Cysteine could not be assayed
due to its non-enzymatic reaction with PMS in the enzyme reaction mixture.
DADH Recognition of Iminoarginine and Iminohistidine. Comparison of the structures of
DADH in complex with iminoarginine and iminohistidine reveals the molecular basis for the
relatively broad substrate specificity of this enzyme. The internal cavity of the substrate-binding
site has a triangular cross-section with a narrow entrance at the top, as shown in Figure 5.5
(DADH residues Tyr53, Gly332 and Gln336 were removed to view the internal site). The bottom
of the cavity extends about 14.9 Å from loop L1 to loop L3 and 14.1 Å from loop L2 to loop L4.
The depth of this substrate-binding pocket is about 10 Å from the Cα atom of Gly54 in loop L1
to the FAD O2 atom. The main chain of iminohistidine is located almost on the same plane and
parallel to the isoalloxazine ring of FAD, however, the side chain of iminoarginine lies almost
perpendicular to the isoalloxazine ring of FAD. The main-chain atoms of iminoarginine are very
157
close to those of im
minohistidinee, and form similar polaar interactioons with the DADH ressidues
Tyr53, Arg222,
A
Tyr249, Arg305
5, and Gly3
332. Since tthe binding pocket interractions witth the
main chaain of the su
ubstrate are conserved,
c
DADH
D
has a broad subsstrate specifiicity for a vaariety
of D-amiino acids, ass shown in Table
T
5.2, an
nd reported previously ((14). The diffferent side chain
interactio
ons with DA
ADH are prresumed to be responsiible for the specificity differences.. The
longer D-arginine
D
sid
de chain fitss well in the large substrrate-bindingg pocket of D
DADH, whille the
smaller D-histidine
D
cannot
c
fill the
t cavity, which
w
is connsistent withh the differeences in cataalytic
activity for
f the two su
ubstrates.
Figure 5.5
5 Comparisson of DAD
DH structuress in complexx with iminooarginine annd iminohistiidine.
DADH active
a
site residues (greeen sticks) line the interioor of the subsstrate bindinng pocket (T
Thr53,
Gly332 and
a Gln336 were
w omitted
d for clarity)). Iminoarginnine (red) annd iminohisttidine (blue)) bind
to the active site in very
v
distinctt conformatiions. FAD iss representedd by its isoaalloxazine riing in
green.
158
Table 5.2 Steady State Kinetics of DADHa
substrates
kcat/Km, M-1s-1
kcat, s-1
Km, mM
D-arginine
(3.4 ± 0.3) × 106
204 ± 3
0.06 ± 0.01
D-lysine
(5.3 ± 0.2) × 105
141 ± 3
0.26 ± 0.01
D-tyrosine
27,600±3,800
23±1
0.8 ± 0.1
D-methionine
14,800±600
154±3
10 ±1
D-phenylalanine
6,900±300
75±3
11±1
D-histidine
3,140 ± 30
35 ± 1
11 ± 1
D-leucine
515±60
6.4± 0.3
12± 1
D-proline
420 ± 10
-b
-
D-tryptophan
245 ± 3
-
-
D-isoleucine
195 ± 3
-
-
D-valine
47 ± 1
-
-
D-alanine
41 ± 1
-
-
D-glutamine
186 ± 3
-
-
D-asparagine
16 ± 1
-
-
D-serine
3.8 ± 0.1
-
-
D-threonine
0.75 ± 0.01
-
-
D-glutamate
-
-
-
D-aspatate
-
-
-
L-arginine
-
-
-
Glycine
-
-
-
D-cysteine
ndc
nd
nd
a
Enzyme activity was measured at varying concentrations of substrateand 1 mM PMS in 20 mM
Tris-Cl, pH8.7, at 25 oC. bCannot be saturated with the substrate, thereby kcat and Km values are
not reported.c Not determined. PMS was reduced by cysteine.
Structural Comparison with Related Proteins. Determination of the crystal structure of
DADH allows a detailed comparison with other mechanistically related flavin-dependent
oxidoreductases. Analysis of the sequences and structures showed that DADH shares low
159
sequence identity (<20%) but similar three-dimensional structures with L-Proline dehydrogenase
β subunit (18.5% sequence identity, RMSD 2.2 Å for 342 Cα atoms, (26)), heterotetrameric
sarcosine oxidase (18.4% sequence identity, RMSD 2.2 Å for 339 Cα atoms, (33)), porcine Damino acid oxidase pDAAO (17.2% sequence identity, RMSD 2.4 Å for 270 Cα atoms, (27)) and
L-amino acid oxidase LAAO (16.4% sequence identity, RMSD 3.0 Å for 245 Cα atoms, (29)).
For a detailed comparison of the overall structure and active site geometry, we have chosen the
structures of the enzyme complexes with dicarboxylate ligands: pDAAO from pig kidney in
complex with iminotryptophan (PDB entry 1DDO, (23)) and LAAO from Rhodococcus opacus
in complex with L-alanine (PDB entry 2JB1, (29)).
pDAAO and LAAO both are homodimeric enzymes formed by two subunits interacting
through their helical domain or substrate binding domain, respectively (27-29). Structural
superimposition indicates that the overall structure of DADH resembles the monomer structures
of pDAAO and LAAO (Figure 5.6). DADH and LAAO share the same topology for the FAD
binding domain in which a central β-sheet is surrounded by α-helixes on one side and a threestranded anti-parallel β-sheet on the other side. However, in the FAD-binding domain of
pDAAO, an α-helix is observed instead of the three-stranded anti-parallel β-sheet located in
DADH (Figure 5.6A). All three enzymes contain a large β-pleated sheet in the substrate binding
domain, while the orientation of this β-sheet in LAAO is significantly different from those in
DADH and pDAAO. Two extra β-strands are observed for LAAO at the Gly57-Cys70 region
(Figure 5.6B). Furthermore, the helical domain involved in dimerization of LAAO is absent
from the structures of DADH and pDAAO.
160
5 Structuraal comparison of DADH
H with pDAA
AO (A) and LAAO (B). DADH is shhown
Figure 5.6
in green cartoon rep
presentation while pDA
AAO and LA
AAO are coolored in ligght blue withh red
regions indicating th
heir major strructural diffferences com
mpared to DA
ADH. FAD is representeed by
sticks in
n green and
d light blue correspond
ding to the compared structures. T
The ligandss are:
iminoarg
ginine in DA
ADH (mageenta), imino
otryptophan in pDAAO
O (orange) aand L-alaninne in
LAAO (o
orange).
The substrate
s
bin
nding sites off these threee enzymes arre all locatedd near the ree face of the FAD
isoalloxaazine ring att the interfacce of the FA
AD-binding and substraate-binding ddomains. Deespite
the opposite orientation of L-alaanine bound to LAAO ccompared to the bindingg mode of liggands
in the DA
ADH and pD
DAAO comp
plexes, simillar protein-liigand interacctions are rettained amonng the
three enzzymes. The specific
s
arran
ngement of active
a
site reesidues is suuggested to bbe responsible for
the strict enantiosellectivity of each enzy
yme (23, 277-29). In thhe pDAAO//iminotryptoophan
complex,, the carboxy
ylate group of
o the ligand
d forms a sallt bridge witth the guaniddinium side chain
of Arg28
83 (Figure 5.7A).
5
Simillar interactio
ons are form
med betweenn L-alanine aand Arg84 iin the
structure of LAAO/L
L-alanine (F
Figure 5.7B)). However, in the DAD
DHred/iminoaarginine com
mplex,
this ionicc interaction
n is replaced
d by interacttions betweeen the ligandd and two aarginines: Arrg222
161
and Arg3
305. The caarboxylate oxygen
o
atom
ms of the liggands also fform polar interactions with
Tyr53, Tyr249
T
and Gly332
G
of DA
ADH and Ty
yr224, Tyr228 and Gly3313 of pDAA
AO, respectiively.
L-alaninee forms simiilar interactiions with the polar residdues Gln2288 and Tyr371 of LAAO. The
side chaains of thesse ligands form
f
distincct interactioons with ennzyme residdues. In DA
ADH
iminoarg
ginine forms strong ionic interaction
ns with Glu887 and a hyydrogen bonnd with Thr550. In
contrast, in pDAAO
O the indole side chaain of iminnotryptophann is surrounnded by seeveral
hydropho
obic residuees (Ala49, Leu51,
L
Ile215, Ile230, aand Tyr228). The short side chain of Lalanine in
nteracts with
h several hyd
drophobic reesidues Trp4426, Ala466 and Trp467 of LAAO. T
These
similaritiies and diffeerences in in
nteractions are
a presumeed to be impportant deterrminants off their
different substrate sp
pecificities.
5 Compariison of the active
a
sites of DADH, pDAAO annd LAAO. T
The active siite of
Figure 5.7
DADH (yellow)
(
is superimpose
s
ed on that of
o pDAAO ((A, lightbluee) and LAA
AO (B, lighttblue)
along th
he FAD iso
oalloxazine ring. Imino
oarginine in DADH (m
magenta carrbon atoms)) and
iminotryp
ptophan in pDAAO
p
(cy
yan carbon atoms)
a
and L
L-alanine in LAAO (cyaan) are show
wn as
sticks. FA
AD is repressented by its isoalloxazin
ne ring.
162
5.6
Discussion
An “Active Site Lid” Controls Substrate Accessibility. Interestingly, the atomic resolution
DADH crystal structure demonstrates two conformations corresponding to the ligand-bound and
ligand-free forms. A loop covering the active site shows a substantial conformational change
between the two forms. A similar loop region described as an “active site lid” was reported in the
structure of pDAAO (residues 216-228, (27)).It is hypothesized that this lid switches between
closed and open conformations to allow substrate binding and product release (27). A similar
“loop-and-lid” structure has been assigned in some of the glucose-methanol-choline (GMC)
family members, including glucose oxidase (residues 76-97, (34)), cholesterol oxidase (residues
95-109, (35)), pyranose 2-oxidase (residues 452-456, (34)), and the flavoprotein domain of
cellobiose dehydrogenase (residues 289-299 (36)). Besides its role as a gate in opening and
closing the enzyme active site, the active site lid may be important in determining the substrate
specificity of DADH. The conformational change of the lid, especially for Tyr53, may allow
DADH to accommodate bulky residues like D-phenylalanine or D-tryptophan. Similar
interactions have been observed in pDAAO (Tyr224, (23)) and flavocytochrome b2 (Tyr254,
(37)). Furthermore, closing of the lid shields the active site and the FAD from solvent. An
increase of the active site hydrophobicity caused by loop closure may facilitate the hydride
transfer step leading to substrate oxidation in pDAAO (23). Similar phenomena have been
described in several NAD(P)H-dependent dehydrogenases (38). Further insight into the function
of this active site lid may be obtained by protein engineering studies on DADH.
Substrate Specificity. DADH is characterized by a broad substrate specificity, being able to
oxidize basic and hydrophobic D-amino acids of various size (14). In the crystal structures,
iminoarginine forms a strong ionic interaction with the side chain of Glu87 and a hydrogen bond
163
with Thr50, whereas the side chain of iminohistidine extends in a different direction and forms
hydrogen bonds with His48, Thr50 and Gln336. Indeed, the steady state kinetics data show a
large kcat/KArg value of 106 M-1s-1 that is 1,000-fold higher than the kcat/KHis value (Table 5.2).
Kinetic data for other D-amino acids indicate that the negatively charged side chain of Glu87 is
the major determinant for the specificity of DADH for the positively charged substrates Darginine and D-lysine. Another example is seen in trypsin: residue Asp189 is responsible for its
narrow selection for positively charged arginine and lysine (39-40). In addition, the hydrophobic
walls (Tyr53, Met240, Val 242 and Tyr249) of the DADH active site pocket create a favorable
environment for the long aliphatic and unbranched parts of the basic D-arginine and D-lysine.
Furthermore, D-tyrosine, D-methionine and D-phenylalanine, which are good substrates of
DADH, may form favorable van der Waals contacts with the hydrophobic walls of the active site.
In contrast, DADH shows low or undetectable activities toward several D-amino acids,
especially the negatively charged D-glutamate and D-aspartate.
The Ser/Ala Switch in the FAD Binding Site. The Ser45-Ala47 region also has two
conformations corresponding to the ligand-free and product-bound conformations in the DADH
structure, while this region has a well-defined single conformation in the structures of
DADHred/iminoarginine and DADHred/iminohistidine. A hydrogen bond (OSer45—N5FAD) in the
ligand-free conformation is replaced by another polar interaction (NAla46—N5FAD) in the productbound conformation (Figure 5.4C). Furthermore, interactions of the FAD N5 atom with a
residue structurally equivalent to Ala46 of DADH are conserved among different FADdependent enzymes, such as pDAAO (Ala49, (27)), yeast D-amino acids oxidase (Gly52, (41)),
L-proline dehydrogenase (Gly48, (26)), flavocytochrome b2 (Ala198, (37)), and L-galactono-γlactone dehydrogenase (Ala113, (42)). Human DAAO shares the same sequence (V47AAGL51)
164
with pDAAO near the si face of FAD ring. However, hDAAO shows a conformational shift in
this region, which has been suggested to be responsible for its low binding affinity for FAD as
well as the slower rate of flavin reduction (32). Furthermore, substitution of the structurally
equivalent Ala113 with Gly in L-galactono-γ-lactone dehydrogenase increased the reactivity of
the reduced flavin with oxygen by about 400-fold (42). Therefore, the Ser/Ala switch in DADH
might be involved in structural stability and also the enzymatic properties of this protein. Further
studies are required to elucidate its function.
Structural Comparison with Related Proteins. Structural comparison of DADH with
pDAAO and LAAO revealed that the active site residues of DADH are more similar to those of
pDAAO (Figure 5.7), in agreement with their specificity for D-amino acids rather than L-amino
acids (14). The spatial arrangement of active residues is critically related to the enzyme
enantiomeric selectivity. A mirror-symmetrical relationship of active sites has been described
among enzymes with opposite stereospecificities and two different modes have been reported.
One is observed between pDAAO and flavocytochrome b2, in which the active sites of the two
enzymes are mirrored through the plane of isoalloxazine (on the re side in pDAAO, on the si side
of flavocytochrome b2) (27). Another situation is that the substrate binding sites are mirrored
through the plane perpendicular to the isoalloxazine ring, which is observed in the comparison
between pDAAO and LAAO (28). The active site of DADH is highly similar to that of pDAAO,
and mirror-symmetrically related to that of LAAO (Figure 5.7). This observation is in agreement
with the kinetic study described previously (14) and in this paper showing that DADH can
oxidize a variety of D-amino acids but not L-arginine. A similar active site arrangement has been
observed in other flavoenzymes like glycolate oxidase (44), L-glutamate oxidase (45) and L-
165
aspartate oxidase (46), suggesting these enzymes employ a common mechanism to control
enantioselectivity.
In addition to their specificity toward different enantiomers of amino acids, DADH, pDAAO
and LAAO display other distinct features. For example, the ligands are held in the active site by
a salt bridge formed with an arginine in pDAAO and LAAO, while two active site arginines are
involved in DADH. The active sites of DADH and pDAAO are covered by a loop termed the
“active site lid”. This feature was not reported in the structure of LAAO, and a funnel-shaped
entrance for substrate binding was proposed instead (47). In addition to the polar interactions
between the ligands and the enzyme active site residues, a hydrophobic environment has been
observed in all three enzymes. In fact, all the enzymes interact with a wide range of hydrophobic
amino acids (14, 23, 48). Overall, the different composition and arrangement of active site
residues determine differences in substrate specificity, while critical interactions for catalysis are
conserved among these mechanistically related enzymes.
Conclusions. Comparison of the crystal structures of DADH and its complexes with
iminoarginine and iminohistidine has highlighted important structural differences that rationalize
the catalytic activities and substrate specificity of the enzyme. The imino products of D-arginine
and D-histidine bind to the active site in very distinct side-chain conformations. Glu87 forms
strong electrostatic interactions with iminoarginine, which is likely responsible for the high
selectivity of DADH for positively charged residues like D-arginine and D-lysine. A loop region
has been designated as an active site lid controlling the substrate accessibility to the active site,
similar to those reported in other flavin-dependent enzymes. Structural comparison of DADH
with other related flavin-dependent enzymes reveals that the spatial arrangement of active site
residues is essential for the differences in enzyme enantioselectivity, while some specific
166
interactions needed for catalysis are conserved among the enzymes. Overall, the high-resolution
structures for DADH described in this study will provide new guidelines for future studies of
similar flavin-dependent enzymes.
5.7
Acknowledgment
We thank Dr. Johnson Agniswamy and Dr. Yuan-Fang Wang for providing help with
refinement and valuable discussions. We are especially grateful for the assistance of Dr. ZhengQing Fu and the staff at the SER-CAT beamline at the Advanced Photon Source, Argonne
National Laboratory, for assistance during X-ray data collection. Use of the Advanced Photon
Source was supported by the U.S. Department of Energy, Office of Science, Office of Basic
Energy Sciences, under Contract No. W-31-109-Eng-38.
5.8
1.
References
Pollegioni, L., Piubelli, L., Sacchi, S., Pilone, M. S., and Molla, G. (2007) Physiological
functions of D-amino acid oxidases: from yeast to humans, Cell Mol Life Sci 64, 13731394.
2.
Kolodkin-Gal, I., Romero, D., Cao, S., Clardy, J., Kolter, R., and Losick, R. (2010) DAmino Acids Trigger Biofilm Disassembly, Science 328, 627-629.
3.
Yoshimura, T., and Esak, N. (2003) Amino acid racemases: functions and mechanisms, J
Biosci Bioeng 96, 103-109.
4.
Dunlop, D. S., Neidle, A., McHale, D., Dunlop, D. M., and Lajtha, A. (1986) The
presence of free D-aspartic acid in rodents and man, Biochem Biophys Res Commun 141,
27-32.
167
5.
Hashimoto, A., Nishikawa, T., Oka, T., and Takahashi, K. (1993) Endogenous D-serine
in rat brain: N-methyl-D-aspartate receptor-related distribution and aging, J Neurochem
60, 783-786.
6.
Snyder, S. H., and Kim, P. M. (2000) D-amino acids as putative neurotransmitters: focus
on D-serine, Neurochem Res 25, 553-560.
7.
Wolosker, H. (2007) NMDA receptor regulation by D-serine: new findings and
perspectives, Mol Neurobiol 36, 152-164.
8.
Olsiewski, P. J., Kaczorowski, G. J., and Walsh, C. (1980) Purification and properties of
D-amino acid dehydrogenase, an inducible membrane-bound iron-sulfur flavoenzyme
from Escherichia coli B, J Biol Chem 255, 4487-4494.
9.
Raunio, R. P., and Jenkins, W. T. (1973) D-alanine oxidase form Escherichia coli:
localization and induction by L-alanine, J Bacteriol 115, 560-566.
10.
David, H. L., Clavel-Seres, S., Clement, F., Lazlo, A., and Rastogi, N. (1989) Methionine
as methyl-group donor in the synthesis of Mycobacterium avium envelope lipids, and its
inhibition by DL-ethionine, D-norleucine and DL-norleucine, Acta Leprol 7 Suppl 1, 7780.
11.
Haas, D., Matsumoto, H., Moretti, P., Stalon, V., and Mercenier, A. (1984) Arginine
degradation in Pseudomonas aeruginosa mutants blocked in two arginine catabolic
pathways, Mol Gen Genet 193, 437-444.
12.
Lu, C. D. (2006) Pathways and regulation of bacterial arginine metabolism and
perspectives for obtaining arginine overproducing strains, Appl Microbiol Biotechnol 70,
261-272.
168
13.
Li, C., and Lu, C. D. (2009) Arginine racemization by coupled catabolic and anabolic
dehydrogenases, Proc Natl Acad Sci U S A 106, 906-911.
14.
Li, C., Yao, X., and Lu, C. D. (2010) Regulation of the dauBAR operon and
characterization of D-amino acid dehydrogenase DauA in arginine and lysine catabolism
of Pseudomonas aeruginosa PAO1, Microbiology 156, 60-71.
15.
Liu, P., Wang, Y. F., Ewis, H. E., Abdelal, A. T., Lu, C. D., Harrison, R. W., and Weber,
I. T. (2004) Covalent reaction intermediate revealed in crystal structure of the
Geobacillus stearothermophilus carboxylesterase Est30, J Mol Biol 342, 551-561.
16.
Otwinowski, Z., and Minor, W. (1997) Processing of X-ray diffraction data collected in
oscillation mode, Methods Enzymol 267, 307-326.
17.
Fu, Z. Q., Rose, J., and Wang, B. C. (2005) SGXPro: a parallel workflow engine enabling
optimization of program performance and automation of structure determination, Acta
Crystallogr D Biol Crystallogr 61, 951-959.
18.
Lamzin, V. S., and Wilson, K. S. (1993) Automated refinement of protein models, Acta
Crystallogr D Biol Crystallogr 49, 129-147.
19.
McCoy, A. J., Grosse-Kunstleve, R. W., Storoni, L. C., and Read, R. J. (2005)
Likelihood-enhanced fast translation functions, Acta Crystallogr D Biol Crystallogr 61,
458-464.
20.
Potterton, E., Briggs, P., Turkenburg, M., and Dodson, E. (2003) A graphical user
interface to the CCP4 program suite, Acta Crystallogr D Biol Crystallogr 59, 1131-1137.
21.
Sheldrick, G. M., and Schneider, T. R. (1997) SHELXL: high-resolution refinement,
Methods Enzymol 277, 319-343.
169
22.
Emsley, P., and Cowtan, K. (2004) Coot: model-building tools for molecular graphics,
Acta Crystallogr D Biol Crystallogr 60, 2126-2132.
23.
Todone, F., Vanoni, M. A., Mozzarelli, A., Bolognesi, M., Coda, A., Curti, B., and
Mattevi, A. (1997) Active site plasticity in D-amino acid oxidase: a crystallographic
analysis, Biochemistry 36, 5853-5860.
24.
Altschul, S. F., Madden, T. L., Schaffer, A. A., Zhang, J., Zhang, Z., Miller, W., and
Lipman, D. J. (1997) Gapped BLAST and PSI-BLAST: a new generation of protein
database search programs, Nucleic Acids Res 25, 3389-3402.
25.
Krissinel, E., and Henrick, K. (2004) Secondary-structure matching (SSM), a new tool
for fast protein structure alignment in three dimensions, Acta Crystallogr D Biol
Crystallogr 60, 2256-2268.
26.
Tsuge, H., Kawakami, R., Sakuraba, H., Ago, H., Miyano, M., Aki, K., Katunuma, N.,
and Ohshima, T. (2005) Crystal structure of a novel FAD-, FMN-, and ATP-containing
L-proline dehydrogenase complex from Pyrococcus horikoshii, J Biol Chem 280, 3104531049.
27.
Mattevi, A., Vanoni, M. A., Todone, F., Rizzi, M., Teplyakov, A., Coda, A., Bolognesi,
M., and Curti, B. (1996) Crystal structure of D-amino acid oxidase: a case of active site
mirror-image convergent evolution with flavocytochrome b2, Proc Natl Acad Sci U S A
93, 7496-7501.
28.
Pawelek, P. D., Cheah, J., Coulombe, R., Macheroux, P., Ghisla, S., and Vrielink, A.
(2000) The structure of L-amino acid oxidase reveals the substrate trajectory into an
enantiomerically conserved active site, EMBO J 19, 4204-4215.
170
29.
Faust, A., Niefind, K., Hummel, W., and Schomburg, D. (2007) The Structure of a
Bacterial l-Amino Acid Oxidase from Rhodococcus opacus Gives New Evidence for the
Hydride Mechanism for Dehydrogenation, Journal of Molecular Biology 367, 234-248.
30.
Dym, O., and Eisenberg, D. (2001) Sequence-structure analysis of FAD-containing
proteins, Protein Sci 10, 1712-1728.
31.
Dixon, D. A., Lindner, D. L., Branchaud, B., and Lipscomb, W. N. (1979) Conformations
and electronic structures of oxidized and reduced isoalloxazine, Biochemistry 18, 57705775.
32.
Kawazoe, T., Tsuge, H., Pilone, M. S., and Fukui, K. (2006) Crystal structure of human
D-amino acid oxidase: context-dependent variability of the backbone conformation of the
VAAGL hydrophobic stretch located at the si-face of the flavin ring, Protein Sci 15,
2708-2717.
33.
Chen, Z. W., Hassan-Abdulah, A., Zhao, G., Jorns, M. S., and Mathews, F. S. (2006)
Heterotetrameric sarcosine oxidase: structure of a diflavin metalloenzyme at 1.85 A
resolution, J Mol Biol 360, 1000-1018.
34.
Wohlfahrt, G., Witt, S., Hendle, J., Schomburg, D., Kalisz, H. M., and Hecht, H. J.
(1999) 1.8 and 1.9 A resolution structures of the Penicillium amagasakiense and
Aspergillus niger glucose oxidases as a basis for modelling substrate complexes, Acta
Crystallogr D Biol Crystallogr 55, 969-977.
35.
Li, J., Vrielink, A., Brick, P., and Blow, D. M. (1993) Crystal structure of cholesterol
oxidase complexed with a steroid substrate: implications for flavin adenine dinucleotide
dependent alcohol oxidases, Biochemistry 32, 11507-11515.
171
36.
Hallberg, B. M., Henriksson, G., Pettersson, G., and Divne, C. (2002) Crystal structure of
the flavoprotein domain of the extracellular flavocytochrome cellobiose dehydrogenase, J
Mol Biol 315, 421-434.
37.
Xia, Z. X., and Mathews, F. S. (1990) Molecular structure of flavocytochrome b2 at 2.4
A resolution, J Mol Biol 212, 837-863.
38.
Stoll, V. S., Kimber, M. S., and Pai, E. F. (1996) Insights into substrate binding by D-2ketoacid dehydrogenases from the structure of Lactobacillus pentosus D-lactate
dehydrogenase, Structure 4, 437-447.
39.
Lu, W., Apostol, I., Qasim, M. A., Warne, N., Wynn, R., Zhang, W. L., Anderson, S.,
Chiang, Y. W., Ogin, E., Rothberg, I., Ryan, K., and Laskowski, M. (1997) Binding of
amino acid side-chains to S1 cavities of serine proteinases, Journal of Molecular Biology
266, 441-461.
40.
Brandsdal, B. O., Aqvist, J., and Smalas, A. O. (2001) Computational analysis of binding
of P1 variants to trypsin, Protein Sci 10, 1584-1595.
41.
Umhau, S., Pollegioni, L., Molla, G., Diederichs, K., Welte, W., Pilone, M. S., and
Ghisla, S. (2000) The x-ray structure of D-amino acid oxidase at very high resolution
identifies the chemical mechanism of flavin-dependent substrate dehydrogenation, Proc
Natl Acad Sci U S A 97, 12463-12468.
42.
Leferink, N. G., Fraaije, M. W., Joosten, H. J., Schaap, P. J., Mattevi, A., and van Berkel,
W. J. (2009) Identification of a gatekeeper residue that prevents dehydrogenases from
acting as oxidases, J Biol Chem 284, 4392-4397.
172
43.
Wagner, M. A., Trickey, P., Chen, Z. W., Mathews, F. S., and Jorns, M. S. (2000)
Monomeric sarcosine oxidase: 1. Flavin reactivity and active site binding determinants,
Biochemistry 39, 8813-8824.
44.
Lindqvist, Y. (1989) Refined structure of spinach glycolate oxidase at 2 A resolution, J
Mol Biol 209, 151-166.
45.
Arima, J., Sasaki, C., Sakaguchi, C., Mizuno, H., Tamura, T., Kashima, A., Kusakabe,
H., Sugio, S., and Inagaki, K. (2009) Structural characterization of L-glutamate oxidase
from Streptomyces sp. X-119-6, FEBS J 276, 3894-3903.
46.
Bossi, R. T., Negri, A., Tedeschi, G., and Mattevi, A. (2002) Structure of FAD-bound Laspartate oxidase: insight into substrate specificity and catalysis, Biochemistry 41, 30183024.
47.
Moustafa, I. M., Foster, S., Lyubimov, A. Y., and Vrielink, A. (2006) Crystal structure of
LAAO from Calloselasma rhodostoma with an L-phenylalanine substrate: insights into
structure and mechanism, J Mol Biol 364, 991-1002.
48.
Massey, V., and Curti, B. (1967) On the reaction mechanism of Crotalus adamanteus Lamino acid oxidase, J Biol Chem 242, 1259-1264.
173
CHAPTER 6
Steady State Kinetic Mechanism and Redutive Half-Reaction of D-Arginine
Dehydrogenase from Pseudomonas aeruginosa
(This chapter has been published verbatim in Yuan, H., Fu, G., Brooks, P., Weber, I. and Gadda,
G. (2010), Biochemistry 49: 9542-9550. The author contributed to this chapter by carrying out
all of the experiments and directing the research of Mr. Phillip Brooks in the initial purification
of the untagged enzyme. The crystallographic figures were prepared by Mr. Guoxing Fu.)
6.1
Abbreviations
dNTPs, deoxynucleotide triphosphate; IPTG, isopropyl-β-D-thiogalactopyranoside; PMS,
phenazine methosulfate; PMSF, phenylmethylsulfonyl fluoride; DEAE, Diethyl-Amino-Ethyl.
6.2
Abstract
D-Arginine dehydrogenase from Pseudomonas aeruginosa catalyzes the oxidation of Darginine to iminoarginine, which is hydrolyzed in solution to ketoarginine and ammonia. In the
present study, we have genetically engineered an untagged form of the enzyme that was purified
to high levels and characterized in its kinetic properties. The enzyme is a true dehydrogenase that
does not react with molecular oxygen. Steady state kinetics studies with D-arginine or Dhistidine as substrate and PMS as the electron acceptor established a ping-pong bi-bi kinetic
mechanism. With the fast substrate D-arginine a dead-end complex of the reduced enzyme and
the substrate occurs at high concentrations of D-arginine yielding substrate inhibition, while the
overall turnover is partially limited by the release of the iminoarginine product. With the slow
substrate D-histidine the initial Michaelis complex undergoes an isomerization involving
multiple conformations that are not all equally catalytically competent for the subsequent
oxidation reaction, while the overall turnover is at least partially limited by flavin reduction. The
174
kinetic data are interpreted in view of the high-resolution crystal structures of the iminoarginineand iminohistidine-enzyme complexes.
6.3
Introduction
D-Arginine dehydrogenase is a novel flavin-dependent enzyme with FAD as a cofactor (1). It
was recently isolated from Pseudomonas aeruginosa, a Gram-negative soil bacterium that is well
known to be an important opportunistic human pathogen. The enzyme from P. aeruginosa has
been cloned and expressed in Escherichia coli (2). It catalyzes the oxidation of D-arginine to
iminoarginine, which is non-enzymatically hydrolyzed to α-ketoarginine and ammonia, as
illustrated in Scheme 6.1 (2-4). Coupled with an anabolic NAD(P)H-dependent L-arginine
dehydrogenase, the products of the D-arginine dehydrogenase oxidation of D-arginine, αketoarginine and ammonia, are converted into L-arginine (1). The proposed physiological
function of D-arginine dehydrogenase in Pseudomonas is to contribute the first stage in a twoenzyme coupled system to racemize D-arginine in the D- to L-arginine conversion (1).
unknown e- acceptor
COO
R
H
NH3
R=
FAD FADH2
R
COO
+ H2O
R
NH2
H
N
NH2
COO + NH4
O
non-enzymatic
NH2+
Scheme 6.1 Oxidation of D-Arginine by D-Arginine Dehydrogenase.
The biochemistry of D-amino acid catabolism has been poorly studied in comparison to that
of L-amino acids. In general, D-amino acids can be metabolized after conversion into the Lenantiomers by racemase (5). Alternatively, D-amino acids can be utilized by D-amino acid
oxidase, which is an FAD-dependent enzyme and plays an important role in microbial
175
metabolism (6). D-Amino acid oxidase was first studied by Krebs in 1935 (7) and is now well
characterized. The enzyme has FAD noncovalently bound to the polypeptide and exhibits
optimal activity toward neutral D-isomers of amino acids and lower efficiency toward basic ones
(8). The reduced FAD is then reoxidized by molecular oxygen to yield hydrogen peroxide (8).
D-Arginine dehydrogenase is characterized by broad substrate specificity, being able to
oxidize D-amino acids of various size and polarity (9). All naturally occurring D-amino acids
except for D-glutamate, D-aspartate and glycine are oxidized by the enzyme, with the best
substrate, D-arginine, displaying the highest kcat/Km with a value of 3 x 106 M-1s-1 (9). D-Arginine
dehydrogenase was first described in 1988 by Hass et al. (2). Recently, the dauA gene encoding
for D-arginine dehydrogenase was cloned and used to express a His-tagged version of the
enzyme that was characterized crystallographically at high resolutions (i.e., ≤1.3 Å) in complex
with either iminoarginine or iminohistidine (1, 9). The physiological electron acceptor of the
enzyme has not been identified yet (1).
In the present study, we have genetically engineered an untagged form of D-arginine
dehydrogenase, purified it to high levels and investigated its steady state kinetic mechanism and
reductive half-reaction with the physiological substrate D-arginine and with D-histidine. The
kinetic data presented are interpreted in view of the high resolution structures of D-arginine
dehydrogenase reported recently (1, 9).
6.4
Experimental Procedures
Materials. The plasmid pCR3 harboring the dauA gene encoding for D-arginine
dehydrogenase was a gift from Prof. Chung-Dar Lu at Georgia State University. E. coli strain
Rosetta(DE3)pLysS was from Novagen (Madison, WI). NdeI, BamHI, alkaline phosphatase, T4
DNA ligase, dNTPs, BSA, lambda DNA/EcoR I + Hind III markers and MgCl2 were purchased
176
from Promega (Madison, WI); cloned Pfu DNA polymerase was from Stratagene (La Jolla, CA);
QIAprep and QIAquick purification kits were from Qiagen (Valencia, CA); Luria-Bertani agar,
Luria-Bertani broth, chloramphenicol, IPTG, lysozyme, sodium hydrosulfite (dithionite), Darginine, phenazine methosulfate (PMS) and PMSF were obtained from Sigma-Aldrich (St.
Louis, MO). Ampicillin, agarose and electrophoresis-grade agar were purchased from ICN
Biomedicals (Aurora, OH). Oligonucleotides were from Sigma Genosys (The Woodlands, TX).
D-Histidine was from Alfa Aesar (Ward Hill, MA). All of the other reagents were of the highest
purity commercially available.
Plasmid Construction. Plasmid pCR3 harboring the dauA gene encoding for D-arginine
dehydrogenase (1) was employed as template to construct a recombinant plasmid for the
expression of untagged enzyme using standard PCR amplification methods. Forward and reverse
oligonucleotide primers were designed incorporating NdeΙ and BamHΙ as the 5’ and 3’ cloning
sites, respectively. The PCR product was then ligated into the expression vector pET20b(+)
cleaved with the same restriction enzymes to ensure directional cloning to remove the N-terminal
His-tag. The resulting plasmid pET/PA3863 was confirmed by sequencing and transformed into
E. coli strain DH5α.
Gene Expression and Enzyme Purification. Permanent frozen stocks of E. coli cells
Rosetta(DE3)pLysS harboring the plasmid pET/PA3863 were used to inoculate 50 mL of LuriaBertani broth medium containing 50 μg/mL ampicillin and 34 μg/mL chloramphenicol and
cultures were grown at 37 °C for 16 hours to be used as a preculture. The starting precultures
(48 mL) were used to inoculate 7.5 L Luria-Bertani broth medium containing 50 μg/mL
ampicillin and 34 μg/mL chloramphenicol. When the cultures reached optical densities of ~0.6 at
177
600 nm, the temperature was lowered to 18 °C and IPTG was added to a final concentration of
0.1 mM. After 18 hours the cells were harvested by centrifugation at 20,000 g for 20 min at 4 °C.
All purification steps were carried out at 4 °C. The wet cell paste was suspended in 0.1 mM
PMSF, 0.2 mg/mL lysozyme, 1 mM EDTA, 10% glycerol and 20 mM Tris-Cl, pH 8.0, in a ratio
1 g wet cell paste to 4 mL of lysis buffer. The suspended cells were then allowed to incubate
with stirring for 30 min on ice with 20 μg/mL RNase and 50 μg/mL DNase in the presence of 10
mM MgCl2. The resulting slurry was sonicated 5 times for 5 min, with 2 min intervals, before
removing the cell debris by centrifugation at 20,000 g for 20 min. Solid ammonium sulfate was
slowly added to the cell free extract to achieve 30% saturation. After incubating for 30 min on
ice, the insoluble fraction was removed by centrifugation and discarded. The supernatant was
brought up to 65% ammonium sulfate saturation, and the pellet fraction was collected by
centrifugation after 30 min of stirring. The resulting pellet was suspended in 20 mM Tris-Cl, pH
8.0, 10% glycerol, and dialyzed over a period of 18 hours with four buffer changes. After
dialysis, the precipitated proteins were removed by centrifugation at 20,000 g for 20 min and the
supernatant was loaded directly onto a DEAE-Sepharose Fast Flow column (3 × 28 cm),
equilibrated with 20 mM Tris-Cl pH 8.0, 10% glycerol. The column was eluted with 2 volumes
of the same buffer, followed by a linear gradient from 0 to 0.5 M NaCl over 1 liter. The fractions
with the highest purity as judged by enzymatic activity and UV-visible absorbance spectroscopy
were pooled together and concentrated with the addition of 65% ammonium sulfate saturation
followed by centrifugation. After centrifugation, the resulting pellet was suspended in 20 mM
Tris-Cl, pH 8.7, 10% glycerol, and dialyzed against four changes of the same buffer. After
removal of the precipitated protein by centrifugation, the enzyme was stored at -20 °C.
178
Enzyme Assay. The concentration of D-arginine dehydrogenase was determined with the
method of Bradford (10) by using the Bio-Rad protein assay kit with BSA as standard. The
enzymatic activity of D-arginine dehydrogenase was measured by monitoring the initial rate of
oxygen consumption with a computer-interfaced Oxy-32 oxygen-monitoring system (Hansatech
Instrument) at 25 °C in which PMS (1 mM) was used as the primary electron acceptor and the
enzymatically reduced PMS was spontaneously reoxidized by molecular oxygen. The reaction
was started with the addition of D-arginine dehydrogenase to 1 mL reaction mixture, with the
final concentration of 9.9 nM enzyme and 20 mM D-arginine. One unit of enzymatic activity
corresponds to the consumption of one μmol of oxygen per min.
First-order rate constants for flavin reduction were determined at varying concentrations of
D-arginine (0.1-0.5 mM) or D-histidine (2.5-100 mM) in 20 mM Tris-Cl, pH 8.7, using a
stopped-flow spectrophotometer thermostated at 25 °C. Equal volumes of the enzyme and Darginine or D-histidine were mixed anaerobically in the stopped-flow spectrophotometer
following established procedures (11) yielding a final enzyme concentration of ~10 µM. The
stopped-flow traces were not changed when the experiment was carried out aerobically.
The determination of the steady state kinetic parameters was carried at varying
concentrations of D-arginine (0.02-2 mM) or D-histidine (1-70 mM) and PMS (0.005-0.5 mM),
in 20 mM Tris-Cl, pH 8.7, 25 °C. The reaction mixture (1 mL) was first equilibrated with the
substrates at the desired concentrations before the reaction was started with the addition of the
enzyme to a final concentration of ~ 0.01 μM. Enzyme assays were conducted in 20 mM Tris-Cl,
pH 8.7. Initial rates of reaction were expressed per molar content of enzyme-bound flavin.
Solvent viscosity effects on steady state kinetic parameters were measured in 20 mM Tris-Cl, pH
179
8.7, 25 °C using glycerol as viscosigen. The values for the relative viscosities of glycerolcontaining solutions were taken from Weast (12) and adjusted for 25 °C.
Data Analysis. Data analysis was carried out by using KaleidaGraph software (Synergy
Software, Reading, PA), Enzfitter software (Biosoft, Cambridge, UK) or the Kinetic Studio
Software Suite (Hi-Tech Scientific, Bradford on Avon, U. K.). Stopped-flow traces were fit to
eq. 1, which describes a single exponential process, where kobs represents the observed first-order
rate constant for flavin reduction at any given concentration of substrate, t is time, A is the
absorbance at 446 nm at any given time, B is the amplitude of the absorbance change, and C is
the absorbance at infinite time. Kinetic parameters for the reductive half-reactions were
determined by using eq. 2, where kobs is the observed first-order rate constant for the reduction of
the enzyme-bound flavin at any given concentration of substrate (S), kred is the limiting firstorder rate constant for flavin reduction at saturating concentrations of substrate, and
app
Kd is the
apparent dissociation constant for binding of the substrate to the enzyme.
A = B exp(−k obs t ) + C
(1)
k red S
app
Kd + S
(2)
kobs =
The steady state kinetic parameters at varying concentrations of both D-arginine and PMS
were determined by fitting the initial rate data to eq. 3, which describes a ping-pong mechanism
with substrate inhibition (13). Data with D-histidine were fit with eq. 4, which describes a pingpong steady state kinetic mechanism. In these equations, e represents the concentration of
enzyme, kcat is the turnover number of the enzyme at saturating concentrations of both the amino
acid and PMS, Ka and Kb represent the Michaelis constants for the amino acid (A) and PMS (B),
respectively, and Ki-A is the substrate inhibition constants for D-arginine.
180
v
kcat ABK i− A
=
e K a BK i− A + K b AK i− A + ABK i− A + K b A2
(3)
v
kcat AB
=
e K a B + Kb A + AB
(4)
The effects of solvent viscosity on the kcat and kcat/KArg values were fit to eq. 5; those on the
kcat/KHis values were fit to eq. 6. In these equations (k)o and (k)η are the kinetic parameters of
interest in the absence and presence of the viscosigen, S is the degree of viscosity dependence,
and ηrel is the relative viscosity of the aqueous buffered solution, YH is the limiting value of the
steady state kinetic parameter of interest at high concentration viscosigen.
(k ) O
= S (η rel − 1) + 1
(k )η
(k ) O
=
( k )η
6.5
1
Y × (η rel − 1)
1+ H
(η rel − 1) + S
(5)
(6)
Results
Expression and Purification of Untagged Enzyme. Recombinant D-arginine dehydrogenase
was expressed in E. coli strain Rosetta(DE3)pLysS and purified via 30-65% ammonium sulfate
precipitation and anionic-exchange chromatography at pH 8.0 in the presence of 10% glycerol.
The purity of the enzyme was confirmed by SDS-PAGE analysis. The UV-visible absorbance
spectrum of the purified enzyme showed bands at 365 nm and 446 nm (data not shown),
consistent with the presence of oxidized flavin bound to the enzyme. Table 6.1 summarizes the
purification procedure.
181
Table 6.1 Purification of Untagged P. aeruginosa D-Arginine Dehydrogenase Expressed in E.
coli Rosetta(DE3)pLysS
total protein total activity (µmol specific activity (µmol
yield
steps
-1
-1
-1
(mg)
O2min )
O2min mg )
(%)
cell-free extract
5,600
91,200
16
100
30-65% saturation of
4,700
100,000
21
110
(NH4)2SO4
DEAE-Sepharose
900
43,000
48
47
Enzymatic activity was measured with 20 mM D-arginine and 1 mM PMS as substrates in airsaturated 20 mM Tris-Cl, pH 8.7 and 25 oC.
Oxygen Reactivity. The reactivity of D-arginine dehydrogenase with molecular oxygen was
investigated by monitoring oxygen consumption over time upon mixing 0.3 µM enzyme with 0.5
mM D-arginine in a Clark-type oxygen electrode at pH 8.7 and 25 °C. As shown in Figure 6.1,
in the absence of any organic electron acceptor there was no detectable consumption of oxygen.
In contrast, upon addition of 1 mM PMS there was a rapid depletion of oxygen in the enzyme
reaction mixture (i.e., with an estimated vo/e value of ~800 s-1 with 0.5 mM D-arginine). Similar
results were obtained with D-lysine, D-histidine, D-serine, D-threonine, D-tyrosine, Dasparagine, D-glutamine, D-alanine, D-valine, D-leucine, D-isoleucine, D-phenylalaine, Dmethionine, D-proline, and D-tryptophan as a substrate (data not shown). These data
unequivocally establish the enzyme as a true dehydrogenase with very poor, if any, reactivity of
the reduced flavin with molecular oxygen.
182
Figure 6.1
6 Lack of reactivity of D-argininee dehydrogeenase with m
molecular oxxygen. The assay
mixture contained
c
20
0 mM Tris-C
Cl, pH 8.7, 0.5 mM D--arginine, at 25 oC. Afteer monitorinng the
backgrou
und for 1 min
n, 0.3 µM D-arginine
D
deehydrogenasse was addedd to enzymee reaction miixture
and the oxygen
o
tracee was monittored. 1.5 min
m later, 1 m
mM PMS w
was added annd the rate oof the
enzymatiic reaction was
w monitoreed.
duction with
h D-Arginin
ne or D-Hiistidine. Thee reductive halfTimee-Resolved Flavin Red
reactionss in which the
t enzyme--bound flavin is reduceed with D-aarginine or D-histidine were
investigaated in a stopped-flow spectrophoto
s
ometer at pH
H 8.7 and 255 °C. With D
D-arginine uunder
condition
ns of pseudo
o-first orderr (i.e., 10 µM
M enzyme aand ≥ 50 µM
M D-argininne after mixxing),
~70% off the decreasse in absorbaance at 446 nm occurredd within thee dead-time oof the instruument
(i.e., 2.2 ms). From the final parrt of the red
duction proc ess that couuld be monittored, similaar kobs
values ~7
700 s-1 weree determined
d with 50, 10
00, or 500 µ
µM D-arginiine. These ddata did not aallow
for an accurate determ
mination of the kinetic parameters
p
k red and Kd, bbut are at least consistentt with
fl
reducttion (kred ≥ 700 s-1) aand suggest a Kd valuee for D-argginine
a fast prrocess of flavin
significan
ntly lower th
han 50 μM.. With D-hisstidine as a substrate, thhe decrease in absorbannce at
446 nm associated with
w the red
duction of th
he enzyme-bbound flavinn was monoophasic at alll the
183
concentraations of su
ubstrate and
d fit best to a single exxponential pprocess (Figgure 6.2A).. The
observed
d rate constants were hyperbolical
h
lly dependennt on the cconcentration of D-histtidine
(Figure 6.2B),
6
allow
wing for the determinatio
d
on of the ratee constant for flavin reduuction (kred = 60 ±
1 s-1) and the appareent equilibriium constan
nt for substraate dissociaation at the aactive site oof the
enzyme (appKd = 10 ± 1 mM). Th
he best fit of the data w
was obtained with the cuurve extrapollating
to a y-inttercept valuee of zero, con
nsistent with
h an irreversiible reductioon of the flavvin.
Figure 6.2
6 Anaerobiic reduction of the D-arg
ginine dehyddrogenase w
with D-histiddine as a substrate
in 20 mM
M Tris-Cl, pH
H 8.7 and 25
5 oC. Panel A shows thee reduction ttraces with 22.5 mM (cyaan), 5
mM (blu
ue), 10 mM (red), 20 mM
m (light green), 50 mM
M (black) annd 100 mM
M (dark greenn) Dhistidine. All traces were
w
fit with
h eq 1. Timee indicated i s after the ennd of the floow, i.e. 2 mss. For
clarity on
ne experimeental point every
e
10 is shown
s
(verttical lines). P
Panel B shoows the obseerved
rate of flaavin reductio
on as a functtion of D-hisstidine conceentration. Daata were fit tto eq 2.
Stead
dy State Kin
netic Mecha
anism. The steady statee kinetic meechanism annd the assocciated
kinetic parameters off the enzymee were deterrmined usingg the methodd of the initial rates (14)). The
rate of ox
xygen consu
umption wass measured at
a varying cconcentrationns of either D-arginine oor Dhistidine and the artificial electrron acceptorr PMS, in 2 0 mM Tris--Cl at pH 8..7 and 25 °C
C. As
illustrated in Figure 6.3, parallell lines were obtained in double recipprocal plots of the enzym
matic
rate versu
us the conceentration of the amino acid
a
substratee, with signnificant substtrate inhibitiion at
concentraations of D-arginine
D
≥0.5
≥
mM. Substrate
S
innhibition waas also seeen with PM
MS at
184
concentrations of D-arginine equal to, or lower than, 50 μM (data not shown). Nonetheless, the
best fit of the data with D-arginine was obtained by using eq. 3, which describes a ping-pong bibi steady state kinetic mechanism with inhibition by D-arginine (13). The best fit of the data with
D-histidine was obtained by using eq. 4, consistent with a ping-pong bi-bi steady state kinetic
mechanism and no substrate inhibition. As summarized in Table 6.2, the overall rate of
enzymatic turnover determined with D-arginine at saturating concentrations of both substrates
(kcat) was ~200 s-1, a value that was 6-fold larger than the kcat value of 35 s-1 determined when Dhistidine was used as substrate. The second-order rate constant for the capture of D-arginine to
yield enzyme-substrate complexes committed to catalysis (i.e., kcat/KArg) was in the 106 M-1s-1
range, which was 850-fold larger than the corresponding value of 4,000 M-1s-1 determined with
D-histidine. These data establish that D-histidine is significantly slower than D-arginine as a
substrate.
Solvent Viscosity Effect. The effects of solvent viscosity on the kcat and kcat/Km values with
D-arginine and D-histidine were investigated to determine whether diffusion-controlled events
influenced the binding of the substrate and the release of the product to and from the enzyme.
With D-arginine as substrate, when the reciprocals of the normalized kcat/Km values determined at
increasing concentrations of glycerol were plotted as a function of the relative viscosity, the data
yielded a line with negligible slope (i.e. 0.004 ± 0.022) (Figure 6.4A). In contrast, a line with
slope of 0.14 ± 0.02 was obtained in a plot of the normalized kcat value (Figure 6.4A). With Dhistidine, the normalized kcat values were independent of solvent viscosity; in contrast, the kcat/Km
values increased with increasing viscosity yielding an inverse hyperbolic pattern in a plot of the
(kcat/Km)o/(kcat/Km)ŋ values versus relative viscosity (Figure 6.4B). These data establish that
185
kinetic stteps involvin
ng substrate binding and
d product rellease are affe
fected in diffferent fashions by
solvent viscosity
v
wheen the enzym
me turns oveer with D-argginine or D-hhistidine.
6 Steady sttate kinetics for the oxid
dation D-argginine (Panell A) or D-hiistidine (Pannel B)
Figure 6.3
catalyzed
d by D-argiinine dehyd
drogenase. In
nitial rates of reactionn were meassured at vaarying
concentraations of bo
oth the amino acid substtrate and PM
MS in 20 mM
M Tris-Cl, ppH 8.7, at 225 oC.
Panel A: Concentratiions of PMS
S were (○) 50
00 µM; (■) 100 µM; (□)) 20 µM; (▲) 10 µM andd (∆)
5 µM. Daata were fit to
t eq. 3. Pan
nel B: Conceentrations off PMS were (■) 100 µM
M; (□) 50 µM
M; (▲)
10 µM; and
a (∆) 5 µM
M. Data weree fit to eq. 4.
Figure 6.4
6 Effects of
o solvent visscosity on th
he steady staate kinetic pparameters fo
for the D-argginine
dehydrog
genase with D-arginine or D-histidin
ne as a subsstrate. Panel A shows thhe normalizeed kcat
(●) and kcat/Km (■) values with D-arginine
D
ass a function of the relatiive solvent vviscosity. Pannel B
shows th
he normalized kcat (●) and
d kcat/Km (■)) values withh D-histidinee as a functiion of the rellative
solvent viscosity.
v
Th
he dashed liines with a slope of 1 iindicate the expected reesults for a fully
diffusion
n-limited reaaction. The values
v
for th
he relative viiscosities off the solvent were taken from
o
Weast (12) and adjussted for 25 C.
C Reaction rates were m
measured at varying conncentrations of Dne and fixed
d 1 mM PMS
S in 20 mM T
Tris-Cl, pH 8.7, at 25 oC
C.
arginine or D-histidin
186
Table 6.2 Kinetic Parameters of D-Arginine Dehydrogenase with D-Arginine and D-Histidine a
kinetic parametersb
D-arginine
D-histidine
kcat, s-1
204 ± 3
35 ± 3
Ka, mM
0.06 ± 0.01
8.8 ± 0.5
Kb, µM
11 ± 2
8.0 ± 0.6
kcat/Ka, M-1s-1
(3.4 ± 0.3) × 106
4,000 ± 300
kcat/Kb, M-1s-1
(1.9 ± 0.3) × 107
(4.4 ± 0.3) × 106
Ki-A, mM
0.9 ± 0.3
noc
kred, s-1
≥ 700d
60 ± 1
Kd, mM
≤ 0.05e
10 ± 1
a
Enzyme activity was measured at varying concentrations of both D-arginine or D-histidine and
PMS in 20 mM Tris-Cl, pH 8.7, at 25 oC. b Kinetic data are readily accounted for with the kinetic
mechanism of Scheme 6.2. c Not observed. d Estimated lower limiting value. e Estimated upper
limiting value. The reaction occurs within the dead time of the spectrometer, thus cannot be
accurately measured by stopped-flow techniques.
6.6
Discussion
Steady State Kinetic Mechanism. The steady state kinetic mechanism of D-arginine
dehydrogenase has been determined with D-arginine and D-histidine as substrates at pH 8.7 and
is consistent with the minimal ping-pong bi-bi mechanism illustrated in Scheme 6.2. After the
formation of the Michaelis complex DADHox•S, the amino acid substrate is oxidized to the
corresponding imino acid with concomitant reduction of the enzyme-bound flavin (DADHred•P).
Release of the imino product from the active site of the enzyme completes the reductive halfreaction. The oxidative half-reaction with the artificial electron acceptor PMS occurs through the
initial formation of a Michaelis complex DADHred•PMS, within which catalysis yields the
DADHox•PMSH2 complex. Turnover is then completed with the release of reduced PMS from
the oxidized enzyme. Evidence in support of a ping-pong kinetic mechanism comes from the
steady state kinetic data determined at varying concentrations of D-arginine or D-histidine and
PMS, and from the substrate inhibition of enzymatic turnover at high concentrations of D 187
arginine. First, the parallel lines observed in the plots of the reciprocal of the initial rates of
reaction as a function of the reciprocal of the substrate concentrations is consistent with the
irreversible step of product release (k5) occurring before binding of PMS to the reduced enzyme
(k7). D-Arginine inhibition of turnover is readily explained with the formation of a dead-end
complex DADHred•Arg when D-arginine at high concentrations binds to the free reduced enzyme
(in blue in Scheme 6.2). In agreement with a ping-pong kinetic mechanism where PMS reacts
with the reduced enzyme after the iminoarginine product of the reaction is released from the
enzyme active site, the KPMS had similar values irrespective of whether D-histidine or D-arginine
was the substrate, namely ~10 µM (Table 6.2).
DADHox-S
S, k1
k3
k2
k4
DADHox
PMSH2
DADHred-P
k5
k11
P
DADHred
DADHox-PMSH2
k8
k9
Arg, k13
k14
DADHred-Arg
PMS,k7
DADHred-PMS
Scheme 6.2 Proposed Steady State Kinetic Mechanism for the Oxidation of D-Arginine
Catalyzed by D-Arginine Dehydrogenase. DADHox, oxidized D-arginine dehydrogenase; S,
substrate; DADHred, reduced D-arginine dehydrogenase; P, imino product; PMSH2, reduced
phenazine methosulfate. Arg, D-arginine; k5, k11 are shown as irreversible because initial rate are
measured in the absence of products; k4, although shown, is close to zero as suggested by
stopped-flow kinetic data (see text); k9 is shown as irreversible based on the oxidation-reduction
potential of PMS/PMSH2 > FAD/FADH2, with values of +80 mV and -200 mV (60),
respectively.
In principle, a kinetic pattern with parallel lines could also be obtained with PMS binding to
a DADHred•P complex formed by an irreversible conversion of the DADHox•S complex (15). If
188
this were the case, D-arginine inhibition of turnover would require the formation of either a
ternary complex of the enzyme with two D-arginines or a quaternary complex of the enzyme,
PMS and two D-arginines. Both these possibilities appear very unlikely since the available
crystallographic structure of the enzyme co-crystallized with D-arginine shows that the active
site cavity of the enzyme is almost entirely occupied with a single iminoarginine, thereby
preventing the binding of a second arginine (9). Inhibition of catalytic turnover due to formation
of a dead-end complex between the substrate and the reduced form of the enzyme was previously
observed in other flavoproteins, such as nitroalkane oxidase (16), aldehyde oxidase (17),
flavocytochrome b2 (18), cellobiose dehydrogenase (19) and thymidylate synthase (20).
Substrate Binding. Binding of the D-amino acid substrate to the oxidized enzyme occurs in
rapid equilibrium. This establishes that the rate constant for the dissociation of the D-amino acid
from the DADHox•S complex destined for catalysis (k2) must be significantly larger than the rate
constant for its oxidation to yield the products of the reaction (k3). Evidence to support this
conclusion comes from the solvent viscosity effects on the second-order rate constants for the
capture of D-arginine or D-histidine onto the enzyme to yield complexes committed to catalysis
(kcat/KArg and kcat/KHis) (Figure 6.4). These data are readily explained if one considers that a line
with a slope between zero and +1 would be expected in a plot of the normalized kcat/Km values as
a function of increasing relative viscosity of the solvent for a reductive half-reaction whose
overall rate was at least partially controlled by the diffusion of the substrate in the active site of
the enzyme (21). Such a linear dependence of the normalized kcat/Km values on solvent viscosity
was not observed with D-arginine dehydrogenase, for which the solvent viscosity effect was
either negligible with a slope not significantly different from zero in the case of D-arginine or
inverse and hyperbolic in the case of D-histidine (Figure 6.4). The lack of a solvent viscosity
189
effect on the kcat/KArg value is somewhat surprising because large kcat/Km values in the range of
106 M-1s-1 like the one determined here with D-arginine are commonly associated with some
degree of diffusion-control for the binding of substrates to enzymes (22-25).
Substrate binding to the oxidized enzyme in rapid equilibrium immediately establishes the
app
(Kd) values determined in the stopped-flow spectrophotometer as the true dissociation
constants reporting on the thermodynamic equilibrium of the free oxidized enzyme and the
Michaelis enzyme-substrate complex (i.e., Kd = k2/k1) (Scheme 6.2). Although an accurate
determination of the Kd value with D-arginine could not be obtained due to the reaction being too
fast to be followed with a stopped-flow spectrophotometer, the rapid kinetic data on the
reductive half-reaction suggest a Kd value for D-arginine lower than 50 μM. Thus, binding of Darginine in the active site of the enzyme is at least 200-fold tighter than that of D-histidine, with
a Kd value of 10 mM. Tighter binding of D-arginine to D-arginine dehydrogenase as compared to
D-histidine is primarily due to the favorable electrostatic interaction of the guanidinium side
chain of D-arginine with the active site residue Glu87; such an interaction is not present with Dhistidine, as suggested by the structures of the enzyme in complex with iminoarginine and
iminohistidine recently reported (Figure 6.5) (9). The importance of an electrostatic interaction
involving a glutamate residue for selective binding of the substrate in the active site of a
flavoenzyme was recently established in choline oxidase through mechanistic and structural
studies (11). In that case, site directed mutagenesis studies or the use of substrate analogs devoid
of charge were consistent with a ~15 kJ mol-1 energetic contribution of the side chain of Glu312
to the ionic interaction with the positively charged choline (11). A comparable energetic
contribution to substrate binding is likely provided by Glu87 in D-arginine dehydrogenase, as
suggested by the at least 200-fold ratio of the Kd values estimated for D-arginine and D-histidine.
190
6 Structural comparisson of activ
ve sites of D
D-arginine ddehydrogenaase/iminoargginine
Figure 6.5
(yellow) and D-argiinine dehydrrogenase/im
minohistidinee (grey). Acctive site residues and FAD
moleculee are represen
nted as stick
ks. The proteein main chaain is shown as green coiil.
Isomerization off the Michaeelis Complex
x with D-Hiistidine. Afteer the initiall formation oof the
Michaeliis complex involving th
he oxidized enzyme an d D-histidinne, the DAD
DHox•His sppecies
isomerizees to produce an enzym
me-substratee complex D
DADHox•Hiss* that is coompetent foor the
subsequeent reaction of flavin red
duction (Sch
heme 6.3). E
Evidence forr this concluusion comes from
the effecct of increassing solvent viscosity on
n the normaalized kcat/K
KHis values ddetermined iin the
presence of glycero
ol. The kcat/K
KHis values increased hhyperbolicallly to a lim
miting value with
increasin
ng viscosity of the solv
vent (Figuree 6.4), conssistent with the presencce of an intternal
equilibriu
um of the emzyme-subsstrate compllex in the reeductive hallf-reaction. M
Moreover, iin the
crystal structure of D-arginine dehydrogen
nase in com
mplex with iiminohistidinne, the ligaand is
present in
i two confformations with
w
respect to the flaavin (Figuree 6.5) (9). One of thee two
conformaations presen
nts the imino group of iminohistidin
i
ne in the sam
me orientation relative tto the
191
flavin 7,8-dimethyl-isoalloxazine that is seen in the crystal structure of the enzyme in complex
with iminoarginine (Figure 6.5) (9). This conformation is likely to be competent for catalysis in
the enzyme-substrate complex. The alternative conformation most likely is not competent for the
subsequent reaction of flavin reduction, as suggested by the relative orientation of groups
participating in the hydride transfer reaction. Thus, the isomerization of the DADHox•His
complex can be readily explained as reflecting the reversible conversion of multiple binding
conformations, not all of which are catalytically competent for the subsequent oxidation reaction
involving the flavin. The isomerization of the DADHox•S complex is not observed with Darginine as substrate. This stems from a more optimal binding mode of D-arginine as compared
to D-histidine, which also exploits the extra interaction of the positively charged side chain with
the active site Glu87 (vide ante) (Figure 6.5) (9). Isomerizations of Michaelis complexes prior to
the reaction of flavin reduction have been previously observed in the wild type and selected
mutant forms of flavocytochrome b2 (26) and in a mutant form of choline oxidase where the
active site Glu312 is replaced with aspartate (11).
DADHox
His, k1
k2
DADHox-His
Kiso
DADHox-His*
k3
k4
DADHred-P
k5
P
DADHred
Scheme 6.3 Isomerization of the Michaelis Complex with D-Histidine. DADHox, oxidized Darginine dehydrogenase; His, D-histidine; DADHred, reduced D-arginine dehydrogenase; P,
iminohistidine.
Flavin Reduction. The oxidation of the amino acid substrate catalyzed by D-arginine
dehydrogenase entails the two-electron, irreversible reduction of the enzyme-bound flavin
without formation of any observable reaction intermediates. Evidence for this conclusion comes
from the rapid kinetic data on the reductive half-reaction determined in a stopped-flow
spectrophotometer, yielding a monophasic decrease in the absorbance of FAD when the enzyme
is mixed with D-histidine as substrate. Irreversibility of the flavin reduction (i.e., k4 being close
192
to zero in Scheme 6.2) is established from the extrapolation to the origin of the hyperbolic
dependence of the kobs values as a function of the concentration of the D-histidine substrate (27).
Lack of observable reaction intermediates implies that if they were formed they would have to
decay at rates that are at least 20-times faster than their rates of formation. Although alternate
mechanisms for flavin reduction cannot be ruled out at this stage, these observations are
consistent with the mechanism for substrate oxidation by hydride transfer that has been
previously proposed for other flavoproteins acting on amino acids, such as D-amino acid oxidase
(28, 29) and sarcosine oxidase (29), α-hydroxy acids, such as flavocytochrome b2 (26, 30, 31) or
alcohol substrates, such as choline oxidase (32, 33), aryl alcohol oxidase (34), and pyranose 2oxidase (35).
The oxidation of D-arginine by D-arginine dehydrogenase is at least 20-times faster than the
oxidation of D-histidine, as indicated by the rate constants for flavin reduction estimated upon
mixing the enzyme and the substrate in the stopped-flow spectrophotometer. The larger kred value
seen with D-arginine likely originates from a better orientation and positioning in the active site
of D-arginine with respect to D-histidine, due to the favorable interaction of its side chain with
Glu87 (Figure 6.5) (9). The importance of substrate orientation and positioning for efficient
oxidation of an organic molecule have been recently established in another flavoprotein, choline
oxidase, where it was shown that the conservative replacement of an active site Glu with Asp
results in a 230-fold decrease in the kred value for the oxidation of the alcohol substrate (11).
Product Release. With D-arginine as substrate, the release of iminoarginine from the active
site of the enzyme is partially rate limiting for the overall turnover of the enzyme. This
conclusion is supported by the effect of increasing solvent viscosity on the normalized rate
constant for the overall turnover number of the enzyme at saturating concentrations of D-
193
arginine and PMS (i.e. kcat). The reaction would be slower in solvents with higher viscosities if a
diffusion-controlled process is the rate-limiting step (36). Since product release is the only
diffusive step during turnover when the enzyme is saturated with substrates, a line with a slope
of 0.14 as observed in Figure 6.4 indicates that the release of the product is partially rate limiting
in turnover. In contrast, enzymatic turnover with D-histidine is not limited by the rate of product
release from the enzyme active site, as suggested by the lack of solvent viscosity effects on the
kcat value with D-histidine. Instead, the overall turnover of the enzyme with D-histidine is at least
partially limited by the kinetic step of flavin reduction, as suggested by the comparison of the kred
value of 60 s-1 determined by using rapid reaction kinetics and the kcat value of 35 s-1 determined
by using the steady state kinetic approach.
Lack of Oxygen Reactivity. D-Arginine dehydrogenase is a true dehydrogenase that reacts
very poorly, if at all, with molecular oxygen. This conclusion is supported by the lack of oxygen
consumption observed upon mixing the enzyme with D-arginine or with 15 other D-amino acids
at pH 8.7 as determined by using a Clark-type oxygen electrode. In contrast, an initial rate of
~800 s-1 was observed upon addition of 1 mM PMS as an electron acceptor to the same reaction
mixture, identifying the enzyme as a true dehydrogenase.
The analysis of the crystal structure of the enzyme and its comparison with that of D-amino
acid oxidase, which readily reacts with oxygen (i.e., kcat/Koxygen = 105 M-1s-1) (37), helps to
rationalize the lack of oxygen reactivity in D-arginine dehydrogenase. First, the methyl side
chain of Ala46 appears to physically block access of oxygen to the reactive C(4a) atom of the
flavin in the active site of D-arginine dehydrogenase (Figure 6.6). The corresponding residue in
D-amino acid oxidase is Gly52 (with the numbering of the enzyme from Rhodotorula gracilis)
(38). When Gly52 is replaced with a bulkier Val the resulting mutant enzyme loses the ability to
194
react witth oxygen du
ue to steric hindrance preventing
p
oxxygen accesss to the redduced flavin (38).
The impo
ortance of permitting ph
hysical accesss of oxygenn to the C(4aa) atom of thhe flavin hass also
been sho
own with sitee-directed mutagenesis
m
in
i L-galactoono-γ-lactonee dehydrogeenase (39, 400). In
the wild type enzymee, which reaacts poorly with
w oxygen,, access to thhe C(4a) atom
m of the flavvin is
blocked by
b the side chain of Ala113. Howeever, replaceement of this residue wiith a smallerr Gly
yields a 400-fold
4
inccrease of the kcat/Koxygen to
t 3.4 x 105 M-1s-1, a vaalue that is tyypically fouund in
oxidases (39).
Figure 6.6
6 Active siite of D-argiinine dehydrrogenase (PD
DB number 3NYE) (9). For clarity,, only
selected amino acid residues aree shown. Th
he FAD cofaactor is show
wn as a sticck representaation.
The side chain of Alaa46 is shown
n in green.
Lack of oxygen reactivity in
n D-argininee dehydroge nase is alsoo most likelyy associatedd with
the absen
nce of positiv
ve charges in
n close prox
ximity of thee C(4a) and N
N(1)-C(2) attoms of the fflavin.
In this reegard, the po
ositive charg
ge of the sub
bstrate/produ
duct of D-argginine dehyddrogenase caan be
immediattely ruled out since thee ping-pong steady statee kinetic meechanism esttablishes thaat the
reaction of
o the reduceed flavin witth the electro
on acceptor occurs on thhe enzyme affter the releaase of
195
the reaction product from the enzyme. The positive charge of Arg305 is the closest to both the
C(4a) and N(1)-C(2) atoms of the flavin, at distances ≥7 Å (Figure 6.6). In some oxidases the Nterminal end of an α-helix providing a partial positive charge either reinforces or substitutes for
the effect of the full positive charge proximal to the flavin N(1)-C(2) atoms, as exemplified by
D-amino acid oxidase (41-43), cholesterol oxidase (44), and polyamine oxidase (45). However,
in D-arginine dehydrogenase the closest α-helix is 6 Å away from the N(1)-C(2) atoms of the
flavin and points in the opposite direction (Figure 6.6). These positive charges, which are typical
in several flavoprotein oxidases (46-52), exert two independent roles. The positive charge close
to the flavin C(4a) atom has been proposed to facilitate the stabilization of the negatively
charged superoxide species that transiently forms in the reaction of the reduced flavin with
oxygen. This has been established in a number of flavoprotein oxidases by using mechanistic and
structural approaches, among which are glucose oxidase (53), monoamine oxidase (54, 55),
monomeric sarcosine oxidase (56) and choline oxidase (57, 58). The positive charge close to the
N(1)-C(2) atoms of the flavin has been proposed to provide electrostatic stabilization of the
anionic hydroquinone form of the reduced flavin, which is a common feature in flavoprotein
oxidases but not in dehydrogenases (59).
Conclusions. In summary, we have engineered an untagged form of D-arginine
dehydrogenase, purified it to high levels and investigated its steady state kinetic mechanism and
reductive half-reaction with the substrates D-arginine and D-histidine. The results of the kinetic
investigation show that the enzyme is a true dehydrogenase that does not react in its reduced
state with molecular oxygen. With both amino acid substrates the enzyme displays a ping-pong
bi-bi kinetic mechanism when PMS is used as electron acceptor. Flavin reduction is partially rate
limiting for the overall turnover of the enzyme with the slow substrate D-histidine but not with
196
the fast substrate D-arginine. With the latter, release of the iminoarginine product of the
oxidation of D-arginine from the active site of the enzyme is partially rate limiting for the overall
turnover of the enzyme. An isomerization of the Michaelis complex is also established with Dhistidine prior to the flavin kinetic step, likely reflecting the conversion of multiple binding
conformations that are not all catalytically competent for the subsequent flavin reduction. The
kinetic investigation of the authentic untagged form of D-arginine dehydrogenase reported
herein, along with the recent structural investigation of the iminoarginine- and iminohistidineenzyme complexes at high resolutions, will provide a solid framework for future mutagenesis
and mechanistic studies aimed at the characterization of the enzyme.
6.7
Acknowledgment
The authors thank Prof. Chung-Dar Lu and Dr. Congran Li for the kind gift of plasmid pCR3
harboring the dauA gene encoding for D-arginine dehydrogenase. We thank Dr. Andrea Pennati
and Mr. Kevin Francis for the critical reading of the manuscript.
6.8
References
1.
Li, C., and Lu, C. D. (2009) Arginine racemization by coupled catabolic and anabolic
dehydrogenases, Proc. Natl. Acad. Sci. U. S. A. 106, 906-911.
2.
Jann, A., Matsumoto, H., and Haas, D. (1988) The fourth arginine catabolic pathway of
Pseudomonas aeruginosa, J. Gen. Microbiol. 134, 1043-1053.
3.
Olsiewski, P. J., Kaczorowski, G. J., and Walsh, C. (1980) Purification and properties of
D-amino acid dehydrogenase, an inducible membrane-bound iron-sulfur flavoenzyme
from Escherichia coli B, J. Biol. Chem. 255, 4487-4494.
197
4.
White, T. A., Krishnan, N., Becker, D. F., and Tanner, J. J. (2007) Structure and kinetics
of monofunctional proline dehydrogenase from Thermus thermophilus, J. Biol. Chem.
282, 14316-14327.
5.
Li, C., Yao, X., and Lu, C. D (2010). Regulation of the dauBAR operon and
characterization of D-amino acid dehydrogenase DauA in arginine and lysine catabolism
of Pseudomonas aeruginosa PAO1, Microbiology 156, 60-71.
6.
Boselli, A., Piubelli, L., Molla, G., Sacchi, S., Pilone, M. S., Ghisla, S., and Pollegioni, L.
(2004) On the mechanism of Rhodotorula gracilis D-amino acid oxidase: role of the
active site serine 335, Biochim Biophys Acta 1702, 19-32.
7.
Krebs, H. A. (1935) Metabolism of amino-acids: Deamination of amino-acids, Biochem.
J. 29, 1620-1644.
8.
Vanoni, M. A., Cosma, A., Mazzeo, D., Mattevi, A., Todone, F., and Curti, B. (1997)
Limited proteolysis and X-ray crystallography reveal the origin of substrate specificity
and of the rate-limiting product release during oxidation of D-amino acids catalyzed by
mammalian D-amino acid oxidase, Biochemistry 36, 5624-5632.
9.
Fu, G., Yuan, H., Li, C., Lu, C. D., Gadda, G., and Weber, I. T (2010). Conformational
Changes and Substrate Recognition in Pseudomonas aeruginosa D-Arginine
Dehydrogenase, Biochemistry.
10.
Bradford, M. M. (1976) A rapid and sensitive method for the quantitation of microgram
quantities of protein utilizing the principle of protein-dye binding, Anal Biochem 72, 248254.
198
11.
Quaye, O., Lountos, G. T., Fan, F., Orville, A. M., and Gadda, G. (2008) Role of Glu312
in binding and positioning of the substrate for the hydride transfer reaction in choline
oxidase, Biochemistry 47, 243-256.
12.
Weast, R. C. (1981) Handbook of Chemistry and Physics, Chemical Rubber Publishing
Company (CRC Press), Boca Raton, FL.
13.
Cook, P. F., and Cleland, W. W. (2007) Enzyme kinetics and mechanism, Garland
Science Publishing, New York, NY.
14.
Gadda, G., and McAllister-Wilkins, E. E. (2003) Cloning, expression, and purification of
choline dehydrogenase from the moderate halophile Halomonas elongata, Appl. Environ.
Microbiol. 69, 2126-2132.
15.
Rungsrisuriyachai, K., and Gadda, G. (2009) A pH switch affects the steady-state kinetic
mechanism of pyranose 2-oxidase from Trametes ochracea, Arch. Biochem. Biophys. 483,
10-15.
16.
Gadda, G., and Fitzpatrick, P. F. (2000) Iso-mechanism of nitroalkane oxidase: 1.
Inhibition studies and activation by imidazole, Biochemistry 39, 1400-1405.
17.
Itoh, K., Asakawa, T., Hoshino, K., Adachi, M., Fukiya, K., Watanabe, N., and Tanaka,
Y. (2009) Functional analysis of aldehyde oxidase using expressed chimeric enzyme
between monkey and rat, Biol Pharm Bull 32, 31-35.
18.
Rouviere, N., Mayer, M., Tegoni, M., Capeillere-Blandin, C., and Lederer, F. (1997)
Molecular interpretation of inhibition by excess substrate in flavocytochrome b2: a study
with wild-type and Y143F mutant enzymes, Biochemistry 36, 7126-7135.
199
19.
Igarashi, K., Momohara, I., Nishino, T., and Samejima, M. (2002) Kinetics of interdomain electron transfer in flavocytochrome cellobiose dehydrogenase from the white-rot
fungus Phanerochaete chrysosporium, Biochem J 365, 521-526.
20.
Wang, Z., Chernyshev, A., Koehn, E. M., Manuel, T. D., Lesley, S. A., and Kohen, A.
(2009) Oxidase activity of a flavin-dependent thymidylate synthase, FEBS J 276, 28012810.
21.
Gavish, B., and Werber, M. M. (1979) Viscosity-dependent structural fluctuations in
enzyme catalysis, Biochemistry 18, 1269-1275.
22.
Mattei, P., Kast, P., and Hilvert, D. (1999) Bacillus subtilis chorismate mutase is partially
diffusion-controlled, Eur J Biochem 261, 25-32.
23.
Shih, M. J., Edinger, J. W., and Creighton, D. J. (1997) Diffusion-dependent kinetic
properties of glyoxalase I and estimates of the steady-state concentrations of glyoxalasepathway intermediates in glycolyzing erythrocytes, Eur J Biochem 244, 852-857.
24.
Lin, Y., Volkman, J., Nicholas, K. M., Yamamoto, T., Eguchi, T., Nimmo, S. L., West, A.
H., and Cook, P. F. (2008) Chemical mechanism of homoisocitrate dehydrogenase from
Saccharomyces cerevisiae, Biochemistry 47, 4169-4180.
25.
Wood, B. M., Chan, K. K., Amyes, T. L., Richard, J. P., and Gerlt, J. A. (2009)
Mechanism of the orotidine 5'-monophosphate decarboxylase-catalyzed reaction: effect
of solvent viscosity on kinetic constants, Biochemistry 48, 5510-5517.
26.
Sobrado, P., and Fitzpatrick, P. F. (2003) Solvent and primary deuterium isotope effects
show that lactate CH and OH bond cleavages are concerted in Y254F flavocytochrome
b2, consistent with a hydride transfer mechanism, Biochemistry 42, 15208-15214.
200
27.
Strickland, S., Palmer, G., and Massey, V. (1975) Determination of dissociation constants
and specific rate constants of enzyme-substrate (or protein-ligand) interactions from rapid
reaction kinetic data, J. Biol. Chem. 250, 4048-4052.
28.
Fitzpatrick, P. F. (2004) Carbanion versus hydride transfer mechanisms in flavoproteincatalyzed dehydrogenations, Bioorg. Chem. 32, 125-139.
29.
Fitzpatrick, P. F. (2010) Oxidation of amines by flavoproteins, Arch. Biochem. Biophys.
493, 13-25.
30.
Mowat, C. G., Wehenkel, A., Green, A. J., Walkinshaw, M. D., Reid, G. A., and
Chapman, S. K. (2004) Altered substrate specificity in flavocytochrome b2: structural
insights into the mechanism of L-lactate dehydrogenation, Biochemistry 43, 9519-9526.
31.
Tabacchi, G., Zucchini, D., Caprini, G., Gamba, A., Lederer, F., Vanoni, M. A., and Fois,
E. (2009) L-lactate dehydrogenation in flavocytochrome b2: a first principles molecular
dynamics study, FEBS J. 276, 2368-2380.
32.
Fan, F., and Gadda, G. (2005) On the catalytic mechanism of choline oxidase, J. Am.
Chem. Soc. 127, 2067-2074.
33.
Gadda, G. (2008) Hydride transfer made easy in the reaction of alcohol oxidation
catalyzed by flavin-dependent oxidases, Biochemistry 47, 13745-13753.
34.
Ferreira, P., Hernandez-Ortega, A., Herguedas, B., Rencoret, J., Gutierrez, A., Martinez,
M. J., Jimenez-Barbero, J., Medina, M., and Martinez, A. T. (2010) Kinetic and chemical
characterization of aldehyde oxidation by fungal aryl-alcohol oxidase, Biochem. J. 425,
585-593.
201
35.
Sucharitakul, J., Wongnate, T., and Chaiyen, P. (2010) Kinetic isotope effects on the
noncovalent flavin mutant protein of pyranose 2-oxidase reveal insights into the flavin
reduction mechanism, Biochemistry 49, 3753-3765.
36.
Blacklow, S. C., Raines, R. T., Lim, W. A., Zamore, P. D., and Knowles, J. R. (1988)
Triosephosphate isomerase catalysis is diffusion controlled. Appendix: Analysis of triose
phosphate equilibria in aqueous solution by 31P NMR, Biochemistry 27, 1158-1167.
37.
Pollegioni, L., Langkau, B., Tischer, W., Ghisla, S., and Pilone, M. S. (1993) Kinetic
mechanism of D-amino acid oxidases from Rhodotorula gracilis and Trigonopsis
variabilis, J. Biol. Chem. 268, 13850-13857.
38.
Saam, J., Rosini, E., Molla, G., Schulten, K., Pollegioni, L., and Ghisla, S. (2010) O2reactivity of flavoproteins: dynamic access of dioxygen to the active site and role of a H+
relay system in D-amino acid oxidase, J. Biol. Chem.285, 24439-24446.
39.
Leferink, N. G., Fraaije, M. W., Joosten, H. J., Schaap, P. J., Mattevi, A., and van Berkel,
W. J. (2009) Identification of a gatekeeper residue that prevents dehydrogenases from
acting as oxidases, J. Biol. Chem. 284, 4392-4397.
40.
Baron, R., Riley, C., Chenprakhon, P., Thotsaporn, K., Winter, R. T., Alfieri, A., Forneris,
F., van Berkel, W. J., Chaiyen, P., Fraaije, M. W., Mattevi, A., and McCammon, J. A.
(2009) Multiple pathways guide oxygen diffusion into flavoenzyme active sites, Proc.
Natl. Acad. Sci. U. S. A. 106, 10603-10608.
41.
Mattevi, A., Vanoni, M. A., Todone, F., Rizzi, M., Teplyakov, A., Coda, A., Bolognesi,
M., and Curti, B. (1996) Crystal structure of D-amino acid oxidase: a case of active site
mirror-image convergent evolution with flavocytochrome b2, Proc. Natl. Acad. Sci. U. S.
A. 93, 7496-7501.
202
42.
Fitzpatrick, P. F., and Massey, V. (1983) The reaction of 8-mercaptoflavins and
flavoproteins with sulfite. Evidence for the role of an active site arginine in D-amino acid
oxidase, J. Biol. Chem. 258, 9700-9705.
43.
Porter, D. J., Voet, J. G., and Bright, H. J. (1977) Mechanistic features of the D-amino
acid oxidase reaction studied by double stopped flow spectrophotometry, J. Biol. Chem.
252, 4464-4473.
44.
Li, J., Vrielink, A., Brick, P., and Blow, D. M. (1993) Crystal structure of cholesterol
oxidase complexed with a steroid substrate: implications for flavin adenine dinucleotide
dependent alcohol oxidases, Biochemistry 32, 11507-11515.
45.
Binda, C., Coda, A., Angelini, R., Federico, R., Ascenzi, P., and Mattevi, A. (1999) A 30angstrom-long U-shaped catalytic tunnel in the crystal structure of polyamine oxidase,
Structure 7, 265-276.
46.
Xia, Z. X., and Mathews, F. S. (1990) Molecular structure of flavocytochrome b2 at 2.4
Å resolution, J. Mol. Biol. 212, 837-863.
47.
Barber, M. J., Neame, P. J., Lim, L. W., White, S., and Matthews, F. S. (1992)
Correlation of x-ray deduced and experimental amino acid sequences of trimethylamine
dehydrogenase, J. Biol. Chem. 267, 6611-6619.
48.
Mattevi, A., Fraaije, M. W., Mozzarelli, A., Olivi, L., Coda, A., and van Berkel, W. J.
(1997) Crystal structures and inhibitor binding in the octameric flavoenzyme vanillylalcohol oxidase: the shape of the active-site cavity controls substrate specificity,
Structure 5, 907-920.
49.
Stenberg, K., and Lindqvist, Y. (1997) Three-dimensional structures of glycolate oxidase
with bound active-site inhibitors, Protein Sci. 6, 1009-1015.
203
50.
Fox, K. M., and Karplus, P. A. (1994) Old yellow enzyme at 2 Å resolution: overall
structure, ligand binding, and comparison with related flavoproteins, Structure 2, 10891105.
51.
Rowland, P., Bjornberg, O., Nielsen, F. S., Jensen, K. F., and Larsen, S. (1998) The
crystal structure of Lactococcus lactis dihydroorotate dehydrogenase A complexed with
the enzyme reaction product throws light on its enzymatic function, Protein Sci. 7, 12691279.
52.
Benson, T. E., Walsh, C. T., and Hogle, J. M. (1997) X-ray crystal structures of the
S229A mutant and wild-type MurB in the presence of the substrate enolpyruvyl-UDP-Nacetylglucosamine at 1.8-Å resolution, Biochemistry 36, 806-811.
53.
Roth, J. P., and Klinman, J. P. (2003) Catalysis of electron transfer during activation of
O2 by the flavoprotein glucose oxidase, Proc Natl Acad Sci U S A 100, 62-67.
54.
Tan, A. K., and Ramsay, R. R. (1993) Substrate-specific enhancement of the oxidative
half-reaction of monoamine oxidase, Biochemistry 32, 2137-2143.
55.
Miller, J. R., and Edmondson, D. E. (1999) Influence of flavin analogue structure on the
catalytic activities and flavinylation reactions of recombinant human liver monoamine
oxidases A and B, J Biol Chem 274, 23515-23525.
56.
Zhao, G., Bruckner, R. C., and Jorns, M. S. (2008) Identification of the oxygen activation
site in monomeric sarcosine oxidase: role of Lys265 in catalysis, Biochemistry 47, 91249135.
57.
Gadda, G., Powell, N. L., and Menon, P. (2004) The trimethylammonium headgroup of
choline is a major determinant for substrate binding and specificity in choline oxidase,
Arch Biochem Biophys 430, 264-273.
204
58.
Finnegan, S., Agniswamy, J., Weber, I. T., and Gadda, G. (2010) Role of Val464 in the
Flavin Oxidation Reaction Catalyzed by Choline Oxidase, Biochemistry 49, 2952-2961.
59.
Fraaije, M. W., and Mattevi, A. (2000) Flavoenzymes: diverse catalysts with recurrent
features, Trends Biochem. Sci. 25, 126-132.
60.
Pudek, M. R., and Bragg, P. D. (1976) Redox potentials of the cytochromes in the
respiratory chain of aerobically grown Escherichia coli, Arch. Biochem. Biophys. 174,
546-552.
205
CHAPTER 7
On the Catalytic Mechanism of D-Arginine Dehydrogenase
7.1
Abstract
The chemical mechanism of leucine dehydrogenation, catalyzed by D-arginine
dehydrogenase (DADH), has been explored with a combination of the effects of pH, substrate,
and solvent KIE and proton inventory by using rapid kinetics technique. The effect of pH on the
reductive half-reaction shows that DADH binds the zwitterionic form leucine. Values for the rate
constant of flavin reduction are strongly-pH dependent indicated that an unprotonated group for
conversion leucine to iminoleucine. These results support a hydride transfer mechanism for the
dehydrogenation reaction in the DADH. The reductive half-reaction with leucine proceeds via
releasing a proton for the substrate with α-NH3+ form, followed by a hydride transfer from Cα to
N5 flavin.
7.2
Introduction
The oxidation of primary amines plays important roles in biology and is crucial to several
biochemical processes, such as neurotransmission (1), cell growth and differentiation (2) and
neoplastic cell proliferation (3, 4). The enzymes that oxidize C-N bonds have been shown to
utilize NADP+, copper and 2,4,5-trihydroxyphenylalanine quinone (TPQ) or flavins (5-8).
Glutamate dehydrogenase serves as the prototype for the NADP+-dependent class of primary
amine oxidizing enzymes (5), copper-amine oxidase for the Cu2+-TPQ enzymes (6) and D-amino
acid oxidase and monoamine oxidase for the flavin-dependent family (7, 9, 10), (11).
Amine oxidation in glutamate dehydrogenase has been proposed to proceed through a direct
hydride transfer from the Cα atom of glutamate to the NADP+. The reaction is triggered by the
206
enzyme-catalyzed deprotonation of the substrate α-NH3+ group in the active site of the enzyme
(12).
Mechanistic studies on copper amine oxidase have elucidated the involvement of a Schiff
base of the TPQ with the amino group of the substrate, which remains bound to the reduced
cofactor after substrate oxidation (13). The oxidation of the reduced TPQ by molecular oxygen
through two one-electron steps and a TPQ semiquinone intermediate complete the turnover of
the enzyme (14).
With D-amino acid oxidase the ability of the enzyme to catalyze elimination of HCl from βCl-alanine and of oxidizing this substrate to β-Cl-pyruvate was initially interpreted in favor of a
carbanion mechanism of amine oxidation (Scheme 7.1a) (15). Lack of a suitable base in the
active site of the enzyme, however, does not support a carbanion mechanism of catalysis (7, 16,
17). The results of mechanistic studies with physiological substrates and of structural studies
with the enzyme in complex with inhibitors are consistent with a hydride transfer mechanism of
catalysis (7, 16, 17). Lack of solvent kinetic isotope effect (KIE) and an inverse
15
N isotope
effect of 0.994 on the kcat/Km value with D-serine as substrate are consistent with a hydride
transfer mechanism from the Cα atom of the anionic amine to the flavin N(5) atom (Scheme
7.1b) (18, 19). What promotes the formation of the anionic amine in the enzyme-substrate
complex is not clear, although the lack of a suitable active site base was taken as independent
evidence that the active form of the substrate must be the anionic amino acid (7, 16, 20).
Computational studies have also supported a hydride transfer mechanism from the anionic amino
acid to the flavin rather than a carbanion mechanism (21). With monoamine oxidase the
inactivation of the enzyme by trans-2-phenylcyclopropylamine was initially used to propose that
oxidation of primary amines involves radical intermediates (Scheme 7.1c) (22-25). A
207
nucleophilic mechanism for amine oxidation in which a flavin C(4a) adduct is formed prior to
proton transfer from the substrate Cα atom to the flavin N(5) atom (Scheme 7.1d) was also
proposed (26, 27),(28). However, the lack of detectable intermediates in stopped-flow reductive
half-reactions (29, 30) with physiological substrates and the insensitivity of the rate constants for
enzyme reduction by benzylamines (31) make radical and nucleophilic proposals in monoamine
oxidase not conclusive. Thus, the mechanism for primary amine oxidation catalyzed by flavindependent enzymes has been controversial for several decades.
(a) Carbanion mechanism
R
N
N
R
N
O
NH
N
R
N
O
NH
N
N
O
NH
N
O
H O
H3C C NH2
COO
H3C C NH2
COO
B:
O
H
H3C C NH2
COO
N
N
R
N
O
NH
N
O
H
R C NH2
COO
N
H
R
N
O
NH
O
NH2
C
COO
(c) Radical mechanism
R
N
N
N
O
NH
O
H
R C NH2
H
(d) Nucleophilic mechanism
R
N
N
O
NH
N
H
R C NH2
H
R
N
N
H
O
R
C
H
NH2
N
O
NH
O
N
H
N
O
NH
O
NH2
H 3C C
COO
(b) Hydride transfer mechanism
R
N
R
N
208
R
N
N
N
H
O
R C NH2
H
O
NH
R
N
N
H
N
R
N
O
NH
O
R C NH2
H
R
C
H
N
N
H
O
NH2
O
NH
Scheme 7.1 Proposed Mechanisms of Amine Oxidation
We have recently sub-cloned, expressed and purified to high levels a novel member of the
class of flavin-dependent, amine-oxidizing enzymes, namely D-arginine dehydrogenase (DADH)
(32, 33). The enzyme from Pseudomonas aeruginosa catalyzes the flavin-linked oxidation of Darginine to iminoarginine, which is nonenzymatically hydrolyzed to 5-guanidino-2-oxopentanoic
acid and ammonia (Scheme 7.2). DADH is a strict dehydrogenase, in that during turnover its
reduced FAD reacts with electron acceptors other than molecular oxygen (33), presumably
ubiquinone. The enzyme displays broad substrate specificity, being able to oxidize all D-amino
acids except for D-glutamate, D-aspartate and glycine (32, 34). In the crystal structure of the
enzyme in complex with iminoarginine resolved to 1.3 Å the side chain of Glu87 forms a strong
ionic interaction with the guanidinium group of the ligand (Figure 7.1) (32), providing a firm
rationale for D-arginine being the best substrate with a kcat/Km value of 106 M-1s-1 (33). Substrate
accessibility to the active site is proposed to be controlled by a mobile loop composed of residues
50-56, which assumes two distinct conformations in the ligand-free and product-bound
crystallographic structures (32). The hydroxyl moiety of Tyr53 points outward to the solvent in
the ligand-free conformation and it is 3.3 Å and 4.2 Å away from the carboxylate and imino
groups of iminoarginine in the product-bound conformation (32). With arginine as substrate
flavin reduction (kred) occurs with a rate constant ≥700 s-1, precluding mechanistic investigations
of the reductive half-reaction catalyzed by the enzyme with this substrate (33). Finally, several
209
ionic and
d hydrogen bond
b
interacctions are esstablished beetween the ccarboxylate of iminoargginine
and the side
s
chains of Tyr53, Arrg222, Arg3055 and Tyr2499, and the m
main chain O atom of G
Gly332,
suggestin
ng that these residues maay play roless in substratee binding (Figure 7.1).
CO
OO
R
H
NH3
R
FAD
D FADH2
COO
H2 O
NH2
R
COO + NH4
O
-
unknown e acceptorr
R=
H
N
NH2
NH2
Scheme 7.2 Oxidatio
on of D-Argiinine Dehyd
drogenase byy DADH
Figure 7.1
7 Active sitte of the DA
ADH structurre with iminooarginine.
In th
he present study, the mechanism
m for aminee oxidation catalyzed by DADH was
investigaated with leu
ucine as subsstrate by usin
ng rapid kineetics and meechanistic prrobes, such aas pH
effects, substrate and
d solvent deu
uterium KIEss and solvennt viscosity eeffects. D-leuucine was chhosen
as the flaavin-reducing
g substrate because
b
it is the slowest substrate off DADH for which the kiinetic
210
parameters kcat and Km had been previously determined by using the steady state kinetic approach
(32). The mechanistic investigation is consistent with amine oxidation by DADH occurring via a
hydride transfer mechanism from an anionic substrate that is formed through the enzymecatalyzed deprotonation of the zwitterionic substrate. Molecular dynamics analysis of the effect
exerted by the enzyme active site on the pKa value of the α-NH3 group of leucine corroborates
the experimental findings. The mechanistic conclusions are interpreted in view of the highresolution structures of DADH reported recently (32).
7.3
Experimental Section
Materials. D-Leucine was from TCI America (Portland, OR). D-Leucine-d10 was purchased
from CDN Isotopes (Canada). Sodium deuterium oxide (99%) was from Isotec Inc.
(Miamisburg, OH). Deuterium oxide (99.9%), glucose and glucose oxidase were obtained from
Sigma-Aldrich (St. Louis, MO). Oxidized DADH was prepared as described previously (33). All
of the other reagents were of the highest purity commercially available.
Rapid Reaction Kinetics. Reductive half-reactions were carried out using an SF-61DX2 HI-
TECH KinetAsyst high performance stopped-flow spectrophotometer, thermostatted at 25 oC.
Unless otherwise stated, the observed rate constants for flavin reduction (kobs) were determined at
varying concentrations of D-leucine in 20 mM sodium pyrophosphate, pH 10.3, under pseudofirst order conditions where the final concentration after mixing of the enzyme was 10 µM and
that of the substrate between 0.5 and 50 mM. The stopped-flow traces obtained anerobically
were identical to those obtained aerobically. Leucine and leucine-d10 were used as substrates to
determine the substrate deuterium KIE. For the determination of the solvent deuterium KIE and
the proton inventories in solvents containing varying mole fractions of D2O, the pD values were
211
adjusted using DCl and NaOD based on the empirical relationship (eq 1) that exists between the
pH-meter reading and the pD value at varying mole fractions of D2O (n) (35). Solvent viscosity
effects were measured in the presence of 9% glycerol as viscosigen, in both the tonometer
containing the enzyme and the syringes containing the organic substrates. The resulting relative
viscosity at 25 oC was 1.26, which is slightly above the value of 1.23 representing a 100%
solution of D2O (36). The pH dependence of the reductive half-reaction of the DADH was
studied in the stopped-flow as described above in the pH range from 7.0 to 11.0. Sodium
phosphate (20 mM final) was used in the pH range from 7.0 to 8.0; sodium pyrophosphate was
used in the pH range from 8.5 to 10.3 and EDTA was used in the pH range from 10.5 to 11.0.
Instability of the enzyme precluded extending the pH studies at pH values lower than 7.0.
(ΔpH ) n = 0.076n2 + 0.3314n
(1)
Steady State Kinetic Parameters. The steady state kinetic parameters were determined by
measuring initial rates of reaction with an oxygen electrode at varying concentrations of leucine
and a fixed, saturating, concentration of 1 mM for PMS as electron acceptor as previously
described (33). All reaction rates are expressed per molar concentration of enzyme-bound flavin.
Data Analysis. Data were fit with the KaleidaGraph software (Synergy Software, Reading,
PA) and the Hi-Kinetic Studio Software Suite (Hi-Tech Scientific, Bradford on Avon, U. K.).
Time-resolved flavin reductions were fit to eq 2, which describes a single exponential
process for flavin reduction; kobs represents the observed first-order rate constant associated with
the absorbance changes at 449 nm at a given concentration of substrate, t is time, A is the
absorbance at 449 nm at any given time, B is the amplitude of the absorbance change, and C is
the absorbance at infinite time.
A = Bexp(−kobst) + C
(2)
212
Reductive half-reaction parameters were determined by using eq 3, where kobs is the observed
first-order rate constant for the reduction of the enzyme-bound flavin at any given concentration
of substrate (S), kred is the limiting first-order rate constant for flavin reduction at saturating
concentrations of substrate, and Kd is the apparent dissociation constant for the dissociation of
the enzyme-substrate complex that undergoes flavin-reduction into free enzyme and substrate.
kobs =
k red S
Kd + S
(3)
Substrate deuteriuem KIE were calculated using eq 4, which applies for a KIE on the kred
value. Solvent deuterium and multiple KIEs were calculated using eq 5, which describes a KIE
on the kred value and a solvent effect on the Kd value. Fi is the fraction of heavy atom, Dkred is the
KIE on kred, and D(Kd) is the KIE or, as in the case under study here (see Results), a solvent effect
due to pH on the apparent Kd value.
kobs =
k red S
(K d + S )[1 + Fi ( Dkred − 1)]
k red S
kobs =
D
K d [1 + Fi ( D
k red
− 1)] + S[1 + Fi ( Dk red − 1)]
(K d )
(4)
(5)
The pH dependence of the kred value was determined by fitting the data with eq 6, where
kred(lim) is the limiting, pH-independent rate constant for flavin reduction at high pH, and pKa is
the apparent value for the ionization of a group that must be unprotonated for flavin reduction.
log k red = log
k red (lim)
10 − pH
1+ − pK a
10
(6)
The pH dependence of the Kd value was determined by fitting the data with eq 7, which
describes a curve with a slope of +1 at high pH and a plateau region that defines a limiting, pHindependent Kd value at low pH, i.e., Kd(lim).
213
log(K d ) = log[ K d (lim) • (1 + 10 pH − pK a )]
(7)
The proton inventory on the kred value was fit with eq 8 (35). Here, (kred)n is the limiting firstorder rate constant for flavin reduction in n mole fraction D2O, (kred)H2O is the limiting first-order
rate constant for flavin reduction in H2O, and ΦT is the isotopic fractionation factor of the
transition state proton.
(kred )n
(kred )H 2O
7.4
= 1 − n + nφT
(8)
Results
Time-resolved Flavin Reduction. The time-resolved reduction of DADH was investigated in
a stopped-flow spectrophotometer at pH 10.3 and 25 °C by monitoring the loss of absorbance of
the oxidized flavin at 449 nm upon mixing the enzyme with the reducing substrate leucine.
Pseudo-first order conditions with ~10 μM enzyme and 0.5-50 mM reducing substrate were
maintained, and the stopped-flow traces could be fit to single exponentials to obtain observed
rate constants (kobs) for flavin reduction at any given substrate concentration (Figure 7.2A). The
kobs value was hyperbolically dependent on the concentration of the reducing substrate (Figure
7.2B), allowing for the determination of the limiting rate constant for flavin reduction (kred) and
the apparent equilibrium constant for the dissociation of the substrate from the ES complexes
that will undergo flavin reduction (Kd). The best fit of the data was obtained with the curve
extrapolating to the plot origin (Figure 7.2B), consistent with flavin reduction being irreversible,
and in agreement with previous results with histidine as the reducing substrate (33).
214
0.15
120
A
B
80
0.05
40
k
obs
, s-1
0.1
0
0.0001
0
0.001
0.01
0.1
time, s
1
0
20
40
60
[D-leucine], mM
Figure 7.2 Effects of pH on the kinetic parameters of RHR (solid circle) and SSK (empty circle)
for DADH with leucine as the reducing substrate. Panel A shows the pH profile of kred and kcat.
Data were fit to eq. 5, yielding kred(lim) = 133 ± 5 s-1 and pKa = 9.6 ± 0.1 (R2 = 0.999). Panel B
shows the pH profile of Kd and Km. Data were fit to eq. 6, yielding Kd(lim) = 3.5 ± 0.3 mM and
pKa = 10.3 ± 0.1 (R2 = 0.89). Panel C shows the pH profile of kred/Kd and kcat/Km. Data were fit to
eq. 7, yielding kred/Kd(lim) = 16,000 ±2,000 M-1s-1 and pKa values of 9.2 ±0.1 and 11.0 ± 0.2.
pH Effects on Flavin Reduction. The effects of pH on the kred and Kd values for DADH with
leucine as the reducing substrate were determined in the pH range 7.0-11.0. This was carried out
with the goal of determining the ionization states and the apparent pKa values of protein groups
that participate in the reaction of amine oxidation. At all pH values tested, and ensuring that
pseudo-first order conditions were maintained, the kinetics of the reductive half-reaction
followed the general trends of Figure 7.2, allowing us to establish kred and Kd values as a
function of pH. The kred increased with increasing pH and reached a plateau at high pH with a
limiting value of 133 ± 5 s-1 (Figure 7.3A). An apparent pKa of 9.6 ± 0.1 for an unprotonated
group participating in flavin reduction was established. The pH profile of the kred value also
established that solvent KIEs are devoid of artifactual contributions that originate from pH
effects at pL ≥10.3, i.e., where the kred is pH-independent. The Kd was pH-independent between
below pH 9.5 and increased with increasing pH values, defining an apparent pKa of ~10.3 for a
group that must be protonated for substrate binding to yield ES complexes that will undergo
flavin reduction (Figure 7.3B).
215
Figure 7.3
7 Effects of pH on the kinetic param
meters of RH
HR (solid circle) and SS
SK (empty ciircle)
for DAD
DH with leu
ucine as the reducing
r
sub
bstrate. Panell A shows thhe pH profilee of kred and kcat.
Data weere fit to eq. 5, yielding kred(lim) = 133 ± 5 s-1 annd pKa = 9.6 ± 0.1 (R2 = 0.999). Panel B
shows the pH profille of Kd and Km. Data weere fit to eq. 6, yielding Kd(lim) = 3.5 ± 0.3 mM and
pKa = 10
0.3 ± 0.1 (R2 = 0.89). Pan
nel C shows the pH proffile of kred/Kd and kcat/Km. Data were fit to
eq. 7, yielding
y
kred/K
/ d(lim) = 16,000 ±2,000 M-1s-1 andd pKa values of 9.2 ±0.1 and 11.0 ± 00.2.
Substtrate Deuteerium KIE on kred. Th
he substrate deuterium KIE on thee kred valuee was
determin
ned at pH 10
0.3 using leu
ucine-d10 and
d leucine as the reducinng substrates, to report oon the
status of the substrate CH bond in
i the transittion state(s) for the aminne oxidationn catalyzed bby the
enzyme. In aqueous solution the D(kred)H2O had
h a value oof 5.1 ± 0.1 (Figure 7.4), consistentt with
cleavage of the subsstrate CH bo
ond being manifested
m
inn the transittion state foor the reactioon of
flavin red
duction. As expected, th
here was no KIE associaated with thee Kd value aand the best fit of
216
the data was
w obtained
d by using eq
q 4. Upon su
ubstitution oof water withh deuterium ooxide there w
was a
small, bu
ut significantt, decrease in
i the magniitude of the KIE with a D(kred)D2O vvalue of 4.7 ± 0.1
(R2 = 0.9
9996).
Figure 7.4
7 Observeed rate con
nstants as a function oof substrate concentratiion. Black ccurve
representts substrate reduction
r
off DADH enzzyme with l eucine as a substrate inn H2O. Red ccurve
representts substrate reduction
r
off DADH enzzyme with leeucine as a ssubstrate in D2O. Blue ccurve
representts substrate reduction off DADH en
nzyme with lleucine-d10 as a substrate in H2O. Data
were fit to
t eqs 4 and 5.
Solveent Deuteriu
um KIE on kred. The solv
vent deuteriuum KIE on tthe kred valuee was determ
mined
at pL 10..3, where thee kred value is pL-indepen
ndent (Figu re 7.3A), to report on thhe status of bbonds
involving
g solvent exchangeable
e
e protons in the transsition state(ss) for the amine oxiddation
catalyzed
d by the enzy
yme. A norm
mal D2O(kred)H value of 1.77 ± 0.01 (R
R2 = 0.9995) was determ
mined
with leuccine as substtrate (Figuree 7.4). The best
b fit of thee data was oobtained by uusing eq 5, w
which
allowed also for a so
olvent effect on the Kd value of 2.88 ± 0.1 to bbe computedd. Such a soolvent
effect on
n the Kd value is primarilly, if not entirely, due too pH becausee Kd is not pH-independeent at
pL 10.3 (Figure 7.3
3). For this reason, the solvent effeect on the Kd value waas not considdered
217
further in
n this study. Upon replaacing leucinee with leucinne-d10 there was a small, but signifi
ficant,
decrease in the D2O(kkred
o 1.60 ± 0.04
4 (R2 = 0.99889).
r )D value to
due to the inncreased visscosity of D2O as
As a control thatt the solventt effect on kred
r was not d
d to H2O, the kred value was
w determin
ned in aqueoous solutionss in the abseence and pressence
compared
9% glycerol, which is isoviscou
us with D2O (36, 37). At pH 10.33, the kred w
was similar w
when
determin
ned in the prresence and absence of glycerol, w
with a (kred)H22O/(kred)glycerool ratio of 1.03 ±
0.03. Theese results esstablished th
hat the solvent deuterium
m KIE on thee kred value rreports on soolvent
exchangeeable proton(s) being in flight in the transition sttate for the rreaction of fllavin reductiion.
Proto
on Inventoryy on kred. The
T proton in
nventory tecchnique was used to gaiin insights oon the
number of
o solvent ex
xchangeablee protons ressponsible forr the solvennt deuterium KIE observved in
the time--resolved reduction of DADH.
D
Thee kred value was determ
mined at pL 10.3 in soluutions
with vary
ying mole fractions
f
of deuterium oxide,
o
and tthe data were analyzedd in a plot oof the
reciprocaal of the KIE
E as a functio
on of the mo
ole fraction D2O (Figuree 7.5) (38). The resultinng kred
proton in
nventory was linear, sug
ggesting thatt a single soolvent exchanngeable protton is in fligght in
the transiition state fo
or the reactio
on of flavin reduction
r
(Fiigure 7.5).
Figure 7.5
7 Proton inv
ventory for the
t reduction
n rate, kred, uusing leucinee at pL 10.3..
218
Multiple Deuterium KIEs on kred. Multiple deuterium KIEs were determined at pL 10.3 to
elucidate whether the substrate and solvent deuterium KIE determined independently on the kred
value for DADH occurred in the same or in multiple kinetic steps. The
D,D2O
kred value,
representing the effect of substituting leucine in H2O with leucine-d10 in D2O, was 7.8 ± 0.1
(calculated by using eq 5). Such a multiple deuterium KIE is lower than the product of the
individual substrate and solvent KIEs, i.e., D(kred)D2O x
D2O
(kred)H, yielding a computed value of
9.0 ± 0.2. This is consistent with the substrate and solvent KIEs on flavin reduction occurring not
in the same transition state (39).
pH Effects on the Steady State Kinetic Parameters. Previous results with arginine or
histidine as substrate for DADH indicated that the enzyme follows a Ping-Pong Bi-Bi steady
state kinetic mechanism (33). Moreover, irrespective of the amino acid substrate the Km value for
PMS, which is used as electron acceptor, is ~10 μM (33). This establishes that the apparent
steady state kinetic parameters determined at a fixed concentration of 1 mM PMS, i.e., ~100-fold
larger than Km, approximate well the true kcat, Km, and kcat/Km values for the amino acid
substrates. Consequently, the effects of pH on the steady state kinetic parameters were
determined here by measuring initial rates of reaction with leucine as the varying substrate for
DADH at a fixed concentration of 1 mM PMS. A high non-enzymatic reactivity of PMS with
oxygen seen at alkaline pHs prevented to extend the profile above pH 10.0. Nonetheless, the pHprofile of the kcat value was similar to that of the kred (Figure 7.3A), with the kcat being between
1.2- to 1.6-times lower than the kred throughout the pH range considered (Table S7.2). This
establishes that flavin reduction primarily contributes to limiting the overall turnover of the
enzyme with leucine as substrate. The Km values between pH 7.0 and 10.0 were remarkably
219
similar to the Kd values determined by using a stopped-flow spectrophotometer (Figure 7.3B),
indicating that with leucine as substrate the Km value reports directly on the apparent affinity of
the enzyme for substrate.
7.5
Discussion
In this paper, deuterium KIEs have been used to obtain sights into the reductive half-reaction
mechanism of leucine oxidation in the reaction catalyzed by DADH. The results of the rapidreaction kinetic analyses described here are consistent with the kinetic mechanism of Scheme
7.1b. Strong evidence for such a mechanism comes from the substrate, solvent deuterium kinetic
IEs and proton inventory determined on the reductive half-reaction presented here. The oxidation
of leucine catalyzed by DADH involves deprotonation of amino group of the substrate and
hydride transfer from α-hydrogen of the substrate to the N5 of the flavin. Two bonds cleavage
occurs in an asynchronous fashion. The substrate KIE on kred provides a direct probe of the status
of the CH bond, whereas the solvent KIE on kred can be used to provide information on the status
of the NH bond.
The oxidation of leucine requires an unprotonated group for catalysis, as suggested by the
effects of pH on the rate constant for flavin reduction. The kred values of DADH exhibited pH
dependence with limiting value of 133 ± 5 s-1 at high pH, and a pKa of about 9.6 was measured
for group that must be unprotonated for efficient catalysis. Based on the crystal structure and
kinetic data, the pKa of 9.6 was assigned to the residue Tyr53 which stabilize the partially
positively charged NH2+ group of transition-state. From the DADH-iminoarginine complex
structrure, in the product-bound conformation, the aromatic ring of Tyr53 moves into the active
site and close to the imino group of the imino acid (32). Therefore, at high pH, the deprotonated
Tyr53 will form the electrostatic interaction with imino group and stabilize the transition-state
220
which facilitates the reductive half-reaction. Tyrosine residues are generally found to be
conserved and have similar function in other flavin-dependent enzymes. The residue 224 in pig
kidney D-amino acid oxidase was suggested to participate in the reductive half-reaction and
could stabilize the second charge-transfer intermediate, interacting with the positively charged
NH2+ group of the product imino acid (40). Tyr254 in flavocytochrome b2 was proposed to take
part in Michaelis complex formation and stabilize transition-state (41). The kinetic properties of
Tyr129 glycolate oxidase provide evidence that the main function of the hydroxyl group of
tyrosine is to stabilize the transition state (42).
DADH binds the protonated amino group of leucine, as suggested by the pH profile of the Kd
values. From the pH profile, the substrate affinity decrease while increasing the pH, defining an
apparent pKa of ~10.3. Therefore, there is a protonated group involved in substrate binding. The
pKa can be assigned to the substrate amino group. At high pH, the binding affinity is decreased.
This indicates that DADH binds the zwitterionic form of leucine, the predominant species in
solution. Alternatively, the pKa of 10.3 can also be assigned to Tyr249. Based on the previously
reported crystal structure of enzyme-iminoarginine complex (32), there are several electrostatic
and hydrogen bond interactions between iminoarginine and the side chain or main chain atoms of
Tyr53, Glu87, Arg222, Tyr249, Arg305, and Gly332. These residues constitute the site of
interaction for binding substrate. Due to the pKa value is around 10.3, two tyrosines are good
candidates. Since Tyr53 forms H-Bond with amino group of the substrate, deprotonation will not
affect Kd. Therefore, the pKa of 10.3 was assigned to the residue Tyr249. The increasing in Kd at
high pH was due to the deprotonation of Tyr249 which forms hydrogen bond with carboxyl
group of the product complex. Since there is only one pKa observed in pH profile of Kd, the pKa
of the amino group should either less than 7.0 or large than 11.0. Computational studied
221
investigated by Hamelberg’s group shows the pKa of amino group of the substrate will increase
while the substrate binds to the enzyme (personal communication). In this case, the pKa of
substrate amino group is larger than 11. Therefore, the DADH still binds the zwitterionic form of
leucine at the pH used in the experiment. The observation that the anionic form of leucine with a
neutral amino group is the reactive species in the reduction is similar to that observed for other
flavoenzymes, such as monomeric sarcosine oxidase (43), monoaime oxidase (22, 44) and
trimethylamine dehydrogenase (45, 46).
The hydride ion transfer from substrate to the enzyme-bound flavin and proton abstraction
from the amine substrate are both rate-limiting steps in the overall turnover of the DADH
enzyme. This conclusion is supported by both steady state kinetics and rapid kinetic data. First,
the pH-profile of the kcat value was similar to that of the kred, with the kcat being between 1.2- to
1.6-times lower than the kred throughout the pH range considered. Independent evidence comes
from the Dkcat value with deuterated leucine significantly larger than unity, with D(kred)H2O value
of 5.1 ± 0.1 and D(kred)D2O value of 4.7 ± 0.1. Thus, the CH bond cleavage is almost fully ratelimiting step. In the reaction catalyzed by DADH enzyme the abstraction of the amino proton is
associated with deuterium kinetic isotope effect significantly larger than unity, with
D2O
(kred)H
value of 1.77 ± 0.01. According to the lack of viscosity effect, the observed solvent isotope effect
is consistent with the cleavage of substrate N-H bond. Moreover, linear relationship observed in
the proton inventory associated with kred values suggests a single proton being in flight in the
catalysis. Thus, the proton form substrate amine group can be removed by the active site residue
or solvent. Therefore, the overall turnover of the enzyme is limited primarily by the substrate CH bond cleavage with minimal contribution of substrate NH bond cleavage.
222
Timing of CH bond and NH bond cleavage. An asynchronous mechanism is proposed for CH
bond and NH bond cleavage based on the kinetic isotope effect determined on the rate constant
for flavin reduction. Since in the asynchronous mechanism, upon slowing down the cleavage of
the other bond a decreasing in the magnitudes of the substrate and solvent kinetic isotope effects
is expected. The decrease in the rate constants from the substrate and solvent kinetic isotope
effect were observed upon replacing H2O with D2O and leucine with deuterium leucine,
respectively. Consistent with asynchronous mechanism, the multiple kinetic isotope effects
D,D2O
kred value of 7.8 ± 0.1 were smaller than the product of the individual substrate and solvent
kinetic isotope effects (D(kred)D2O x D2O(kred)H), with a value of 9.0 ± 0.2.
There are forward and reverse commitments to catalysis of the enzyme-substrate complex at
pH 10.3. This conclusion is supported by the asynchronous mechanism for the cleavage of NH
and CH bonds of the substrate. Since the kcat being 1.41-times on average lower than the kred
throughout the pH range, the kcat value at pH 10.3 can be estimated about 81 ± 3 s-1 according to
kred with a value of 113 s-1. Thus, based on the eq 9, kNH can be calculated as 290 ± 10 s-1. At pH
10.3, the observed substrate and solvent kinetic isotope effects are 5.1 and 1.77, respectively.
Therefore, the D2OkNH and DkCH can be calculated with values of 3.7 and 6.7 by eq 10 and 11.
k cat =
k NH k CH
k NH + k CH
D 2O
D 2O
k red =
D
D
k red =
k NH +
1+
k CH +
1+
(9)
k NH
k CH
k NH
k CH
(10)
k CH
k NH
k CH
k NH
(11)
223
Conclusions. In summary, the results of the mechanistic investigation with deuterated
substrate and solvent presented in this study indicate substrate NH bond cleavage and CH bond
cleavage that corresponds to reduction of the flavin cofactor involves two steps. The oxidation of
leucine proceeds via the releasing of a proton from the substrate with α-NH3+ form. Oxidation
this species occurs by an unprotonated group with a pKa of 9.6 for direct hydride transfer from
the amino acid Cα to the flavin.
7.6
Support Information
Table S7.1 Reductive Half-Reaction Kinetic Parameters for DADH with Leu as Substrate a
app
pH
kred, s-1
Kd, leu, mM
R2
7.0
0.33 ± 0.01
2.6 ± 0.1
0.999
7.5
0.93 ± 0.01
3.5 ± 0.1
0.999
8.0
3.5 ± 0.1
3.0 ± 0.3
0.995
8.5
8.4 ± 0.3
3.2 ± 0.4
0.991
9.0
26 ± 1
4.8 ± 0.4
0.997
9.5
50 ± 2
4.3 ± 0.5
0.994
10.0
91 ± 1
6.8 ± 0.2
0.999
10.3
113 ± 1
9.7 ± 0.3
0.999
10.5
119 ± 2
9.1 ± 0.3
0.999
10.7
125 ± 5
15 ± 1
0.997
11.0
129 ± 4
16 ± 1
0.998
a
o
Activity assays of DADH were performed in 20 mM buffer pH for 6.0 to 11.0, at 25 C.
Kinetic parameters were fit to eq. 2.
Table S7.2 Steady State Kinetic Parameters for DADH with Leu as Substrate a
pH
kcat, s-1
Km D-leu, mM
kcat/ Km D-leu, M-1s-1
kred/kcat
0.26 ± 0.01
3.1 ± 0.2
80 ± 4
1.26 ± 0.06
7.0
0.64 ± 0.01
3.6 ± 0.3
180 ± 10
1.45 ± 0.02
7.5
2.3 ± 0.1
2.6 ± 0.1
870 ± 50
1.54 ± 0.09
8.0
7.0 ± 0.2
3.4 ± 0.2
2,100 ± 200
1.20 ± 0.06
8.5
17 ± 1
4.7 ± 0.6
3,700 ± 500
1.54 ± 0.09
9.0
31 ± 1
3.8 ± 0.3
8,200 ± 700
1.61 ± 0.08
9.5
73 ± 3
6.6 ± 0.5
11,100 ± 900
1.25 ± 0.05
10.0
a
o
Conditions: 20 mM buffer, pH from 7 to 10 and 25 C.
R2
0.997
0.997
0.999
0.997
0.993
0.996
0.993
224
7.7
References
1.
Edmondson, D. E., Mattevi, A., Binda, C., Li, M., and Hubalek, F. (2004) Structure and
mechanism of monoamine oxidase, Curr Med Chem 11, 1983-1993.
2.
Thomas, T., and Thomas, T. J. (2001) Polyamines in cell growth and cell death:
molecular mechanisms and therapeutic applications, Cell Mol Life Sci 58, 244-258.
3.
Pegg, A. E. (1988) Polyamine metabolism and its importance in neoplastic growth and a
target for chemotherapy, Cancer Res 48, 759-774.
4.
Seiler, N., Atanassov, C. L., and Raul, F. (1998) Polyamine metabolism as target for
cancer chemoprevention (review), Int J Oncol 13, 993-1006.
5.
Green, J., and Large, P. J. (1984) Regulation of the key enzymes of methylated amine
metabolism in Candida boidinii, J Gen Microbiol 130, 1947-1959.
6.
Wilce, M. C., Dooley, D. M., Freeman, H. C., Guss, J. M., Matsunami, H., McIntire, W.
S., Ruggiero, C. E., Tanizawa, K., and Yamaguchi, H. (1997) Crystal structures of the
copper-containing amine oxidase from Arthrobacter globiformis in the holo and apo
forms: implications for the biogenesis of topaquinone, Biochemistry 36, 16116-16133.
7.
Mattevi, A., Vanoni, M. A., Todone, F., Rizzi, M., Teplyakov, A., Coda, A., Bolognesi,
M., and Curti, B. (1996) Crystal structure of D-amino acid oxidase: a case of active site
mirror-image convergent evolution with flavocytochrome b2, Proc Natl Acad Sci U S A
93, 7496-7501.
8.
Fitzpatrick, P. F. (2010) Oxidation of amines by flavoproteins, Arch Biochem Biophys
493, 13-25.
225
9.
Binda, C., Newton-Vinson, P., Hubalek, F., Edmondson, D. E., and Mattevi, A. (2002)
Structure of human monoamine oxidase B, a drug target for the treatment of neurological
disorders, Nat Struct Biol 9, 22-26.
10.
De Colibus, L., Li, M., Binda, C., Lustig, A., Edmondson, D. E., and Mattevi, A. (2005)
Three-dimensional structure of human monoamine oxidase A (MAO A): relation to the
structures of rat MAO A and human MAO B, Proc Natl Acad Sci U S A 102, 1268412689.
11.
A number of flavoproteins are active on N-methylated amino acids, with oxidation of the
bond between the amino N atom and the methyl group rather than the Cα atom. Sarcosine
oxidase serves as the prototype for this class of flavin-dependent enzymes, which also
includes N-methyltryptophan oxidase, N,N-dimethylglycine oxidase, and NikD. The
various proposed mechanisms of amine oxidation that have been considered for this class
of flavoenzymes, which are equivalent to those in Scheme 7.1 c-d, have been recently
reviewed by Dr. Fitzpatrick (see ref. 8 ).
12.
Baker, P. J., Waugh, M. L., Wang, X. G., Stillman, T. J., Turnbull, A. P., Engel, P. C.,
and Rice, D. W. (1997) Determinants of substrate specificity in the superfamily of amino
acid dehydrogenases, Biochemistry 36, 16109-16115.
13.
Mure, M., Mills, S. A., and Klinman, J. P. (2002) Catalytic mechanism of the topa
quinone containing copper amine oxidases, Biochemistry 41, 9269-9278.
14.
Kishishita, S., Okajima, T., Kim, M., Yamaguchi, H., Hirota, S., Suzuki, S., Kuroda, S.,
Tanizawa, K., and Mure, M. (2003) Role of copper ion in bacterial copper amine oxidase:
spectroscopic and crystallographic studies of metal-substituted enzymes, J Am Chem Soc
125, 1041-1055.
226
15.
Walsh, C. T., Schonbrunn, A., and Abeles, R. H. (1971) Studies on the mechanism of
action of D-amino acid oxidase. Evidence for removal of substrate -hydrogen as a proton,
J Biol Chem 246, 6855-6866.
16.
Umhau, S., Pollegioni, L., Molla, G., Diederichs, K., Welte, W., Pilone, M. S., and
Ghisla, S. (2000) The x-ray structure of D-amino acid oxidase at very high resolution
identifies the chemical mechanism of flavin-dependent substrate dehydrogenation, Proc
Natl Acad Sci U S A 97, 12463-12468.
17.
Todone, F., Vanoni, M. A., Mozzarelli, A., Bolognesi, M., Coda, A., Curti, B., and
Mattevi, A. (1997) Active site plasticity in D-amino acid oxidase: a crystallographic
analysis, Biochemistry 36, 5853-5860.
18.
Denu, J. M., and Fitzpatrick, P. F. (1994) Intrinsic primary, secondary, and solvent
kinetic isotope effects on the reductive half-reaction of D-amino acid oxidase: evidence
against a concerted mechanism, Biochemistry 33, 4001-4007.
19.
Kurtz, K. A., Rishavy, M. A., Cleland, W. W., and Fitzpatrick, P. F. (2000) Nitrogen
Isotope Effects As Probess of the Mechanism of D-Amino Acid Oxidase, J Am Chem Soc
122, 12896-12897.
20.
Mizutani, H., Miyahara, I., Hirotsu, K., Nishina, Y., Shiga, K., Setoyama, C., and Miura,
R. (1996) Three-dimensional structure of porcine kidney D-amino acid oxidase at 3.0 A
resolution, J Biochem 120, 14-17.
21.
Tilocca, A., Gamba, A., Vanoni, M. A., and Fois, E. (2002) First-principles molecular
dynamics investigation of the D-amino acid oxidative half-reaction catalyzed by the
flavoenzyme D-amino acid oxidase, Biochemistry 41, 14111-14121.
227
22.
McEwen, C. M., Jr., Sasaki, G., and Jones, D. C. (1969) Human liver mitochondrial
monoamine oxidase. II. Determinants of substrate and inhibitor specificities,
Biochemistry 8, 3952-3962.
23.
Paech, C., Salach, J. I., and Singer, T. P. (1980) Suicide inactivation of monoamine
oxidase by trans-phenylcyclopropylamine, J Biol Chem 255, 2700-2704.
24.
Silverman, R. B. (1995) Radical Ideas About Monoamine Oxidase, Acc. Chem. Res. 28,
335-342.
25.
Hanzlik, R. P., Kishore, V., and Tullman, R. (1979) Cyclopropylamines as suicide
substrates for cytochromes P-450, J Med Chem 22, 759-761.
26.
Edmondson, D. E., Binda, C., and Mattevi, A. (2007) Structural insights into the
mechanism of amine oxidation by monoamine oxidases A and B, Arch Biochem Biophys
464, 269-276.
27.
Brown, L. E., and Hamilton, G. A. (1970) Some model reactions and a general
mechanism for flavoenzyme-catalyzed dehydrogenations, J Am Chem Soc 92, 7225-7227.
28.
The nucleophilic mechanism was initially proposed by Hamilton and involved the
abstraction of the proton from the substrate Cα atom by an active site base occurring after
the initial nucleophilic attack of the substrate N atom to the flavin C(4a) atom (see ref.
27). The recent proposal by Edmondson and coworkers where the flavin N(5) atom
accepts the proton from the substrate Cα as the substrate N atom forms the flavin-adduct
originates from the lack of an appropriate base in the crystallographic structure of the
enzyme (see ref. 26).
29.
Husain, M., Edmondson, D. E., and Singer, T. P. (1982) Kinetic studies on the catalytic
mechanism of liver monoamine oxidase, Biochemistry 21, 595-600.
228
30.
Miller, J. R., and Edmondson, D. E. (1995) Mechanistic probe of Monoamine oxidase B
catalysis: rapid-scan stopped flow and magnetic field independence of the reductive halfreaction, J Am Chem Soc 117, 7830-7831.
31.
Walker, M. C., and Edmondson, D. E. (1994) Structure-activity relationships in the
oxidation of benzylamine analogues by bovine liver mitochondrial monoamine oxidase
B, Biochemistry 33, 7088-7098.
32.
Fu, G., Yuan, H., Li, C., Lu, C. D., Gadda, G., and Weber, I. T. (2010) Conformational
changes and substrate recognition in Pseudomonas aeruginosa D-arginine dehydrogenase,
Biochemistry 49, 8535-8545.
33.
Yuan, H., Fu, G., Brooks, P. T., Weber, I. T., and Gadda, G. (2010) Steady State Kinetic
Mechanism and Reductive Half-Reaction of D-Arginine Dehydrogenase from
Pseudomonas aeruginosa, Biochemistry.
34.
Li, C., Yao, X., and Lu, C. D. (2010) Regulation of the dauBAR operon and
characterization of D-amino acid dehydrogenase DauA in arginine and lysine catabolism
of Pseudomonas aeruginosa PAO1, Microbiology 156, 60-71.
35.
Schowen, K. B., and Schowen, R. L. (1982) Solvent isotope effects of enzyme systems,
Methods Enzymol 87, 551-606.
36.
Lide, D. R. (2000) Handbook of Chemistry and Physics, CRC Press, Boca Raton, FL.
37.
Karsten, W. E., Lai, C., and Cook, P. F. (1995) Inverse solvent isotope effects in the
NAD-Malic enzyme reaction are the result of the viscosity difference between D2O and
H2O: Implications for solvent isotope effect studies, J Am Chem Soc 117, 5914-5918.
38.
Venkatasubban, K. S., and Schowen, R. L. (1984) The proton inventory technique, CRC
Crit Rev Biochem 17, 1-44.
229
39.
Cleland, W. W. (1991) in enzyme mechanism from isotope effects, CRC Press, Boca
Raton, FL.
40.
Pollegioni, L., Fukui, K., and Massey, V. (1994) Studies on the kinetic mechanism of pig
kidney D-amino acid oxidase by site-directed mutagenesis of tyrosine 224 and tyrosine
228, J Biol Chem 269, 31666-31673.
41.
Sobrado, P., and Fitzpatrick, P. F. (2003) Solvent and primary deuterium isotope effects
show that lactate CH and OH bond cleavages are concerted in Y254F flavocytochrome
b2, consistent with a hydride transfer mechanism, Biochemistry 42, 15208-15214.
42.
Macheroux, P., Kieweg, V., Massey, V., Soderlind, E., Stenberg, K., and Lindqvist, Y.
(1993) Role of tyrosine 129 in the active site of spinach glycolate oxidase, Eur J Biochem
213, 1047-1054.
43.
Zhao, G., and Jorns, M. S. (2002) Monomeric sarcosine oxidase: evidence for an
ionizable group in the E.S complex, Biochemistry 41, 9747-9750.
44.
McEwen, C. M., Jr., Sasaki, G., and Lenz, W. R., Jr. (1968) Human liver mitochondrial
monoamine oxidase. I. Kinetic studies of model interactions, J Biol Chem 243, 52175225.
45.
Basran, J., Sutcliffe, M. J., and Scrutton, N. S. (2001) Deuterium isotope effects during
carbon-hydrogen bond cleavage by trimethylamine dehydrogenase. Implications for
mechanism and vibrationally assisted hydrogen tunneling in wild-type and mutant
enzymes, J Biol Chem 276, 24581-24587.
46.
Basran, J., Sutcliffe, M. J., and Scrutton, N. S. (2001) Optimizing the Michaelis complex
of trimethylamine dehydrogenase: identification of interactions that perturb the ionization
230
of substrate and facilitate catalysis with trimethylamine base, J Biol Chem 276, 4288742892.
231
CHAPTER 8
General Discussion and Conclusions
The dissertation is focused on flavoenzyme mechanisms and uses choline oxidase and Darginine dehydrogenase as model systems to gain an understanding of enzyme catalysis.
Mechanistic and structural studies have been carried out by using pH and kinetic isotope effects,
proton inventories and site-directed mutagenesis on the selected active site residues, as well as
X-ray crystallography. These allowed investigating the mechanism of alcohol and amine
oxidation. In the choline oxidase, the catalytic role of a serine at position 101 has been addressed
which is important to both the activation of the alcohol substrate and the subsequent hydride
transfer reaction. In addition, the mechanism of D-arginine dehydrogenase was investigated and
discussed to gain more insight of the amine oxidation.
Choline oxidase catalyzes the two-step, four-electron oxidation of choline to glycine betaine,
with betaine aldehyde as an intermediate. It has been previously studied using biochemical (1-4),
site-directed mutagenic (5-12), structural (13-15), mechanistic (16-20) and computational
approaches (21). The results show that the oxidation of choline to betaine aldehyde occurs
through a hydride transfer mechanism. Catalysis is triggered by the removal of the hydroxyl
proton of choline by an active site base and the hydride was subsequently transferred from the αcarbon of the substrate to the N5 atom of the flavin. In the wild-type enzyme, the abstraction of
the hydroxyl proton is fast and occurred in the dead-time of the stopped-flow instrument, i.e. ~
2.2 ms. Therefore, the proton abstraction was not observed and there was one phase for the flavin
reduction in the stopped-flow experiment.
232
From the X-ray structure of the wild-type choline oxidase resolved to 1.86 Å, Ser101 is less
than 4 Å from the N(5) atom of FAD and within hydrogen bonding distance (i.e., <3 Å) of the
oxygen atom of DMSO, an additive that was used in the crystallization experiments (15). In
order to probe the role of Ser101, the variants in which the serine was replaced with alanine,
threonine, cysteine or valine were prepared by site-directed mutagenesis and investigated with
rapid kinetics and steady state kinetics.
Firstly, the steady state, transient kinetic and crystal structure studies were used to investigate
Ser101Ala variant of choline oxidase. This work suggests that the OH group of Ser101 facilitates
the reduction of the enzyme-bound flavin by choline and betaine aldehyde but impedes the
reaction with O2. Compared to the wild-type enzyme, the mutant shows no significant structural
differences except for the lack of a hydroxyl group on residue 101 and a more planar
configuration of the flavin isoalloxazine ring (22). The serine residue in position 101 optimizes
enzyme turnover number with suboptimal oxidation half-reactions and faster reductive halfreactions. Thus, the residue at position 101 is important to balance the oxidative and reductive
half-reactions and catalyze the reaction by compromise.
Then the rapid reaction kinetic studies of Ser101Ala, Ser101Thr, Ser101Cys and Ser101Val
variants were investigated. Two phases were observed for flavin reduction giving a fast phase
and a slow phase which is different from the single phase observed in the wild-type enzyme. Use
of stopped-flow spectrophotometry revealed two distinct processes that were ascribed to
successive deprotonation and oxidation of bound choline. With solvent isotope effect and
substrate isotope effect, the results show that the side chain hydrogen bonding is critical for both
proton abstraction and hydride transfer. Greatest insight into the catalytic mechanism is
developed by comparing the rate constant (k5) for hydride transfer with hydrophilicity of the side
233
chain at position 101 in a series of variant enzymes in which serine is replaced by alanine,
threonine, cysteine and valine. This work convincingly demonstrates that general polarity as well
as hydrogen bonding influenced by residue 101 plays a dominate role in catalysis. This study
represents the first instance in which the hydrophilic character of the serine residue proximal to
the C4a and N5 flavin atoms in a flavin-dependent enzyme, besides its hydrogen bonding
capability, has been shown to facilitate the transfer of the hydride ion from the activated alkoxide
to the N5 atom of flavin. The conclusions provide a significant contribution to our developing
understanding of side chain participation in catalytic turnover of flavoproteins.
D-arginine dehydrogenase catalyzes the oxidation of D-amino acids to the corresponding
imino acids, which are hydrolyzed to keto acids and ammonia. Non-covalently linked FAD is
used as a cofactor in catalysis and the native electron acceptor still remains unknown (23). Darginine dehydrogenase, with L-arginine dehydrogenase, has been shown to be a 2-component
amino acid racemase in the D- to L- inversion in Pseudomonas aeruginosa (23).
The steady state kinetic parameters with various amino acids and crystal structures of DADH
indicate that the negatively charged side chain of Glu87 is responsible for the high selectivity of
DADH for positively charged substrates via electrostatic interactions. The other residues in the
active site, such as Tyr53, Met240 and Val242, form hydrophobic walls for specificity of the
long aliphatic and unbranched substrate. Therefore, the electrostatic and hydrophobic
interactions are responsible for selecting D-arginine and D-lysine as the best substrate of DADH.
D-tyrosine, D-phenalanine and D-methionine may form favorable Van der Waals interactions
with the hydrophobic wall of the active site as good substrate.
Moreover, two distinct
conformations with a mobile loop composed of residue 50-56 as observed in the ligand-free and
product-bound crystallographic structures substrate is proposed to control substrate accessibility
234
to the active site. The conformational change of the lid, especially Tyr53, may allow DADH to
accommodate bulky residues like D-phenylalanine or D-tryptophan. Thus, the electrostatic and
hydrophobic interactions bring the substrate into the active site and the lid switch to closed form.
Closing of the lid may facilitate the hydride transfer step by shielding the active site and the FAD
from solvent to increase the overall hydrophobicity of the active site. Further insight into the
function of this active site lid may be obtained by protein engineering studies on DADH. For
example, replacement of Glu87 to a positive charged residue may allow the DADH narrowly
selecting for negatively charged D-amino acid substrates. Replacement of Tyr53 with a large size
residue, tryptophan, may result in big substrates being unable to bind to the active site.
Conversely, replacement of Tyr53 with small size residues may decrease the rate for the hydride
transfer since the active site would be more exposed to the solvent.
The steady state kinetic mechanism of D-arginine dehydrogenase was investigated and
discussed with D-arginine or D-histidine as a substrate. The kinetic investigations show that with
both substrates the enzyme displays a ping-pong kinetic mechanism when PMS is an electron
acceptor. Meanwhile, a dead-end complex of the reduced enzyme with high concentration Darginine yielding substrate inhibition and an isomerization involving multiple conformations of
the initial Michaelis complex with D-histidine were proposed. The kinetic data along with the
structure with iminoarginine or iminohistidine complex will provide a guideline for future
mutagenesis and mechanistic studies aimed at the characterization of the enzyme. The
mechanism for amine oxidation catalyzed by DADH was investigated with leucine as substrate
by using rapid kinetics and mechanistic probes, such as pH effects, substrate and solvent
deuterium KIEs and solvent viscosity effects. Substrate KIE on kred provide a direct probe of the
status of the CH bond, whereas the solvent KIE on kred can be used to provide information on the
235
status of the NH bond. The pH profile suggests kred is pH independent at high pH; in contrast, Kd
is pH independent at low pH. Substrate kinetic isotope effect on kred suggests hydride transfer is
at least partially rate-limiting step. The solvent isotope effect on kred, lack of viscosity effect and
proton inventory studies indicate that a single proton is in flight in the transition state for the
reduction of flavin. The multiple isotope effect on kred is smaller than the product of individual
isotope effect. Based on these results, the mechanism of amine oxidation by DADH is proposed.
The study shows that the oxidation of leucine catalyzed by DADH involves deprotonation of
amino group of the substrate and hydride transfer from α-hydrogen of the substrate to the N5 of
the flavin. The cleavages of the two bonds occur in an asynchronous fashion. The mechanistic
investigation is consistent with amine oxidation by DADH occurring via a hydride transfer
mechanism from an anionic substrate that is formed through the enzyme-catalyzed deprotonation
of the zwitterionic substrate. Molecular dynamics analysis of the effect exerted by the enzyme
active site on the pKa value of the α-NH3 group of leucine corroborates the experimental
findings.
Comparative studies of the DADH crystal structure and the crystal structure of D-amino acid
oxidase have been carried out. Based on the definition, the enzyme with oxygen as an electron
acceptor and hydrogen peroxide as product are called oxidase. Those using other electron
acceptors instead of molecular oxygen are dehydrogenases. Using the oxygen electrode, there is
no oxygen consumption determined with DADH. However, the addition of an artificial electron
acceptor, phenazine methosulfate (PMS), results in the rapid depletion of oxygen since the
reduced PMS is easily oxidized by oxygen. Thus, the DADH is a strict dehydrogenase. The lack
of oxygen reactivity in DADH has been interpreted based on the crystal structures. First, the
accessibility of oxygen may be blocked by residue Ala46 which is right in front of the C(4a)
236
atom of flavin, the site of oxygen reactivity in flavin dependent enzymes. In D-amino acid
oxidase from Rhodotorula gracilis, the corresponding residue Gly52 was investigated by
replacing to valine. The study showed the loss of oxygen reactivity because steric hindrance
prevents the oxygen from reacting with the reduced flavin. However, in L-galactono-γ-lactone
dehydrogenase, the Ala113Gly variant showed a 400-fold increase in kcat/Km,ox compared to the
wild-type enzyme. The oxygen reactivity is very similar to the typical of oxidases reinforcing the
necessity of physical access of oxygen to the reduced flavin. Thus, based on similarities in the
positioning of alanine residues in both the enzymes, the lack of oxygen reactivity in DADH can
due to the physical barrier provided by the Ala46 residue. Therefore, it would be interesting to
replace the residue, glycine, to a smaller residue and study the oxygen reactivity.
The positive charges in the proximity of the C(4a) or N(1)-C(2) atoms of flavin are also
important for oxygen reactivity. This has been established in a number of flavoprotein oxidases,
such as glucose oxidase (24), choline oxidase (1, 25) and monoamine oxidase (26, 27). However,
in the crystal structure of DADH, the positive charged residue or N-terminal end of an α-helix
are all far away from this region. Also, the ping-pong steady state kinetic mechanism rule out the
positive charged substrate/product since the product will be released before the electron acceptor
binds to the active site.
Overall, in this dissertation, the studies on choline oxidase and D-arginine dehydrogenase
have provided insight into the oxidation of amines or alcohols by flavoenzymes. The importance
of a hydrophilic residue with capability of hydrogen bonding proximal to the C(4a)-N(5) flavin
locus in a flavin-dependent enzyme has been established and represents the first instance in
which the hydride ion transfer is affected by the character of the side chain. The high-resolution
crystal structures as well as the kinetic data have highlighted the important structure differences
237
that rationalize the catalytic activities and substrate specificity of the enzymes. The steady state
kinetic and hydride transfer mechanistic studies will provide useful information on the
mechanistic properties on the amine oxidation by other flavin-dependent enzymes.
8.1
References
1.
Gadda, G., Powell, N. L., and Menon, P. (2004) The trimethylammonium headgroup of
choline is a major determinant for substrate binding and specificity in choline oxidase,
Arch Biochem Biophys 430, 264-273.
2.
Fan, F., Ghanem, M., and Gadda, G. (2004) Cloning, sequence analysis, and purification
of choline oxidase from Arthrobacter globiformis: a bacterial enzyme involved in
osmotic stress tolerance, Arch Biochem Biophys 421, 149-158.
3.
Ghanem, M., Fan, F., Francis, K., and Gadda, G. (2003) Spectroscopic and kinetic
properties of recombinant choline oxidase from Arthrobacter globiformis, Biochemistry
42, 15179-15188.
4.
Gadda, G., Fan, F., and Hoang, J. V. (2006) On the contribution of the positively charged
headgroup of choline to substrate binding and catalysis in the reaction catalyzed by
choline oxidase, Arch Biochem Biophys 451, 182-187.
5.
Finnegan, S., and Gadda, G. (2008) Substitution of an active site valine uncovers a
kinetically slow equilibrium between competent and incompetent forms of choline
oxidase, Biochemistry 47, 13850-13861.
6.
Ghanem, M., and Gadda, G. (2005) On the catalytic role of the conserved active site
residue His466 of choline oxidase, Biochemistry 44, 893-904.
238
7.
Ghanem, M., and Gadda, G. (2006) Effects of reversing the protein positive charge in the
proximity of the flavin N(1) locus of choline oxidase, Biochemistry 45, 3437-3447.
8.
Quaye, O., Nguyen, T., Gannavaram, S., Pennati, A., and Gadda, G. (2010) Rescuing of
the hydride transfer reaction in the Glu312Asp variant of choline oxidase by a substrate
analogue, Arch Biochem Biophys 499, 1-5.
9.
Rungsrisuriyachai, K., and Gadda, G. (2008) On the role of histidine 351 in the reaction
of alcohol oxidation catalyzed by choline oxidase, Biochemistry 47, 6762-6769.
10.
Rungsrisuriyachai, K., and Gadda, G. (2010) Role of asparagine 510 in the relative
timing of substrate bond cleavages in the reaction catalyzed by choline oxidase,
Biochemistry 49, 2483-2490.
11.
Quaye, O., and Gadda, G. (2009) Effect of a conservative mutation of an active site
residue involved in substrate binding on the hydride tunneling reaction catalyzed by
choline oxidase, Arch Biochem Biophys 489, 10-14.
12.
Quaye, O., Cowins, S., and Gadda, G. (2009) Contribution of flavin covalent linkage
with histidine 99 to the reaction catalyzed by choline oxidase, J Biol Chem 284, 1699016997.
13.
Finnegan, S., Agniswamy, J., Weber, I. T., and Gadda, G. (2010) Role of valine 464 in
the flavin oxidation reaction catalyzed by choline oxidase, Biochemistry 49, 2952-2961.
14.
Orville, A. M., Lountos, G. T., Finnegan, S., Gadda, G., and Prabhakar, R. (2009)
Crystallographic, spectroscopic, and computational analysis of a flavin C4a-oxygen
adduct in choline oxidase, Biochemistry 48, 720-728.
239
15.
Quaye, O., Lountos, G. T., Fan, F., Orville, A. M., and Gadda, G. (2008) Role of Glu312
in binding and positioning of the substrate for the hydride transfer reaction in choline
oxidase, Biochemistry 47, 243-256.
16.
Fan, F., and Gadda, G. (2005) Oxygen- and temperature-dependent kinetic isotope effects
in choline oxidase: correlating reversible hydride transfer with environmentally enhanced
tunneling, J Am Chem Soc 127, 17954-17961.
17.
Fan, F., and Gadda, G. (2005) On the catalytic mechanism of choline oxidase, J Am
Chem Soc 127, 2067-2074.
18.
Fan, F., Germann, M. W., and Gadda, G. (2006) Mechanistic studies of choline oxidase
with betaine aldehyde and its isosteric analogue 3,3-dimethylbutyraldehyde, Biochemistry
45, 1979-1986.
19.
Gadda, G. (2003) pH and deuterium kinetic isotope effects studies on the oxidation of
choline to betaine-aldehyde catalyzed by choline oxidase, Biochim Biophys Acta 1650, 49.
20.
Gadda, G. (2003) Kinetic mechanism of choline oxidase from Arthrobacter globiformis,
Biochim Biophys Acta 1646, 112-118.
21.
Xin, Y., Gadda, G., and Hamelberg, D. (2009) The cluster of hydrophobic residues
controls the entrance to the active site of choline oxidase, Biochemistry 48, 9599-9605.
22.
Finnegan, S., Yuan, H., Wang, Y. F., Orville, A. M., Weber, I. T., and Gadda, G. (2010)
Structural and kinetic studies on the Ser101Ala variant of choline oxidase: catalysis by
compromise, Arch Biochem Biophys 501, 207-213.
23.
Li, C., and Lu, C. D. (2009) Arginine racemization by coupled catabolic and anabolic
dehydrogenases, Proc Natl Acad Sci U S A 106, 906-911.
240
24.
Roth, J. P., and Klinman, J. P. (2003) Catalysis of electron transfer during activation of
O2 by the flavoprotein glucose oxidase, Proc Natl Acad Sci U S A 100, 62-67.
25.
Finnegan, S., Agniswamy, J., Weber, I. T., and Gadda, G. Role of valine 464 in the flavin
oxidation reaction catalyzed by choline oxidase, Biochemistry 49, 2952-2961.
26.
Miller, J. R., and Edmondson, D. E. (1999) Influence of flavin analogue structure on the
catalytic activities and flavinylation reactions of recombinant human liver monoamine
oxidases A and B, J Biol Chem 274, 23515-23525.
27.
Zhao, G., Bruckner, R. C., and Jorns, M. S. (2008) Identification of the oxygen activation
site in monomeric sarcosine oxidase: role of Lys265 in catalysis, Biochemistry 47, 91249135.