2,
Mechanistic and Physical Studies of Methane
Monooxygenase from Methylococcuscapsulatus(Bath)
by
Katherine E. Liu
B.A., Cornell University (1989)
Submitted to the Department of Chemistry
in Partial Fulfillment of the Requirements for the
Degree of
Doctor of Philosophy
at the
Massachusetts Institute of Technology
February, 1995
© 1995 Massachusetts Institute of Technology
All rights reserved
Signature of Author
Department of Chemistry
November
Certified by
--
--
--
V
U
I/
14, 1994
Professor Stephen J. Lippard
Thesis Supervisor
/q~
Accepted by
Professor Dietmar Seyferth
Chairman, Departmental Committee on Graduate Students
e-;.{ence
.
.
:
-. .; 1QO
D
2
This doctoral thesis has been examined by a Committee of the
Department
of Chemistry as follows:
Professor Alan Davison, Committee Chair
/'--
l'
6/O ProfessorStephenJ. Lippard
Arthur Amos Noyes Professor of Chemistry
PA fessor JoAnne Stubbe
John G. Sheehan Professor of Chemistry
Z n
A n
P
Professor 4wliam H.
me Johnson
3
Acknowledgments
There are many people who helped make it possible for me to write
this thesis. First and foremost, I must thank Steve Lippard for providing me
the opportunity to join his research group at MIT. He allowed me to work
independently, yet provided me much support and encouragement over the
last five years. Studying the MMO system has led us down a variety of paths,
and he was never hesitant to let me pursue new directions. I would also like
to thank JoAnne Stubbe, William Orme-Johnson, and Alan Davison for
serving on my thesis committee.
I was fortunate to collaborate with may people away from MIT on
several projects. Vickie DeRose and Brian Hoffman are responsible for many
exciting EPR discoveries. Marty Newcomb fought along side us against the
mechanistic dogma we encountered.
My trips to the National Tritium
Labeling Facility were made more enjoyable by Barrie Wilkinson, from Heinz
Floss' group, Phil Williams, and Hiromi Morimoto. Ah, the excitement of
what sandwich to order for lunch! Southern hospitality was provided by
Vincent Huynh, Dale Edmondson, and Danli Wang at Emory University in
Atlanta, GA, who introduced me to rapid freeze-quench techniques.
Someday we will have to go Tapas Bar Hopping!
Back at MIT, I wish to thank Ann Valentine for enthusiastically
accepting to continue the mechanistic work on MMO. I also appreciate her
help ill purifying protein and carrying out some kinetic experiments at the
end. I am sure she saved me many months in producing this thesis! Thanks
goes to Amy Rosenzweig, who worked with me on those grueling protein
purfications in the early days, and to Thanos Salifoglou for his great
fermentation efforts. Axel Masschelein expedited the stopped-flow work by
assembling our equipment and designing our system software. I am also
grateful to the other members of the MMO subgroup who have given me
support along the way: David Coufal, Sonja Komar-Panicucci, and Andrew
Feig. The above mentioned people made both my professional and personal
life at MIT more enjoyable.
There are many more names to include who have given me fond
memories to take away with me. Fishing trips and Thanksgiving dinners
were made special by Mike Keck and Ted Carnahan. I must thank Pieter Pil
and Steve Brown for toughening me after the move to the second floor, for
4
participating in numerous waterfights (let's get Pieter!), and for providing
daily banter upstairs, along with Megan McA'Nulty and Joyce Whitehead.
Many thanks go to JoAnne Yun and Maria Bautista, who fed me the
Hlawaiian pizza that made it possible for me to finish this thesis, followed by
the 52's that made it possible for me to celebrate finishing this thesis. I am
also grateful to Susanna Herold for her boomerangs and Kevin Ward for his
brownies. Other fun times were provided by Jonathan Wilker (see you at
Boston '96!), Rajesh Manchanda, and Deborah Zamble (thanks for help with
the move, too). David Goldberg has provided much support during the last
five years, and I am thankful for his friendship.
I wish to express my appreciation toward my family: my father, Sally,
Kim and Dave, David and Cheryl, and Mike and Wendy, all of whom, I am
sure, were convinced I dropped off the face of the planet while writing this
thesis. Believe it or not, I will actually leave MIT!! Finally, I must thank the
person who has had by far the deepest impact on both my personal and
professional life, John Protasiewicz. I am positive that I could not have made
it this far without him. I would express amazement at how one person could
have provided such a constant stream of love and support, but I know it all
comes so natural for such a warm and kind individual (I am afraid, though,
that you will never fully understand how much I love you).
Mechanistic and Physical Studies of Methane
Monooxygenase from Methylococcus capsulatus (Bath)
by
Katherine E. Liu
Submitted to the Department of Chemistry on November 14, 1994
in partial fulfillment of the requirements for the Degree of
Doctor of Philosophy in Chemistry
ABSTRACT
In Chapter 1, progress in understanding the sMMO systems of both M.
Physical
capsulatus (Bath) and M. trichosporium OB3b are reviewed.
parameters of the hydroxylase component determined from a variety of
spectroscopic techniques are discussed. Interactions of the hydroxylase with
the protein B and reductase components are outlined.
Included in this
discussion is how the other proteins affect both the physical characteristics
and the reactivity of the hydroxylase. Studies to determine how the enzyme
reacts with dioxygen in the presence and absence of substrate are presented,
and mechanistic proposals for the hydroxylation reaction are given.
The reduction potentials of the hydroxylase component of the soluble
methane monooxygenase from Methylococcus capsulatus (Bath) are
investigated through potentiometric titrations in Chapter 2. The potentials
were determined by EPR spectroscopic quantitation of the mixed valent form
of the hydroxylase as a function of added sodium dithionite in the presence of
electron transfer mediators. Addition of substrate has little effect on the
potentials, but protein B causes a dimunition of both potentials. The presence
of both protein B and reductase inhibits reduction of the diiron center in the
absence of substrate. When the substrate propylene is added to this system,
reduction is greatly facilitated. These results reveal aspects of how both
protein B and substrate can regulate electron transfer into and out of the
hydroxylase component of methane monooxygenase.
Chapter 3 describes studies the sMMO system by ENDOR spectroscopy.
Information about the coordination sphere of the iron atoms both in the
absence and presence of substrates and inhibitors is presented. Most
significantly, these studies identify a proton resonance from the bridging
hydroxide ligand present in the mixed-valent form of the hydroxylase.
In Chapter 4, studies in which five mechanistic probes were used as
substrates to investigate the possible formation of radical intermediates in the
catalytic cycle of the sMMO system are outlined. No ring-opened products
were observed, which suggests that the hydroxylation reaction may not
proceed through substrate radical formation. A lower limit of 1013 s- 1 was
calculated for the rate constant of a radical rebound process to account for the
6
products recovered in the reaction. The possibility of a stereoelectronic
barrier to ring opening in the active site of the enzyme is addressed through a
semi-quantitative analysis of the rate constant with Marcus theory. Kinetic
isotope effect experiments were also carried out. No intermolecular isotope
effect was observed, which is consistent with C-H bond breaking not being
involved in the rate determining
step of the enzymatic reaction.
Intramolecular kinetic isotope effects of kH/kD = 5 were obtained, indicating
that the hydroxylation step of the reaction involves a substantial C-H bond
stretching component.
These results are discussed in terms of several
possible detailed mechanisms for the MMO hydroxylase reaction.
In Chapter 5, the use of chiral, tritiated hydrocarbons as substrates with
MMO is described. These studies indicate that the stereoselectivity of the
hydroxylation reaction depends on whether hydroxylation occurs at the
proton or deuterium atom of the chiral carbon. A significant intermolecular
kinetic isotope effect of kH/kD = 4 was obtained in most cases. The
intramolecular kinetic isotope effects were much lower, and even inverted in
some cases. An exchange mechanism in which the product alcohol reacts
further in the active site prior to dissociation is invoked to reconcile the
different stereoselectivity and the kinetic isotope effects observed with two
enantiomers of the same substrate. Mechanistic implications for the overall
hydroxylation reaction are discussed.
The reactivity of the reduced hydroxylase protein (Hred) toward
dioxygen is outlined in Chapter 6. Included in this work are the effects of
protein B and the reductase on the regioselectivity, product yield, and rate
constant of the hydroxylation reaction. Two intermediates in this reaction,
designated L and Q, were observed. Kinetic parameters from optical stoppedflow and rapid freeze-quench M6ssbauer data indicate that Hred reacts with
dioxygen initially to form L, which is subsequently converted to Q. The decay
of Q occurs with concomitant production of the product alcohol and oxidized
hydroxylase (Hox). The Mdssbauer parameters for both intermediates are
unusual and do not match the values of any known diiron carboxylatebridged model compounds. Based on an isomer shift of 0.66 mm/s, XmaxO
of
625 nm, and an 1 8 -sensitive resonance Raman feature at 902 cm -1, L is
assigned as a diiron(II) peroxide species. Intermediate Q exhibits a low
isomer shift -=0.15 mm/s and is consistent with an Fe(IV) species. Proposed
hydroxylation mechanisms are discussed in view of these results.
Thesis Supervisor: Dr. Stephen J. Lippard
Title: Arthur Amos Noyes Professor of Chemistry
7
Table of Contents
Page
Acknowledgments
Abstract .
....................................................
.......................................................
................................................... ..........................................................................
Table of Contents.......................................................................
List of Tables .................................................................................................
List of Figures...............................................................................................
3
5
7
12
14
Chapter 1. Studies of the Soluble Methane Monooxygenase Protein
System: Structure,Component Interactions, and Hydroxylation Mechanism
Introduction ...
...................................................
20
Structural Studies of sMMO Hydroxylase ......................................................
21
Component Interactions ......................................................
24
Mechanism of Hydroxylation ......................................................
30
Reaction of Hred with Dioxygen......................................................
30
Reaction of Q with Substrate......................................................
35
Conclusions ................................................................................................................... 40
References .......................................................................................................................41
Tables.........................................................................................................
............ 48
Figures ..........................................................................................................
52
Chapter 2. Redox Properties of the Hydroxylase
Component of Methane Monooxygenase from Methylococcus
capsulatus (Bath) -- Effects of Protein B, Reductase, and Substrate
Introduction.....................................................................................................................
64
Experimental
......................................................
66
Protein Isolation......................................................
Redox Titrations ......................................................
66
66
EPRSpectroscopy.
67
.....................................................
Results...........................................................................................................
Redox Titrations ......................................................
Component Interactions ......................................................
Discussion.....
.................................................
Hydroxylase Potentials......................................................
Literature Comparisons ......................................................
68
68
69
70
70
72
8
72
74
Structural Implications....................................................
Component Interactions....................................................
References
....................................................
78
Tables ...............................................................................................................................81
Figu res ............................................................................................................
83
Chapter 3. Electron Nuclear Double Resonance Studies of
the Dinuclear Iron Center in the Hydroxylase Component of
Methane Monooxygenase from Methylococcus capsulatus (Bath)
Introduction ....................................................................................................................
98
Basics of EPR and ENDOR Spectroscopy ....................................................
Experimental
....................................................
99
101
Native Samples....................................................
101
15 N
Enrichment ....................................................
Enrichment ....................................................
102
102
D20 Exchange....................................................
103
5 7Fe
H2
17 0
Exchange ....................................................
103
Sample Preparation ....................................................
Addition of Protein B....................................................
Addition of Inhibitors/Substrates ....................................................
ENDOR Spectroscopy....................................................
Results and Discussion....................................................
EPR Spectrum....................................................
1H ENDOR Spectra....................................................
103
104
104
105
105
105
105
14 N
and 15 N ENDOR Spectra ....................................................
57 Fe ENDOR Spectra ....................................................
107
109
Samples with Isotopically-Labeled DMSO ................................................... 109
Other Samples....................................................
110
Conclusions...............................................................................................................
111
R eferences ......................................................................................................................112
Tables .....................................................................................................................
1......15
Figures .......................................................................................
......................... 116
Chapter 4. Radical Clock Substrate Probes and Kinetic Isotope Effect
Studies of the Hydroxylation of Hydrocarbons by Methane Monooxygenase
Introduction ......
..
Experimental
........
.......................................................... ...........
....... 147
149
9
Isolation and Purification of Proteins.................................................
149
Synthesis and Characterization of Substrates and Products .......................150
Enzymatic Reactions.................................................
152
Instrumentation ................................................. ...........
...... 152
Product Identification and Quantitation .................................................. 153
Isotope Effect Studies..................................................
154
Results............................................................................................................. 155
Reactions with 1.................................................
156
Reactions with 2.................................................
157
Reactions with 3.................................................
157
Reactions with 4.................................................
157
Reactions with 5..................................................
158
Discussion...................................................................................................
159
Analysis of Products Formed..................................................
159
Deuterium Isotope Effects..................................................
161
Mechanistic Considerations .................................................
163
Formation of the Dioxygen-Activated Diiron Center........................163
The Hydroxylation Reaction..................................................
165
Comparisons with Related Work..................................................
172
Conclusions ..................................................................................................................173
Acknowledgments
..................................................
174
Referen ces............................. .....................................................................
175
Tables ..............................................................................................................................182
Figures ..............................................................................................................
.... 185
Chapter 5. Tritiated Chiral Alkanes as
Substrates for Methane Monooxygenase from Methylococcus
capsulatus (Bath): Probes for the Mechanism of Hydroxylation
Prefa ce............................................................................................................................211
Introduction
...............................
211
Experim enta l.................................................................................................................214
Bacterial Growth and Fermentation ............................
Reagent Synthesis...............................
...................... 214
216
(S)-[l- 1 H,1-2H]-Ethanol ........................................
216
(S)-[1-1H,1-2 H]-Ethyl Tosylate
216
.................................
(S)-[1-2 H,1- 3 H]-Ethane .............. .................
Enzymatic Reactions...............................
217
217
10
Derivitization
.......................................................
218
Product Analysis.......................................................
Results .......................................................
Analysis of Lypholysates .......................................................
Analysis of Derivatives.......................................................
219
219
219
220
Discussion
.......................................................
223
Analysis of Product Distributions .......................................................
223
Alkanes are not Highly Restricted in the Active Site............................
223
H,T Vs D,T Products and Exchange at the -Carbon ............................
224
Kinetic Isotope Effects.......................................................
227
Mechanistic Implications.......................................................
229
Comparisons with Literature.......................................................
231
Concluding Remarks...................................................................................................232
References ................................................................................................. ...................
234
Tables ..............................................................................................................................237
Figures ...........................................................................................................................243
Chapter 6. Intermediates in, and Kinetic
Studies of the Reaction of Reduced Hydroxylase of
MMO from Methylococcus capsulatus (Bath) with Dioxygen.
Introduction ............................................................................................................... 290
Experimental .................................................................................................................292
Bacterial Growth and Protein Purification ....................................................
292
Generation of H red............................................................................................293
Single Turnover Reactions.........................................................................294
Kinetic Studies........................................................
294
Product Distributions and Yields........................................................294
KNl of Nitrobenzene ........................................................
297
Stopped-Flow Spectrophotometry ........................................................ 297
Rapid Freeze-Quench Studies ........................................................
299
EPR Spectroscopy........................................................
299
M6ssbaiuer Spectroscopy........................................................
299
Resonance Raman Spectroscopy........................................................
300
Results............................................................................................................... 300
Studies of Nitrobenzene Hydroxylation .......................................................
300
Product Distributions and Yields in Single Turnover Expts .............300
Kinetic Analysis of the Hydroxylation of Nitrobenzene ....................301
11
Intermediates Formed in the Reaction of Hred with Dioxygen..................302
Spectroscopic Characterization of Intermediates................................
302
Kinetic Analysis ...................................................
303
Discussion
. ................................................... 305
MMO Component Interactions....................................................
305
Spectroscopic Studies of Intermediates ....................................................
Conclusions .........................................................................................................
310
318
Acknowledgments
...................................................
References
...................................................
319
320
Tables .............................................................................................................................324
Figu res...........................................................................................................................328
Biography
...............................
381
12
List of Tables
Chapter 1.
Table 1.
Physical Properties of Hox from M. trichosporium OB3b
and M. capsulatus(Bath)...................................................
Table 2. Reduction Potentials of the MMO Hydroxylase from
48
M. capsulatus (Bath) and M. trichosporium OB3b.......................................49
Table 3. Product Distributions from Reactions of
MM() with Nitrobenzene..................................................
50
'Table4. Spectroscopic and Kinetic Parameters for Intermediates
in the Reactions of Hred with Dioxygen .................................................. 51
Chapter 2.
Table 1. Reduction Potentials of the MMO Hydroxylase
from M. capsulatus (Bath) and M. trichosporiumOB3b
Under Various Conditions ..................................................
Table 2. Reduction Potentials of the Mediators
Employed in the Titrations ..................................................
81
82
Chapter 3.
Table 1.
1H Hyperfine Coupling Constants for
Table 2.
14 N
Hmx and Hmv with DMSO..................................................
and
15 N
115
Coupling Constants of Hmv with
and without DMSO..................................................
115
Chapter 4.
Table 1. Instrumental Conditions for GC
and GC/M S Analyses...................................................
Table 2. Retention Times of Reactants and Products in the
182
Reaction of 1 - 5 with MMO Hydroxylase ............................................... 183
Table 3. Expected Products and Observed Ratios for Reactions
with MMO from M. capsulatus(Bath).................................................. 184
Chapter 5.
Table 1. Component Amounts Used In Enzymatic
Reactions at NTLF...................................................
237
13
Table 2.
Results of MMO Reactions with (S) and
(R)-[1-2 H ,1-3 H ]-Ethane ................. .................................................
238
Table 3. Results of MMO Reactions with (S) and
Table 4.
Table 5.
(R)-j[1-2 H ,1-3 H ]-Butane .............. ......................................................
239
Results of MMO Reactions with (R), (S), and
Racemic [2- 3 H]-Butane .....................................................
Foul Sets of Conditions for which Quantitative
240
Exchange Reactions are Calculated..................................................... 241
Chapter 6.
Table 1. Product Distributions from Reactions of MMO
with Nitrobenzene ......................................................
324
Table 2. Mossbauer Parameters of Species Detected in Rapid Freeze
Quench Samples from the Reaction of Hred with Dioxygen................325
Table 3.
Rate Constants for Formation and Decay of Intermediate
Table 4.
Q Under a Variety of Conditions...................................................... 326
Activation Parameters from Arrhenius and Eyring
Plots of the Reaction of Hred with Dioxygen..........................................
327
14
List of Figures
Chapter 1.
Figure 1.
Active Site Structure of Hox
as Determined by X-ray Crystallography ............................................ 52
Figure 2.
Possible Structures for a Diiron(III) Peroxide
Unit in Intermediate L.............................................
54
Proposed Route for Conversion of Intermediate L
to Intermediate Q.............................................
56
Proposed Catalytic Cycle for the Hydroxylation
of Hydrocarbons by MMO.............................................
58
Possible Mechanisms for Hydroxylation by sMMO.............................60
Figure 3.
Figure 4.
Figure 5.
Chapter 2.
Figure 1.
X-band EPR Spectrum of Mixed-Valent
Hydroxylase
.............................................
Figure 2.
X-band EPR Spectrum of Reduced
Hydroxylase
.............................................
Figure 3.
Figure 4.
Figure 5.
Figure 6.
Figure 7.
83
85
EPR Spectra of the Hydroxylase at Different
Reduction Potentials from a Reductive Titration..................................
87
Plot Quantitating the Amount of Hydroxylase
Present in Each Oxidation State with Conditions A..............................
89
Plot Quantitating the Amount of Hydroxylase
Present in Each Oxidation State with Conditions B and C...................91
EPR Spectra of the Hydroxylase at Various Potentials
in the Presence of Protein B and Reductase...........................................
93
X-band EPR Spectrum of Mixed-Valent Hydroxylase
Formed Under Conditions B.............................................
95
Chapter 3.
Figure 1. Energy Level Diagram of a Free Electron
Figure 2.
Figure 3.
in an Applied Magnetic Field.............................................
116
Energy Level Diagram of a Paramagnetic Center
Further Split by the Influence of a Nuclear Spin.................................
118
Examples of ENDOR Spectra
for Differing A Strengt
hs.........................................................................120
15
Figure 4
Q-band EPR Spectra of Hmv...................................................................122
Figure 5.
Q-band EPR Spectra of Hmv in the
Presence of DMSO....................................
Figure 6.
1H
Figure 7.
1H
124
ENDOR Spectra of Hmv at
Various g-Values................................
126
ENDOR Spectra of Hmv
at g = 1.94....................................
Figure 8.
1H
Figure 9.
14 N
Figure 10.
1 4N
ENDOR Spectrum of Hmv
in the Presence of DMSO at g = 1.87...........................
and
15N
128
130
ENDOR Spectra
of H m v........................................................................................................132
and
15 N
ENDOR Spectra
of Hmv in the Presence of DMSO...........................
Figure 11.
57 Fe ENDOR
Spectra of the Fe(III)
Site at Various g Values..................................
Figure 12.
57 Fe
134
136
ENDOR Spectra of the Fe(II)
Site at Various g Values ........................................
138
Figure 13. ENDOR Spectra of Hmv in the
Presence of Isotopically-Labeled DMSO...........................
140
Figure 14. Mims Pulsed ENDOR Spectra of Hmv
142
in the Presence of Isotopically-Labeled DMSO...................................
Figure 15.
2H
ESEEM Spectra of Hmv
in the Presence of DMSO....................................
144
Figure 2.
Reactions after Abstraction of a Hydrogen
Atom from 1.......................................
Substrates 1 through 5.......................................
185
187
Figure 3.
Typical GC Traces from Reactions of MMO
Chapter 4.
Figure 1.
Figure 5.
with Probes 1 and 2.......................................
Formation of Product Alcohol
over Time....................................
Formation of 2a over Time...........................
Figure 6.
Product Formation of Reactions of MMO
Figure 4.
w ith 4.......................................
189
191
193
195
16
Proposed Catalytic Cycle for the Hydroxylation
of Hydrocarbons by MMO..............................................
Figure 8. Possible Reactions of a Diferric Peroxide Intermediate
in the Catalytic Cycle..............................................
Figure 9. Possible Modes of Peroxide Binding..............................................
Figure 10. Possible Mechanisms for the Hydroxylation
Reaction..............................................
Figure 11. Reaction Pathways after Abstraction of a Hydrogen
Atom from 5..............................................
Figure 7.
197
199
201
203
206
Figure 12. Plot of kr as a Function of the Exergonicity of the
Ring-Opening Reaction.............................................
208
Chapter 5.
Figure 1.
Figure 2.
Figure 3.
Possible Reaction of MMO with
2H,1-3H]-Ethane......................................................
(S)-[1-
243
Reaction Apparatus Used at NTLF..............................................
Possible Reaction Products from Reactions with
245
MMO and (S)-[1-2 H,1- 3 H]-Ethane ...................................................... 247
Figure 4.
1H NMR
Spectrum of the Lypholysate from the
Reaction of MMO with (S)-[1-2 H,1- 3 H]-Ethane ...................................249
Figure 5.
3 H NMR
Spectrum of the Lypholysate from the
Reaction of MMO with (S)-[1-2 H,1- 3 H]-Ethane
Figure 6.
.............................251
Reaction Scheme of the Derivitization
Procedure...................................................................................................253
Figure 7.
3H
NMR Spectrum of the Derivitized Lypholysate from
the Reaction of MMO with (S)-[1-2 H,1- 3 H]-Ethane .............................255
Figure 8.
3H
NMR Spectrum of the Derivitized Lypholysate from
the Reaction of MMO with (S)-[1-2 H,1-3 H]-Ethane (Zoom) ...............257
Figure 9.
3H
NMR Spectrum of the Lypholysate from the
Reaction of MMO with (S)-[1-2 H,1-3 H]-Butane ....................................259
Figure 10.
3H
NMR Spectrum of the Derivitized Lypholysate from
the Reaction of MMO with (S)-[1-2 H,1- 3 H]-Butane .............................261
Figure 11.
3H
NMR Spectrum of the Derivitized Lypholysate from the
Reaction of MMO with (S)-[1-2 H,1- 3 H]-Butane (Zoom) .....................263
Figure 12. 3H NMR Spectrum of the Derivitized Lypholysate from the
Reaction of MMO with (S)-[2-3 H]-Butane (Zoom) .............................. 265
17
Figure 13. 3 H NMR Spectrum of the Derivitized Lypholysate from the
Reaction of MMO with Racemic [2-3 H]-Butane (Zoom) ....................267
Figure 14. Possible Alcohol Products after Reactions
of (R)-[2-3 H]-Butane with MMO ...............................................
Figure 15. Hypothetical Orientation of
in the Active Site...............................................
Figure 16. Hypothetical Orientation of
Figure 17.
Figure 18.
Figure 19.
Figure 20.
269
(S)-[1-2 H,1-3 H]-Ethane
271
(S)-[1-2 H,1-3 H]-Butane
in the Active Site...............................................
273
Hypothetical Orientation in the Active Site of a Product
Alcohol which is Undergoing and Exchange Reaction.......................275
Exchange Reaction Interconverting
277
the H,T Products ...............................................
Exchange Reaction Interconverting
279
the D,T Products ...............................................
Illustration of the Dependence of C2/C1
on k H/k D ...................................................................................................281
Figure 21. Effects of the Exchange Reaction on the
Intermolecular Kinetic Isotope Effect....................................................283
Figure 22. Two Possible Mechanisms for Hydroxylation
by MMO...............................................
Figure 23. Precursors to the Four Products Observed
by 3H NMR Spectroscopy...............................................
285
287
Chapter 6.
Figure 1.
Schematic of the Rapid Freeze-Quench Apparatus ............................328
Figure 2.
HPLC Traces of Single Turnover Reactions of Hred
Figure 4.
330
with Nitrobenzene and Dioxygen.........................................
Plot of the Yield of Single Turnover Reactions
as a Function of Protein B Concentration ......................................... 332
Kinetic Analysis of the Catalytic Reaction
Figure 5.
of the MMO System with Nitrobenzene .........................................
Kinetic Traces at 404 nm of Single Turnover
Figure 3.
334
Reactions with and without Protein B...................................................
336
Figure 6.
Plot of Rate Constants of Single Turnover Reactions
Figure 7.
as a Function of Protein B Concentration ......................................... 338
M6issbauerAnalysis of Intermediate L.......................................340
18
Figure 8.
Mdssbauer Spectrum of a 3 s Rapid
Figure 9.
Freeze-Quench Sample....................................................
342
344
Mbssbauer Analysis of Intermediate Q................................................
Figure 10. Mtssbauer
Spectra of Hox and Hred.................................................... 346
Figure 11. High Field M6ssbauer Spectra
..............
of fIred, L, and Q......................................
349
Figure 12. Kinetic Trace of Intermediate L at 625 nm ............................................351
353
Figure 13. Optical Spectrum of Q over Time ....................................................
Figure 14. EPR Spectrum of a 155 ms Rapid
Freeze-Quench Sample....................................................
355
Figure 15. EPR Spectrum of a 8 s Rapid
357
Freeze-Quench Sample....................................................
Figure 16. Resonance Raman Spectra of Rapid Freeze-Quench
Samples from the Reaction of Hred with Dioxygen.............................359
Figure 17. Resonance Raman Spectra of L
with Excitation at 647 nm........................................................................361
Figure 18. Plots of the Percent M6ssbauer Area of
L, Q, and Hox as a Function of Time ................................................... 363
Figure 19. Protein B Dependence of the Kinetic Traces of the
365
Reaction of Hred with Dioxygen at 420 nm...........................................
'Figure 20. Plot of the Rate Constants for Q Formation and
Decay Vs Dioxygen Concentration ....................................................
Figure 21. Plot of the Rate Constants for Q
Formation and Decay Vs pH ....................................................
367
369
Figure 22. Kinetic Trace of the Reaction of D20-Exchanged
Hred with Dioxygen at 420 nm....................................................
Figure 23. Arrhenius and Eyring Plots for the Rate Constants for
the Formation and Decay of Q....................................................
371
373
Figure 24. Kinetic Trace of the Reaction of Hred with Dioxygen
in the Presence of Methane at 420 nm...................................................
375
377
Figure 25. Proposed Structures for Intermediates L and Q..................................
Figure 26. Proposed Catalytic Cycle for the Oxidation of
Hydrocarbons by MMO....................................................
379
19
CHAPTER
1.
Studies of the Soluble Methane Monooxygenase Protein
System: Structure, Component Interactions, and Hydroxylation Mechanism.
20
Introduction
Methanotrophic bacteria have the fascinating ability to use methane as
their sole source of carbon and energy.1 These organisms play a crucial role in
the atmospheric balance of this greenhouse gas and in the overall global carbon
cycle by annually consuming billions of tons of methane. 2' 3
Because these
bacteria are capable of converting methane into methanol at ambient pressure
and temperature, methanotrophs have received much attention in the search for
alternative methods for methanol synthesis,4 which are currently not cost
efficient and require high temperatures and pressures. 5 In addition to methane,
these organisms can metabolize a wide variety of other substrates,6
-8
which has
spurred interest in their bioremedial applications. 91 3
Methanotrophs rely on the enzymatic system methane monooxygenase
(MMO) for their first metabolic step, as shown in equation (1).1,14 The MMO
systems in two organisms, a Type I methanotroph 15 Methylococcuscapsulatus
(Bath) and a Type II methanotroph
CH4 + 0: + H+ + NADH
15
Methylosinus trichosporiumOB3b, have
-
CH3OH + H 2 0 + NAD+
been the focus of intense research in recent years.
(1)
Progress toward
understanding the structure of the system, the interactions between the three
components, and the hydroxylation mechanism are reviewed here.
In most methanotrophs, two distinct forms of MMO can exist in the cell,
with the dominant species depending on the bioavailability of copper.16 The
particulate MMO (pMMO) is a membrane bound system which has been difficult
to study because of its instability outside the cell. Recent progress has been made
in the isolation and characterization of pMMO from M. capsulatus (Bath), the
activity of which was found to depend on the copper concentration within the
21
membrane. 171 8 EPR and magnetic susceptibility studies indicate that pMMO
contains an exchange-coupled trinuclear Cu cluster.18 Since the EPR signal of the
chemically reduced tricopper cluster changed to that of the oxidized signal upon
exposure to dioxygen, this unit is believed to occur at the active site of pMMO.1 8
Under conditions of copper stress, some methanotrophs can express a
cytosolic, soluble form of MMO (sMMO),19-2 2 the properties of which form the
focus of this review. The sMMO systems from both M. capsulatus (Bath) and M.
trichosporiumOB3b have been well studied, and comprise three separate protein
components which have all been purified to homogeneity.2 3' 24 The hydroxylase
component, a 251 kD protein, contains two copies each of three subunits in an
ca2P2Y2
configuration. The c subunit of the hydroxylase contains a dinuclear iron
center 2 5 responsible for dioxygen activation and for hydroxylation of the
substrate.
26
The 38.6 kD reductase contains an FAD and an Fe2S2 cofactor27
which enable it to relay electrons from NADH to the diiron center in the
hydroxylase.2 8 2 9 The third component, protein B, contains no cofactors.3 0 This
small, 15.5 kD protein interacts with hydroxylase and reductase in several ways.
Structural Studies of sMMO Hydroxylase
The hydroxylase component of sMMO belongs to a family of proteins
which contain a non-heme, carboxylate-bridged, dinuclear iron unit at their
active site.31 -34
Included in this family are the proteins hemerythrin,
ribonucleotide reductase, and purple acid phosphatase. Understanding how the
diiron unit is tuned in each protein to exhibit such diverse functionality, ranging
from the reversible binding of dioxygen in hemerythrin to activation of dioxygen
for converting methane to methanol in MMO, is a primary goal of research in this
area. Emerging features of the non-heme systems can be compared to related
structural and functional information about the analogous heme proteins
hemoglobin and cytochrome P-450, which have been more extensively studied.
22
The diiron center in the sMMO hydroxylase has been investigated by a
variety of physical techniques.
Properties of the hydroxylase from both M.
capsulatus(Bath.)and M. trichosporiumOB3b in three oxidation states are listed in
Table 1. The enzyme is isolated in its native Fe(III)Fe(III) oxidation state (Hox).35
M6ssbauer parameters are consistent with the presence of two high spin iron(III)
atoms,36 37 and are slightly perturbed with increased pH.3 7
Because of
antiferromagnetic coupling between the Fe(III) atoms, the ground state of Hox is
EPR silent. 3 6 3 7 EPR signals at g = 4.3 and 2.0, together which account for < 5 %
of total iron, are present in spectra of the enzyme in all oxidation states and are
assigned to adventitious Fe(III) and a protein-associated free radical,
respectively. The S = 2 excited state of Hox displays an integer spin EPR signal at
g = 8 (J = -8 cm- ' , where H-= -2JS1S2 ).3 7 Since the oxo-bridged diiron(III) units in
hemerythrin and the R2 subunit of ribonucleotide reductase exhibit coupling
constants of -135 and -110 cm 1, respectively, the small J value suggested that an
oxo-bridge is not present in Ho. 37 This hypothesis was in accord with previous
EXAFSspectral studies36
38
in which no short Fe-O distance was found and with
the lack of any optical features above 300 nm in any oxidation state.36 A 2.2 A
resolution x-ray crystal structure of Hox from M. capsulatus(Bath)3 9 is consistent
with the occurrence of a bridging hydroxide ligand (vide infra), and the active
site structure derived from this analysis is illustrated in Figure 1. Two additional
bridges, a bidentate glutamate, and an exogenous acetate from the crystallization
conditions, were also identified. The Fe...Fedistance was determined to be 3.4 A.
Each iron atom is ligated by one histidine nitrogen atom. The remaining
coordination sites are occupied by two monodentate glutamate residues on one
iron atom, whereas the second iron atom contains one monodentate glutamates
and a water ligand in its coordination sphere.
23
The mixed-valent Fe(II)Fe(III) oxidation state (Hmv) is readily accessible
by one-electron reduction of the dinuclear center. Mbssbauer data indicate the
presence of one Fe(III) and one Fe(II) center.37 Hmv gives rise to a rhombic EPR
signal with gav = 1.8326,35 (J = -30 cm-1),36 ' 3 7 which is characteristic of the EPR
signals from the other mixed-valent non-heme carboxylate bridged diiron
proteins such as hemerythrin
(J = -15 cm-1).31 ENDOR spectroscopic studies of
Hm,, from M. capsulatus (Bath) provided direct evidence for a hydroxide bridge
and a terminal water ligand of the diiron unit.4 0 The hyperfine parameter of the
hydroxide proton was highly unusual, displaying an extremely large and
anisotropic coupling with the iron center. Exchange with D2 0 demonstrated that
the proton was derived from solvent. Spectra with semimetazido hemerythrin
displayed a similar feature, and confirmed the assignment since it is known from
a variety of other methods to contain a bridging hydroxide ligand.4 1 4 4 A pulsed
EPR report of Hmv from M. trichosporium OB3b was consistent with these
conclusions.4 5 Extensive ESEEMand ENDOR studies of Hmv have been carried
out.4 6-4 8 These experiments provided extensive characterization of the diiron
core examining
14 N, 15 N, 1 H, 2 H, 1 3C,
and
57 Fe
nuclei in native Hmv. The
inhibitor DMSO4 8 provided information about the geometry of the substrate
binding site in Hmv.46' 4 8 Mims pulsed ENDOR of Hmv treated with d6 -DMSO
suggested that the methyl groups bind the diiron unit asymmetrically.4 6
Further reduction of Hmv results in formation of the fully reduced,
Fe(II)Fe(II)protein (Hred). Mbssbauer features of Hred from both organisms are
best fit as two unique quadrupole doublets.3 7 4 9 For M. trichosporiumOB3b, the
doublets were interpreted as indicating the presence of two distinct Fe(II) sites in
the diiron center.3 7 For M. capsulatus (Bath), one doublet was attributed to
dinuclear Fe(II) centers in the protein that react with dioxygen in a nonproductive manner.4 9 Hred from M. trichosporiumn
OB3b and both forms of Hred
24
from M. capsulatus (Bath) exhibit an EPR feature at g = 1526,36arising from a
ferromagnetically coupled, S = 4 spin system, with the two lowest spin levels
being nearly degenerate.5 0 Identification of nitrogen donor ligands was
demonstrated through pulsed ENDOR and ESEEM spectra of Hred from M.
capsulatus (Bath).51 These studies provided the first advanced EPR study of a
non-Kramers doublet spin system, and suggest the possibility of investigating
similar signals from other proteins in this family. EXAFS spectroscopy did not
exhibit Fe...Fe backscattering in Hred.36
This observation could result from
thermal disorder, or it may suggest that, as a consequence of the reduction of Hox
to Hred, the Fe...Fe distance has increased significantly. MCD experiments
indicate that both iron atoms are five-coordinate and are ferromagnetically
coupled with J = 0.3 to 0.5 cm-1.5 2 Comparison with deoxyhemerythrin,
antiferromagnetically
coupled
(J -12 to -36 cm-1)41'
42 5 3 5 4
which is
and contains
a
hydroxide bridge, led to the hypothesis that the hydroxide bridge present in Hox
and Hmv is converted to a water bridge in Hred.52
Component Interactions
Although the structure of the hydroxylase is now reasonably well
understood, less is known about the interactions among the three component
proteins of MIMO. Despite the fact that the physical properties of the M.
capsulatus (Bath) and M. trichosporium OB3b hydroxylase are very similar,
preliminary work with the other components indicates that significant
differences exist. The interactions among the component proteins are quite
complex, as manifest by the regulation of electron transfer to the hydroxylase, the
product yields and regioselectivity of the hydroxylation reaction, and the
detailed kinetic behavior of the systems.
In the M. capsulatus (Bath) system, all three components are necessary to
obtain turnover with NADH as the reductant.5
5
With the M. trichosporiumOB3b
25
system, protein B is apparently not necessary.26 Instead, in this latter system
protein B increases the initial rates of the catalytic hydroxylation reaction.2 6
Catalysis can be achieved by means of a shunt pathway with hydrogen peroxide
and Hox,,alone from both organisms.56-58 The efficiency of the shunt pathway
varies significantly, however. With M. trichosporiumOB3b, alcohol yields greater
than those obtained with the complete reconstituted system were observed.5 6
Upon addition of protein B, however, the initial rates were diminished, behavior
opposite to that observed with the catalytic system.5 9 Reasons for the different
effects of protein B on the two reactions with Hox are unclear. With Hox from M.
capsulatus(Bath), activities of only = 10 % of the values observed under optimal
catalytic conditions were observed with the H 2 0 2 shunt pathway, assuming
specific activities greater than 200 mU/mg. 5 7
Since such poor yields were
observed, the effect of protein B on that system was not investigated.
Investigations of the reduction potentials of the diiron center in the
hydroxylase described by equation (2) as well as the effects of adding the other
Fe(III)Fe(III)
E1
°
-
Fe(II)Fe(III)
E2
-
Fe(II)Fe(II)
(1)
proteins on the potentials, again demonstrated inconsistent behavior for the two
organisms, as shown in Table 2. For M. capsulatus (Bath), values of +350 and -25
mV vs NHE E1° and E20 obtained from redox titrations measuring appearance
and disappearance of the EPR signal of Hmv were initially reported. 3 5 To
investigate problems attaining high yields of Hmv, these potentials were
remeasured. Two sets of electron transfer mediators were used, yielding
potentials of +48 and -135 mV vs NHE for one set of conditions,6 0 and +100 and
-100 mV using the second set of conditions. 6 1 Substrate had little effect on the
potentials, whereas addition of protein B altered them to values of +50 and -170
26
mV. The presence of both protein B and reductase prevented the detection of
EPR signals associated with Hmv or Hred at potentials as negative as - 200 mV.6 0
When the titration of the three components was carried out in the presence of
substrate, only Hred was detected, but at potentials as high as 100 mV. This
result indicates that, with the complete reconstituted system including substrate,
two-electron transfer occurs. Reduction Hox,,directly to Hred indicates that E20
has shifted to a value greater than E1°. No EPR signal indicative of Hmv was
observed implying that Hmv is not physiologically significant. The inhibition of
hydroxylase reduction by protein B and reductase is consistent with kinetic work
i.nwhich this component prevented NADH oxidation by the MMO system in the
absence of substrate.6 2 Addition of substrate similarly led to efficient electron
transfer to the hydroxylase as measured by the consumption of NADH.
E1° and E2 values of +76 and +21 mV, respectively, were measured by
similar methods for Hox from M. trichosporiumOB3b.6 3 Addition of protein B
lowered the potentials to -52 mV and -115 mV. The regulation of electron
transfer to the hydroxylase with protein B and reductase observed with the M.
capsulatus(Bath) MMO was not seen with this system. Instead, it was reported
that the potentials of Hox and of Hox with added protein B were shifted slightly
:more positive in the presence of reductase, and the reduction was not substratedependent.
As indicated by the negative shifts in the reduction potentials of HoX,61'6 3
protein B can affect with the diiron center in Hmv from both MMO systems.
Consistent with this interpretation are EPR studies of Hmv from both organisms
which indicate that in the presence of protein B, the EPR signal moves from gav
1.83 to gayv 1.75.46,64
The distribution of product formed by hydroxylation of nitrobenzene and
isopentane with MMO from M. trichosporiumOB3b were investigated,5 6
59
and
27
the results with nitrobenzene are listed in Table 3. Protein B had a dramatic
effect on the regioselectivity of hydroxylation.5 9 This behavior was attributed to
changes in the hydroxylase structure upon complexation with protein B, causing
the substrate molecule to interact differently at the active site of the enzyme.
Similar studies of the substrate nitrobenzene with the M. capsulatus(Bath) system
were carried out,4 9 and the results of this work are included in Table 3. Very
different regioselectivity was observed with the native reconstituted system, with
product distributions that curiously matched those obtained by the hydrogen
peroxide shunt pathway with M. trichosporiumOB3b. With the M. capsulatus
(Bath) system the presence of protein B in single turnover reactions with
chemically reduced Hred had no effect on the regioselectivity of the reaction.
This result implies that the reductase has a significant effect on the distribution of
products from the hydroxylation reaction. Of interest are the KM values for
nitrobenzene in the various systems. KM value of = 5 mM were obtained with
both the catalytic system from M. capsulatus(Bath)4 9 and with the H 2 0 2 shunt
pathway with. M. trichosporiun OB3b.5 6 As indicated above, these two systems
display similar product distributions in the hydroxylation reaction. The catalytic
system from M. trichosporium OB3b yielded a KM value = 100 PM, 56 however,
and the regioslectivity also differed from that observed with the other two
systems.
Yields of single turnover reactions with chemically reduced Hre,,dfrom M.
trichosporium OB3b were not sensitive to the presence of the other two
components. 26 From these and other studies with this system, it was concluded
that protein B does not affect the reactivity of the hydroxylase with dioxygen,
and that it has little effect on the hydroxylation reaction once Hred is formed.6 4
In contrast to this work, experiments with MMO from M. capsulatus (Bath)
indicated that total yield from single turnover reactions of chemically reduced
28
Hred was significantly affected by the other components of the system (see
Chapter 6).4 9 Maximal yields occurred when greater than 1.5 equivalents of
protein B were added, either prior to or following reduction of Hox to Hred.
Combining both protein B and reductase with Hred resulted in slightly
diminished yields (see Chapter 6), and adding protein B and reductase to Hox
prior to reduction resulted in extremely low levels of hydroxylation. This latter
finding is consistent with the redox studies in which, even at quite negative
potentials, Hred was not detected by EPR spectroscopy with protein B and
reductase present.6 0 Adding reductase alone to Hox or Hred did not have a
substantial effect on the product yields, which were quite low since protein B
was not present.
Protein B from M. capsulatus(Bath) affects the hydroxylation in two ways.
Studies of single turnover reactions reveal that protein B increases the
hydroxylation efficiency reflected in increased yields and rate constants. As
determined by the redox studies, it regulates electron transfer so as to occur only
in the presence of substrate, which prevents the wasteful cconsumption of
reducing equivalents and protects the enzyme from inactivation. The redox
results also indicate that incubation of Hox with protein B and reductase
produces a more reactive Hred species. The reactive Hred species formed in this
manner could presumably be produced by chemical reduction of Hox with all
three protein components as well as substrate present.
The presence of
nitrobenzene, the substrate in these experiments, prevented the mediated
reduction of Hox to Hred, however. 6 0
Addition of protein B and reductase to
Hred may not produce the same reactive species as formed when Hox is reduced
in the presence of protein B and reductase. For example, structural alterations
produced along with the change in oxidation state from Hox to Hred could arise
29
from kinetic effects imposed by the other two components which influence the
coordination sphere of the iron atoms.4 9
Protein B from M. capsulatus(Bath) not only increased the product yields,
but it also influenced the rate constant in single turnover reactions of Hred with
nitrobenzene. 496 5 The pseudo-first order rate constant increased by up to 33fold when Hred was titrated with protein B. Neither addition of reductase to Hox
or Hred, nor of protein B and reductase to Hred, could similarly affect the rate
constant. These observations again imply either that protein B alone activates the
hydroxylase or that protein B and reductase do not significantly affect Hred.
Complex formation among the three protein components of the M.
trichosporium OB3b MMO has been demonstrated. 6 4
In this work, it was
hypothesized that protein B binds tightly to Hox to form an activated complex
but that, after formation of Hred, the binding affinity is diminished.6 3 6 4 Results
reported for M. capsulatus (Bath) are consistent with initial binding of Hox to
protein B to form an activated complex,4 9 because increased yields and rate
constants are seen when protein B is added to Hred. If the Hred-protein B
complex of M. capsulatus (Bath) is the species influencing reactivity, then the
addition of more protein B, which would ensure that Hred is in the protein Bbound complex, would be expected to increase the rate constants. Since H is a
dimer (227Y2'),two equivalents of protein B and reductase are believed to bind to
one H molecule. Concentrations of protein B above two equivalents per Hred
may diminish the rate constant of the reaction by sterically hindering substrate
binding to Hred. The total product yield could still remain high since nearly all
of the Hred is in the active protein B-bound form. If there are a finite number of
binding sites for additional molecules of protein B beyond two equivalents,
excess protein B would have little effect once all of these extra sites were fully
occupied, except to block possibly the reductase binding site. This reasoning was
30
also invoked in work with M. trichosporiumOB3b to explain the titration curve
describing the effect of protein B on the initial rates of the catalytic system with
Iox and reductase.6 4
The M. capsulatus (Bath) reductase alters the distribution of products
formed in the reaction of Hred and dioxygen with nitrobenzene, 49 but protein B
plays a wider variety of roles in this MMO system. As mentioned earlier, it
regulates electron transfer from the reductase to the hydroxylase so as to occur
only in the presence of substrate,6 0 6 2 it shifts the reduction potentials of the
substrate/reductase/hydroxylase
complex,6 0 and it affects kobs and yields in
single turnover reactions of Hred.4 9 Protein B from M. trichosporium OB3b
changes product distributions with Hred,5 9 increases the initial velocity of the
complete reconstituted system,6 4 but reduces the initial velocity with the H 2 0 2
shunt system. 64 Reductase played no major role in the regioselectivity of the
hydroxylation reaction in this case.5 9 With hydroxylase from both organisms,
protein B perturbs the EPR signal of Hmv4 6' 64 and lowers the reduction
potentials. 61
63
A unifying explanation for these observed functions has yet to be
determined. Some of the differences may arise from variations in the proteins
themselves.
For example, appreciable differences in protein B from the two
MMO's have been reported by the inability of antibodies raised toward one
protein B to bind the other.6 6 More structural work on component interactions is
essential to understand fully the fundamental reasons for the observed effects
and to reconcile differences between the two MMO systems.
Mechanism of Hydroxylation
Reaction of Hred with Dioxygen.
Hred is the form of the hydroxylase which reacts with dioxygen.
26
Recently, intermediates were detected in this reaction.67-70In all cases, dioxygen
was mixed with Hred in the presence of two equivalents of protein B under
31
pseudo-first order conditions in dioxygen. Reactions were carried out at 4 °C in
the absence of substrate to accumulate detectable quantities of the intermediates.
Spectroscopic parameters and kinetic constants for two intermediates, designated
L and Q, are listed in Table 4. The proposed reaction summary is given in
equation (3). Magnetic Mdssbauer4 9' 6 7 and EPR spectroscopy4 9 68 indicated that
k,
Hrc.d
k2
L
k3
>
Q
(3)
ox
both intermediates are diamagnetic. In the M. trichosporiumOB3b MMO system,
rapid freeze-quench EPR studies revealed that the g = 15 EPR signal of Hred
decays with a first order rate constant of = 22 + 5 s-O upon exposure to
dioxygen.68 Based on this observation, a proposed intermediate named "P" was
suggested as a precursor to Q (vide invfa). M6ssbauer analysis of rapid freezequench samples prepared from the M. capsulatus(Bath) system identified such an
intermediate,6 9 but this compound was named "L" since it was unknown at that
time whether it corresponed to the immediate precursor to Q. Kinetic Mossbauer
data for L from M. capsulatus(Bath) indicates that L forms with a first order rate
constant of = 25 s-1,68and concomitantly with the decay of Hred. The minimal
mechanism therefore dictates that intermediate L represents the initial species
formed in the reaction of Hred with dioxygen.
The M6ssbauer spectrum of L consists of a symmetric doublet which was
fit with parameters
= 0.66 + 0.02 mm/s and AEQ = 1.51 + 0.03 mm/s.
69
The
isomer shift is larger than values (0.45 - 0.55 mm/s) generally seen for
carboxylate-bridged diiron(III) clusters but significantly smaller than isomer shift
values for diiron(II) clusters (1.1 - 1.3 mm/s).3 1 -33 Although parameters
matching the values of L have not been observed in model compounds, the
32
magnitude of the isomer shift is close to parameters seen for diiron(III) clusters,
and L was accordingly proposed to be a diiron(III) peroxide complex.6 9 Such
species have been assigned to products formed in reactions of several diiron(II)
model complexes with dioxygen and result from a two electron transfer from the
Fe atoms to the dioxygen ligand.3 4 The optical absorption at 625 nm ( = 500 M1 cm-1)
reported for L70 ,71 is close to reported electronic spectral maxima (max
600 or 604 nm) for several diiron(III) peroxide model complexes.7 2 -7 4 Similarly,
the resonance Raman spectrum of L with excitation at 647 nm displayed a band
at 902 cm-1,70 which matches u(O-O) values reported for several model
compounds.
2-7 7 4
From the sharp, symmetrical shape of the Mbssbauer signal, the iron
atoms were interpreted
as being in nearly equivalent coordination
environments.6 9 To account for the high isomer shift value of 0.66 mm/s for L,
six-coordinate iron atoms with considerable peroxide-to-iron charge transfer or
seven coordinate iron atoms were invoked.6 9 These structures could arise from
the two pentacoordinate iron atoms in Hred transferring two electrons to oxygen
to form a peroxide ligand. The two resulting Fe(II) atoms may bind the peroxide
ligand either a t-r12, 11,1 1 or an 12,112 fashion. This latter binding mode has been
observed for oxyhemocyanin7 5 and in a dicopper(II) peroxide complex.7 6 Figure
2 illustrates possible structures for L which are compatible with its spectroscopic
parameters. 49
The time dependence revealed by rapid freeze-quench M6ssbauer
experiments with M. capsulatus(Bath)4 9 indicated that decay of L proceeds with
the concomitant formation of another intermediate, named compound "Q". This
intermediate, observed both in the M. trichosporium OB3b67'6 8 and the M.
capsulatus (Bath)4 9' 6 9 MMO systems by M6ssbauer and optical spectroscopy,
decays faster in the presence of substrates. This behavior indicates that this
33
intermediate is on the kinetic reaction pathway for hydroxylation.49,68 The
M6ssbauer spectrum of Q from M. trichosporiumOB3b contained a doublet with 6
= 0.17 mm/s and AEQ= 0.53 mm/s. 6 7 This doublet was symmetric and the iron
atoms were assumed reside in equivalent coordination environments. In the M.
capsulatus (Bath) system, two unresolved doublets of equal intensity were
distinguished and fit to the following parameters: 6 = 0.21 + 0.02 mm/s and AEQ
= 0.68 ± 0.03 mm/s
for doublet
1, and 6 = 0.14 + 0.02 mm/s
and AEQ = 0.55 ± 0.03
mm/s for doublet 2.69 The average isomer shift and quadrupole splitting of the
two doublets, 0.18 mm/s and 0.62 mm/s, respectively, agrees with parameters
obtained from the corresponding spectrum of Q from the M. trichosporiumOB3b
organism.
In the M. capsulatus (Bath) system, however, the presence of two
distinct signals for Q indicates that the iron atoms are in inequivalent sites.
From the average isomer shift value of 0.18 mm/s, it was suggested that Q
is an iron(IV) oxo species.6 7 Figure 3 illustrates possible structures for Q. In M.
capsulatus(Bath), the inequivalent Fe sites could arise from rearrangement of the
protein ligands in Q compared to their positions as revealed in the x-ray crystal
structure of Hox.39 Alternatively, the oxo ligand in the proposed Fe(IV) species
may not be bound symmetrically between the Fe atoms. Equivalent iron sites in
the M. trichosporiumOB3b system could arise from symmetrical bridging of the
oxygen atoms between the two Fe(IV) atoms.4 9
Cleavage of the 0-0 bond in L followed by release of H 20 most likely
results in the formation of Q.4 9 The second iron atom in the MMO hydroxylase
active site can stabilize this unit through redox charge delocalization in a similar
fashion to the stabilization of a high valent iron oxo species in cytochrome P-450
by oxidation of the porphyrin ring to a 7r-cation radical.7 7 A large change in
entropy
(ASt
-:
147 Jmole- 1 K - 1) was reported
for the conversion
of L to Q,
consistent with the release of a water molecule during this step.4 9 The rate
34
constant for the growth of Q was independent of dioxygen concentration.4 9' 6 8
Since formation of intermediate L is rapid and irreversible and formation of Q is
the slow step, the rate constant for Q is not expected to be dependent on
dioxygen concentration under experimentally practical conditions. Conversion
of intermediate L, a diiron(III) peroxide species, to Q, a putative diiron(IV) oxo
species, implies proton transfer to the core such that a molecule of water is
produced from the second oxygen atom of the peroxide. The lack of a pH
dependence or a significant kinetic deuterium isotope effect on the rate constant
for this process indicates that proton transfer is not overall rate determining in
the conversion.4 9
The widely accepted hydroxylation mechanism for the heme protein
cytochrome P-450 invokes an Fe(IV) oxo species which abstracts a hydrogen
atom from the substrate molecule.7 7 Since cytochrome P-450 cannot hydroxylate
methane, a similar Fe(IV) oxo intermediate in MMO could be significantly more
reactive.
Very recently, Fe(IV) intermediates have been ruled out in the
mechanism for generating the tyrosyl radical in the R2 subunit of ribonucleotide
reductase. 78
8 0
Instead, a diiron(III) oxygen radical species was invoked as the
species which causes the oxidation and deprotonation of the tyrosyl residue. A
similar compound has been suggested for Q,34 but such a species would be
expected to be paramagnetic.
In addition, the O-H bond of tyrosine is
significantly weaker than the C-H bond in methane, and the active abstracting
species in R2 may not be strong enough to break the C-H bond in methane. It
has been speculated that a cysteinyl radical helpts to homolyze this bond. 3 4 If
such a species were to couple magnetically with the oxyl radical in Q, it could
account for the observed diamagnetism.
In the M. trichosporiumOB3b system, a third intermediate, T, with ,nmax
at
325 nm (£ = 6000 M-1cm-1) was observed in the presence of the substrate
35
nitrobenzene. 68 This species was assigned as the product, 4-nitrophenol, bound
to the dinuclear iron site, and its absorption was attributed primarily to the 4-
nitrophenol moiety. No analogous intermediate was found with the M.
capsulatus(Bath) system in the presence of nitrobenzene.
A proposed cycle49 for hydroxylation by MMO is illustrated in Figure 4.
Substrate first binds to the complete system containing all three protein
components.
Addition of NADH next effects two-electron reduction of the
hydroxylase from the oxidized Fe(III)Fe(III)to the fully reduced Fe(II)Fe(II)form,
bypassing the inactive Fe(II)Fe(III) state. The fully reduced hydroxylase then
reacts with dioxygen in a two-electron step to form intermediate L, a diiron(III)
peroxide complex. The possibility that L itself is sufficiently activated to carry
out the hydroxylation reaction cannot be ruled out. Intermediate L is then
converted to Q as shown in Figure 3. Substrate reacts with Q, and product is
released with concomitant formation of the diiron(III) form of the hydroxylase
which enters another cycle in the catalysis.
Reaction of Q with Substrate
Figure 5 presents possible mechanisms for substrate hydroxylation by
intermediate Q with the assumption that this species is a diiron(IV) oxo complex.
In mechanism. (A), direct insertion of the oxygen atom of Q into a C-H bond
occurs. Concerted 2 + 2 cycloaddition of the C-H bond to Q to form a metalcarbon bond followed by reductive elimination of the alcohol is outlined in
mechanism (B), whereas (C) illustrates heterolysis of the C-H bond by Q
followed by recombination to afford product In C(1), a proton is formed and in
C(2) a carbocation is generated. Mechansim (D) involves homolysis of the C-H
bond by Q followed by return of the hydroxyl group from iron to the alkyl
radical. This process is analogous to the oxygen rebound step in the postulated
hydroxylation
mechanism of cytochrome P-450. Finally, mechanism (E)
36
illustrates abstraction of a hydrogen atom from the substrate by hydroxyl or
another radical within the active site to form an alkyl radical which then adds to
the iron-bound oxygen atom of Q. An alternative mechanism in which L reacts
in a concerted fashion with substrate via electrophilic attack to form a carbonoxygen bond followed by release of product is outlined in mechanism (F). In F,
the 0-0 bond does not cleave the C-H bond.
The catalytic hydroxylation mechanism for MMO has been frequently
compared to that of its heme analog, cytochrome P-450.8 56
81
A widely accepted
mechanism for P-450 hydroxylations invokes a high valent iron oxo, or ferryl,
species which is then postulated to abstract a hydrogen atom from bound
substrate 7 7 in a manner similar to that outlined in Figure 5D. The resulting
substrate radical can then recombine with the iron-bound hydroxyl radical in the
rebound step.
Evidence for the presence of substrate radical intermediates in cytochrome
P-450 includes studies with radical clock substrate probes from which the
rebound rate constant was calculated to be
' 83 Such substrates
2 x 1010s-1.82
rearrange rapidly upon the abstraction of a hydrogen atom in the enzyme active
site, affording skeletally modified hydroxylation products.8 3 -8 5 The presence of
such rearranged products is taken as evidence that a substrate radical
intermediate participates in the mechanism. Studies of MMO from M. capsulatus
(Bath) and M. trichosporiumOB3b using several radical clock substrate probes,
including two much faster probes than those reported previously for
experiments with cytochrome P-450, were reported. 5 7 For M. capsulatus (Bath)
MMO no products consistent with the formation of a substrate radical were
detected. This result implied that reactions with those substrates did not
generate substrate radical or carbocation intermediates, excluding pathways D
37
and E in Figure 5. Alternatively, any rebound reaction occurring would have a
rate constant
> 1013 s-1 .
Using an identical radical clock substrate probe which did not rearrange
upon hydroxylation with M. capsulatus(Bath), rearranged product was detected
with MMO from M. trichosporiumOB3b.5 7 From the ratio of unrearranged to
rearranged products, a rebound rate constant was calculated to be = 6 x 1012s- 1 at
30 C for this system. A separate study with another radical clock substrate
probe with MMO from M. trichosporiumOB3b reported products consistent with
both radical and cationic substrate intermediates.8 6
Evidence in support of radical intermediates with MMO from both M.
capsulatus (Bath) and M. trichosporiumOB3b was reported from experiments in
which substrate radicals were trapped during turnover. 8 7
88
The amount of
trapped radical was not quantitated in these experiments, however. In other
reports, no diffusable radical species were detected in reactions with MMO from
M. trichosporiumOB3b.59
With M. trichosporium OB3b, epimerization
with exo, exo, exo, exo-2,3,5,6-
d4 -norbornane upon hydroxylation occurs,8 1 which parallels results for
cytochrome P-450 hydroxylation with this substrate. 8 9
The extent of
epimerization with MMO, however, was significantly less, being 2% following
hydrogen atom abstraction from the endo position compared to 18% with cytochrome P-450, and 5% after abstraction at the exo position as compared to 14%
with cytochrome P-450.90 Allylic rearrangements with 3,3,6,6-d4 -cyclohexene
occurred in 20% of the MMO hydroxylation products compared to 33% for
cytochrome P-450. These two experiments suggest that, with M. trichosporium
OB3b, a rebound reaction must occur with a greater rate constant that with
cytochrome P-450, in accord with the radical clock substrate work.
38
Since work with the radical clock substrate probes indicated important
differences in the hydroxylation mechansims for M. capsulatus (Bath) and M.
trichosporiumOB3b, work with (R) and (S)-[1-2 H,1-3 H]ethane with both enzymes
was carried out.91 ' 92 With M. trichosporium OB3b, approximately 65% of the
products displayed retention of stereochemistry.9 1 A rebound rate constant of 3
x 1012 s-1 was calculated, assuming a free energy change of 0.5 kcal mole- 1 for
rotation about the C-C bond.9 2
This value is in good agreement with that
obtained from the radical clock substrate probe analysis.57
Preliminary
experiments with the M. capsulatus(Bath) system indicate that this MMO is prima
facia more complex, with the apparent enantioselectivity of the reaction being
dependent upon whether hydroxylation occurs at the deuterium atom or at the
hydrogen atom of the chiral carbon.9 2 In some cases, > 90 % retention of the
stereochemical center was observed. An enantioselective exchange mechanism
in which the alcohol product loses the tritium label was invoked to account for
the differences in apparent sterechochemistries.
Kinetic isotope effects in the hydroxylation reaction have been measured
for MMO from both organisms. With both systems, no significant intermolecular
kinetic isotope effect was found for most susbstrates, 5 7
93'94
although a value of
kH/kD = 12 was reported with CH4 vs CD4.62 Substantial intramolecular isotope
effects (kH/kID = 5) with M. capsulatus (Bath) were found with several
substrates.5793
Studies with M.
trichosporium OB3b have revealed
intramolecular isotope effects of kH/kD = 4 to 5.81,91 These results indicate that
C-H bond breaking is not the rate-determining step in the overall enzymatic
reaction.
The large intramolecular isotope effect does indicate that the
hydroxylation step involves a substantial C-H bond-breaking component. The
magnitude of the intramolecular isotope effects for both MMO's can be compared
to values of ktH/kD = 7-14 seen with cytochrome P-450 hydroxylations.
7 7 95
39
Some fundamental differences between MMO and cytochrome P-450 have
emerged. The ability of high concentrations of hydrogen peroxide to effect a
shunt pathway for the M. trichosporiumOB3b MMO has been cited as evidence
for a hydroxylation mechanism paralleling that of cytochrome P-450.56
It is
noteworthy, however, that oxo transfer reagents such as iodosylbenzene, which
support catalysis with cytochrome P-450, cannot effect turnover with MMO. In
,addition, the mechanism for olefin epoxidation by MMO from M. trichosporium
OB3b8 1 differs from the proposed mechanism for cytochrome P-450.9 6 With
cytochrome P-450, the 1-trans-proton of propylene exchanges with solvent
protons during turnover, from which an epoxidation mechanism involving
oxametallocycles and iron carbene intermediates was proposed. In the reaction
of propylene with MMO, however, no such exchange occurred.
In summary, mechanistic studies have revealed intriguing differences
between MMO from M. capsulatus(Bath) and MMO from M. trichosporiumOB3b.
With M. capsulatus (Bath), radical clock substrate probes indicated either that a
substrate radical is not produced or that it reacts with a rate constant >1013 s-1.
Reactions with chiral ethane were primafacia quite complex. Further analysis of
the data suggested the operation of an enantioselective exchange mechanism.
With MMO from M. trichosporium OB3b, radical involvement was suggested
from several experiments, and a rebound rate constant of 3 to 6 x 1012s-1 was
calculated for this system.
The different behavior between the two MMO systems has several
implications. For each case, two reaction pathways may operate in parallel, one
involving a substrate radical (Figure 5D), and one not (Figure 5A-C). The degree
to which the radical vs. non-radical mechanism is followed may be imposed by
the steric requirements
of the substrate in the active site. Alternatively,
abstraction of a hydrogen atom to form a substrate radical may occur in both
40
cases, but the rate constants for rebound may differ between the two organisms.
For example, in the M. capsulatus(Bath) system, the rebound rate constant may
be so large that only an extremely small amount of the substrate radical has a
sufficient lifetime for rotation about the C-C bond to occur before recombination
with the bound hydroxyl group.
Conclusions.
Physical studies of the hydroxylase have established the structural nature
of the diiron core in its three oxidation states, Ho, Hmv, and Hred. Although the
active site structure
of hydroxylase from M. trichosporium OB3b and M.
capsulatus (Bath) are similar, fundamental differences were observed for other
aspects of the MMO system. The interactions between the other components,
protein B and reductase, vary substantially.
More structural information is
necessary to understand how each of the components affects the others with
respect to its physical properties and role in the hydroxylation mechanism and to
reconcile the different properties seen in the two MMO systems. The kinetic
behavior of intermediates in the hydroxylation reaction cycle and the physical
parameters of intermediate Q appear similar. The reaction of Q with substrate,
however, varies. Formation of radical substrates are more obvious with the M.
trichosporium(OB3bsystem. In comparison to the cytochrome P-450 system, the
hydroxylation mechanism with both MMO systems either has a rebound rate
constant which is much larger, and/or it takes place by an alternative pathway.
41
References
(1)
Hanson, R. S. In Advances in Applied MicrobiologyAcademic Press: New
York, 1980; Vol. 26; pp 3.
(2)
Enhalt, D. H. In MicrobialProducationand Utilization of Gases;Schlegel, H.
G., Gottschalk, G. and Pfennig, N., Eds.; Goltze Publishers: Gottingen, 1976; pp
13-22.
(3)
Enhalt, D.; Schmidt, U. Pure Appl. Geophys. 1978, 116, 452-464.
(4)
Dalton, H.; Leak, D. J. In Gas Enzymol.; H. Degn, Ed.; D. Reidel Publishing
Co.: London, 1985; pp 169-186.
(5)
Burch, R.; Squire, G. D.; Tsang, S. C. J. Chem. Soc., Faraday Trans. I 1989, 85,
3561-3568.
(6)
Colby, J.; Stirling, D. I.; Dalton, H. Biochem. J. 1977, 165, 395-402.
(7)
Dalton, H. In Adv. Appl. Microbiol. Academic Press: 1980; Vol. 26, pp 71-87.
(8)
Green, J..;Dalton, H. J. Biol. Chem. 1989, 264, 17698-17703.
(9)
Adriaens, P. App. Environ. Microbiol. 1994, 60, 1658-1662.
(10)
Jahng, D.; Wood, T. K. App. Environ. Microbiol. 1994, 60, 2473-2482.
(11)
Fox, B. G.; Borneman,
J. G.; Wackett, L. P.; Lipscomb,
J. D. Biochem. 1990,
29, 6419-6427.
(12)
Lindstrom,
J. E.; Prince, R. C.; Clark, J. C.; Grossman,
M. J.; Yeager, T. R.;
Braddock, J. F.; Brown, E. Appl. Environ. Microbiol. 1991, 57, 2514-2522.
(13)
Pritchard, P. H.; Costa, C. F. Envrion. Sci. Tehnol. 1991, 25, 372-379.
(14)
Anthony, C. The Biochemistry of Methylotrophs; Academic Press: New York,
1982, p 296-379.
(15)
Bowman, J. P.; Sly, L. I.; Nichols, P. D.; Hayward,
A. C. International Journal
of Systematic Bacteriology 1993, 43, 735-753.
(16)
Bedard, C.; Knowles,
(17)
Akent'eva,
R. Microbiol. Rev. 1989, 53, 68-84.
N. F.; Gvozdev,
R. I. Biokhimiya 1988, 53, 91-96.
42
(18)
Nguyen,
H. T.; Shiemke, A. K.; Jacobs, S. J.; Hales, B. J.; Lidstrom,
M. E.;
Chan, S. I. J. Biol. Chem. 1994, 269, 14995-15005.
(19)
Stainthorpe,
A. C.; Salmond,
G. P.; Dalton, H.; Murrell, J. C. FEMS
Microbiol. Lett. 1991, 70, 211-216.
(20)
Stainthorpe, A. C.; Lees, V.; Salmond, G. P. C.; Dalton, H.; Murrell, J. C.
Gene 1990, 91, 27-34.
(21)
Nakajima, T.; Uchiyama, H.; Yagi, O.; Nakahara, T. Biosci. Biotech. Biochem.
1992, 56, 736-740.
(22)
Koh, S.-C.; Baumeen, J. P.; Sayler, G. S. App. Environ. Microbiol. 1993, 59,
960-967.
(23)
Fox, B. G.; Froland, W. A.; Jollie, D. R.; Lipscomb, J. D. In Methods In
Enzymol.Academic Press: New York, 1990; Vol. 188; pp 191-202.
(24)
Pilkington, S. J.; Dalton, H. In Methods In EnzymologyAcademic Press: New
York, 1990; Vol. 188; pp 181-190.
(25)
Prior, S. D.; Dalton, H. FEMS Microbiol. Lett. 1985, 29, 105-109.
(26)
Fox, B. (.; Froland, W. A.; Dege, J. E.; Lipscomb, J. D. J. Biol. Chem. 1989,
264, 10023-10033.
(27)
Colby, J.; Dalton, H. Biochem. J. 1979, 177, 903-908.
(28)
Lund, J.; Woodland, M. P.; Dalton, H. Eur. J. Biochem. 1985, 147, 297-305.
(29)
Lund, J.; Dalton, H. Eur. J. Biochem. 1985, 147, 291-296.
(30)
Green, J.; Dalton, H. J. Biol. Chem. 1985, 260, 15795-15801.
(31)
Que, L., Jr.; True, A. E. In Prog. Inorg. Chem.; Lippard, S. J., Ed.; John Wiley
& Sons: New York, 1990; Vol. 38; pp 97-200.
(32)
Kurtz, D. M. Chem. Rev. 1990, 90, 585-606.
(33)
Wilkins, R. G. Chem. Soc. Rev. 1992, 171-178.
(34)
Feig, A. L.; Lippard,
S. J. Chem. Rev. 1994, 94, 759-805.
43
(35)
Woodland,
M. P.; Patil, D. S.; Cammack, R.; Dalton, H. Biochim. Biophys.
.Acta 1986, 873, 237-242.
(36)
DeWitt, J. G.; Bentsen, J. G.; Rosenzweig, A. C.; Hedman, B.; Green, J.;
Pilkington, S.; Papaefthymiou,
G. C.; Dalton, H.; Hodgson, K. O.; Lippard, S. J. J.
Am. Chem. Soc. 1991, 113, 9219-9235.
(37)
Fox, B. (G.; Hendrich,
M. P.; Surerus, K. K.; Andersson,
K. K.; Froland,
W.
A.; Lipscomb, J. D.; Miinck, E. J. Am. Chem. Soc. 1993, 115, 3688-3701.
(38)
Prince, R. C.; George, G. N.; Savas, J. C.; Cramer, S. P.; Patel, R. N. Biochim.
Biophys. Acta 1988, 952, 220-229.
(39)
Rosenzweig, A. C.; Frederick, C. A.; Lippard, S. J.; Nordlund,
P. Nature
1993, 366, 537-543.
(40)
DeRose, V. J.; Liu, K. E.; Kurtz, D. M., Jr.; Hoffman,
B. M.; Lippard,
S. J. J.
Am. Chem. Soc. 1993, 115, 6440-6441.
(41)
Maroney, M.; Kurtz, D. M., Jr.; Nocek, J. M.; Pearce, L. L.; Que, L., Jr. J. Am.
Chem. Soc. 108, 6871-6879.
(42)
Pearce, L. L.; Kurtz, D. M., Jr.; Xia, Y.-M.; Debrunner,
P. G. J. Am. Chem.
Soc. 1987, 109, 7286-7293.
(43)
Scarrow, R. C.; Maroney, M. J.; Palmer, S. M.; Que, L., Jr.; Roe, A. L.;
Salowe, S. P.; Stubbe, J. J. Am. Chem. Soc. 1987, 109, 7857-7864.
(44)
McCormick,
J. M.; Reem, R. C.; Solomon, E. I. J. Am. Chem. Soc. 1991, 113,
9066-9079.
(45)
Thomann, H.; Bernardo, M.; McCormick, J. M.; Pulver, S.; Andersson, K.
K.; Lipscomb,
(46)
DeRose,
J. D.; Solomon, E. I. J. Am. Chem. Soc. 1993, 115, 8881-8882.
V. J.; Liu, K. E.; Lippard,
S. J.; Hoffman,
B. M.. manuscript
in
preparation.
(47)
Bender, C. J.; Rosenzweig, A. C.; Lippard, S. J.; Peisach, J. J. Biol. Chem.
1994, 269, 15993-15998.
44
(48)
Hendrich, M. P.; Fox, B. G.; Andersson, K. K.; Debrunner, P. G.; Lipscomb,
Jr.D. J. Biol. Chem. 1992, 267, 261-269.
(49)
Liu, K. E.; Wang, D.; Valentine, A. M.; Huynh, B. H.; Edmondson, D. E.;
Salifoglou, T.; Lippard, S. J. manuscript in preparation.
('50)
'1990,
(51)
Hendrich, M. P.; Miinck, E.; Fox, B. G.; Lipscomb, J. D. J. Am. Chem. Soc.
112, 5861-5865.
Lippard,
(52)
B. M.; Sturgeon,
Hoffman,
B. E.; Doan,
P. E.; DeRose,
V. J.; Liu, K. E.;
S. J. J. Am. Chem. Soc. 1994, 116, 6023-6024.
Pulver, S.; Froland, W. A.; Fox, B. G.; Lipscomb, J. D.; Solomon, E. I. J. Am.
Chem. Soc. 1993, 115, 12409-12422.
(53)
Reem, R. C.; Solomon, E. I. J. Am. Chem. Soc. 1984, 106, 8323-8325.
(54)
Reem, R. C.; Solomon, E. I. J. Am. Chem. Soc. 1987, 109, 1216-1226.
(55)
Colby, J.; Dalton, H. Biochem. J. 1978, 171, 461-468.
{(56) Andersson, K. K.; Froland, W. A.; Lee, S.-K.; Lipscomb, J. D. New J. Chem.
1991, 15, 411-415.
(57)
Liu, K. E.; Johnson,
C. C.; Newcomb,
M.; Lippard,
S. J. J. Am. Chem. Soc.
1992, 115, 939-947.
(58)
Jiang, Y.; Wilkins, P. C.; Dalton, H. Biochim. Biophys. Acta 1993, 1163, 105-
112.
(59)
Froland, W. A.; Andersson, K. K.; Lee, S.-K.; Liu, Y.; Lipscomb, J. D. J. Biol.
Chem. 1992, 267, 17588-17597.
S. J. J. Biol. Chem. 1991, 266, 12836-12839, 24859.
(60)
Liu, K. E.; Lippard,
(61)
Liu, K. E.; Lippard, S. J., manuscript in preparation.
(62)
Green, J..; Dalton, H. Biochem. J. 1989, 259, 167-172.
(63)
Paulsen,
K. E.; Liu, Y.; Fox, B. G.; Lipscomb,
J. D.; Miinck, E.; Stankovich,
M. T. Biochem. 1994, 33, 713-722.
(64)
Fox, B. GC.;Liu, Y.; Dege, J.; Lipscomb, J. D. J. Biol. Chem. 1991, 266, 540-550.
45
(65)
Liu, K. E.; Feig, A. L.; Goldberg, D. P.; Watton, S. P.; Lippard, S. J. In The
Activation of Dioxygen and Homogeneous Catalytic Oxidation; Barton, D. H. R.,
Martell, A. E. and Sawyer, D. T., Eds.; Plenum Press: New York, 1993; pp 301-320.
(66)
Tsien, H.-C.; Hanson, R. S. Appl. Environ. Microbiol. 1992, 58, 953-960.
(67)
Lee, S.-K.; Fox, B. G.; Froland, W. A.; Lipscomb, J. D.; Miinck, E. J. Am.
(Chem. Soc. 1993, 115, 6450-6451.
(68)
Lee, S.-K.; Nesheim, J. C.; Lipscomb, J. D. J. Biol. Chem. 1993, 268, 21569-
21577.
(69)
Liu, K. E.; Wang, D.; Huynh, B. H.; Edmondson,
Lippard,
(70)
D. E.; Salifoglou, A.;
S. J. J. Am. Chem. Soc. 1994, 116, 7465-7466.
Liu, K. E.; Valentine, A. M.; Qui, D.; Edmondson,
D. E.; Spiro, T. G.;
Lippard, S. J., manuscript in preparation.
(71)
This absorption was detected using monochromatic methods at 15 nm
resolution.
The 625 nm value may therefore not reflect the true absorbance
maximum for L.
(72)
Brennan, B. A.; Chen, Q.; Juarez-Garcia, C.; True, A. E.; O'Connor, C. J.;
Que, L., Jr. Inorg. Chem. 1991, 30, 1937-1943.
(73)
Nishida, Y.; Takeuchi, M.; Shimo, H.; Kida, S. Inorg. Chim. Acta 1984, 96,
115-119.
(74)
Nishida, Y.; Takeuchi, M. Z. Naturforsch. 1987, 42B, 52-54.
(75)
Magnus, K. A.; Hazes, B.; Ton-That, H.; Bonaventura, C.; Bonaventure, J.;
Hol, W. G. J. Proteins 1994, 19, 302-309.
(76)
Kitajima, N.; Fujisawa, K.; Fujimoto, C.; Moro-oka, Y.; Hashimoto,
S.;
Kitagawa, T.; Toriumi, K.; Tatsumi, K.; Nakamura, A. J. Am. Chem. Soc. 1992, 114,
:1.277-1291.
46
(77)
Ortiz de Montellano, P. R. In CytochromeP-450 Structure, Mechanism, and
Biochemistry;Ortiz de Montellano, P. R., Eds.; Plenum Publishing Corp.: New
York, 1986; pp 217-271.
(78)
Bollinger, J. M., Jr.; Tong, W. H.; Ravi, N.; Huynh, B. H.; Edmondson, D.
E.; Stubbe, J. J. Am. Chem. Soc. 1994, 116, 8015-8023.
(79)
Bollinger, J. M., Jr.; Tong, W. H.; Ravi, N.; Huynh, B. H.; Edmondson, D.
E.; Stubbe, J. J. Am. Chem. Soc. 1994, 116, 8024-8032.
(80)
Ravi, N.; Bollinger, J. M., Jr.; Huynh, B. H.; Edmondson,
D.; Stubbe, J. J.
Am. Chem. Soc. 1994, 116, 8007-8014.
(81)
Rataj, M. J.; Kauth, J. E.; Donnelly, M. I. J. Biol. Chem. 1991, 266, 18684-
18690.
(82)
Bowry, V. W.; Lusztyk, J.; Ingold, K. U. J. Am. Chem. Soc. 1989, 111, 1927-
1928.
(83)
Bowry, V. W.; Ingold, K. U. J. Am. Chem. Soc. 1991, 113, 5699-5707.
(84)
Ortiz de Montellano, P. R.; Stearns, R. A. J. Am. Chem. Soc. 1987, 109, 3415-
3420.
(85)
Bowry, V. W.; Lusztyk, J.; Ingold, K. U. J. Am. Chem. Soc. 1991, 113, 5687-
5698.
(86)
Ruzicka, F.; Huang, D.-S.; Donnelly, M. I.; Frey, P. A. Biochem. 1990, 29,
1696-1700.
(87)
Wilkins, P. C.; Dalton, H.; Podmore, I. D.; Deighton, N.; Symons, C. R. Eur.
J. Biochem. 1992, 210, 67-72.
(88)
Dalton, H.; Wilkins, P. C. Faraday Discuss. 1992, 93, 163-171.
(89)
Groves, J. T.; McClusky,
G. A. Biochemical and Biophysical Research
Communications 1978, 81, 154-160.
(90)
Groves, J. T.; Subramanian,
D. V. J. Am. Chem. Soc. 1984, 106, 2177-21.81.
47
(91)
Priestley, N. D.; Floss, H. G.; Froland, W. A.; Lipscomb, J. D.; Williams, P.
G.; Morimoto, H. J. Am. Chem. Soc. 1992, 114, 7561-7562.
(92)
Wilkinson, B.; Liu, K. E.; Valentine, A. M.; Morimoto, H.; Williams, P. G.;
Lippard, S. J.; Floss, H. unpublished results.
(93)
Choi, S.-Y.; Eaton,
P. E.; Hollenberg,
P. F.; Liu, K. E.; Lippard,
S. J.;
Newcomb, M.; Put, D. A.; Upadhyaya, S. P., manuscript in preparation.
(94)
Shimoda,
(95)
Atkinson, J. K.; Hollenberg, P. F.; Ingold, K. U.; Johnson, C. C.; Le Tadic,
.M.-H.; Newcomb,
(96)
M.; Okura, I. J. Mol. Catal. 1992, 72, 263-267.
M.; Putt, D. A. Biochem. 1994, 33, 10630-10637.
Groves, J. T.; Avaria-Neisser,
G. E.; Fish, K. M.; Imachi, M.; Kuczkowski,
R. L. J. Am. Chem. Soc. 1986, 108, 3837-3838.
48
o\
o,
0"
dt
ON
d
'4
m
\kO
m
\I
m
\C
c
\
t
ct
0)
-5
o\
LrL
00
Lo
CZ
o
rocd
r-q
I-'
z
0)
cu
E
z
rt
Z,
3n
C
" C * tM
O 00 N
N
!e
Z.ct
to)
;
.
M
CZ
0)
Q
.
E
C)
k
cn
0)
O
tn
00
t
00
'N
N
N
N
.b
o
otn
Z;
\IC n
C4
t-
N
00
eN
rN
t
to
to
\D
tO
N"
\t
'IC \I
N
N
\IC
O
O
UO
-,4
0I.
.Z1.
0
rU
CD 0N
_r)
0·kFU
°
)0
©
ON
;I -:z
£0
C, m
Vo
C 00
ti
I
I
-.
---
.
00N
.
.
=rz
-
:r.
'"a
h ,.Q:,.o
-o
aE oo
d cn °..:
'z~
-
tE
-
(0
E1
u r..
E
E
E
,-q
**
E E
(,
'5; co
U
)
O
E
Ei
CZ
:0
0
W
x
+
N
S
N
>
.:n
r .
C
-
O- nX
$;
E
tc
;
CO
0
Z
0
z
t
49
Table 2. Reduction Potentials (mV vs NHE, pH 7.0) of the MMO
Hydroxylase from M. capsulatus (Bath)60,6 1 and M. trichosporium
OB3b. 6 3
Component(s)
Conditionsb
E1°
E20
H
1
48 ± 5
-135 +5
H, propylene
1
30 ±+5
-156 ± 5
H, protein B, R
1
a
a
H, protein B, R, propylene
1
< 100 + 25
= 100 + 25
H
2, 3
100 ± 15
-100 + 15
H, protein B
2, 3
50 ± 15
-170 ± 15
H (M. trich.)
4
76 + 15
21 + 15
H, protein B (M. trich.)
4
-52 + 15
-115 + 15
H, protein B, R (M. trich.)
4
> 50 + 15
> -170 + 15
H, R (M. trich.)
4
> 50 + 15
>-170 + 15
H, protein B, R, propylene
4
> 50
> -170
15
(M. trich.)
a No reduction was observed in these cases.
bSee appropriate references for specific experimental conditions.
15
50
o
r.
C
~1
U
I0
r-.
o
A
2
oC.
0
8
o
Lt)
d
0
III
m
C)
z0
1-4
C-
C.Z
o
o
0
C7\
ON
00
W
4Z.
V
00
C
O
Lr)
cr)
r
n
A
L-)
a.
z
0
C.
:
rS
Z
0
.5
.1
Z
oC
L.
or)
C
P6
C: C)
Wu
ad
S
C
u
cj,
U
ac:~o
uGd
75
U
aa
c 7O
o
US
X CZ
o
4e
x
0
o~~~~~~~~
o
X
r
X
;
4
3
Gd
=CO
Gd
''
gE
ou
r
X03
I.,x
o
E
(t
tA
O
U
d
51
Table 4. Spectroscopic and Kinetic Parameters for Intermediates in the
Reaction of Hred with Dioxygen.
optical
kmax, nm
-1
(E, M-lcm
M6ssbauer
)
M.
M.
trichosporiuma
capsulatus (Bath)b
Q
L
Q
330 (7500)
=625 (500)
420 (3600)
430 (7500)
kform,
S- 1
1
kdecay,
1-
0.05
350 (1800)
22
0.5
0.07
site 1:
6, mm/s
0.17
0.66
0.21
AEQ, mm/s
0.53
1.51
0.68
site 2:
Raman
6, mm/s
0.14
AEQ, mm/s
0.55
kform, s - 1
25
0.3
kdecay, s - 1
0.3
0.03
cm-1
904
aSee refs. 67 and 68. bSee refs. 49, 69, 70.
52
00
C;
U
z
0.)
C
0.)
0.)
U
0.
'S
C
0.
Ul
'-4
CI
53
CN
5q
.z
el
NA
,I
N
C,,
I
I
5-
54
.
cl
I-
0o
a)
I-
t-4
-0
·..
55
0
0
z
Z
I
C
0
/
t
o
C14
-
_
N
I
::.L
I\
O
0 Z0
0
0
0
I
Z
0--~f--_-©
Z
_,11
©
I)-/
'O
,0
tl
It
0
.0
0I\
0
oflo
Iv
_I
Z
O.
- L.
z0 0-0
Joz
Z
0
Z
56
Figure 3. Proposed route for conversion of intermediate L to intermediate Q.
The resulting diiron(IV) oxo species could bind oxygen with one iron, or the
oxygen could be bound symmetrically by both iron atoms.
57
'l'
\0
-
O-Feo
-0
N \0
N
2H+
H2 0
0
O
H
\
jo,0
O-FeV
-\.
N
H
o
o
O
O
O
N
0
N
O
H
O
I
F
N
Fe
'eIII-0O
/
\o
/
o /0 \
I
FeIV-O
OO-- Fe FelV<
or / \-O
0
,Q"
0
58
Figure 4. Proposed catalytic cycle for the hydroxylation of hydrocarbons by
MMO.
59
R-H.
ROH + H2 0
FeIII'
FeIII
H
-1
R-H
R-H
FeII,- Fe"'
J
shunt
H
pathway
NADH
0.31 s -l
2H +
R-H
Fe2 III(02 -2)
I
"Ll"'~.....~25
s'.
NA)' -+ HT
,
e'
Fe"
1
R-H:
02
I
H
w ~.lt.
60
V1
u4-
C
u
CY
0
U
o
cC
0
0)
x
0)
0
x
C
u)
o,.4
)
oo4
cu
o
.4-Z
oo
C
o
'4-
0'
C
C
I
o
z
E
.o
a
.- 4
a)
._
c
C
U
,C
a
tC
S0)
C,
0)
.4o
5-4
0)
Ur
C
U
o
o
U
U
Ca
--
Cd
C
Lta
U
=cC
61
,-i
\
of
0-/
,0-4
\
©-0
Co
0*
i
CD
I
.O-=
z-
0u
-a0
I
.-
wo
\
0.*
C0-
t~~~~~e
~
=
O-
O.-*
0-L
-
l-
4)
O-=
O.,ll
wo~
0-3:
.0-0
.
+
.
...
a-.
-IQ
0=
wo
U
W-0
O
O-=~
b-31
W-4e
=
U
cu
u
62
0-4
00-
_
0m..
w
0-*
0-_
L
wo
b-3:
~~ ~+
ct
0-=
C4
OSC-4
E
Ie >+
-
_
a;
tr
=0
cc( e3
o-- 4's
;>
O=
0
&-4
0 /
O-
LL
I
a)
C4
Q
-..
I
0=:
0-=
/
To=,
.. "0
,,/A/
v
63
CHAPTER 2.
Redox Properties of the Hydroxylase
Component of Methane Monooxygenase from Methylococcus
capsulatus (Bath) -- Effects of Protein B, Reductase, and Substrate
64
Introduction
Methane monooxygenase (MMO) from the methanotrophic bacterium
Methylococcus capsulatus (Bath) catalyzes the synthesis of methanol from methane
and dioxygen, which is the first step in the metabolic pathway of the organism. 1
MMO exhibits an extremely broad substrate specificity, reportedly inserting oxygen
into a wide variety of alkanes, alkenes, ethers, and alicyclic, aromatic, and
heterocyclic compounds. 2 In addition, MMO from Methylosinus trichosporium
OB3b oxidizes haloalkenes, including trichloroethylene (TCE).3 This reactivity
parallels that of the well-studied heme analogue, cytochrome P-450,4' 5 although the
mechanism of dioxygen activation may be more complex.
Soluble MMO from both organisms consists of three components, all of
which are necessary for activity.6 The hydroxylase component has a Mr of 251 kDa
and comprises three subunits in an o2P2Y2 configuration.
7
The hydroxylase is the
site of substrate binding and contains up to two non-heme dinuclear iron cores, the
number of which depends on conditions of growth, harvesting, and purification. 8
The fully reduced form of the hydroxylase is thought to be responsible for dioxygen
activation. 8 The 38.6 kDa reductase contains one FAD and one [2Fe2S] cluster with
reduction potentials of -260 and -220 mV vs. NHE, respectively. 9 The oxidized form
of the antiferromagnetically coupled [2Fe2S] cluster has no EPR signal, whereas the
reduced form has a resonance at gav = 1.96.10 The reductase accepts electrons from
NADH and subsequently transfers them to the hydroxylase.1
1
The third
component, protein B, has a Mr of 15.5 kDa, contains no prosthetic groups and has
been proposed to regulate electron transfer between the reductase and hydroxylase. 12
The hydroxylase has been previously characterized, in part by our laboratory,
with the aid of X-ray crystallography as well as optical, M6ssbauer, EXAFS, EPR, and
ENDOR spectroscopy. 13- 17 The native oxidized Fe(III)Fe(III) form is EPR-silent due
to antiferromagnetic coupling in the bridged diiron core, whereas the mixed valent
65
Fe(II)Fe(III) species exhibits an EPR signal with gav = 1.83 indicative of a non-heme,
bridged diiron center (Figure 1). The fully reduced Fe(II)Fe(II) form gives rise to a
signal at g = 15 characteristic of its integer spin (Figure 2). The EPR spectra of the
protein in all three oxidation states usually contain a minor signal at g = 4.3 assigned
to rhombic high spin Fe(III) and a free radical signal at g = 2.01. Each of these signals
generally account for less than 5% of total iron present in the sample and arise from
trace levels of mononuclear iron and protein associated radical, respectively. The
optical spectrum is essentially featureless above 300 nm in any oxidation state.
The dinuclear iron core in MMO has features in common with those in other
iron oxo proteins such as hemerythrin, ribonucleotide reductase, and purple acid
phosphatase.
Accurate models for the structural, physical, and chemical properties
of the diiron centers in these proteins have been a focus of attention in our
laboratory.
1 8 -2 0
The iron core in each protein functions differently. In hemerythrin
the iron center reversibly binds dioxygen, in purple acid phosphatase it hydrolyzes
phosphate esters, in ribonucleotide reductase it is required to generate a tyrosyl
radical, whereas in MMO it catalyzes the selective oxidation of methane to
methanol.
A compelling question yet to be answered from studies with both the
models and the proteins themselves is just how nature has selectively tuned each
active site to carry out its particular function. The highly specialized reactivity most
likely results from modifications in the coordination environment of the bridged
pair of iron atoms.
An early study of the hydroxylase reduction potentials reported a large
separation between the two reduction potentials separating the Fe(III)Fe(III),
Fe(]:I)Fe(III),and Fe(II)Fe(II) states, which suggested that high yields of the mixedvalent form could be produced.2 1
In practice, however, only = 25 % of this state was
observed. 1 4 To reconcile this behavior, the reduction potentials of the hydroxylase
were reinvestigated by monitoring the EPR spectrum of the mixed valent form at
66
various potentials. To probe factors that might control the functional properties of
the dinuclear iron center, the effects of the reductase and protein B components as
well as substrate on the potentials were studied.
Moreover, we discovered that
addition of substrate to the complex formed by the hydroxylase, protein B and the
reductase increased the redox potentials. Most of the work reported here was
previously published. 2 2 A recent study of the redox properties of hydroxylase from
M. trichosporiun OB3b reported different potential values obtained under
conditions purported to be closer to the natural system.2 3 Accordingly, additional
titrations were carried out under similar experimental conditions and are reported
herein.
Experimental
Protein Isolation.
Growth of M. capsulatus (Bath) and purification of the
hydroxylase were carried out as reported elsewhere.14
22
Protein B was prepared as
reported.2 4 For experiments with reductase, an unresolved mixture of reductase
and protein B obtained from the DEAE cellulose column described in the
hydroxylase isolation scheme was used without further purification.
Specific
activity (propylene) for the hydroxylase ranged from 150 to 250 mU/mg with 2.2 to
2.5 Fe/mole, but all preparations resulted in nearly identical EPR spectra.
Redox Titrations.
Titrations were performed in a manner similar to the
method described, elsewhere,2 1 under one of three slightly varying sets of conditions
referred to as A, B and C. Solutions in all conditions typically contained about 5 ml
of 100 M hydroxylase
in 50 mM MOPS (MOPS = 3-(N-morpholino)propanesulfonic
acid, Sigma), pH 7.0. Under conditions A, the following were added (100 jiM each) to
mediate electron transfer from the reductant to the hydroxylase:
N,N,N',N'-
tetramethyl-1,4-phenylenediamine dihydrochloride, 2,3,5,6-tetramethyl-1,4phenylenediamine, 2,6-dichloroindophenol (sodium salt, hydrate), duroquinone,
phenazine methosulfate, thionin, methylene blue, potassium indigotetrasulfonate,
67
Indigo Carmine, anthroquinone-1,5-disulfonic
acid (sodium salt hydrate),
anthroquinone-2-sulfonic acid (sodium salt, monohydrate), benzyl viologen
dichloride (all from Aldrich).
The mediators methylene blue and 2,6-
dichloroindophenol were omitted in conditions B. Conditions C were identical to B
except that the mediator concentrations were lowered to 20 gM.
The resulting solution was repeatedly evacuated and back-filled with
dioxygen-free argon, then transferred with a gas-tight syringe into an apparatus
described elsewhere.25 To perform experiments with substrate present, propylene
(Matheson, polymer grade) saturated with buffer was continuously passed through
the titration vessel. A redox combination electrode (Pt vs. Ag/AgCl, Corning) was
used to monitor the potential.
With constant stirring, small volumes of typically 1
to 3 l of a 0.1 M sodium dithionite (Aldrich) solution were used to lower the
potential.
After a 10 min period, or 90 min for conditions C, determined in time
course studies to be sufficient for the system to reach equilibrium, a 300 jgl sample
was removed with a gas-tight syringe and then injected into an argon-filled, quartz
EPR tube.
The tube was immediately frozen in cold isopentane (-140 C) and
immersed in liquid N 2.
For experiments performed with protein B, hydroxylase was incubated with
two equivalents of the protein prior to reduction. In studies of the complete system,
a mixture of protein B and reductase was added either before addition of reductant
or after the fully reduced form of the hydroxylase was produced. The amount of
protein B/reductase mixture used was determined by adjusting to achieve maximal
hydroxylase activity, each component = 200 M, or twice the concentration of the
hydroxylase as estimated by HPLC analysis.
EPR Spectroscopy. All EPR spectra were recorded at X-band on a Bruker
Model ESP 300 spectrometer with an Oxford Instruments EPR 900 liquid He cryostat
set at 9 K. The g < 2 signal was quantitated under non-saturating conditions by
68
double integration of the first derivative spectrum for comparison to a frozen
solution of copper perchlorate
(mM
CuSO 4 , 2 M NaC10 4 , 0.01 M HC1). Transition
probabilities were corrected for g-value anisotropy. 2 6
The percentage of mixed
valent species was calculated from the concentration of spin relative to the
concentration of iron present in the sample.
To compare different titrations the
mixed valent signals were normalized to the maximum signal observed for each
individual titration.
Results
Hydroxylase Titrations. Three titrations under conditions A were carried out
on the hydroxylase component of MMO. Representative EPR spectra are shown in
Figure 3.
The mixed valent signal increased in intensity as the potential was
lowered, reached a maximum at approximately -50 mV vs. NHE, then decreased as
the potential was lowered. At roughly -100 mV the fully reduced, g = 15, signal was
detected and grew in intensity as the potential was further diminished.
The
reversibility of the system was confirmed by back titration at several potentials. The
maximum intensity of the mixed valent signal corresponded to 85% of the total iron
present.
This yield most likely reflects the maximum attainable because of the
inaccuracies introduced by using a copper standard for the integrations.
The
reduction potentials of the hydroxylase under all conditions investigated are
summarized in Table 1.
A plot of the relative intensities of the mixed valent signal versus potential is
shown in Figure 4 for conditions A data.
Titrations of the hydroxylase were
additionally performed under both conditions B and C, and intensities of the mixedvalent EPR signals are plotted in Figure 5. The potentials for stepwise reduction of
the hydroxylase diiron(III) core, shown in eq. 1, were obtained by fitting the data in
this plot of mixed valent intensities vs. potential to eq. 2. This expression was
derived from the two individual Nernst equations governing the reductions and
69
the mass balance equation. The best fit, where n = 1, F is the Faraday constant, R is
E] °
Fe(III)Fe(III)
E2°
-
Fe(II)Fe(III)
2
Fe(II)Fe(II)
Fe(II)Fe(III) = 100 / [exp (-nF/RT) (E - E20)} + exp {(nF/RT) (E - E1°)} + 1 ]
(1)
(2)
the gas constant, and T = 298 K, yielded potentials of 48 and -135 mV vs NHE for E1°
and E20 , respectively, with estimated errors of + 5 mV. The E1 and E20 values can
also be used to compute the relative amounts of all three redox species present at a
given potential. The results of such a calculation are also depicted in Figure 4. The
disproportionation constant for the mixed valent species was computed to be 8.04 x
10-4 . Fits of the data from titrations under B and C conditions both yielded values of
100 mV and -100 mV vs. NHE for E1° and E2 ° , with estimated errors of ± 15 mV, as
shown in Figure 5.
A plot of the EPR signal intensity versus potential of the mixed valent
hydroxylase measured under conditions A in the presence of the substrate
propylene as a substrate is also shown in Figure 4. A fit of these data to eq 2 revealed
only a slight diminution of the reduction potentials, to values of 30 vs 48 mV and
-156 vs -135 mV for E1° and E2 ° , respectively. Titrations of hydroxylase with protein
B and were performed under conditions B and C. In these experiments, E1° and E2°
were lowered to 50 mV and -170 mV ± 10 mV, also shown in Figure 5. Although
shifts in the potential induced by substrate and protein B are small, they do suggest
some perturbation of the environment at the diiron center.
Component Interactions. Addition of protein B and reductase under
conditions A produced a significant change, illustrated
in Figure 6A. For all
potentials investigated down to -200 mV, no mixed valent nor any fully reduced
signal was detected, implying that no reduction occurred. When protein B and
70
reductase were added to the fully reduced form of the hydroxylase, the EPR signal of
the latter persisted. A dramatic change resulted when propylene was introduced to
all three MMO components. Reduction of the hydroxylase to the diiron(II) state was
observed upon initial introduction of reductant, even at potentials as high as 150
mV. Representative spectra are displayed in Figure 6B.
Discussion
Hydroxylase Potentials. As reported previously,2 1 the redox potentials for the
stepwise reduction of the hydroxylase component of MMO can be determined by
monitoring the appearance and disappearance of the g < 2 EPR signal intensity
characteristic of the mixed valent form. Our values (El1 = 48 to 100 and E20 = -135 to
-100 mV) for the potentials differ from ones reported early2 1 (350 and 25 mV,
respectively). We have no obvious explanation for the discrepancy, owing in part to
the lack of experimental detail given in the previous study. For example, values for
activity and iron content of the hydroxylase were not specified, nor was it clear that
the potentials were measured by an electrode. By using the present potential values
for the hydroxylase and E°' of 80 mV for phenazine methosulfate we calculate the
maximum attainable yield of the mixed valent species at this potential to be 20 to
600%. The redox potentials determined in this study could explain why, at 100 mV,
the spin concentration of the M. capsulatus (Bath) hydroxylase was only 0.27 per
protein molecule in previous reports. 2 1
Finally, the present results are more in
accord with an earlier report for M. capsulatus (Bath) that formation of the mixed
valent hydroxylase occurs "at redox values around 0 mV" and that the fully reduced
form can be obtained "below - 100 mV".2 7
Incubation with low concentrations of mediators (20 gM, conditions C vs. 100
pM, conditions B) relative to 100 jgM hydroxylase resulted in nearly identical
potentials.
A longer time of 90 min, however, was necessary for the solutions to
reach equilibrium under for conditions C. This behavior is consistent with the M.
71
trichosporium OB3b work, in which conditions most like set C were employed, and
an incubation time of 1 h was used.
Although the separation between the potentials (E2°- E10 ) remained the same
within experimental error (183 vs. 200 mV) with different mediators present, the
potentials changed from 48 to 100 mV and from -135 to -100 mV. The increase of the
potentials suggests reduction may be mediator-dependent.
A recent report of the
determination of M. trichosporium OB3b hydroxylase redox potentials suggested
that methylene blue and 2,6-dichloroindophenol may perturb the diiron unit and
alter the EPR signal as well as the potentials.
We have determined that the
lineshape of the signal is not perturbed by addition of these two mediators.
The
Fe(II)Fe(III) EPR signal in the presence of these two mediators is given in Figure 1,
and the signal in their absence is illustrated in Figure 7. No shift in the g-values of
the signal was seen, but a small feature at higher field was observed in their absence.
Examination of the signal lineshape does not conclusively exclude possible
perturbation of the active site by the mediator, making it very difficult to determine
which mediator(s), if any, actually do interact with the diiron unit. The increase in
potential cannot be explained by loss of the redox buffering capacity in the potential
regions covered by the omitted mediators. Table 2 lists all the mediators employed,
together with their reduction potentials. 2 5 Omission of methylene blue and 2,6-
dichlorophenolindophenol
removes buffering in the potential
ranges of
approximately 41 to -19 mV and 247 to 187 mV vs. NHE, respectively, which are not
those where the hydroxylase potentials lie. Whereas the reduction does seem
mediator-dependent, identifying the nature for the dependence has proven difficult.
To combat problems associated with mediators, the reduction potentials could be
determined from a set of titrations in which each mediator is systematically omitted,
thereby determining the effect of each mediator on the potential. In addition, direct
electrochemical measurement of the redox potentials at an electrode surface would
72
eliminate possible complications with mediators, but use of this methodology with
proteins has other signficant problems such as protein denaturation and slow rates
of electron transfer. These type of experiments were attempted, but no reduction of
the hydroxylase was observed.2 8
Literature Comparisons. The E1° and E2 values obtained in this work (100
and -100 mV) differ from the potentials found in studies with hydroxylase from M.
trichosporium OB3b (76 and 21 mV) under similar conditions. 2 3
Our values are
closer to these numbers than was previously reported (350 and 25 mV).2 1 The
reason for the discrepancy with M. trichosporium OB3b work may be experimental
variations, or possibly differences in the MMO systems from the two organisms.
Examples of variable behavior between the two systems in other areas include
results with radical clock probes,22 unpublished work with chiral ethane as a
substrate, and details of the results of kinetic studies of intermediates in the catalytic
reaction cycle. Alternatively, the reduction potentials may depend on different
mediators employed. In the M. trichosporiumOB3b work, two additional mediators
were substituted for methylene blue and 2,6-dichloroindophenol,
but since these
two were not readily attainable, we could not exactly duplicate their experimental
conditions. As mentioned above, it appears that the electron transfer mediators are
non-innocent, which will complicate comparisons of hydroxylase potential values
obtained under different conditions.
Structural Implications.
Several synthetic model complexes containing
diiron oxo and related units are known, but only a few of these display reversible or
quasi-reversible redox reactions connecting the Fe(II)Fe(II), Fe(II)Fe(III), and
Fe(III)Fe(III) states
2 9-32
A comparison of the structures and redox properties of the
dinuclear iron centers in the proteins and model compounds leads to some
interesting observations.
Hemerythrin contains an oxo bridge in the fully oxidized
state, whereas MMO does not.1 4
The diiron center in hemerythrin has reduction
73
potentials
of El'' = 110 and E2° = 310 mV vs NHE.2 9 Since E1° is less than E2° , the
mixed valent form of the protein is unstable with respect to disproportionation to
the diiron(II) and diiron(III) forms. 3 3 A mixed valent form of the R2 protein of
ribonucleotide reductase, which also contains an oxo-bridged dinuclear iron center,
has been observed only in very small quantities (< 10 %).29 The difficulty in
obtaining the mixed-valent form of R2 suggests that E1° is less than E20 for this form
of the protein.
The mixed valent form of the oxo-bridged model complex
[Fe2 O(O 2CCH 3) 2(Me 3TACN)2 ] 2 + , where Me 3TACN is the facially capping ligand
1,4,7-trimethyl--1,4,7-triazacyclonane,
disproportionation.
30
is also unstable
with
respect
to
Thus the oxo-bridged models and proteins may share the
common feature that E1° < E2 °. The presence of a hydroxo bridge in the oxidized
and mixed valent hydroxylase may cause E1° > E20°. Binding of reductase and protein
B could cause significant perturbations shifting E2 > E1 °.
A diiron complex in which the metals are connected by two phenoxide
bridges has been described which undergoes quasi-reversible redox reactions similar
to those in eq. 1 with E 1 = -9 and E2 = -259 mV vs NHE in dimethylformamide
solution. 3 2
Although it is not valid to compare the absolute values of these
potentials to those of MMO as determined here, the difference between the two, 250
mV, is reasonably close to the difference between the corresponding
potentials,
protein
200 mV. In this complex, both iron atoms are octahedrally coordinated
to oxygen donor ligands. These observations indicate that the structure of the core
has a direct influence on the redox behavior.
In particular,
an increase in the
number of nitrogen ligands relative to oxygen ligands has been correlated to higher
reduction potentials of diiron complexes.3 4 The influences of protein B, reductase,
and substrate on the hydroxylase reduction potentials could, for example, result
from a change in the number or protonation of an oxygen ligand(s).
74
Component Interactions. Work with hydroxylase from M. trichosporium
OB3b also indicates that, in the presence of protein B, the g values, and thus the
lineshape, of the mixed valent signal change, from 1.94, 1.86, and 1.75 to 1.87, 1.77,
and 1.62, respectively. 3 5 Protein B additionally caused the reduction potentials of
this hydroxylase to shift more negative values, from 76 and 21 mV to -52 and -115
mV vs. NHE, respectively, for E1 ° and E2.2 3 For M. capsulatus (Bath) hydroxylase,
we have observed a dimunition of the hydroxylase reduction potentials with
protein B to values of 50 and -170 mV vs. NHE, as shown in Table 2. In both
systems, it is evident that binding protein B must perturb the diiron center, probably
by altering the coordination environment of the iron atoms. Additional evidence
that protein B alters the coordination sphere of the diiron center is provided by the
shifting of hyperfine coupling constants in the ENDOR spectrum of the mixed
valent state. 36 To elucidate the nature of this perturbation, more structural work of
the hydroxylase/protein
B complex is imperative.
Addition of protein B and the reductase to the oxidized hydroxylase
component blocks the mediated transfer of electrons to the dinuclear iron core. Use
of NADH as the reductant in place of sodium dithionite yielded similar inhibition,
and suggests that the electron transfer pathway through either the mediators or the
reductase is the same.
These results account for previous observations that a
mixture of hydroxylase and reductase oxidizes NADH with no substrate turnover,
but that addition of protein B to this mixture results in NADH consumption only
when substrate
is oxidized.l
2
The same report stated that the rate of NADH
oxidation by a mixture of hydroxylase and reductase was independent of substrate
and. occurred at only 40% of the rate in the complete system (hydroxylase, protein B,
reductase, and substrate) to produce either hydrogen peroxide or water. These
observations indicate that protein B is necessary to obtain high electron transfer
rates, and to couple the transfer with substrate turnover. Studies of MMO from M.
75
trichosporium OB3b indicate that only a large excess (17 equivalents) of protein B
can inhibit mediated reduction of hydroxylase in the absence reductase.3 5 Our
observations for the M. capsulatus (Bath) enzyme reveal that mediated electron
transfer is completely inhibited upon addition of protein B and reductase at ratios
that maximize hydroxylase activity, approximately 2:2:1 molar quantities of
reductase:protein B:hydroxylase. These results, together with those from the
previous studies, 12 indicate that both protein B and the reductase may be necessary
to inhibit efficiently reduction of the hydroxylase.
This inhibition could either be
thermodynamic, for example by alteration of the iron coordination environment in
some way and hence the redox potentials, or kinetic, through blocking access of
reductant.
The finding that the addition of protein B and reductase to the fully
reduced hydroxylase does not lead to reoxidation of the enzyme implies that the
inhibition of electron transfer is kinetically controlled. This explanation is attractive
in view of evidence that stable complexes form among the three protein
components.
35
filtshould be noted that electron transfer from sodium dithionite to
the redox mediators is believed to be fast relative to the rate of direct electron
transfer to the proteins,2 5 and so the potential changes are indeed buffered. Because
we limited our redox titration to potentials greater than - 210 mV (Figure 5), no EPR
signal characteristic of the reduced reductase (g = 1.96)was observed.
Perhaps the most interesting discovery of the present investigation is that,
upon addition of propylene to the three components of MMO, electrons are
transferred to the hydroxylase even at the highest potentials monitored, > 150 mV.
Electron transfer into the dinuclear iron center of the complete system is thus tightly
regulated, being allowed to occur only when substrate is present.
Since no mixed
valent is observed, E20 has shifted such that E2° > E1°, which indicates that a change
in the coordination sphere of the diiron unit has taken place. E2 is estimated from
the amplitude of the diiron(II) signal at varying potentials to be 100
25 mV vs.
76
NHE. This estimate should be considered to be only a rough approximation since
the nature of the g = 15 EPR signal makes it difficult to quantitate accurately.
These findings are consistent with the steady-state kinetic analysis of the
MMO complex reported previously, where electron transfer was observed to occur
only after substrate was bound.3 7 Such an increase in reduction potentials is similar
to behavior seen with cytochrome P-450, where small changes in the coordination
sphere of the heme iron due to the binding of the substrate can raise the reduction
potential from -300 to -170 mV.3 8 Parallel work with MMO from M. trichosporium
OB3b failed to demonstrate a similar mode of regulation.2 3 Instead, it was reported
that the potentials of the hydroxylase/reductase,
B, and the hydroxylase/reductase/protein
the hydroxylase/reductase/protein
B/substrate complexes were "slightly
more positive" than the values for hydroxylase alone. These discrepancies may be
attributed either to procedural or protein variations.
The radically different behavior of the MMO complex in the presence and
absence of substrate has several implications. As stated earlier, binding of protein B
and reductase prevents reduction of the hydroxylase, probably kinetically. Binding
of substrate to the hydroxylase not only reverses this effect, but also renders the
hydroxylase
more electron deficient and hence easier to reduce.
This
thermodynamic effect would occur only when all three components are present,
and the greater driving force could explain the faster rates of NADH oxidation
observed when protein B is available. The increased affinity for electrons of the
complete MMO system in the presence of substrate could be due in part to enhanced
electrophilicity of the diiron(III) core through proton transfer or hydrogen bond
formation involving one or more ligands in the coordination sphere.
hydroxylase/protein
B/reductase
Since the
complex3 5 appears to be responsible for the
inhibition of electron transfer without substrate, the enhancement with substrate
might involve a conformational change at the dinuclear iron center.
More
77
structural work involving the component interactions is necessary to reach firm
conclusions in this area.
Regulation of electron transfer as just described would be beneficial to the
organism in many ways.
The obvious advantage of the kinetic inhibition is that
when no substrate is present, no NADH consumption takes place, thereby
preventing the wasteful consumption of reducing equivalents. 12 If reduction were
allowed to occur in the absence of substrate, there would be a greater chance that the
enzyme might be inactivated through generation of the highly reactive, high energy
species required for the oxygenation reaction. By moving E2 > E2°, the enzyme
prevents formation of the catalytically inactive mixed-valent form.
From an
energetic standpoint, the shift to higher reduction potentials with substrate present
is consistent with the core becoming more electron deficient.
This potential shift
would increase the reactivity of an activated iron core. Without protein B and
reductase, such a species would be lower in energy owing to the lowered reduction
potentials.
In single turnover experiments in which hydroxylase is chemically
reduced to the diiron(II) form, low yields are generally observed in the absence of
the other components.3 9 This argument implies that the higher energy activated
hydroxylase iron core that is generated in the presence of protein B and reductase
may be required for efficient substrate oxidation.
78
References
(1)
Hanson, R. S. In Advances in Applied Microbiology Academic Press: New
York, 1980; Vol. 26; pp 3.
(2)
Green, J.; Dalton, H. J. Biol. Chem. 1989, 264, 17698-17703.
(3)
Fox, B. G.; Borneman, J. G.; Wackett, L. P.; Lipscomb, J. D. Biochemistry 1990,
29, 6419-6427.
(4)
Ortiz de Montellano, P. R. Cytochrome P-450 Structure, Mechanism, and
Biochemistry; Plenum Publishing Corp.: New York, 1986, pp. 217-271.
(5)
Dawson, J. H.; Eble, K. S. Adv. Inorg. Bioinorg. Mech. 1986, 4, 2-64.
(6)
Fox, B. G.; Lipscomb, J. D. Biochem. Biophys. Res. Comm. 1988, 154, 165-170.
(7)
Colby, J.; Dalton, H. Biochem. J. 1978, 171, 461-468.
(8)
Fox, B. G.; Froland, W. A.; Dege, J. E.; Lipscomb, J. D. J. Biol. Chem. 1989, 264,
10023-10033.
(9)
Lund, J.; Dalton, H. Eur. J. Biochem. 1985, 147, 291-296.
(10)
Colby, J.; Dalton, H. Biochem. J. 1979, 177, 903-908.
(11)
Green, J.; Dalton, H. Biochem. J. 1989, 259, 167-172.
(12)
Green, J.; Dalton, H. J. Biol. Chem. 1985, 260, 15795-15801.
(13)
Rosenzweig,
A. C.; Frederick, C. A.; Lippard, S. J.; Nordlund,
P. Nature 1993,
366, 537-543.
(14)
Dewitt,
J. G.; Bentsen, J. G.; Rosenzweig,
Pilkington, S.; Papaefthymiou,
A. C.; Hedman,
B.; Green, J.;
G. C.; Dalton, H.; Hodgson, K. O.; Lippard, S. J. J. Am.
Chem. Soc. 1991, 113, 9219-9235.
(15)
DeRose, V. J.; Liu, K. E.; Kurtz, D. M., Jr.; Hoffman, B. M.; Lippard, S. J. J. Am.
Chem. Soc. 1993, 115, 6440-6441.
(16)
Fox, B. G.; Surerus, K. K.; Miinck, E.; Lipscomb, J. D. J. Biol. Chem. 1988, 263,
1053-1056.
79
(17)
M. P.; Fox, B. G.; Andersson,
Hendrich,
K. K.; Debrunner,
P. j.; Lipscomb,
J. D.
J. Biol. Chem. 1992, 267, 261-269.
(18)
Lippard, S. J. Angew. Chem. Int. Ed. Engl. 1988, 27, 344-361.
(19)
Tolman, W. B.; Liu, S.; Bentsen, J. G.; Lippard, S. J. J. Am. Chem. Soc. 1991,
113, 152-164.
(20)
Feig, A. L.; Lippard,
(21)
Woodland,
S. J. Chem. Rev. 1994, 94, 759-805.
M. P.; Patil, D. S.; Cammack, R.; Dalton, H. Biochim. Biophys. Acta
1986, 873, 237-242.
(22)
Liu, K. E.; Ilippard, S. J. J. Biol. Chem. 1991, 266, 12836-12839, 24859.
(23)
Paulsen,
K. E.; Liu, Y.; Fox, B. G.; Lipscomb,
J. D.; Miinck, E.; Stankovich,
M. T.
Biochem. 1994, 33, 713-722.
(24')
Liu, K. E.; Johnson,
C. C.; Newcomb,
S. J. J. Am. Chem. Soc. 1992,
M.; Lippard,
115, 939-947.
(25)
Dutton, L. P. Methods in Enzymol. 1978, 54, 411-435.
(26)1 Asa, R.; Vanngard, T. J. Mag. Res. 1975, 19, 308-315.
(27)
Dalton, H.; Leak, D. J. In Gas Enzymol.; H. Degn, Ed.; D. Reidel Publishing Co.:
London, 1985; pp 169-186.
(28)
This work was attempted in collaboration with Professor Frasier Armstrong
at the University of California, Irvine.
(29)
Que, L., Jr.; True, A. E. Prog. Inorg. Chem. 1990, 38, 97-200.
(30)
Hartman,
Papaefthymiou,
J. A.; Rardin,
R. L.; P., C.; Pohl, K.; Nuber,
B.; Weiss,
J.;
G. C.; Frankel, R. B.; Lippard, S. J. J. Am. Chem. Soc. 1987, 109, 7387-
7396.
(31)
Snyder, B. S.; Patterson,
G. S.; Abrahamson,
A. J.; Holm, R. H. J. Am. Chem.
Soc. 1989, 111, 5214-5223.
(32)
Stassinopoulos,
A.; Schulte, G.; Papaefthymiou,
Chem. Soc. 1991, 113, 8686-8697.
G. C.; Caradonna,
J. P. J. Am.
80
(33)
Armstrong,
F. A.; Harrington,
P. C.; Wilikins, R. G. J. Inorg. Biochem. 1983, 18,
83-'91.
(34)
Rosenzweig, A. C.; Feng, X.; Lippard, S. J. In Applications
of Enzyme
Biotechnology; Kelly, J. W., Eds.; Plenum Press: New York, 1991.
(35)
Fox, B. G.; Liu, Y.; Dege, J. E.; Lipscomb, J. D. J. Biol. Chem. 1991, 266, 540-550.
(36')
DeRose,
V. J.; Liu, K. E.; Lippard,
S. J.; Hoffman,
B. M. manuscript
in
preparation.
(37)
Green, J.; Dalton, H. Biochem. J. 1986, 236, 155-162.
(38)
Raag, R.; Poulos, T. L. Biochem. 1989, 28, 917-922.
(39)
Single turnover experiments reported in Chapter 6 are maximized with two
equivalents of protein B, although protein B actually diminished the hydroxylase
reduction potentials.
In this case, the increase in reactivity could arise from kinetic
effects. For example, steric modifications of the active site imposed by protein B
could occur which cause the substrate to bind tighter, thereby coupling oxidation of
Hred with substrate hydroxylation more efficiently. Reductase did not affect the rate
constant or yield of the reaction. Mediated reduction of Hox was inhibited by the
substrate of the reaction, nitrobenzene. This limitation prevented the production of
Hred
from Hox in the presence of protein B, reductase, and substrate to test the
reactivity of this putative highly reactive species.
81
Table 1. Reduction Potentials (mV vs NHE, pH 7.0) of the MMO
Hydroxyla.se from M. capsulatus (Bath) Under Various Conditions (H =
hydroxylase, R = reductase). Also included are the potentials for M.
trichosporium OB3b hydroxylase23 ,b and hemerythrin, 2 9 both sets of
which were obtained under other conditions.
Component(s)
Conditions
E10
E2°
H
A
48 ±+5
-135 + 5
H, propylene
A
30 ±+5
-156 + 5
H, protein B, R
A
a
a
H, protein B, R, propylene
A
< 100 + 25
= 100 + 25
H
B, C
100 + 15
-100 + 15
H, protein B
B, C
50 + 15
-170 ± 15
H (M. trich.)
-52 + 15
-115 + 15
H, protein B (M. trich.)
-76 + 15
-21 ± 15
H, protein B, R (M. trich.)
> 50
15
> -170 + 15
H, R (M. trich.)
> 50 + 15
> -170
H, protein. B, R, propylene
> 50 + 15
>-170 + 15
110
310
(A4. trich.)
Hemerythrin
a No reduction was observed under these conditions.
15
82
Table 2. Reduction Potentials of the Mediators Employed
in the Titrations (pH 7.0).25
Mediator
E° (mV vs. NHE)
N,N,N',N'-tetra-
276
methylphenylenediamine
diaminodurene
260
2,6-dichloroindophenol
217
duroquinone
140
phenazine methosulfate
80
thionin
56
methylene blue
11
potassium indigotetrasulfonate
-46
indigodisulfonate
-125
anthroquinone-1,5-disulfonate
-174
anthroquinone-2'-sulfonate
-225
benzyl viologen
-358
83
oH
XEa
o
o
0c C,
cr
C)
o
tc
c_
a
*C
,o
*I
-
-
_
c:
84
oN
=
o
t
No
C)
Nu
_o
kX
10
0(z
85
ca
o
Q)
4-
bC
Ui
e3
Ct
x
C
ti
Q.
"t
o
o
C:_
cn
o
o
U
ca
c:
C
Q)
to
CC
o
co
7
-
"o
as
cl
X
C)
c)
a
X
.c©4e
E
i0
!
Xbt3
cu
.
86
0
0
Lf
1n
co
ri
N
*
m
o
re
cn
II')
,-
r~o
_
___ __
0
If,
87
CD
-o
UC)
obo
Co
Sa
0,
0,
0
70
0
e
0)
c0
0)
or
-
0t
c0
0)
I.
0
o
0)
X
X
U.
0)
88
0
to
h,
0
0
o
E[PXP
0
89
¢
._
"0
o
'0
a)
cn
S
c)U
a
"0
"0
o
ao
vv
o
0
*
.C
a
a)
3
a)
5-
a)U
n
a
a)
c"
a)
U)
"0
u
c
oa)
0
-_e m
h
a
-k
E
a
©
c
o0
.
o 0
u
.,
ca
cl
0
"0
'
a)
a)
I
-
o
U:
.
ct
'0
ct
cl
_s
oc
o
vo
a)
._
U:
*4
a)
k
c
'0
cU
.-
,
"0
U
0.-4
'4
o-
bO
._
a)
u0
u
7
0
ac"
.
oi
*_
o
X
u)
ca)
a)
._
oOC
"0
*.=
au)
a
aGJ
U)
I-
a)
:8
©
'0
o
a
U)
0
0
"
'0
u
.
v
o
0
0
L
=Z
90
n
I
I
I
a
I
I
I
I
N
\I
%% I..
,
..
d
't- 1.,1 e
a
ow
0
4
0
0
0
v
0
· · · ·· · ·
0>
°~C)I
· · · · ·
~~~~0ro-
0
0
0c; C
0
0
a
8
I
I
0
I
0(D
I
0
(%) luaseJd unowtV
I
0
a
0
91
o
.
tt
C
I
rl
.-
I
a
cr
o
4,4
4.-
Cl
a)
x
o
o
.4-
"o
0.)
Cl)
C
o
4c,
oo
:.
U)
o
.4-
,.
"0
u')
o
,.
CZ
I.
U
-"0
o
s
.E
"0
C
"0S
C)
UcU
:o
u
P.l C
o
..
C
0.)
4
92
V-4
z
eu
I._
o
cN
ce
00)
10
tn
0
10
% uasald unourV
10
N
0
93
Figure 6. EPR spectra of the hydroxylase at various potentials in the presence
of protein B and reductase. Set A were taken from samples without substrate,
and set B was recorded from samples with propylene present.
Conditions:
100 iM hydroxylase, 1 mW microwave power, 1 x 104 receiver gain, center
field 3050 gauss, sweep width 6000 gauss, temp 9 K. The radical signal at g = 2
was omitted for clarity.
94
A.
B.co
xX0
1250
2450
3650
B (gauss)
4850
95
.:
a)
-
.>
ca
a)
)1
U
a)
o
xa)
cEa
a)
"6
a)
o
a
ca
o0
.5
a)
CZ
c4
ao
a)
'-,
©
a
E
tiS
x0
©
O
;E
aa)
I
oo
oa
Q~
o
a
. .
I-l
5
0
._
w
ca
a)
ca
Ub
o
._-
o
U
L
U)
o
a)
x
CS,
a)
Z
xr
._
96
0
00
U,
#1
CD
cn
eQ
D
)
ci
Xo
X0
"O
t
97
CHAPTER
3.
Electron Nuclear Double Resonance Studies of
the Dinuclear Iron Center in the Hydroxylase Component of
Methane Monooxygenase from Methylococcus capsulatus (Bath)
98
Introduction
Electron nuclear double resonance (ENDOR) spectroscopy is a valuable
technique used in the structural characterization of proteins. 1 This method
indirectly affords an NMR spectrum of nuclei that interact with the electron
spin of a paramagnetic center. Information is provided through the observed
ENDOR frequencies, and directly gives the electron-nuclear coupling
constants.
The transition of the nucleus of interest is not actually observed,
but its interaction with the paramagnetic center is detected through a change
in the EPR signal intensity. ENDOR is a broad-band method very specific for
nuclei that having a hyperfine interaction with the electron spin system being
probed.
The hydroxylase component of methane monooxygenase from
Methylococcus
capsulatus
(Bath) is isolated in its native Fe(III)Fe(III)
oxidation state (Hox).2 ' 3 Because of antiferromagnetic coupling between the
two Fe(III) atoms (J = -32 cm- 1, H-ex = -2J Si S2), this form of the enzyme is
EPR silent liquid helium temperatures.2' 3 The mixed-valent Fe(II)Fe(III)
oxidation state (Hm,,) is readily accessible by one-electron reduction of the
ciinuclear center, and gives rise to a characteristic EPR signal with gay =
1.83.2,3 Further reduction to the fully reduced Fe(II)Fe(II) hydroxylase (Hred)
results in a ferromagnetically coupled, even spin system exhibiting an EPR
feature at g = 15.2-4 EPR signals at g = 4.3 and 2.0, both of which account for <
5 % of total iron, are present in spectra of all oxidation states and are assigned
to adventitious Fe(III) and protein associated free radical, respectively.
In
collaboration
with Professor Brian Hoffman and Dr. Victoria DeRose at
Northwestern
University, we have applied the ENDOR technique to study
the diiron center in Hmv probing the active site under a variety of conditions
to complement parallel structural studies of the diiron center.
99
To obtain information about the coordination sphere of the Fe atoms
of Hmv, ENDOR spectra of native and treated samples, including D2 0exchanged,
measured.
1 5 N-enriched, 5 7 Fe-enriched,
and H2170-exhanged protein, were
Hmvrin the presence of protein B was also studied to monitor
protein interactions.
Effects of the inhibitor DMSO5 on the diiron unit were
determined by comparison to ENDOR spectra of untreated samples.
Isotopically labelled DMSO was employed to obtain additional information
about its binding. ENDOR spectra of samples treated with the substrates CH 4
and CD4 were also recorded. For comparison, the 1 H ENDOR spectrum of
Hmv from Methylosinuls trichosporium OB3b was collected.
Basics of EPR and ENDOR Spectroscopy. 6 '7 An unpaired
electron is
characterized by a quantum number ms and a magnetic moment, R-e,which
are related by equation (1) (g = 2.0023 for a free electron,
= electronic Bohr
magneton = 0.927 x 10-20 erg G-1 ). The energy of the electron magnetic
quantum levels when a magnetic field H is applied are given in equation (2).
Two levels can arise where ms = +1/2, as illustrated in Figure 1. Transitions
between the two levels are governed by the selection rule Ams = +1 and occur
upon absorption of microwaves according to equation (3), where h = Planck's
constant and u = frequency. With the application of a constant microwave
ge = -g
E =-
I
(1)
ms
e H=g
hu =g H
Pm
H
(2)
(3)
frequency, absorption will occur at a characteristic magnetic field strength. By
scanning the magnetic field under these conditions and detecting microwave
power absorption, an EPR spectrum is obtained. Saturation of the EPR signal
100
can be achieved at a static magnetic field. The populations of the two energy
levels can thus become equalized, thereby causing the absorption of
microwaves to cease.
The interaction of a nuclear spin with an unpaired electron is next
described. The interaction energy in this case contains three terms, shown in
equation (4). Division of eq. (4) by Planck's constant, h, yields equation (5).
The nuclear Larmor frequency,
n, and the electron resonant frequency, ')e,
are defined in eqs. (6) and (7). Substitution of t'n and ue into eq. (5) leads to eq.
(8), which defines additional splitting of the electronic energy levels as
illustrated in Figure 2. The magnitude of the parameter
E=g3 H m -gnn H m+hAms
E/h = (g
m
H ms) / h - (gn En H mI) / h + A ms mI
A 1/2 relative to
Un
(4)
(5)
Un = (gn Pn H) / h
(6)
Ue = (g P H) / h
(7)
E/h = tems - Un mI + A ms mI
(8)
can vary. The case where Vn > A 1/2 is shown in Figure 2A, and the reverse,
'un < A 1/2, is illustrated in Figure 2B. EPR transitions (selection rules: Ams =
+1, AmI = 0) occur between the energy levels 1 to 3 and 2 to 4 in Figure 2A and
between 1 to 4 and 2 to 3 in Figure 2B. The resulting spectrum in both cases
will display microwave absorption at two field values.
The ENDOR spectrum for the case in Figure 2A is obtained and
analyzed as follows. Saturation the EPR signal involving transitions between
levels 1 and 3 will result in equal populations of those two energy levels. A
second radiofrequency pulse, referred to as the nuclear frequency, is then
applied to the system. The nuclear frequency is scanned with the goal of
101
matching
n+
IA1/2. At the nuclear frequency
n + IA /2, for example,
energy is absorbed, flipping the nuclear spin mi (Ams = 0, AmI = + 1 for I = 1/2)
and causing a transition from level 3 to level 4 to occur. Since the population
of level 3 is now less than the population of level 1, electron transitions
between level 1 and level 3 resume with the absorption of the microwave
frequency, and an increase in the EPR signal is observed. This change in EPR
signal intensity
is termed the ENDOR effect.
In Figure 3A, ENDOR
resonances are "Larmor-centered" and expected to occur at two nuclear
frequencies,
n ± IA 1/2. In Figure 3B the ENDOR spectrum
centered," with resonances at A 1/2 + 'n.
is "hyperfine-
In practice, only one resonance of
this doublet might be observed.
Experimental
Native Samples.
Growth
of native M. capsulatus (Bath) and
purfication of hydroxylase and protein B were carried out as described
elsewhere. 2' 8 Specific activites and iron content were in the ranges reported.
Growth of M. trichosporium OB3b and purification of its hydroxylase were
carried out as reported. 9 Its specific activity for conversion of propylene to
propylene oxide was monitored with the use of protein B and reductase from
M. capsulatus (Bath). Cross-reactivity between the two systems has been
previously reported. 1 0 ll The measured activity was 180 mU/mg of protein, a
value considerably less than expected, 9 and could be due to the use of protein
B and reductase from a different organism.
Reactions of M. trichosporium
OB3b hydroxylase with hydrogen peroxide produced product amounts in the
reported range. 1 2 The measured iron content, 2.2 mol Fe/mol hydroxylase,
was lower than published values, 9 for which we have no obvious
explanation.
102
15N
enriched
Enrichment. M. capsulatus (Bath) cells were grown in an
medium 13 in order to obtain correspondingly
1 5N
15 N-enriched
hydroxylase samples. The usual media for fermentations was altered only in
that KNO 3 (99 % 15 N, Cambridge Isotope Laboratories) was used to inoculate
five 5 ml cultures with cells from frozen stocks. The head space in each tube
was then replaced with methane. The cultures were allowed to incubate at 45
°C and 250 rpm for 4 days.
The methane was replaced daily.
Once the
cultures were opaque, they were used to inoculate five 50 ml cultures. The
head spaces of these cultures were again replaced with methane. The 50 ml
cultures were allowed to incubate overnight and subsequently used to
inoculate the fermentor.
The total volume of the fermentation was 10 1.
After 48 hours (O.D. at 540 nm = 8.9) the cells were centrifuged and 51 g of cell
paste was collected. Half of this material was then cracked and purified by the
usual methods, yielding 247 mg of hydroxylase (2.4 Fe/mole,
SPA 190
mU/mg).
57 Fe
Enrichment. This growth was performed in the general manner
described above except that the normal medium 13 was changed such that the
source of Fe was Na 5 7 FeEDTA, which was prepared
as follows. 2
5 7 Fe
foil (7
mg, 0.125 mmole, Cambridge Isotope Labs) was dissolved in 1 M of ultrapure
HNO 3. Next, 0.5 ml of H 20 and 80 mg (0.2 mmole) of Na 4EDTA2H 2O were
added. After stirring for 5 min, 4.1 mg (0.1 mmole) of NaOH was added. The
resulting mixture was allowed to stir for 6 hours, after which it was
centrifuged to pellet the precipitate. The dark green supernatant (1 ml) was
added to 100 ml of 20 mM sodium phosphate buffer (pH 7.0) and autoclaved
as required in the normal medium preparation for fermentation.
A 10 1
fermentation was carried out starting from the frozen stocks as described
above. The yield of cell paste was 50.3 g. Two 25 g batches of cells were
103
cracked to yield a total of 120 mg hydroxylase (2.0 Fe/mole, SPA = 163
mU/mg).
D 2 0 Exchange. Deuterium exchange of the hydroxylase was achieved
in the following manner. Liquid chromatography was carried out at 4 C with
a flow rate of 1 ml/min, and the buffer system was 25 mM MOPS (MOPS = Nmorpholinoporopane
sulfonic acid) in D2 0, pD 7.0 (as measured with a pH
electrode) with 50 mM NaCl unless otherwise specified. Purified hydroxylase
(= 50 mg) was loaded on a 1.5 x 5 cm Q Sepharose anion exchange column that
was previously equilibrated with 25 ml buffer. The bound protein was next
washed for 30 min and the wash buffer was discarded. Fresh buffer (100 ml)
was circulated over the column for either 8 or 15 hours. The hydroxylase was
then eluted with buffer containing 0.3 M NaC1.
H21 7 0 Exchange. Since this reagent is quite expensive ($300/ 1 ml, 20 %
enriched, Isotech), the D 20 procedure was not followed. Instead, hydroxylase
(= 50 mg) was concentrated to about 100 l and then diluted with 1 ml of
HE2 1 70.
The resulting mixture was allowed to incubate for at least 12 hours
before concentrating the sample for preparation of Hmv.
Sample Preparation. Hydroxylase,
50 mg per sample, was first thawed
and then concentrated to a volume of 2 ml at 4000 rpm using centriprep
centrifugal concentrators with a molecular weight cut-off of 40,000. The 2 ml
concentrates were then placed in centricon concentrators with the same
molecular weight cut-off. The sample volume was decreased to
100 g1l,
resulting in a concentration of at least 1 mM. A small volume of a mediator
stock solution that contained phenazine methosulfate, methylene blue, and
potassium indigotetrasulfonate (10 mM in each, dissolved in H 20) was added
to achieve a :L mM concentration of each mediator in the sample.
The
solution was placed in a vial, sealed with a septum, and then made anaerobic
104
by repeatedly evacuating and back-filling with dioxygen-free argon. To reduce
each batch, a small volume of a 0.1 M solution of sodium dithionite was
added to achieve a concentration of = 1 mM sodium dithionite in the sample.
The solutions were allowed to equilibrate for 30 min, and at the same time
were brought into anaerobic chamber (Vacuum Atmospheres). The samples
were then loaded into the appropriate tubes, capped with a septum, brought
out of the box, and immediately frozen in liquid N 2. All tubes were stored in
the liquid N 2 refrigerator before being sent out on dry ice to Northwestern
University for ENDOR spectral studies.
Addition of Protein B. Protein B was added to the hydroxylase either
before the addition of reductant, or after equilibration of the hydroxylase with
sodium dithionite.
protein
Samples were made that had from 2 to 4 equivalents of
B (3.8 to 7.6 mg, 2.2 x 10- 7 to 4.4 x 10-7 mole) relative
to hydroxylase
(50
mg, 2.0 x 10- 7 mole).
Addition of Inhibitors/Substrates. Dimethylsulfoxide, DMSO, was
added before the sample solution was made anaerobic. A concentration range
of 50 mM to 0.2 M of DMSO was studied to monitor ENDOR spectral changes
for native Hm, to fully bound Hmv + DMSO. Treatment of the protein with
isotopically enriched DMSO (d6-DMSO, 98 %, Cambridge isotope Labs;
13C-
DMSO, 98 %, Isotech) was carried out in an identical manner.
Samples that contained CD4 or CH 4 were prepared as follows. The
appropriate gas was syringed through the septum of the vial containing the
protein solution in place of argon during the last two cycles of evacuation and
back-filling.
The reaction mixtures were allowed to equilibrate for 5 min
before sodium dithionite was added to initiate the reduction of the protein.
After this point, the procedures for incubating the sample and loading and
freezing the quartz ENDOR tubes were identical to the ones outlined above.
105
ENDOR Spectroscopy. Continuous wave (CW) ENDOR spectra were
recorded by Dr. DeRose at Q band (35 GHz microwave frequency) on a locally
designed instrument.
Through evacuation of the liquid helium cryostat
headspace, temperatures were maintained at 2 K. The dispersion mode EPR
signal was detected with 100 kHz magnetic field modulation.
Results and Discussion
EPR Spectrum. The dispersion-detected rapid-passage EPR signal of
Hmv is given in Figure 4. The spectrum was simulated with the g values of gi
= 1.94, g2 = 1.87, and g3 = 1.74, and displays
some heterogeneity
around
g2. At
Q band, g strain is enhanced which prevents the transition at g3 from being
resolved.
In a previous EPR report of Hmv from M. trichosporium OB3b, a
signal with g values of 1.98, 1.87, and 1.70 corresponding
signal was detected. 5
to = 25 % of the total
This component, thought to be the result of
contamination from protein B, was not observed in our studies.
The EPR spectrum
simulation.
of Hmv + DMSO is given in Figure 5 along with a
The EPR signal of Hmv in the presence of DMSO is sharpened,
which agrees with a previous report of Hmv from M. trichosporium OB3b.5 ' 14
Some heterogeneity was observed in the sample, however. The simulation
in Figure 5 was created from two species, corresponding to 75 % and 25 % of
the sample signal. The major species displays g values of 1.95, 1.86, and the
minor component exhibits g vaues of 1.93, 1.84, and 1.82.
1H
ENDOR Spectra. Three classes of proton resonances are observed
for Hmv, as illustrated in Figure 6, depending on the strength of their
hyperfine couplings.
The spectrum in Figure 7A partially resolves at least
seven weakly coupled doublets with A < 4 mHz (see Table 1). Most of the
resonances persist in a D 20 exchanged sample, shown in Figure 7B, and are
isotropic, meaning that A varies little at different g values.
This class
106
represents protons that are probably two or more bond lengths away from the
diiron center.
The second class exhibits intermediate
strength coupling
constants of A = 7.8 mHz which are highly isotropic. This group of protons is
most likely solvent-derived since the intensity of the resonances decreases
dramatically after exchange with D20, shown in Figure 6. These properties
resemble the previously observed behavior of H 20 and OH- bound to the
[4Fe-4S] cluster of aconitase.15 Accordingly, these protons were assigned to a
terminal H 20 or OH- ligand on one of the Fe atoms.16
The third class of protons are also exchangeable with D 20, but have
strong, anisotropic coupling constants. At gl, a doublet with A = 14 MHz is
visible, as denoted by the brace in Figure 6. Increasing the magnetic field
causes the signal to split into three doublets, marked by the bracket in Figure
6. Maximal coupling is detected at g2 with A
30 MHz. These features were
simulated as arising from a single proton, the resonances of which display
hyperfine anisotropy at field values away from the edges of the EPR envelope.
'The unusual properties of this proton class do not match behavior expected
for an exchangeable proton from a terminal ligand,16 such as histidine, nor do
they resemble characteristics of a group hydrogen-bonded to an Fe ligand.17
1H
ENDOR of semimet azidohemerythrin
resonance.
16
contains a similarly behaved
It is widely accpeted that Hr contains a OH- bridge in the
semimet oxidation state.1 8 -21
We therefore assigned the third set of
resonances in Hmv as arising from the hydrogen atom of a bridging OHligand.
Hmv samples from M. trichopsorium OB3b displayed very similar
resonances in the 1 H ENDOR, and the same assignment of the coordination
environment of the Fe atoms was made.1 6 Subsequent ESEEM studies with
M. trichopsorium OB3b confirmed this work.2 2
Analyses from the X-ray
107
crystal structure,2 3 and both EXAFS and electronic spectroscopy2 of Hox from
M. capsulatus (Bath) are consistent with the 1 H ENDOR assignments for Hmv.
Figure 8 shows an ENDOR spectrum of Hmv in the presence of DMSO.
Some differences from the Hmv signals shown in Figure 5 were detected. The
resonances overall are sharper, and the A values of some were shifted, as
reported in Table 1. The signals of the terminal H 2 0 ligand and the OHbridge persist in the presence of DMSO, as shown in Figure 8, indicating that
these ligands remain bound.
14N
and
1 5N
ENDOR
Spectra.
ENDOR spectra of Hmv at g = 1.94. The
Figure 9 displays the
1 5N
since they lack quadrupolar splitting by the
resonances
14 N
14 N
and
15N
are easier to assign
nucleus. Accordingly,
15 N
data were analyzed first, which facilitated assignment of the more complex
14N
patterns in the
spectra.
Table 2 lists all of the hyperfine coupling
constants extracted from the spectra.
Four peaks at = 5.5, 9.0, 14.0, and 18.1 MHz are detected in the
ZENDOR spectrum
in Figure 9B. The resonances
as illustrated
MHz are attributed to remote
15 N
and
13C
1 5N
at 5.5 and 14
nuclei that are coupled to the
electron spin by the "distant ENDOR effect" and resonate at their Larmor
frequency. 2 4
25
The signal at 18.1 MHz corresponds to the u+ (+ = A/2 + vn)
peak of a hyperfine-centered
doublet with A = 25 MHz (A/2 = 12.5 MHz,
n=
5.6 MHz), and the corresponding nitrogen atom is referred to as N1. The
more weakly coupled nitrogen (N2) at 9.0 MHz iss assigned as the u+
resonance of a Larmor-centered doublet with and A = 6.8 MHz. The ESEEM
signals from both the coordinated and non-coordinated nitrogen atoms from
a histine ligand were previously reported.2 6 In accord with that work and
other structural data, N1 and N2 are interpreted as histidine nitrogen atoms,
one coordinated to each Fe atom.
The hyperfine coupling of a nitrogen atom
108
to the Fe(III) site of Hmv is predicted to have twice the magnitude of the
coupling to a Fe(II) site based on spin exchange interactions.2 7 N1 and N2 are
accordingly assigned to coordinate the Fe(III) and Fe(II) sites, respectively.
Figure 10OB
illustrates the spectrum of an
15N
Hmv sample that was
treated with DMSO. Unlike the spectrum of the untreated sample, five
resonances are distinguished.
attributed to
15 N
and
13C
The signals present at 5.5 and 14 MHz are
nuclei, respectively. The resonance at 9.4 MHz is the
u+ peak of N2, with A = 7.8 HHz. The two remaining signals, at 15.5 and 17.4
MHz, are both assinged to N1. The peak at 17.4 MHz is the u)+transition of a
hyperfine-centered doublet with A = 23.8 MHz, and the feature at 15.5 is a
Larmor-centered doublet with A = 20.0 MHz (A/2 = 10.0 MHz). These two
new signals, identified as Nla and Nib, correspond to nitrogen coordinated to
the Fe(III) site in Hmv. The two hyperfine coupling values probably arise from
two distinct species, in accord with the sample heterogeneity seen in the EPR
signal. DMSO appears to shift the hyperfine coupling of N2 only slightly,
from 6.8 to 7.8 MHz, whereas it signficantly alters the coupling of N1
coordinated to the diiron unit. This behavior differs dramatically from that
derived from a previous ENDOR spectroscopic
study of Hmvxfrom M.
trichosporium (O)B3bin which addition of DMSO left N1 unchanged and N2
was reported to dissociate from the iron center.5
The
14N
ENDOR spectrum of Hmv is given in Figure 9A, and was
and
15 N
resonances are related by their nuclear moments according to un( 15 N)/,n(
14 N)
assigned with the aid of the
15 N
data interpretations.
The
14 N
= A( 15 N)/A(14N) = 1.4.6 Values of 5.6 MHz for un(1 5 N), 25 MHz for A(15 N1)
and 7 MHz for A(1 5 N2) correspond to un( 14 N) = 4 MHz, A(1 4 N1) = 18 MHz and
A(1 4 N2) = 5 MHz. As mentioned earlier, the nuclear quadrupole interaction
(3P) of
14 N
further splits the spectrum and leads to u+ = A/2 + un ± 3P/2. The
109
resonances at 12 and 14 MHz thus arise from N1 with 3P = 2 MHz. The N2
ligand displays signals at 7 and 9 Mhz, with 3P = 1.9 MHz. In the presence of
DMSO (Figure 10A) the u+ peak of Nla is predicted to have A = 16.9 MHz.
The resonances at 12.1 and 13 MHz were therefore assigned to Nla with
3P(Nla) = 0.9 MHz. Similarly, the u+ transition of Nib is expected to occur at
11.1 + 3P/2 MHz. Setting 3P = 1.9 MHz allowed the resonances at 10.2 and 12.1
MHz to be attributed to N1b. For N2, A(15N) = 5.5 MHz was observed, and
A(1 4 N) = 6.7 is calculated. Transitions for o+ are then predicted to be at 6.7 ±
3P/2 for the
14 N
spectra. Two possible sets of resonances in Figure 10A satisfy
these criteria. A value of 3P = 3.7 MHz predicts that the set at 5 and 8.6 MHz
correspond to N2. Alternative assignments for N2 are the resonances at 6.8
and 5.5 MHz, with 3P = 1.3 MHz.
57 Fe
ENDOR Spectra.
5 7 Fe
ENDOR spectra of Hmv with and without
added DMSO are given in Figures 11 and 12, respectively. At gl, two sets of
Larmor-centered resonances are observed. Hyperfine coupling values for Fel,
A = 60 MHz (Figure 11A) and Fe2, A= 38 MHz (Figure 11A), were determined.
Fel displayed isotropic hyperfine coupling across the EPR envelope, whereas
Fe2 exhibited highly anisotropic behavior.
Analysis of the spin exchange
coupling between the Fe(III) and Fe(II) spins of the diiron center leads to the
prediction
that: Fel is the Fe(III) center and Fe2 the Fe(II) center. 2 7 Addition of
DMSO did not significantly perturb the hyperfine parameters of either Fel or
Fe2, as seen in Figures 11B and 12B. Analysis of the
5 7 Fe
data is thus
consistent with that reported for Mbssbauer and ENDOR studies of
from M. trichosporium
57 Fe
Hmv
OB3b.1 4
Samples with Isotopically-Labeled DMSO. ENDOR spectra of samples
of Hmv and DMSO labelled at the methyl groups, either d6- or
1 3 C-DMSO, are
illustrated in Figure 13. Both samples display Larmor-centered peaks, but
110
neither spectrum is sufficiently resolved for easy interpretation because the
coupling is very weak.
When nuclei are weakly coupled, Mims pulsed
ENDOR can be used instead of continuous wave ENDOR to resolve the
resonances.
24
Figure 14 contains such spectra of the labelled samples. The
resonances are Larmor-centered with A
-
1.2 MHz for
13C
and A
-
0.8 MHz
for 2 H.
Simulation of electron spin echo envelope modulation (ESEEM)
spectra can be used to quantitate the number of interacting deuterons. 2 4 The
d 6 -DMSO Hmv, sample displays modulation in its 3-pulse ESEEM spectrum
from the substrate molecule, as shown in Figure 15A. The spectra in Figures
15B and 15C were simulated with contributions from three and six deuterons,
respectively. The amplitude pattern in Figure 15C does not match that of the
experimental spectrum.
The Fourier transforms of spectra corresponding to
the ESEEM spectra in Figure 15A - C are given in Figures 15D - F. Spectra in
Figures 15E and 15F illustrate that amplitudes in the second and third
harmonics of the deuteron frequency increase with contributions from a
greater number of deuterium nuclei. Since the simulation in which three
deuterons contribute exhibits properties most similar to the sample spectrum,
it was concluded that only one methyl group of DMSO is coupled to the
diiron center. This result indicates that DMSO is bound in an asymmetric
manner with respect to the Fe atoms. This behavior is consistent with the
observation from
15 N
ENDOR spectrum that DMSO signficantly affects the
coordination environment of the Fe(III) atom.
The methyl group is most
likely interacting with the Fe(III) atom of Hmv.
Other Samples. The H 20 and OH- ligands are both predicted to be
exchangeable, but no signals have yet been observed with H 21 70 exchanged
samples. It is possible that exchange is extremely slow, and the conditions
111
under which it was performed were not stringent enough. A higher level of
enrichment of H 2 1 7 0 (> 20 %) may also facilitate the observation of an 170
resonance. Similarly, no new signals were observed with the substrates CD4
or CH 4 added to Hmv. The EPR spectrum of Hmv is altered in the presence of
protein B, and ENDOR analysis of these samples is in progress.
methods to study Hred have been developed using the
14 N
and
15 N
ESEEM
samples
described in this report. 28
Conclusions
ENDOR spectroscopy has provided valuable structural information
about Hmv. From 1 H couplings, a terminal H 20 ligand and a bridging OHgroup were identified. Hyperfine coupling constants of the nitrogen atoms
from the imidazole ligands and the Fe atoms have been determined.
The
effects of the inhibitor DMSO on the characteristics of the ligands bound to
the diiron center have been studied, and information about how DMSO binds
has been obtained.
These studies provide a method to investigate the
interactions with other substrates to map the binding site of Hmv. This
information will be useful in correlating structural studies of the enzyme
with mechanistic information about the hydroxylation reaction.
112
References
(1)
Hoffman, B. M. Acc. Chem. Res. 1991, 24, 164-170.
(2)
DeWitt, J. G.; Bentsen, J. G.; Rosenzweig, A. C.; Hedman, B.; Green, J.;
Pilkington, S.; Papaefthymiou,
G. C.; Dalton, H.; Hodgson, K. O.; Lippard, S. J.
J. Am. Chem. Soc. 1991, 113, 9219-9235.
(3)
Fox, B. G.; Froland, W. A.; Dege, J. E.; LIpscomb, J. D. J. Biol. Chem. 1989,
264, 10023-10033.
(4)
Woodland,
M. P.; Patil, D. S.; Cammack,
R.; Dalton,
H. Biochim.
Biophys. Acta 1986, 873, 237-242.
(5)
Hendrich,
M. P.; Fox, B. G.; Andersson,
K. K.; Debrunner,
P. G.;
Lipscomb, J. D. J. Biol. Chem. 1992, 267, 261-269.
(6)
Kevan, L.; Kispert, L. D. Electron Spin Double Resonance Spectroscopy;
John Wiley and Sons: New York, 1976, p 1-57.
(7)
Swartz,
H. M.; Bolton, J. R.; Borg, D. C. Biological Applications
of
Electron Spin Resonance;John Wiley and Sons: New York, 1972, p 1-118.
(8)
Liu, K. E.; Johnson, C. C.; Newcomb, M.; Lippard, S. J. J. Am. Chem. Soc.
1993, 115, 939-947.
(9)
Fox, B.
.; Froland, W. A.; Jollie, D. R.; Lipscomb, J. D. In Methods In
Enzymology, Academic Press: New York, 1990; Vol. 188; pp 191-202.
(10)
Stirling, D. I.; Dalton, H. Eur. J. Biochemistry, 1979, 96, 205-212.
(11)
Dalton, H.; Higgins, I. J. Antonie van Leewenhoek 1987, 53, 23-28.
(12)
Andersson,
K. K.; Froland, W. A.; Lee, S.-K.; Lipscomb, J. D. New J.
Chem. 1991, 15, 411-415.
(13)
Pilkington, S. J.; Dalton, H. In Methods In Enzymology, Academic
Press: New York, 1990; Vol. 188; pp 181-190.
(14)
Fox, B. G(.; Hendrich, M. P.; Surerus, K. K.; Andersson,
K. K.; Froland,
XV.A.; Lipscomb, J. D.; Miinck, E. J. Am. Chem. Soc. 1993, 115, 3688-3701.
113
(15)
Werst,
M. W.; Kennedy,
M.-C.; Beinert, H.; Hoffman,
B. M.
Biochemistry, 1990, 29, 10533-10539.
(16)
DeRose, V. J.; Liu, K. E.; Kurtz, D. M.; Hoffman, B. M.; Lippard, S. J. J.
Am. Chem. Soc. 1993, 115, 6440-6441.
(17)
Babcock, G. T.; El-Deeb, M. K.; Sandusky,
P. O.; Whittaker,
M. M.;
Whittaker, J. W. J. Am. Chem. Soc. 1992, 114, 3727-3734.
(18)
Maroney, M. J.; Kurtz, D. M., Jr.; Nocek, J. M.; Pearce, L. L.; Que, L., Jr. J.
Am. Chem. Soc. 1986, 108, 6871-6879.
(19)
Pearce, L. L.; Kurtz, D. M., Jr.; Xia, Y.-M.; Debrunner, P. G. J. Am. Chem.
Soc. 1987, 109, 7286-7293.
(20)
Scarrow, R. C.; Maroney, M. J.; Palmer, S. M.; Que, L., Jr.; Roe, A. L.;
Salowe, S. P.; Stubbe, J. J. Am. Chem. Soc. 1987, 109, 7857-7864.
(21)
McCormick, J. M.; Reem, R. C.; Solomon, E. I. J. Am. Chem. Soc. 1991,
113, 9066-9079.
(22)
Thomann,
K. K.; Lipscomb,
(23)
H.; Bernardo, M.; McCormick, J. M.; Pulver, S.; Andersson,
J. D.; Solomon, E. I. 1993, 115, 8881-8882.
Rosenzweig, A. C.; Frederick, C. A.; Lippard, S. J.; Nordlund,
P. Nature
1993, 366, 537-543.
(24)
Biological Magnetic Resonance, Volume 13:
Molecules;
Hoffman,
B. M.; DeRose,
V. J.; Doan,
EMR of Paramagnetic
P. E.; Gurbiel,
R. J.;
Houseman, A. L. P.; Telser, J., Ed.; Plenum Press: New York, 1993, p 151-218.
(25)
Houseman,
L. L. P.; Oh, B.; Kennedy,
M. C.; Fan, C.; Werst, M. M.;
Beinert, H.; Markley, J. L.; Hoffman, B. M. Biochemistry, 1992, 31, 2073.
(26)
Bender, C. J.; Rosenzweig, A. C.; Lippard, S. J.; Peisach, J. J. Biol. Chem.
1994, 269, 15993-15998.
(27)
DeRose, V. J.; Liu, K. E.; Lippard, S. J.; Hoffman, B. M., manuscript
preparation.
in
114
(28)
Hoffman,
B. M.; Sturgeon,
B. E.; Doan,
P. E.; DeRose,
Lippard, S. J. J. Am. Chem. Soc. 1994, 116, 6023-6024.
V. J.; Liu, K. E.;
115
Table 1. 1H Hyperfine Coupling Constants for Hmv and Hmv with DMSO.
Type of Proton
Aobs(MHz)
Aobs (MHz) at gl with DMSO
weakly coupled
0.2
0.2
0.4
0.8
0.9
1.4
1.2
1.6
1.8
3.0
2.3
3.4
2.5
3.4
intermediately
coupled
7.4
6.0
7.0
strongly
Table 2.
coupled
14 N
and
15 N
14, 30, 24 (gl, g2, g3)
14, 30, 24 (gl, g2, g3)
Coupling Constants of Hmv With and Without DMSO.
1 4N
Values for
are predicted from 15N data.
Nitrogen Atom
Aobs (MHz) at gl
1 5 N1
25
Aobs (MHz) at gl with DMSO
23.8 (NIa)
20.0 (Nlb)
1 5 N2
7
7.8
1 4 N1
17.9
16.9 (Nla)
14.3 (Nlb)
1 4 N2
5
5.6
116
Figure 1. Energy level diagram of a free electron in an applied magnetic field.
Energy is the verical axis and magnetic field is the horizontal axis.
117
ms = + 1/2
-
(1)
Ii?
h =g
-
Magnetic Field
H
ms = - 1/2
(2)
118
Figure 2. Energy level diagram of a paramagnetic center further split by the
influence of a nuclear spin. (A) represents the case where un > A 1/2 and (B)
shows the splitting for
n <
A 1/2. The vertical axis is energy and the
horizontal axis is magnetic field.
119
(4)
1/2 e + 1/2 un + 1/4 A
+ A/2
1/2
e-
1/2 un - 1/4 A
-1/2 oe + 1/2 un - 1/4 A
- A/2
-1/2 Be- 1/2 iin + 1/4 A
(4 )
-
1D
- I
1/·,
12+
.
e
IA A
./A
. 1
-.+ I/Z
n
n
1/2 'e
-
1/4 A - 1/2 'n
-1/2 ue + 1/4 A - 1/2 on
2- n
-1/2 e -1/4 A + 1/2 n
m I = - 1/2
120
Figure 3. ENDOR spectra for (A)
n > A 1/2 and (B) On< A /2.
121
A.
IAI
1n
MHz
B.
2u n
IA 1/2
MHz
122
Figure 4. Rapid-passage
dispersion-detected
35 GHz EPR spectrum of Hmv is
shown in the upper trace, and the calculated derivative is given in the lower
trace. Data collection parameters were 35.295 GHz microwave frequency, 2 K,
100 kHz modulation amplitude, 12 dB power, 64 ms time constant, 50 gain, 4
min sweep time. The specrum was simulated (dotted line) with gl,g2,g3 =
1.935, 1.865, 1.750.
123
...
.......
........ ........ ........ ........ ........ ........
_
__
13000
__
I
14000
Mognetic Field [G]
__
I
15000
124
Figure 5. The EPR spectrum of Hmv in the presence of DMSO is shown in the
upper trace, and the calculated derivative is given in the lower trace. The
simulated spectrum (dotted line) was generated from 25 % species (1 gl,g2,g3 =
1.95, 1.87, 1.75) and 75 %/ species 2 (gl,g2,g3 = 1.94, 1.87, 1.81).
125
1.86
1
1.94
12500
136000
13500
Madnetic FlVd Q]
14000
14500
126
Figure 6.
1H
ENDOR spectra of Hmv at gl (1.94), g2 (1.87), and g3 (1.74). The
three classes of protons are denoted by the dotted arrow (weakly coupled), the
double-headed arrow (intermediately coupled), and the single arrows
(strongly coupled).
127
1
1.94
'I,
1,
I,
0
1.87
0
4.
1.74
_ __
-20
_ _ _ _
-10
10
0
Frequency
(MHz)
20
128
Figure 7. 1 H ENDOR spectra of Hmv at gl = 1.94. Trace B is from a sample that
was exchanged into D2 0.
Experimental conditions were altered from
conditions described for Figure 5 to obtain better resolution of the weakly
coupled protons. Experimental parameters: 35.1 GHz microwave frequency; 2
K; 12 dB power;
average of 470 (A) and 176 (B) scans.
129
-10
-6
2
-2
V-VH
H
6
A
HO
B
DO
2
2
10
130
Figure 8.
1H
ENDOR spectrum of Hmv in the presence of DMSO at g2 = 1.87.
The three classes of protons are denoted by the dotted arrow (weakly coupled),
the double-headed arrow (intermediately coupled), and the single arrows
(strongly coupled).
131
I
I
I
_
-20
-15
·
·
·
-10
6
5
-5
·
10
v - v(H)(MHz)
15
1
20
25
132
Figure 9.
14 N
(A) and
15 N
(B) ENDOR spectra of Hmv. N1 and N2 transitions
are indicated by vertical lines. The u+ resonances are relatively intense (solid
markings) and the u- transitions (dotted lines) are extremely weak and often
not detected. The hyperfine-centered doublets of N1 are centered at A/2 as
denoted by filled triangles and are separated by 2un. Larmor-centered doublets
of N2 are centered at n (circles) and separated by A. Inverted triangles
indicate
1 3C
and
15 N
transitions at their respective Larmor frequencies due to
the distant ENDOR effect. Conditions: 0.3 G modulation amplitude; 12 dB
power; (A) 35.02 GHz average of 320 scans; (B) 35.09 GHz, average of 190 scans.
133
V
rN
T -
l
1
-i
m
-s
NI
A
N2
_,
_
15
V
I
2
4
6
e
--
10
12
14
Froeqwncy [MHz]
16
I'a
2'0
2:2
N
134
Figure 10.
DMSO.
14 N
(A) and
15 N
(B) ENDOR spectra of Hmv in the presence of
N1 and N2 transitions are indicated by vertical lines.
The v+
resonances are more intense (solid markings) and the
- transitions (dotted
lines) are extremely weak.
doublets of N1 are
The hyperfine-centered
centered at A/2 as denoted by filled triangles and are separated by 2n.
Larmor-centered doublets of N2 are centered at Un(circles) and separated by A.
Inverted triangles indicate
13C
and
15 N
transitions at their respective Larmor
frequencies due to the distant ENDOR effect. Conditions: 0.3 G modulation
amplitude;
12 dB power; (A) 35.10 GHz average of 260 scans; (B) 35.90 GHz,
average of 50 scans.
135
V,
A.
14
N
N2....... .
N1
,!
A14
KA..-t
i
I
NiaL
I1
.-
-I
t
,
i
,
,
I
v
v
B. 15N
.2
4
a
a
lo0
12
Frequency
*14
Hz)
16
I
20
136
Figure 11.
5 7Fe
ENDOR spectra of the Fe(III) site at various g values for (A)
Hmv and (B) Hmv and DMSO.2 3 The doublet is centered around A/2 and split
by
n (< 1 MHz).
Simulations are shown by dotted lines.
Spectrometer
conditions: 35.2 GHz, 0.3 G modulation amplitude, average of 50-200 sweeps.
137
B
A
1.75
...........
...
....
______
..
-
.....
?/
28
2.
/ .
32
.
36
Frequency
.
.
.
40
(MHz)
1.81
1.85
1.87
1.90
1. 93
44
1.80
2r
2
...................
3ue2 36 ( 40
Frequency (Iz)
1.95
138
Figure 12.
57 Fe
ENDOR spectra of the Fe(II) site at various g values for (A)
Hmv and (B) Hmv and DMSO as indicated by arrows. Spectra from unlabeled
samples are denoted by dotted lines. Spectrometer conditions: 35.2 GHz, 0.3 G
modulation amplitude, average of 50-200 sweeps.
139
57
Fe
A
1
1.87
1.94
_
1'2
8
16
20
24
Frequency (MHz)
I
B
1.78
1.82
,
I-
1.92
I
.95
4
-
I
-
1
1
12
16
Frequency
(MHz)
2
2'0
24
140
Figure 13. ENDOR spectra of Hmv at gl = 1.94 in the presence of (A) unlabeled
DMSO, (B) d6 DMSO, and (C)
13 C
DMSO. Larmor frequencies are as indicated,
and the background resonances are due to
14 N.
Experimental conditions: 35.1
GHz (A) and (C), 35.0 GHz (B); 300 scans (A), 135 scans (B), and 200 scans (C).
141
A
(CHS)2SO
v (D)
B
(C*D33 )2SO
v'v((13C)
C)
C
I
I
6
7
I
8
9
I
I
I
I
I
I
i
10 11 12 13 14 15 16 17.18
Frequency MHz)
-r
!
19
(*CH3)
3 2SO
142
Figure 14.
Mims pulsed ENDOR2 7 spectra of Hmv in the presence of
isotopically labeled DMSO at g = 1.85. The spectra are centered at the
respective Larmor frequences of
maximum hyperfine
GHz,
13 C
and 2H. Arrows indicate 1.2 MHz as the
coupling for 1 3C and 0.8 MHz for 2 H. Conditions:
9.434
= 400 ns, ic/2 tp = 16 ns, tp(rf) = 60 s, sequence repetition rate 50 ms,
average of 760 transients (13C) and 500 transients (2H).
143
1
C
-1.4
-1
-. 6
-. 2
.2
.6
1
1.4
Frequency (MHz)
y
2
H
1
-1
-I
-16
' -.-22 ' ..2
Frequency (MHz)
,.66
i
I
1.4
.4
144
Figure 15. 2 H ESEEM spectra of Hmv in the presence of DMSO at g = 1.85. The
time domain data and simulations are given in (A) through (C). The Fourier
transformed frequency domain (D) spectrum and simulations are listed in (D)
through (F). Conditions: 9.434 GHz, 3-pulse ESEEM sequence with z = 192 ns,
7r/2 tp = 16 ns, time T between first and second pulses stepped in 32 ns
increments (256 total points), 50 ms repetition rate, 140 transients averaged.
More details of these data can be found elsewhere.2 3
145
C-l----- - --5
- -- --uL -- Exerment
----------
A
B
v
--
-
Simulation - 3 2H
-A
---L-
C
Simulation - 6 2H
__
I
I
·
20'00
-
6000
T+
.
10000
(ns)
D
Experiment
E
Simulation
-
3 2H
F
Simulation - 6 2H
;L
2
r
1
5
Frequency (MHz)
10
146
CHAPTER 4.
Radical Clock Substrate Probes and Kinetic Isotope Effect
Studies of the Hydroxylation of Hydrocarbons by Methane Monooxygenase
147
Introduction
Methane monooxygenase (MMO) is a protein system employed by
methanotrophic bacteria to oxidize methane to methanol, a process that provides
these organisms with their sole source of carbon and energy (eq. 1).1 The soluble
CH4 + 02 + NADH + H+
-
CH30H + H 2 0 + NAD+
(1)
MMOs from Methylococcus capsulatus (Bath)2 and Methylosinus trichosporium
OB3b,3 two of the better studied systems, comprise three proteins. Included are
a hydroxylase enzyme that binds substrate and dioxygen, a reductase that
contains Fe2S2 and FAD moieties and accepts electrons from NADH, and a
coupling protein that regulates electron transfer between the reductase and
hydroxylase
potentials
7
4
and also modulates the substrate specificity5 6 and redox
of the latter. Recently, much attention has been focused on the
hydroxylase enzyme which, in addition to methane, is capable of hydroxylating a
wide variety of substrates. 8 ' 9 X-ray crystallographic studies1 0 of the diferric
hydroxylase and ENDOR spectroscopic results1 l for the the mixed-valent state
reavealed that the active site contains a non-heme dinuclear iron center in which
the metals are each coordinated to a histidine residue. One iron atom has two
terminal monodentate glutamate ligands, and the other has one terminal
monodentate glutamate and a water molecule in its coordination sphere. In the
x-ray structure, the iron atoms are bridged by a hydroxide, a bidentate
glutamate, and an exogenous acetate ligand.
The catalytic mechanism of the MMO hydroxylase has frequently been
compared to that of its widely studied heme analogue, cytochrome P-450.8' 12 -14
In P-450 catalyzed hydroxylations, it is postulated that a high valent iron-oxo, or
ferryl, moiety is generated which abstracts a hydrogen atom from the substrate to
148
form a hydrocarbon radical that recombines with the coordinated hydroxyl
radical before dissociating from the active site.15- 18 Especially important in
establishing the presence of radical intermediates in this mechanism were studies
with substrate probes that, following hydrogen atom abstraction to form a
radical, isomerize to form, ultimately, a rearranged alcohol product.19 -21 This
(1),
process is illustrated for a specific probe, trans-2-phenylmethylcyclopropane
in Figure 1. Attempts to employ this methodology in studies of the
hydroxylation of 1,1-dimethylcyclopropane with the MMO hydroxylase from M.
trichosporium OB3b were consistent with radical formation, although cationic
intermediates
were also invoked. 1 2
Other experimental evidence for the
involvement of radical intermediates in the hydroxylation chemistry of MMO
from the same organism has been reported. 8' 13 22
In the present article we describe our investigations of the MMO hydroxylase from both organisms in which five substrate probes, 1- 5 illustrated in
Figure 2, have been used to address the possibility of radical intermediates and
other aspects of the hydroxylation mechanism. Specifically, reactions with trans1,2-dimethylcyclopropane (3) and bicyclo[2.1.0]pentane (4), both of which were
used to study the rate constants for rebound with cytochrome P-450,19
20
and
*withthree additional radical clock substrates, trans-2-phenylmethylcyclopropane
(1) and 2,2-diphenylmethycyclopropane (2), and methylcubane (5), were carried
out to determine whether substrate radicals are produced during hydroxylation.
If, for example, hydrogen atom abstraction from probe 1 at the methyl position
were to occur, the resulting cyclopropylcarbinyl radical (1U., Figure 1) would
either
be trapped
before
ring opening
to give
trans-(2-
phenylcyclopropyl)methanol (la) or would ring open to the 1-phenylbut-3-enyl
radical (1R). Subsequent trapping of radical 1R would afford 1-phenylbut-3-en1-ol (lb). Substrate 1 has been employed previously as a hypersensitive radical
149
probe in oxidations by resting cells and crude enzyme preparations of
Pseudomonas oleovorans monooxygenase. 2 3 In that work, the only oxidation
product detected was the ring-opened alcohol, leading to the conclusion that
radical intermediates had been formed and that the oxygen rebound step had a
rate constant of < 4 x 109 s-1 .
Experimental Section
Isolation and Purification of Proteins. M. capsulatus(Bath) was grown as
reported previously.24
25
The hydroxylase protein was purified and assayed fol-
lowing the described procedure, 2 5 except that Q Sepharose was substituted for
DEAE Sepharose in the initial ion exchange chromatography step. Coupling
protein B was prepared from an Escherichia coli strain harboring a plasmid
containing this gene from M. capsulatus(Bath),26 and then purified according to
published procedures. 2 7 The reductase was purified as described2 7 except that
an HPLC TSKG3000SWGcolumn was employed in the final purification step.
Specific activities with propylene were measured at 200 to 250, 8000 to 8500, and
4000 to 4200 mU/mg for the hydroxylase, coupling protein, and reductase,
respectively.
Growth of M. trichosporiumOB3b and purification of its hydroxylase were
carried out as reported. 2 8 Its specific activity for conversion of propylene to
propylene oxide was monitored2 5 with the use of coupling protein and reductase
from M. capsulatus (Bath), for which cross-reactivity with the M. trichosporium
C)B3bhydroxylase has been previously reported.29
30
The measured activity was
180 mU/mg of protein, a value considerably less than expected,2 8 possibly due to
the use of coupling protein and reductase from a different organism. Reactions
of 15 iiM solutions of the M. trichosporiumOB3b hydroxylase with 10 mM hydrogen peroxide produced 0.19 mM propylene oxide in a 3 min reaction, however,
which compares favorably with literature reports that 44 gM hydroxylase and 10
150
mM H 2 0 2 yielded approximately 0.4 mM propylene oxide in the same time period.1 4 The measured iron content, 2.2 mol Fe/mol hydroxylase, was less than
found by others, 4.3 mol/mol protein,2 8 for which we have no obvious explanation.
Synthesis and Characterization of Substrates and Products. The substrate trans-2-phenylmethylcyclopropane
(1) was prepared as previously re-
ported 3 1 by reducing commercially available trans-2-phenylcyclopropanecarboxylic acid or its ethyl ester to alcohol la, trans-(2-phenylcyclopropyl)methanol,
with LiAlH4 , converting the alcohol to the mesylate, and reducing the mesylate
with LiEt3BH. As an alternative, conversion of alcohol la to the corresponding
chloride by treatment3 2 with Ph3P and CC14followed by reduction with LiAlH4
gave hydrocarbon 1 and ring-opened product 4-phenylbut-l-ene; treatment of
the mixture with m-chloroperoxybenzoic acid to epoxidize the alkene byproduct, followed by silica gel chromatography, gave pure 1. Deuterated
analogues of 1 were prepared by the mesylate route with LiAlD4 as the source of
deuterium.
Their NMR and mass spectra were consistent with the expected
amount of deuterium incorporation for each analogue.
The substrate 2,2-diphenylmethylcyclopropane (2) was prepared by a sequence similar to that described above for 1. Reaction3 3 of ethyl diazoacetate in
the presence of CuSO4 with 1,1-diphenylethene gave ethyl 2,2-diphenylcyclo-
propanecarboxylate that was reduced with LiAlH4 to (2,2-diphenylcyclopropyl)methanol
(2a). Conversion of 2a to the mesylate followed by LiEt3 BH
reduction gave 2. Methylcubane (5) was prepared as reported, along with its
mono-, di-, and trideuterated analogs.3 4
Authentic samples of the products obtained by the MMO-catalyzed oxidation of substrates 1, 2, and 5 were prepared in the following manner. trans-(2I'henylcyclopropyl)methanol
la and its dideuterated analogue were obtained
151
from the synthesis of 1, and alcohol 2a was available in the preparation of 2, as
indicated above. Monodeuterated la was prepared by LiA1D4 reduction of trans2-(phenylcyclopropyl)methanal which, in turn, was obtained from the acid by se-
quential treatment with H3 BSMe2 and pyridinium chlorochromate. 35 1Phenylbut-3-en-l-ol (lb) and 1,1-diphenylbut-3-en-l-ol (2b) were prepared by
reactions of allylmagnesium chloride with benzaldehyde and benzophenone, respectively. Cubylmethanol (5a), along with acetylated products of reactions with
5, cubylmethanol acetate (5b), 2-methylcubanol acetate (5c) and 4-methylcubanol
acetate (5d) were prepared from cubanecarboxylic acid.34 Dideuterated 5a was
prepared as described elsewhere.3 4
An authentic sample of trans-2-(p-hydroxyphenyl)methylcyclopropane
(lc) was prepared by the following sequence. Reaction3 3 of p-vinylanisole with
ethyl diazoacetate
gave a mixture of ethyl cis- and trans-2-(p-
methoxyphenyl)cyclopropane
carboxylates which was separated by column
chromatography and crystallization. The trans isomer, identified by NMR spectroscopy, was reduced with LiAlH4, and the resulting alcohol was converted to
its mesylate which was then reduced with LiEt3BH. The resulting trans-2-(pmethoxyphenyl)methylcyclopropane was demethylated 3 6 with NaSEt in DMF to
give the desired product which was purified by preparative GC on a Carbowax
20M column. Authentic samples of the possible aromatic hydroxylated products
from 2 were not prepared; no components consistent with these products were
observed by G(Cmass spectral analysis of the oxidation mixtures.
trans-1,2-Dimethylcyclopropane (3), trans-2-methylcyclopropylmethanol
(3a), 4-penten-2-ol (3b), and 2-methylbut-3-en-l-ol (3c)were obtained from Wiley
Organics (Coschocton, Ohio).
according to a literature method.
Bicyclo[2.1.0]pentane (4) was synthesized
37
152
Enzymatic Reactions. Reactions with MMO from M. capsulatus (Bath)
were carried out at 45 °C with the reconstituted enzyme system in the following
manner. Solutions were prepared containing 1.2 mg (10 gtM)of hydroxylase, 0.2
mg (24 gM) of coupling protein, and 0.4 mg (20 gM) of reductase, all in 25 mM
MOPS (MOPS = 3-[N-morpholino]propanesulfonic
acid), pH 7.0, in a final
volume of 0.5 mL. A 0.5 gL (0.5 mg for 5) portion of substrate was then added by
means of a microsyringe to achieve a maximum concentration of - 8 mM, most of
the substrates not being completely soluble prior to conversion to products. The
reaction mixture was placed in a 10 mL glass vial and sealed with a Suba-Seal
septum.
After addition of a 25 jiL portion of a 0.1 M stock solution of ethanol-
free NADH to initiate the hydroxylation reaction, the vial was placed in a shakerincubator maintained at 45 °C. Control reactions were performed in an identical
manner except that no NADH was added. Reactions were allowed to proceed
for 15 min, a time sufficient to afford at least 30 mol product per mol enzyme in
those cases where authentic standards were available.
Studies with the hydroxylase from M. trichosporiumOB3b and substrate 1
were carried out at 30 °C in one of two ways. Coupling protein and reductase
from M. capsulatus(Bath) were incubated with the hydroxylase in the presence of
NADH. The protein and NADH concentrations were the same as described in
the preceding paragraph. Alternatively, reactions were performed by using 10
mM hydrogen peroxide and M. trichosporiumOB3b hydroxylase only, a known
shunt for this enzyme.14 Substrate was added to the reaction mixture, which was
then incubated as described above. Again, at least 30 turnovers were completed
before the reaction was terminated.
Instrumentation. Products from enzyme oxidations were analyzed with a
Hewlett Packard (HP) model 5890 gas chromatograph equipped either with an
FID detector or interfaced to an HP model 5971A mass spectrometer (EI, 70 eV).
153
For GC and GC /MS analysis of reactions with 1- 5, either a cross-linked FFAP
(HP) or a cross-linked methyl silicon (HP) capillary column was employed,
depending on which column gave better resolution of the reaction components.
The column dimensions were 10 m x 0.5 mm x 1.0 jgm for GC analysis and 25 m x
0.2 mm x 0.33 pm for GC/MS analysis. Slightly different instrumental conditions
were used to analyze the different reactions, as indicated in Table 1.
Product Identification and Quantitation. Products were isolated by
thorough extraction of the reaction mixture with diethyl ether or ethyl acetate.
The septum of the reaction vial was removed and immediately replaced after
addition of 5 mL of extracting solvent. The mixture was incubated for at least 10
min, the organic layer was then removed, and the entire procedure was repeated
twice. The organic layers were combined, dried over anhydrous K2 CO 3, and
then concentrated by slow evaporation under Ar to a volume of -100 gL. After
the addition of an appropriate internal standard (vide infra), a 0.2 to 0.5 jILsample of concentrated extract was injected into the gas chromatograph. The retention times and mass spectra of the products were compared to those of authentic
standards. GC retention times for reactants and products are listed in Table 2.
Since 5a and methylcubanols are reportedly unstable, products formed in
reactions of 5 were acetylated in the following manner to prevent decomposition.
A small amount of pyridine and acetic anhydride (2 - 3 drops each) was
combined with the concentrated extracts, and the resulting solution ( 0.5 ml)
was stirred for 30 min. After the addition of a 1 ml portion of H 2 0 the mixture
was extracted three times with 2 ml volumes of methylene chloride. The CH 2C12
extracts were combined, dried over K2 CO3, concentrated, and analyzed as
described above.
Components of the hydroxylation reaction mixture for 1-3 and 5 were
quantitated by constructing standard curves in the following manner. Solutions
154
containing known concentrations of a given component and a constant amount
of an internal standard dissolved in methylene chloride or diethyl ether were
injected into the GC. The ratio of the peak area corresponding to the component
with that of the standard was then plotted against the concentration of added
component in the sample. A linear plot was obtained and subsequently used as a
calibration curve to determine the concentration of the component in extracts
obtained from reactions with the enzyme. Plots were constructed in this manner
for the alcohols la-3a and 5a, the rearranged alcohols lb-3b and 3c, 2-(p-hydroxyphenyl)methylcyclopropane
c, and for the substrates 1-3. Since no lb was
detected in reactions of MMO from M. capsulatus(Bath) with 1 (vide infra), mass
balance determinations were made to ensure that the amount of unreacted substrate and products corresponded to the amount of substrate added to the reaction. In addition, reactions containing 0.8 mM of lb as well as the usual quantity
(8 mM) of 1 were carried out to determine if the latter would survive the
hydroxyation
conditions and/or compete with 1 for the enzyme. Control
experiments were carried out without addition of NADH, and the substrate was
recovered following the usual extraction and concentration procedures.
Isotope Effect Studies. Reactions with MMO from M. capsulatus (Bath)
were investigated with 1 and 5 and with their mono-, di-, and trideuterated
derivatives at the methyl position. A V/K intramolecular isotope effect,
measured under conditions in which substrate was not saturating, was investigated through studies of the mono- and dideuterated derivatives. The product
alcohols la and 5a from reactions with each substrate were analyzed by GC/MS.
For 1 the isotope effect was then calculated by comparison of the intensities, after
statistical corrections, at m/z 148 and 149 for alcohols arising from the
monodeuterated substrate and at m/z 149 and 150 for alcohols derived from the
dideuterated substrate. For 5, the intensities at m/z (M - 19) 116 and 115 from the
155
monodeuterated substrate and at m/z (M - 19) 117 and 116 for alcohols from the
dideuterated substrate were analyzed.
V/K intermolecular isotope effects were evaluated by comparison of the
rate of formation of la and 5a in MMO catalyzed reactions with the undeuterated
and the trideuterated derivatives. For 1, the experiment was performed in two
ways, as a competitive reaction and as separate reactions with each substrate.
Competition experiments were performed with equal amounts (0.5 gl) of both
derivatives present in the reaction vial. The mass spectral intensity at m/z 148
(la) representing reaction of the undeuterated substrate, was divided by the
intensity at m/z 150 representing reaction with the trideuterated substrate, and
the ratio was taken as the intermolecular isotope effect. In separate reactions
with the two derivatives of 1, 2-(p-hydroxyphenyl)methylcyclopropane
(c), a
product of the enzymatic reaction (vide infra), was used as an internal standard
for the purpose of quantitation. The amount of la relative to lc was calculated
for each derivative.
The ratio of alcohol to phenol for the undeuteurated
derivative was divided by the same ratio for the trideuterated derivative to
provide the intermolecular isotope effect. Only separate reactions were
performed with substrate 5, and the internal standard trans-2methylcyclopropanol was used to quantitate the amount of product formed.
Results
Table 3 reports the results of the catalytic hydroxylation of substrates 1 - 5
by MMO from M. capsulatus (Bath). For substrates 1 - 5, the ratio of experimentally observed alcohols, corresponding to hydroxylation without and with rearrangement, is indicated. In addition, the table includes the known rearrangement rates of the appropriate radical intermediates. Chromatographic retention
times and typical GC traces for components of the reaction mixtures are provided
in Table 2 and Figure 3, respectively.
156
Reactions with 1. The hydroxylation of 1 yielded two products as
revealed by both GC and GC/MS analysis. They were identified as the alcohol
la and 2-(p-hydroxyphenyl)methylcyclopropane
(lc)38 by comparison of their
retention times and mass spectra with those of authentic standards. Quantitation
of the peaks revealed that the two products are formed in approximately equal
concentrations and at equal rates over a 20 min reaction period, as illustrated in
Figure 4. To ensure that all added material was accounted for in the analysis,
mass balance studies were carried out. The amount of material recovered in
control reactions run in the absence of reductant corresponded to 84.4 + 0.6% of
the substrate initially added, the remainder apparently being lost through the
various extraction manipulations.
When NADH was present to initiate and
sustain enzyme turnover, 85.0 ± 0.5% of the material added was recovered, 5.0 ±
0.2% of which was la and 5.1 + 0.1%, c. No rearranged alcohol lb was detected.
The possibility that lb formed but was unstable under the hydroxylation
conditions was investigated through analysis of a reaction containing 8 mM of 1
and 0.8 mM of lb. In the presence of NADH, 78.6 + 1.3% of lb was recovered,
compared to 78.2 + 1.0% in control reactions without NADH.
This result
indicates that lb is unaffected during the hydroxylation reaction. The yields of
la and lc remained unchanged.
Hydroxylation studies of the mono- and dideuterated derivatives of substrate 1 gave values for kH/kD of 5.15 ± 0.38 and 5.03 + 0.42, respectively, for the
intramolecular isotope effect. The intermolecular isotope effect was determined
by comparing results for 1 with those for its trideuterated derivative. A value of
1.18 + 0.15 was obtained for reactions that were performed separately whereas
competitive reactions performed with both derivatives present in the reaction
mixture yielded a value of 0.89
0.10 (see Experimental Section). In the
competition study, the ratio of phenols (lc-do:lc-d 3) was 1.1.
157
Reactions with 2. GC analysis of hydroxylation reactions with 2 revealed
only a single product which, by comparison to an authentic standard, was
identified as the alcohol 2a. No peaks that might arise from rearranged alcohol
2b or aromatic hydroxylation products were detected. A plot showing the rate of
product formation is given in Figure 5.
Reactions with 3. Analysis of the extracts from reactions with trans-1,2dimethylcyclopropane (3) revealed only one product, the concentration of which
increased linearly with time over a 10 min period, as plotted in Figure 4. It was
identified as the unrearranged alcohol, trans-2-methylcyclopropylmethanol (3a),
through comparison of its GC retention time with that of an authentic standard.
No 4-penten-2-ol (3b) or 2-methylbut-3-en-l-ol (3c)was detected.
Reactions with 4. The substrate bicyclo[2.1.0]pentane (4) afforded 3
products upon reaction with MMO (Figure 6), each of which had its parent ion at
nz/z 84 as determined by GC/MS analysis.
No authentic standards were
available for the expected products, but one product could be positively
identified as endo-2-hydroxybicyclo[2.1.0]pentane (4a) by comparison of its mass
spectrum with one published previously. 19 The second major peak in the gas
chromatogram was assigned as exo-2-hydroxybicyclo[2.1.0]pentane (4b).
Unfortunately no mass spectral data for this compound were available for
comparison. The peak had approximately the same intensity (-85%) in the total
ion chromatogram of the reaction extracts as that of 4a. Its mass spectrum
differed from those of 4a and 3-cyclopenten-l-ol (4c), for which mass spectral
data are also available.3 9 Hydroxylation of 4 by cytochrome P-450 yielded only
4a and 4c.1 9 ' 2 0 The failure of this enzyme to afford 4b was attributed to
constraints at the active site. It may be that the active site of MMO is less discriminating, which could account for the formation of both 4a and 4b.
158
The third product from reactions with 4 was present in only trace amounts
and was assigned as 3-cyclopenten-l-ol (4c). It has been previously reported that
4a undergoes thermal rearrangement to 4c.19' 20 In particular, when 4a isolated
from the hydroxylation of 4 by MMO was injected into the GC/MS, the ratio of
4c to 4a was found to be ~1:30. Compound 4b is expected to behave similarly.
The intensity of the peak corresponding to 4c was also found to increase relative
to those for 4a and 4b when the reactions were run at 45 °C for longer times (Fig.
S3). Increasing the injector and oven temperatures of the gas chromatograph
produced a similar effect. The ratio of 4c to 4a was 1:40, or, including both 4a
and 4b, 1:80, and increased with longer incubation times at 45 °C to about 1:15. It
is therefore likely that a significant amount, if not all, of 4c obtained in these experiments is due to thermal rearrangement
of 4a and 4b. The amount of 4c pre-
sent in the enzymatic reactions is small enough that its presence could be solely
the result of thermal rearrangement of 4a and 4b.19
Reactions with 5. In reaction extracts of 5, GC/MS analysis revealed the
presence of only one product, 5a, which was stable for up to a week under the
extraction and GC/MS conditions employed in this work. GC/MS analysis of
the acetylated extracts contained only unrearranged 5b. No products indicative
of radical rearrangement (5e) or hydroxylation at another position (5c or 5d)
were observed, either in the extracts or following acetylation. The mono- and
dideuterated
derivatives
of 5 gave kH/kD ratios of 5.29 + 0.32 and 4.21 ± 0.46,
respectively, revealing a substantial intramolecular isotope effect. Separate
reactions performed with the undeuterated and trideuterated derivatives present
resulted in a value of 0.87+ 0.15. The magnitudes of both the intermolecular and
:intramolecuar isotope effects for 5 are in good agreement with those obtained for
1.
159
Experiments were also carried out with 1 and the hydroxylase from M. trichosporium OB3b. In contrast to reactions with MMO from M. capsulatus(Bath),
three products were observed. They were identified as the unrearranged alcohol
la, 2-(p-hydroxyphenyl)methylcyclopropane (c), and the rearranged product lb
(Figure 3). The ratio of unrearranged to rearranged alcohol was 32:1 in reactions
performed with coupling protein and reductase from M. capsulatus (Bath) and
22:1 for reactions using hydrogen peroxide.
Discussion
Analysis of Products Formed. In this work, five hydrocarbon substrates
were employed as mechanistic probes in oxidations catalyzed by the MMO of M.
capsulatus (Bath). In each case, if a hydrogen atom were abstracted from the
probe, the resulting radical could either rebound to form the unrearranged product or ring open to afford the rearranged product (Figure 1). In no case was rearrangement observed. For trans-1,2-dimethylcyclopropane
(3) only a single
product formed, corresponding to the unrearranged alcohol (3a, eq. 2). No allylic
alcohol was detected (3b, eq. 3), implying that no more than 1% of this product
Phi-HAt
CH3
3
CH3
CH3
rearrangeCH 3
',,A
CH
CH
ment
nt
U
.
CH
trapping
CH
CH 33
CH20H
(2)
OH
trapping
3
3a
CH
)
3
3b
formed during the hydroxylation reaction. From the known rate constant for
:rearrangement of the corresponding cyclopropylcarbinyl radical (Table 3),21and
assuming that this rate constant is the same for substrate free in solution and
bound to the enzyme (vide infra), we estimate that the rebound rate constant of a
160
radical intermediate would have to be at least 4 x 1010s-1 at 45 °C. A similar
analysis2 1 ' 4 0-4 2 of the results for reactions with compounds 4, 5, 1, and 2, raises
rate constant
the rebound
limits to 3 x 1011, 8 x 1012, 4 x 1013, and 1 x 1013 s - 1 ,
respectively. The latter two values exceed the rate constant of decomposition of a
transition state and are incompatible with a discrete radical intermediate.
The rate constants for ring opening of radicals derived from 1 and 2 (eq. 4)
have been measured indirectly by competition against PhSeH trapping.4 1 '4 2 At
room temperature, the ring openings have rate constants of 3 x 1011s-1 and
Ph ~trapping
-H.
R
2 OH
R
'CH
CH 2 '
R=H, 1
R = H, la
R = Ph, 2
R = Ph, 2a
Ph. ,A
R
R
CH,
P
rearrangeJ"CH,
Ph ...
ment
trapping
ph .,
R
OH
,
R = H, lb
(4a)
(4
'
R = Ph, 2b
5 x 1011s-1, respectively, which places these reactions among the fastest reported
radical rearrangements involving a bond breaking step. With such a velocity,
detectable amounts of ring-opened products are expected to be produced from
these radicals even when the fastest possible trapping reactions compete.
Therefore, the absence of ring-opened alcohols lb and 2b suggests that probes 1
and 2 were converted directly to alcohols la and 2a without formation of
radicals. On the other hand, a small amount of rearranged product lb was observed in the studies of MMO from M. trichosporiumOB3b. From the ratio of unrearranged to rearranged alcohol the rate constant for rebound was calculated to
be 6-10 x 1012s-1. The observed differences in the results for MMO from the two
different sources will be discussed later.
161
In the case of the chiral substrate probe 1, oxidation occurred not only at
the methyl group but also at the phenyl ring to form 2-(p-hydroxyphenyl)methylcyclopropane
(c) (Table 3).38 The fact that la and lc are produced in
equal quantities over the course of the reaction suggested that the two products
might arise from different enantiomers of the probe. Studies with partially
resolved (-90 "%o)
1 indicate that, with R,R 1, twice as much of la is produced than
lc. With S,S 1, however, c is present at twice the levels of la. The analysis of
these results is complicated by the possibility that formation of la and lc may
proceed with different rate constants. In addition, a particular rate constant may
vary with each enantiomer of the probe. It is therefore difficult in the absence of
this quantitative information to draw firm conclusions from this experiment. A
system in which only one product is produced would be better for studying the
chiral discrimination of the site.
Nevertheless, it would appear at least
qualitatively that the enzyme can discriminate between the two enantiomers to
achieve regioselective, if not regiospecific, hydroxylation.
Deuterium Isotope Effects. Studies of the reaction of deuterated derivatives of 1 and 5 with MMO from M. capsulatus (Bath) failed to reveal any
intermolecular isotope effect. Both individual and competitive reactions with the
undeuterated and trideuterated derivatives of 1 yielded values of (kH/kD)obs of
approximately 1.0 and in good agreement with one another, considering the
error limits. Similarly, the ratio in the separate experiments with undeuterated
and trideuterated 5 was 0.87. For 1, the ratio of undeuterated and trideuterated
phenols observed in the competition experiment was 1.1. When combined with
the ratio of 0.89 for the undeuterated and dideuterated alcohols la, this result
shows that the overall rates of oxidation of the undeuterated and trideuterated
substrates were equal. Thus, there was essentially no kinetic discrimination for
substrates by the enzyme, and this behavior suggests that C-H bond breaking is
162
not rate-limiting in the overall hydroxylation reaction. Such a result is consistent
with some,13' 4 3 but not all,44 previous studies of MMO hydroxylation reactions
using deuterated substrates.
Given that the overall oxidation rates of 1-doand 1-d3 were the same, the
product distributions should be considered in more detail. If oxidation of each of
the enantiomers of substrate 1 could have occurred at either the methyl group or
aromatic ring, and if rotation can occur in the active site, then a kinetic isotope
effect should have been observed in oxidations of the undeuterated versus
trideuterated probes with substantially more phenol product being formed in the
latter case. This result is predicted from the intramolecular kinetic isotope effects
that were found (vide infra). That such was not the case reinforces speculation
that products la and c arise from different enantiomers of the probe. This effect
could be masked, however, if reoriention of the substrate radical prior to
hydroxylation is hindered, since the two possible positions for hydroxylation are
very far apart and hydroxylation at the other site would most likely involve
substrate movement.
In contrast to the lack of an intermolecular isotope effect, a significant intramolecular isotope effect of 4.8 to 5.1 (statistically corrected) was observed in
oxidations of both the mono- and dideuterated probes 1 and 5. In each case, the
observed isotope effects are equal to the primary kinetic isotope effects (kH/kD)
divided by a secondary kinetic isotope effect that results from the fact that an
additional deuterium atom remains bound after oxidation of a C-H bond. The
data do not permit an extraction of the secondary kinetic isotope effect, but it is
clear that the (kH/kD)obs ratio signals a substantial C-H bond stretching
component in the transition state of the rate-limiting step in the oxidation
reaction. The value of (kH/kD)obs= 5.1 for 1 and 4.8 for 5 may be compared with
the intramolecular isotope effect of 7.8 observed in the oxidation of probe 1 with
163
P. oleovorans monooxygenase. 2 3 Recently, a value for kH/kD = 4.2 ± 0.2 was
reported for MMO from M. trichosporiumOB3b with the substrates (S) and (R)-[12 H,1-3 H]ethane,
in good agreement with the present results.4 5 Cytochrome P-
450 dependent enzymes oxidize a number of substrates with intramolecular
isotope effects of 7-14.2 0 Intramolecular isotope effects for hydrogen atom
abstractions by the reactive tert-butoxy radical studied at temperatures in the
vicinity of 25 C range from 1.2 to 5.4, and isotope effects for hydrogen
abstractions by less reactive peroxy radicals are greater.4 6
Mechanistic Considerations
Formation of the Dioxygen-Activated Diiron Center. Figure 7 outlines a
reasonable working hypothesis for the catalytic mechanism of MMO. Substrate
first binds to the complete system containing all three protein components.
Addition of NADH then effects a two-electron reduction of the hydroxylase from
the oxidized Fe(III)Fe(III) to the fully reduced Fe(II)Fe(II) form, bypassing the
inactive4 7 Fe(II)Fe(III) state.7'4 4 The fully reduced hydroxylase next reacts with
dioxygen in a two-electron step to form a diiron(III) peroxide complex which,
either itself or, upon transformation via two more electron transfer steps at the
diiron center, is sufficiently activated to attack the hydrocarbon substrate. The
nature of this transformed species, if it exists, is currently unknown, but it could
involve a high valent iron oxo moiety (Figure 8), by analogy with that proposed
for cytochrome P-450.1 5 The second iron atom in the MMO hydroxylase active
site would stabilize such a unit through redox charge delocalization, as shown in
Figure 8, in much the same fashion that cytochrome P-450 is thought to facilitate
the formation of a high valent iron oxo species through oxidation of the porphyrin ring to a r-cation radical. Alternatively, the internal two-electron transfer
step could produce an active site radical (R'.), such as the hydroxyl radical or a
ligand-based radical involving an amino acid residue, and an iron(IV) center that
164
would again be stabilized by tautomerization and electron transfer involving the
other iron atom (Figure 8). Although there is presently no direct evidence for
such a radical species in the MMO hydroxylase active site, the combination of a
redox active metal with a protein derived radical to effect oxygen-activated redox
transformations is known in the related iron enzyme ribonucleotide reductase,4 8
in the heme peroxidase prostaglandin H synthase,4 9 and in the copper enzyme
galactose oxidase.5 0 A third possibility is that the diiron(III) peroxide unit itself
might be sufficiently activated to carry out the hydroxylation reaction. As
depicted in Figure 8, the peroxide ligand might perhaps be coordinated to the
diiron(III) unit in an r 2,q 2 fashion, analogous to the binding mode thought to be
important in the dioxygen transport dicopper enzyme hemocyanin. 51,52 Other
modes of peroxide binding to one or both iron atoms are possible, as indicated in
Figure 9. Several important model studies have provided spectroscopic evidence
for the existence of diiron(III) peroxides;5 3-6 0 and, related species,5 3'
6 1' 6 2
including a high valent iron oxo intermediate, 6 3 have been implicated in
hydroxylation
and epoxidation reactions.
None has yet been structurally
characterized, however.
Once the activated iron center, whatever it may be, is discharged, product
is released with concomitant formation of the diiron(III) form of the hydroxylase
that enters another cycle in the catalysis (Figure 7). A peroxide shunt, analogous
to that found in the P-450 enzyme, has been reported for the M. trichosporium
(OB3b) MMO14 and confirmed by us in the present work. As indicated in Figure
7, the use of hydrogen peroxide permits hydroxylation reactions to occur in the
absence of the coupling and reductase proteins. Addition of hydrogen peroxide
to hydroxylase isolated from the M. capsulatusorganism, however, displays only
5 to 10% of the activity measured with the complete, three protein system.6 4
165
The Hydroxylation Reaction. Figure 10 presents six possible mechanisms
for substrate hydroxylation following generation of the dioxygen-activated
diiron center. In all cases except the last, 0-0 bond cleavage occurs prior to the
C-H bond-breaking step. The six possibilities are (A) direct insertion of the
oxygen atom of a high valent iron oxo species into a C-H bond; (B) concerted
addition of the C-H bond to the high valent iron oxo species to form a metalcarbon bond followed by reductive elimination of the alcohol; (C) heterolytic
attack of a high valent iron oxo species on the R-H bond followed by
recombination. to afford product; (D) homolytic attack of a high valent iron oxo
species on the R-H bond followed by return of the hydroxyl group from iron to
the alkyl radical, a so-called oxygen rebound step; (E) abstraction of a hydrogen
atom from the substrate by hydroxyl or another radical within the active site to
form an alkyl radical which then adds to the iron-bound oxygen atom; and (F)
electrophilic attack of an rI2,n 2-peroxide or related species (Figure 9) on the
substrate to form a carbon-oxygen bond followed by release of product.
Mechanisms I) and E involve transient formation of a substrate-derived alkyl
radical, which could rearrange as shown in Figure 1, and it is these pathways
that are directly addressed by the present investigation. The carbocation shown
in C (2) could also rearrange, and its involvement in the hydroxylation
mechanism would similarly be revealed by our studies.
Experiments reported here and previously 9'
13
with MMO from M. tri-
chosporium OB3b indicate that a radical is formed in the catalytic cycle and, a
paper has recently appeared reporting results that are consistent with a radical
intermediate.4
5
No ring-opened products are observed during the hydroxylation
of substrate probes 1 - 5 with MMO from M. capsulatus(Bath), however. There
are several possible explanations for these results. The simplest is that radicals
and carbocations do not form and that pathways C (2), D and E can be excluded
166
for the reaction mechanism for MMO from this organism. Alternatively, radicals
could form but recombine rapidly with the high valent iron-oxo moiety before
they have a chance to ring open. Although this alternative is reasonable for
probes 3, 4 and possibly 5, it is not a viable explanation for 1 and 2 since it would
require a radical rebound rate constant greater than 4 x 1013s- 1. This conclusion
is only valid, however, provided that the hydroxylase active site does not perturb
the ring opening rate constant of probes 1 and 2. Moreover, there are other
potential problems arising in probe substrate oxidations within enzyme active
sites, as delineated in a recent report of cytochrome P-450 dependent enzyme oxidations. 20 In that study, two "poorly behaved" probes gave much less than the
expected amount of rearranged product.
In addition to decreasing the rate
constant for radical rearrangement owing to steric effects, the low yields of
rearranged products might arise from a possible change in mechanism, for
example the onset of an insertion reaction, an increase in the rate constant of
oxygen rebound, or reaction of the rearranged radical with another site on the
enzyme.
Two of these potential problems can be excluded as the reason for the failure to detect ring-opened alcohols in the present study. Because the rearrangement of radical 1U. to 1R. (Figure 1) is fast enough to compete with any other
process, it is not possible that acceleration of the oxygen rebound step without a
diminution in the rate constant for the radical rearrangement could subvert the
rearrangement. Furthermore, if ring-opened radicals were to react with another
position in the enzyme active site, then the probe would most likely have deactivated the enzyme as a suicide substrate inhibitor. Although an extensive analysis of enzyme activity was not conducted for these substrates, there did not appear to be a noticeable loss of activity over the 15 min time course in these studies.
167
The other possibility, that constraints of the enzyme active site alter the
rate constant for radical rearrangement, can be considered for each substrate. For
probe 5, abstraction of a hydrogen atom at the methyl position leads to
cubylmethyl radical 5U, as shown in Figure 11. Rearrangement of this species
yields 5R*. Inspection of 5R. reveals that the substrate radical has undergone
rearrangement,
yet the radical is left at the methyl position.
No skeletal
rearrangement of carbon atoms occurs during this rearrangement, nor is there
any requirement for substrate reorientation for a subsequent rebound reaction.6 5
Steric constraints of the active site should be considered a negligible influence on
the rate constant for radical rearrangement.
For probes 1 and 2, it is possible that the orientation of the phenyl group in
enzyme-bound probe 1 is controlled by the steric confinement of the active site
such as to preclude co-alignment of the 7r-system of the phenyl ring with the
bond of the cyclopropane ring that is cleaved.
The imposition of such a
stereoelectronic effect on the ring opening reaction might result in a marked
reduction in the rate constant for ring opening of the radical derived from 1. In
the extreme case where the it-system of the phenyl ring is oriented perpendicular
to the
cyclopropyl ring bond that is opened, the contribution of a phenyl
resonance effect to the cyclopropylcarbinyl bond breaking step would be
eliminated, leaving only an inductive effect.
The manifestation of an active site constraint-enforced stereoelectronic effect on the kinetics of ring opening can be estimated semi-quantitatively via
Marcus theory..40°
66- 69
It has been shown4 2 that the rate constants for reactions of
the parent system, ring opening of the cyclopropylcarbinyl radical (6) to the 3butenyl radical (7) and cyclization of radical 7, can be used to predict rate
constants for ring openings of substituted cyclopropylcarbinyl radicals. In brief,
the rate constant for ring opening of 6 at 25 °C (kr = 5 x 107 s- 1 for cleavage of one
168
bond)
70 -73
and that for cyclization
of 7 (kr = 8000 s-1)7 4 -7 6 give AG ° = -5.2
kcal/mol; AGt has been measured to be 6.9 kcal/mol.
Solving the Marcus
equation (eq 5) with these values gives AGti = 9.32 kcal/mol, the intrinsic
AGt = AGti + AG°/2 + (AG) 2 /16AGti
(5)
activation energy expected for a thermoneutral cyclopropylcarbinyl radical ring
opening.
This value of AGit can be used in eq 5 to calculate AG$ for
cyclopropylcarbinyl ring openings of varying exergonicity. Figure 12 shows a
plot of log kr as a function of the exergonicity of the reaction.
Comparisons of standard bond dissociation energies (BDE)can be used to
estimate the exergonicity of the cyclopropylcarbinyl ring-opening reactions. The
BDE for a benzylic hydrogen in toluene is 10-11 kcal/mol smaller than that of a
primary C-H bond in ethane or propane. 7 7 78 For the benzylic position in ethylbenzene, the BDE is reduced by approximately 2.5 kcal/mol from that of
toluene. 7 7 78 Thus, the ring opening of the radical derived from 1 is expected to
be about 12.5-13.5 kcal/mol more exergonic than the value of the radical of 6,
giving a value of AG0 = -18.2 kcal/mol. If we assume that the full overlap of the
r-system of the phenyl ring is available in the transition state for ring opening of
the radical derived from 1, then the predicted rate constant for ring opening of
this radical at 25 °C is 1 x 1011s-1. Given the crude nature of the analysis, this
value compares favorably with the observed rate constant of 3 x 1011s-1 .
If the phenyl group in the radical derived from 1 is locked such that its nsystem cannot overlap fully with the breaking bond in the cyclopropyl ring, the
resonance stabilization would be a function of cos2 0, where 0 is the dihedral angle of the breaking bond and the tc-systemof the phenyl ring. When the latter is
orthogonal to the breaking bond (0 = 90°), the effect of the phenyl group would
169
be reduced to approximately that of the inductive effect of an alkyl group. In this
extreme case, the appropriate model reaction for the Marcus analysis would be
one in which the exergonicity of the ring-opening reaction is only about 2.5-3
kcal/mol greater than the value for 6 based on the differences in BDE between 1°
and 2 alkane C-H bonds.7 7 This value gives a predicted rate constant at 25 C of
about 4 x 108 s-1 .
Using the two limiting cases discussed above, corresponding to phenyl
stabilization between 3 and 13 kcal/mol, the stereoelectronic effect of an enforced
position of the phenyl group in the radical derived from 1 is shown on the plot in
Figure 12 for selected dihedral angles. When 0 is 60°, the expected rate constant
for ring opening at 25 C is reduced to 2 x 109 s,
which is about an order of
magnitude less than the oxygen rebound rate constants found in P-450
oxidations. 2 0 Because it is possible that an oxygen rebound rate constant in the
MMO hydroxylase is faster than that found in the P-450 studies, further
reduction of the rate constant for ring opening of radical derived from 1 could
lead to undetectably small amounts of ring-opened alcohol product lb. Thus the
absence of lb from oxidation of probe 1 could result from an extreme alteration
in the rate constant for ring opening of the derived cyclopropylcarbinyl radical in
the enzyme active site.7 9 80
Although the Marcus analysis demonstrates that a stereoelectronic reduction in the rate constant for ring opening of a radical derived from 1 is a possible
explanation for the absence of ring-opened alcohol lb, it should be recalled that
no ring-opened product formed in the hydroxylation of the closely related substrate 2. A significant reduction in the rate constant for ring opening of the
radical derived from this substrate probe due to steric effects in the active site is
less likely. In this case, both phenyl groups must be held such that their
-
systems are virtually orthogonal to the breaking bond in order to reduce
170
resonance stabilization in the transition state for opening, a much less probable
situation. Moreover, such an arrangement would require that the phenyl groups
approach coplanarity with one another, a geometry that would increase steric
interactions between them. The results from probe 2 therefore suggest that a
radical or carbocation is not formed and that the hydroxylation mechanism
involves one of the alternative pathways A, B, C (1), or F in Figure 10. Moreover,
a recent measurement of the ring-opening kinetics of several constrained, arylsubstituted cyclopropylcarbinyl radicals revealed that severe dihedral angles do
not appreciably diminish the rate constants for ring opening. 8 1 This study
strongly implies that non-bonding steric effects in the active site are unlikely to
diminish the ring-opening rate constant.
There is another factor to be considered with regard to the results with
probe 1. The discussion of steric effects on the rate constant for ring opening of
substrate derived radical ignored the fact that two enantiomers of this substrate
were allowed to react with the chiral enzyme. If sterics-enforced stereoelectronic
effects were operative, it is unlikely that both enantiomeric radicals derived from
the enantiomers of the probe would suffer similar ring-opening rate retardations.
The formation of aryl hydroxylated product in amounts comparable to the
amount of alcohol la produced complicates the analysis, however. As indicated
above, it is likely that the two different products are derived from the two different enantiomers of the probe substrate.
Aside from steric factors within the active site, other considerations to explain the lack of rearranged substrate with M. capsulatus (Bath) must be addressed. For example, one cannot rule out that the hydroxylation reaction may
proceed through two different mechanisms from a common intermediate. One
pathway, D or E, could involve a substrate radical, thus accounting for the rearranged products observed with MMO from M. trichosporiumOB3b. An alterna-
171
tive pathway could involve A, B, C, or F. The specific pathway chosen would
then depend on the energetics of the substrate. It is possible, however, that the
frequency with which each pathway is followed may differ between the two
organisms. Alternatively, MMO from M. capsulatus(Bath) and M. trichosporium
OB3b may exhibit different hydroxylation mechanisms, which would explain the
presence of ring-opened products for the latter.
We now return to the intramolecular deuterium isotope effects of 5.1 and
4.8 observed for the hydroxylation of deuterated 1 and 5. Although an isotope
effect of this magnitude is generally associated with processes that involve a
relatively linear transition state,8 2 such as would occur for oxygen rebound,
mechanisms containing intramolecular cyclic transition states cannot be
excluded. In fact, dimethyldioxirane, a reagent believed to oxidize hydrocarbons
via an insertion process, 83 oxidized cyclododecane-do and -d2 4 with an isotope
effect of 5.84 Thus our observed value of 5.1 would be consistent with pathways
A, B, C (1), or F, all of which involve transition states for C-H bond cleavage that
resemble the dioxirane oxygen insertion process and avoid the formation of a
substrate
radical or carbocation in the catalytic cycle.8 2
The previous
observation of an NIH shift in the hydroxylation of aromatic substrates with
MMO from M. capsulatus(Bath) would support such a mechanism.8' 85
In summary, the failure of substrate probes 1, 2, and 5 to form rearranged
alcohol products upon hydroxylation with the M. capsulatus(Bath) MMO is best
interpreted in terms of a mechanism that does not involve the formation of a
radical or carbocation, namely, A, B, C (1), or F. It is unlikely that the enzyme
active site induces a stereoelectronic barrier to ring opening. The fact that the
same amount (85%) of the probe, both converted and unconverted, can be
recovered from the enzymatic reaction mixture under hydroxylation conditions
or in control reactions omitting NADH indicates that ring-opened substrate
172
radical has not formed and been lost in the extraction procedure. In addition,
these probes do not lead to inactivation of the enzyme over the time course of the
reaction, indicating that ring-opened product has not formed and reacted with
the enzyme. The kinetic isotope results are consistent with a concerted pathway
for C-H bond insertion for the key hydroxylation step.
Comparisons with Related Work. The MMO mechanism is almost al-
ways drawn in analogy to that of cytochrome P-450.8' 13 14 In particular, a catalytic cycle involving an iron-oxo intermediate (Figure 8) that abstracts H. from
the alkane (mechanism D, Figure 10) has been proposed by several authors. The
ability of high concentrations of hydrogen peroxide to effect a shunt pathway for
the M. trichosporium OB3b MMO has been cited as evidence for heterolytic
cleavage of the 0-0 bond, 1 4 although it is noteworthy that, unlike the situation
for cytochrome P-450, other oxo transfer reagents could not be substituted for
H 20 2 . Hydrogen peroxide does not afford comparable levels of product in the
M. capsulatus (Bath) system.6 4' 86
intermediates
for the hydroxylation
Various authors have invoked radical
reaction.
For example, exo, exo, exo, exo-
2,3,5,6-d4-norbornane epimerizes upon hydroxylation by M. trichosporiumOB3b
MMO, 13 a result that has its parallels in cytochrome P-450 chemistry. 8 7 The
extent of epimerization with MMO, however, was significantly less, being 2%
following hydrogen atom abstraction from the endo position compared to 18%
with cytochrome P-450, and 5% after abstraction at the exo position as compared
to 14% with cytochrome P-450. Similarly, allylic rearrangements occurred in 20%
of the MMO hydroxylation products from 3,3,6,6-d4-cyclohexene compared to
33% for cytochrome
P-450.1 3' 8 8
Mechanistic studies of MMO from M.
trichosporium OB3b with the free radical probe 1,1-dimethylcyclopropane and
with (S) and (R)-[1-2 H,1-3 H]ethane were also consistent with radical formation,
but cationic intermediates could not be ruled out.12' 4 5 It has also been claimed
173
that radical intermediates can be trapped during methane hydroxylation by the
MMO of M. capsulatus(Bath).2 2 The mechanism for olefin epoxidation by MMO
from M. trichosporiumOB3b differs from that believed to occur with cytochrome
13 With cytochrome P-450, the 1-trans-proton of propylene exchanges with
1P-450.
solvent protons during turnover,8 9 leading to the hypothesis that the epoxidation
mechanism involves oxametallocycles and iron carbene intermediates. In the
reaction of propylene with MMO, however, no such exchange occurred.1 3
In view of these and the present results, it would appear that no single
unifying mechanism can account for all of the data. Instead, it seems likely to us
that more than one mechanism might be operative, possibly from a common
branchpoint. The frequency with which a specific pathway is followed would
depend on factors such as the steric demands of the substrate at the active site,
the organism from which the MMO is isolated, and even the temperature at
which the reaction is carried out. There are differences in the coupling proteins
obtained from the M. capsulatus (Bath) and M. trichosporium OB3b organisms, 9 0
91
which could alter the properties of the system. Moreover, it should be recalled
that the M. capsulatus(Bath) methanotroph was isolated from thermal waters and
has its optimal monooxygenase activity at 45 C, whereas the MMO from M. trichosporium OB3b is typically assayed at ambient temperature (23 C). Further
studies, involving additional mechanistic probes and the application of other
methodologies, are currently in progress to provide further insight into these issues.
Conclusions. The major findings of this investigation may be summarized as follows:
(1) The absence of ring-opened products in the hydroxylation of probes 1,
2, and 5 suggests that the mechanism for MMO from M. capsulatus (Bath) may
not involve radicals or carbocations, at least for these substrates.
174
(2) Isotope effect studies with deuterated probes 1 and 5 indicate that the
hydroxylation step is not rate determining in the overall MMO mechanism but
that substantial C-H bond activation is involved in this step.
(3) The observation that equimolar quantities of phenol and alcohol are
produced in the hydroxylations of racemic mixtures of 1 and its trideuterated
analogue implies that chiral discrimination in the enzyme active site can dictate
the regiospecificity of the reaction.
(4) A semi-quantitative Marcus theory analysis was used to evaluate how
the ring-opening rate constant of phenyl substituted cyclopropanes can be tuned
by changing the dihedral angle between the phenyl and cyclopropyl rings.
(5) Several detailed proposals have been set forth for two discrete arcs of
the MMO catalytic cycle (Figure 7), the activation of the reduced, diiron(II) form
by dioxygen (Figure 8), and the subsequent hydroxylation step (Figure 10).
These considerations should serve as a valuable reference point for the interpretation of mechanistic results from diiron oxo and related monooxygenases as
well as their model compounds.
(6) Analysis of the present and literature results for the hydroxylation of
hydrocarbons by MMO suggests that multiple mechanisms may be operative,
possibly from a common branch point.
Acknowledgments.
Substrate probes and authentic standards were
provided by C. C. Johnson and Professor M. Newcomb at Wayne State
University. Professor Newcomb also contributed the Marcus analysis of the ringopening rate constants of the phenyl substituted cyclopropanes. Most of this
'work was published previously.
providing cell paste.
Thanks also to Athanasios Salifoglou for
175
References
(1)
Anthony, C. The Biochemistry of Methylotrophs; Academic Press: New York,
1982, p 296-379.
(2)
Colby, J.; Dalton, H. Biochem. J. 1978, 171, 461-468.
(3)
Fox, B. G.; Surerus, K. K.; MUinck, E.; Lipscomb, J. D. J. Biol. Chem. 1988,
263, 1053-1056.
(4)
Green, J.; Dalton, H. J. Biol. Chem. 1985, 260, 15795-15801.
(5)
Froland, W. A.; Andersson, K. K.; Lee, S.-K.; Liu, Y.; Lipscomb, J. D. In
IUCCP Symposium on Applications of Enzyme Biotechnology; Plenum Press: College
Station, TX, 1991; pp 39-54.
(6)
Froland, W. A.; Andersson, K. K.; Lee, S.-K.; Liu, Y.; Lipscomb, J. D. J. Biol.
Chem. 1992, 267, 17588-17597.
(7)
Liu, K. E.; Lippard, S. J. J. Biol. Chem. 1991, 266, 12836-12839.
(8)
Green, J.; Dalton, H. J. Biol. Chem. 1989, 264, 17698-17703.
(9)
Fox, B. G.; Borneman, J. G.; Wackett, L. P.; Lipscomb, J. D. Biochemistry
1990, 29, 6419-6427.
(10)
Rosenzweig, A. C.; Frederick, C. A.; Lippard, S. J.; Nordlund,
P. Nature
1993, 366, 537-543.
(11)
DeRose, V. J.; Liu, K. E.; Kurtz, D. M., Jr.; Hoffman,
B. M.; Lippard,
S. J. J.
Am. Chem. Soc. 1993, 115, 6440-6441.
(12)
Ruzicka, F.; Huang, D.-S.; Donnelly, M. I.; Frey, P. A. Biochemistry 1990, 29,
1696-1700.
(13)
Rataj, M. J.; Kauth, J. E.; Donnelly, M. I. J. Biol. Chem. 1991, 266, 18684-
18690.
(14)
Andersson, K. K.; Froland, W. A.; Lee, S.-K.; Lipscomb, J. D. New J. Chem.
1991, 15, 411-415.
176
(15)
McMurry, T. J.; Groves, J. T. In Cytochrome P-450 Structure, Mechanism, and
Biochemistry; Ortiz de Montellano, P. R., Eds.; Plenum Publishing Corp.: New
York, 1986; pp 1-28.
(16)
Ortiz de Montellano, P. R. In CytochromeP-450 Structure, Mechanism, and
Biochemistry; Ortiz de Montellano, P. R., Eds.; Plenum Publishing Corp.: New
York, 1986; pp 217-271.
(17)
Mansuy, D.; Battioni, P. In Activation and Functionalizationof Alkanes; Hill,
C. L., Eds.; Wiley: New York, 1989; chapter VI.
(18)
Guengerich,
F. P. In Biological Oxidation Systems, Vol 1; Reddy, C. C.,
Hamilton, G. A. and Madyastha, K. M., Eds.; Academic Press: San Diego, 1990;
pp 51-67.
(19)
Ortiz de Montellano,
P. R.; Stearns, R. A. J. Am. Chem. Soc. 1987, 109, 3415-
3420.
(20)
Bowry, V. W.; Ingold, K. U. J. Am. Chem. Soc. 1991, 113, 5699-5707.
(21)
Bowry, V. W.; Lusztyk, J.; Ingold, K. U. J. Am. Chem. Soc. 1991, 113, 5687-
5698.
(22)
Deighton, N.; Podmore, I. D.; Symons, M. C. R.; Wilkins, P. C.; Dalton, H.
J. Chem. Soc., Chem. Comm. 1991, 1086-1088.
(23)
Fu, H.; Newcomb, M.; Wong, C.-H. J. Am. Chem. Soc. 1991, 113, 5878-5880.
(24)
Woodland,
(25)
DeWitt, J. G.; Bentsen, J. G.; Rosenzweig, A. C.; Hedman,
M. P.; Dalton, H. J. Biol. Chem. 1984, 259, 53-59.
IPilkington, S.; Papaefthymiou,
B.; Green, J.;
G. C.; Dalton, H.; Hodgson, K. O.; Lippard, S. J. J.
Am. Chem. Soc. 1991, 113, 9219-9235.
(26)
Wu, W. W.; Rosenzweig, A. C.; Lippard, S. J., unpublished results.
(27)
Pilkington, S. J.; Dalton, H. In Methods In Enzymology Academic Press:
New York, 1990; Vol. 188; pp 181-190.
177
(28)
Fox, B. G.; Froland, W. A.; Jollie, D. R.; Lipscomb, J. D. In Methods In
EnzymologyAcademic Press: New York, 1990; Vol. 188; pp 191-202.
(29)
Stirling, D. I.; Dalton, H. Eur. J. Biochemistry 1979, 96, 205-212.
(30)
Dalton, H.; Higgins, I. J. Antonie van Leewenhoek 1987, 53, 23-28.
(31)
Gajewski, J. J.; Squicciarini, M. P. J. Am. Chem. Soc. 1989, 111, 6717-6728.
(32)
Falorni, M.; Lardicci, L.; Giacomelli, G. J. Org. Chem. 1986, 51, 5291-5294.
(33)
Blatchford, J. K.; Orchin, M. J. Org. Chem. 1964, 29, 839-843.
(34)
Choi, S.-Y.; Eaton,
P. E.; Hollenberg,
P. F.; Liu, K. E.; Lippard,
S. J.;
Newcomb, M.; Putt, D. A.; Upadhyaya, S. P., manuscript in preparation.
S. U.; Rao, C. G. Synthesis 1979, 704-705.
(35)
Brown, H. C.; Kulkarni,
(36)
Feutrill, G. I.; Mirrington,
(37)
Gassman, P. G.; Mansfield, K. T.; Taylor, G. N.; Wiberg, K. B. Org. Syn.
R. N. Tetrahedron Lett. 1970, 11, 1327-1328.
1969, 49, 1-6.
(38)
It is possible that the ortho and meta isomers of c coeluted with lc in the
GC analysis and were thus not detected. Separation of enzymatic reaction
mixtures with 1 by HPLC (C-18 reverse-phase) yielded only 3 peaks which
corresponded to 1, la, and c. It is unlikely that the ortho and meta isomers of lc
would coelute with lc in both GC and HPLC analysis.
(39)
Singy, G. A.; Buchs, A. Helv. Chim. Acta 1974, 57, 1158-1169.
(40)
Newcomb,
M.; Manek, M. B.; Glenn, A. G. J. Am. Chem. Soc. 1991, 113, 949-
(41)
Newcomb,
M.; Manek, M. B. J. Am. Chem. Soc. 1990, 112, 9662-9663.
(42)
Newcomb,
M.; Johnson, C. C.; Manek, M. B.; Varick, T. R. J. Am. Chem. Soc.
958.
1992, 114, 10915-10921.
(43)
Shimoda, M.; Okura, I. J. Mol. Catal. 1992, 72, 263-267.
(44)
Green, J.; Dalton, H. Biochem. J. 1989, 259, 167-172.
178
(45)
Priestley, N. D.; Floss, H. G.; Froland, W. A.; Lipscomb, J. D.; Williams, P.
G.; Morimoto, H. J. Am. Chem. Soc. 1992, 114, 7561-7562.
(46)
Howard, J. A.; Scaiano, J. C. Landolt-Bornstein Numerical Data and
Functional Relationshipsin Scienceand Technology;Springer-Verlag: Berlin, 1984;
Vol. 13, subvol. d, p 125-127.
(47)
Fox; B. G.; Froland, W. A.; Dege, J. E.; Lipscomb, J. D. J. Biol. Chem. 1989,
264, 10023-10033.
(48)
Bollinger, J. M., Jr.; Edmondson, D. E.; Huynh, B. H.; Filley, J.; Norton, J.
R.; Stubbe, J. Science 1991, 253, 292-298.
(49)
Kulmacz, R. J.; Ren, Y.; Tsai, A.-L.; Palmer, G. Biochemistry 1990, 29, 8760-
8771.
(50)
Babcock, G. T.; El-Deeb, M. K.; Sandusky,
Whittaker,
(51)
P. O.; Whittaker,
M. M.;
J. W. J. Am. Chem. Soc. 1992, 114, 3727-3734.
Magnus, K. A.; Hazes, B.; Ton-That, H.; Bonaventura, C.; Bonaventura, J.;
Hol, W. G. J. Proteins 1994, 19, 302-309.
(52)
Ling, J.; Nestor, L. P.; Czernuszewicz,
R. S.; Spiro, T. G.; Fraczkiewicz,
R.;
Sharma, K. D.; Loehr, T. M.; Sanders-Loehr, J. J. Am. Chem. Soc. 1994, 116, 76827691.
(53)
Murch, B. P.; Bradley, F. C.; Que, L. J. J. Am. Chem. Soc. 1986, 108, 5027-
5028.
(54)
Menage, S.; Brennan, B. A.; Juarez-Garcia, C.; Miinck, E.; Que, L., Jr. J. Am.
Chem. Soc. 1990, 112, 6423-6425.
(55)
Kitajima, N.; Fukui, H.; Moro-oka, Y. J. Am. Chem. Soc. 1990, 112, 6402-
6403.
(56)
Brennan, B. A.; Chen, Q.; Juarez-Garcia, C.; True, A. E.; O'Connor, C. J.;
Que, L., Jr. Inorg. Chem. 1991, 30, 1937-1943.
179
(57)
Kitajima, N.; Tamura, N.; Amagai, H.; Fukiu, H.; Moro-oka, Y.; Mizutani,
Y.; Kitagawa, T.; Mathur, R.; Heerwegh, K.; Reed, C. A.; Randall, C. R.; Que, L.,
Jr.; Tatsumi,
(58)
K. J. Am. Chem. Soc. 1994, 116, 9071-9085.
Dong, Y.; Menage, S.; Brennan, B. A.; Elgren, T. E.; Jang, H. G.; Pearce, L.
L.; Que, L., Jr. J. Am. Chem. Soc. 1993, 115, 1851-1859.
(59)
Hayashi, Y.; Suzuki, M.; Uehara, A.; Mizutani, Y.; Kitagawa, T. Chemistry
Letters 1992, 91-94.
(60)
Sawyer,
D. T.; McDowell,
M. S.; Spencer, L.; Tsang, P. K. S. Inorg. Chem.
1989, 28, 1166-1170.
(61)
Tung, H.-C.; Kang, C.; Sawyer, D. T. J. Am. Chem. Soc. 1992, 114, 3445-3455.
(62)
Barton, D. H. R.; Beviere, S. D.; Chavasiri,
W.; Csuhai, E.; Doller, D.
Tetrahedron 1992, 48, 2895-2910.
(63)
Leising, R. A.; Brennan, B. A.; Que, L., Jr. J. Am. Chem. Soc. 1991, 113, 3988-
3990.
S. J., unpublished
results.
(64)
Liu, K. E.; Lippard,
(65)
Eaton, P. E.; Yip, Y. C. J. Am. Chem. Soc. 1991, 113, 7692-7697.
(66)
Lowry, T. H.; Richardson, K. S. Mechanism and Theroy in Organic Chemistry;
3rd ed.; Harper & Row: New York, 1987, p 222-227.
(67)
Marcus, R. A. Ann. Rev. Phys. Chem. 1964, 15, 155.
(68)
Applications of Marcus theory to predict the rate constants of group
transfer reactions of radicals are generally quite accurate when charge transfer
states are unimportant in the radical reactions. 69 A Marcus approach was used
successfully to estimate rate constants for hydrogen atom transfer reactions.4 0 .
(69)
Fox, G. L.; Schlegel, H. B. J. Phys. Chem. 1992, 96, 298-302.
(70)
Maillard, B.; Forrest, D.; Ingold, K. U. J. Am. Chem. Soc. 1976, 98, 7024-7026.
(71)
Mathew, L.; Warkentin, J. J. Am. Chem. Soc. 1986, 108, 7981-7984.
(72)
Newcomb,
M.; Glenn, A. G. J. Am. Chem. Soc. 1989, 111, 275-277.
180
(73)
Bechwith, A. L. J.; Bowry, V. W.; Moad, G. J. Org. Chem. 1988, 53, 1632-
1641.
(74)
The reported 7 5 rate constant for cyclization of the 3-butenyl radical has
been corrected using a more recently determined value7 6 for the rate constant for
reaction of Bu3 SnH with a primary radical.
(75)
Effio, A.; Griller, D.; Ingold, K. U.; Beckwith, A. L. J.; Serelis, A. K. J. Am.
Chem. Soc. 1980, 102, 1734-1736.
(76)
Johnston, L. J.; Lusztyk, J.; Wayner, D. D. M.; Abeywickreyma,
A. N.;
Beckwith, A. L. J.; Scaiano, J. C.; Ingold, K. U. J. Am. Chem. Soc. 1985, 107, 45944596.
1(77) McMillen, D. F.; Golden, D. M. Ann. Rev. Phys. Chem. 1982, 33, 493-531.
(78)
Bordwell, F. G.; Cheng, J.-P.; Harrelson, J. A., Jr. J. Am. Chem. Soc. 1988,
::Z10,1229-1231.
(79)
For a related treatment of stereoelectronic effects on the kinetics of radical
reactions, see Tanko, J. M.; Kamrudin, N.; Blackert, J. F. J. Org. Chem. 1991, 56,
6395-6399.
(80)
Another possible explanation for formation of cyclic alcohol la, reversible
ring opening of the radical from 1 and selective trapping of the cyclic radical, can
be excluded readily. From the expected exergonicity of the ring opening reaction
(i.e. AG° = - 18.2 kcal/mol), the equilibrium constant for the ring opening at room
temperature
is about 2 x 1013. With a ring opening
rate constant
of 3 x 1011 s-1,
the ring closure of the acyclic radical must have a rate constant on the order of
0.01 s -1 . That the benzylic radical could have a lifetime of 100 s at room
temperature is without precedence, and the turnover rate of MMO exceeds this
value.
(81)
Martin-Esker,
A. A.; Johnson, C. C.; Horner, J. H.; Newcomb, M. J. Am.
Chem. Soc. 1994, 116, 9174-9181.
181
R. A. M. J. Chem. Soc. (B) 1970, 785-790.
(82)
O'Ferrall,
(83)
Adam, W.; Curci, R.; Edwards,
(84)
Murray,
R. W.; Jeyaraman,
J. O. Acc. Chem. Res. 1989, 22, 205-211.
R.; Mohan, L. J. Am. Chem. Soc. 1986, 108, 2470-
2472.
(85)
Tsuda, M.; Oikawa,
S.; Okamura,
Y.; Kimura, K.; Urabe, T.; Nakajima,
M.
Chem. Pharm. Bull. 1986, 34, 4457-4466.
(86)
Jiang, Y.; Wilkins, P. C.; Dalton, H. Biochim. Biophys. Acta 1993, 1163, 105-
112.
(87)
Groves, J. T.; McClusky, G. A. Biochemical and Biophysical Research
Communications
1978, 81, 154-160.
(88)
Groves, J. T.; Subramanian, D. V. J. Am. Chem. Soc. 1984, 106, 2177-2181.
(89)
Groves, J. T.; Avaria-Neisser,
G. E.; Fish, K. M.; Imachi, M.; Kuczkowski,
R. L. J. Am. Chem. Soc. 1986, 108, 3837-3838.
(90)
Rosenzweig,
A. C.; Feng, X.; Lippard,
S. J. In IUCCP Symposium
on
Applications of Enzyme BiotechnologyPlenumPress: College Station, TX, 1991; pp
182-203.
(91)
Tsien, H.-C.; Hanson, R. S. Appl. Environ. Microbiol. 1992, 58, 953-960.
182
CZ
._
X
-
._
.-
.-
-b4
_.
.
I~
E
0
Co
Co
ir
Lr)
C
Cl
:
N
m)
i
I
I--
o
I
.,
.
I-
UV
w'
)
k
!.~~
I'
C)
F
.6.
M
-
c' c,
E
.-;::
,_
z,
E_
m
C
U
u-o3
F
0On
U
C
0o
:t:
r.
U
t
Lf
*-4
-*4
o
L
C,)
F4
,.
m
EU
H6
C4
e
-~
c
c3
t
U)ti'
C)
3
4r3
ir
183
._
c3
tL
L6
ON
00
r-
O
r4
Cl
00
\6d
r
NCl
._
U
._o
Q)
0
r.
0
S
u
L0
LO)
Lt)
U
U)
U
Lf)
U
u
o
I-
r.
E
a)
U
CU
ca
NL
d
r
C6)
c
C
r-
Cl~
cj
O
)
Cl~
~ %
cN £d
o00
Pen
U
C
4
p
L)
c
CN
E
r
S
:
0
e)
en
I-oo
E
Wo
I
.ca
:S
Cl
a)
--
Cr; r
r;
k Lf
< C%)
0
;
C
Cl
OQ)
o~I
_-
0
._
S
0
E.
-
V)
*'a)
"
¢ r-q
ur-i
Ni
Ca
.40
OBa
k
-
184
Table 3. Expected Products and Observed Ratios for Reactions with MMO from
A. capsulatus(Bath).
Rate Constant
Rearranged Observed Ratio of
Unrearranged Alcohol(s) Unrearranged to
For Radical RearSubstrate Product(s)
Expected Rerranged Alcohol
rangement, s -1
at 45 °C
OH
Phi. OH Phib
Ph
la
1
HO
>100:1
4 x 101'
>25:1
5 x 10
both
>100:1
3b: 2.0 x 108
3c: 2.4 x 108
80:1
3x10 9
lb
+ic
1c
Ph
Ph
Ph\ . OH HO
Ph _A
Ph¢
Ph
2
2a
2b
OH
A'/,
,
/,.,OH
3b
3a
3
HOu3
3c
OH
HO
4
4c
4a
4b
X-'OH
o OH
>25:1
3 x 101°
25 °C
5
5a
55a 5e~~~~~~~~~~~~~~~~~~~~~~
5e
185
r4
E
0
0o
,.
to
ca
o
CZ
0
u
r.
w4
186
·c0
:1
m
.0
P.4
P-4
ta
CZ
4
.--
A
4C4
E
A)
187
Figure 2. Substrates
1 through
5.
188
Phl,
Ph* jj
Ph~
2
2
1
$0\////8 '..
3
4
5
189
Ca
,.
Ct
0)
to
-0
U
u
U
.~,
190
0
IF
aw
k
I
C
ea
I.
.0.
rO
To
,
Ch
I-I
in
VD
(C)
0
*r
0VA
w
0
C6P
CZ
&.4
Q
P=
QJ
w =
t!=
> ov
in0. .
I[
(n
"V
m:
rc
-
en
AD
_
w
C,
0,
C,
r
~t
.Z
flX
m
1-
=t
_
rj
tn
C
Q
191
cn
U)
C
U
C
0.)
,.o
U)
C
u
u
un
.
©
o
To
To
u
E
*4
ca
c,
o
oD
._
-4
192
ut
tN-.
ts
tNq
c0
N
0
0
m
-1
O
p3npo:cdIalWN
0
0
0
193
Ea;
cu
0o
0o
.f
i*
O4
194
Ln
m
00
r-
("4
b.
pnpojd alOwN
195
.u
C
CI
U)
0)
Uo
0)
.C
c)
o
196
xNi
",
t'e
cN0
To
Tq
li!sualuj UoI
0o
197
Figure 7. Proposed catalytic cycle for the hydroxylation of hydrocarbons by
MMO.
198
ROH + H 2 0
R-H
FeII
H/
FeI"'
I
H
R-H
Fe I I I
H 20 2
0/
J
FeI
shunt
pathway
I
H
R-H
Fe"'
FemII
I
H
+ bound peroxide
or equivalent
ADH
R-H
Fel
Fell
NAD+ + H +
02
H
199
u
u
u
.-
a:
a)
-4
ut)
X.
..
u
-4-
ca
.4'4
200
o= v~
\
0=~~
/-
O Qg
0= (
-4
md
.i.
w,
",/ \
O. o
wl
0-=
"o
·
·
-
>/
0= ,
0= o
+ o)Q:
CO
x
~~ it~~4)
0
o0-\
\
/ 0-9
o
_
0-
\
/
IV
201
o.
0I
:.4
202
O0
(U
0/0
\\
00\
O-- O
0
0©
\o
/
/
0
W
wo
W-4
©o
!
/\
Wr
203
0
u.Ld_
o
-
I
x
Td
U)
._..
X
©
.
o
4d
a)
E
EU
03
so
on
©
r.
E
M
IC.
204
0-=
g,
(s
_/
0-=
1
re
:zw
0-=
b-:z
\
/
b-
of
0<
C
P
0-tM
0*
i
/
=*0-
/
0+
0.*
+=
O
. O-Zd
0-A
i
Q,
LI
0-::=-
0-:
0-
>
94 w
_/
0-=
0.4/
0-4
/
x
O-
>
>0-=
0=
- ./
V-A,(U
wo
0-4
=
w~
O-A
(-
u
u
205
-z
_s -
O
*0
/ 0--
/A
0-=)~
-I
©o
0
O
0-
I
0--o
-
+
."E
0+0!
P-4
0k
~~~~~~~
I
0O-
c
"4
t'"
C4
*=
=
0
-- ,o
00 0
To4
=
=
0
0
206
ui
E
0
E
0
.o
C
tc
0
0C)
-
.4.
o
;.4
207
©
0
!ToL
II
-V
O
0~~~~~f
©~
InI
Ln
-
n
208
a
o ~a
o
C
at
et
o
o
E
u
"0
'-4
0
ad
ad
U
cu
U
U
,.U
'4-
a
'4-
U
0
a
"
a
P,
.4-
4
u ad
c-
ad
U
o
d
Q)
U yU
U
.4-h
o,
Q0
-i
5
'-4
I0
"
-"4
"
adn
0
w
209
0
C%4J
U-)
I
0
Lt
0
1.
-ol
210
CHAPTER
5.
Tritiated Chiral Alkanes as
Substrates for Methane Monooxygenase from Methylococcus
capsulatus (Bath): Probes for the Mechansim of Hydroxylation
211
Preface.
This chapter describes work which is ongoing at the time of this
writing.
Most of the results are preliminary, and the following discussion
should not be taken to represent the final conclusions from this series of
experiments. In particular, the exchange process currently remains under
investigation.
The work described here was carried out with pure
hydroxylase and an unresolved mixture of protein B and reductase.
Direct
evidence for an exchange process was obtained when a significant amount of
CD 3 CDHOH was recovered from reactions with C2D6. We have ruled out the
possibility that the exchange arises from an artifactual property of the
unresolved mixture of protein B and reductase since no scrambling of
CD3 CD20H takes place in the absence of pure hydroxylase (see discussion).
Introduction.
The methanotrophic bacteria Methylococcus capsulatus (Bath) and
Methylosinus trichosporium OB3b rely on methane monooxygenase (MMO)
to convert methane into methanol in the first step of their biosynthetic
pathways.
1
This reaction provides the organisms with their sole source of
carbon and energy according to equation (1). MMO exists in both a particulate
and a soluble form.1 The former resides in the cell membranes under normal
CH4 + 02 + NADH + H+
-
CH30H + H20 +NAD+
(1)
growth conditions and contains copper at its active site. 2 Under low copper
concentrations, the methanotroph switches on the synthesis of the soluble
MMO. 3
Soluble MMO's from both Methylococcus capsulatus (Bath)4 and
Methylosinus trichosporiumOB3b5 comprise three proteins, a hydroxylase
component (MMOH) that binds the hydrocarbon substrate and dioxygen,6 a
212
reductase (MMOR) containing Fe2S 2 and FAD cofactors which enable it to
accept electrons from NADH and transfer them to the hydroxylase, 4 ' 7 and a
third component, protein B (MMOB), which regulates electron transfer
between the reductase and hydroxylase. 8 In addition to methane, the MMO
system is capable of oxidizing a wide variety of substrates.9 -1 1
X-ray crystallographic studies of the resting, diferric form of the
hydroxylase,
12
ENDOR spectroscopy of the mixed-valent state, 13 and other
physical methods 61' 4 indicate that the M. capsulatus (Bath) MMOH active site
contains a non-heme dinuclear iron center, the structure of which is given on
the following page.
As shown in this drawing, the iron atoms can be
additionally bridged by an exogenous acetate ligand introduced under the
crystallization conditions for the x-ray structural investigation.1 2
CH3
l
water O
Glu 114
o Glu 243
Glu
209
Glu 209
Fe 2
Fe,
N
His 147
acetate
H
0
N
0
His 246
Glu 144
The heme enzyme cytochrome P-450 displays similar reactivity to
MMO with the exception that it will not hydroxylate methane.1 5 Cytochrome
P-450 oxidations are thought to proceed through a high valent iron oxo, or
ferryl, intermediate.
15 -18
This species is postulated to abstract a hydrogen
atom from the hydrocarbon substrate, resulting in a substrate radical and a
coordinated hydroxyl radical. The hydroxyl radical and substrate radical next
recombine in a "rebound" reaction to afford the product alcohol, which dissociates from the active site leaving the resting state of the enzyme.
Radical
clock probes which rearrange upon formation of a substrate radical have been
213
useful in establishing the presence of substrate radicals in the reaction cycle of
cytochrome P-450.19 -21
The radical clock results reported in Chapter 4 differ for MMO systems
isolated from the two organisms M. capsulatus (Bath) and M. trichosporium
OB3b. With M. capsulatus (Bath), no evidence for the formation of substrate
radicals was found with all substrates tested. Based on the known rate
constants for rearrangement of the various probes, the lower limit for a
rebound reaction was estimated to be 1013s-1 at 45 °C. For M. trichosporium
OB3b, however, evidence for substrate radicals was detected in the form of
rearranged product alcohols.
observed,
a rate constant
Based on the amount of rearrangement
of 6 to 9 x 1012 s- l at 30 C was estimated
for a
rebound reaction with this enzyme.
Evidence for intermediate radical formation has also been obtained in
other studies with MMO from M. trichosporium OB3b,2 2 -24 including a report
with the substrates (R)-[1-2 H,1- 3 H]-ethane and (S)-[1-2 H,1- 3 H]-ethane. These
results were interpreted as evidence that the hydroxylation reaction with
these substrates proceeds through a mechanism analogous to that widely
accepted for cytochrome P-450 hydroxylation, as illustrated in Figure 1. Upon
abstraction of a hydrogen atom from the chiral substrate, the ethyl radical
results.
This substrate radical has the option to rebound and form the
unrearranged product alcohol. Alternatively, since the carbon radical is now
planar, rotation about the C-C bond prior to recombination in the rebound
reaction leads to a rearranged
configuration.
product
alcohol with the opposite
Both alcohol products can then be derivatized with a chiral
acid, and 3H NMR analysis of the diastereomeric esters used to quantitate the
amount of retention or inversion that occurred during the hydroxylation
reaction. Results with MMO from M. trichosporiumOB3b indicated that, for
214
both R and S chiral ethane substrates, = 65 % of the product alcohols displayed
retention whereas = 35 % displayed inversion.2 2 Based on a value of 0.5 kcal
for rotation about the C-C bond, a rate constant of 3 x 1012 -1 can be calculated
for the rebound reaction rate constant for the substrate radical. This value
agrees well with that estimated from the radical clock work.2 5
Although the radical clock substrate probes, which displayed slightly
different reactivity with MMO from the different organisms, gave results in
agreement with the chiral ethane experiment, we were interested in
determining whether the differences between the M. capuslatus (Bath) and M.
trichosporiumOB3b enzymes would persist with a substrate purported more
representative
22
of methane. Accordingly, experiments employing the chiral
ethane methodology were carried out with MMO from M. capsulatus (Bath)
in collaboration with Dr Philip Williams and Dr. Hiromi Morimoto at the
National Tritium Labeling Facility and with Professor Heinz Floss and Dr.
Barrie Wilkinson at the University of Washington.
This group of
collaborators is the same who participated in studies with the M.
trichosporium OB3b MMO. Accordingly, the substrates (S)-[1-2 H,1-3 H]-ethane,
(R)-[1-2 H,1-3 H]-ethane, (S)-[1-2H,1-3 H]-butane, (R)-[1-2 H,1-3H]-butane, (S)-[23 H]-butane
and (R)-[2-3 H]-ethane were hydroxylated with MMO from M.
capsulatus (Bath) and the results and analysis are reported here.
Experimental.
Synthesis of the substrate precursors was performed by Dr. Barrie
Wilkinson. NMR data were collected by Dr. Philip Williams at NTLF.
Bacterial Growth and Protein Purification.
Methylococcus capsulatus (Bath) cells were grown as described
previously.
26
Hydroxylase was purified as reported elsewhere, and the iron
content and specific activities were in the reported ranges.1 4 ' 2 5 In the
215
enzymatic reactions, an unresolved mixture of protein B and reductase from
the DEAE cellulose column in the hydroxylase purification method was
added to the hydroxylase. The ratio of this protein B-reductase mixture
relative to hydroxylase was determined by maximizing hydroxylase activity
with propylene as substrate.
Preliminary assays with unlabelled (proteo) ethane and butane were
carried out to determine conditions under which 1 to 2 mmole of product
alcohol could be produced.
MMOH was concentrated to = 300 gM and
incubated with the MMOB-MMORmixture in a 5 ml reaction flask which
was capped with a septum. The resulting volume was adjusted to 400 Rl with
25 mM MOPS, pH 7.0 buffer. A syringe was inserted through the septum and
2 ml of the head space gas was removed and replaced with 2 ml of the
substrate gas. The mixture was incubated for 30 s at 45 C in a shaking
incubator, after which time 100 Rl of a 0.1 M ethanol-free solution of NADH
was added to initiate the reaction. The flask was returned to the incubator
and the reaction was allowed to proceed for 5 min. A 5 Rl portion of the
solution was next injected into a GC equipped with a Porapak Q column to
quantitate the concentration of alcohol product present in the reaction
solution. The temperature settings were 200 °C for the injector, 225 °C for the
FID detector, and 180 °C for the oven. The area of the product peak was then
compared to a previously constructed calibration curve of alcohol peak area
vs the concentration of sample injected. Conditions which afforded at least 1
mmol of alcohol product per 1 ml of reaction volume was produced were
subsequently employed at NTLF. Under these conditions, the specific activity
calculated for the hydroxylase is low, 40 mU/mg, owing to product
inhibition. Quantities of the proteins used in various runs at NTLF are listed
in Table 1.
216
Reagent Synthesis. 2 7
(S)-[1-1H,1- 2H]-Ethanol. [1-2H]-Acetaldehyde (0.56 ml, 10 mmole) was
added to a stirred solution of (R)-alpine hydride (10 mmole) in 20 ml dry THF
at -70 °C under argon. The temperature was maintained at -70 °C for 4 h, after
which time the mixture was allowed to warm to room temperature. After 15
to 18 h, the solution was cooled to 0 C and ethanolamine (0.74 ml, 12.3
mmole) was added.
After 30 min a white precipitate formed. The solvent
was removed by distillation, during which the precipitate dissolved upon
warming. Just before the end of the distillation, 5 ml of benzene was added
to, and distilled from, the mixture.
The benzene addition and distillation
steps were repeated twice. The final volume of the collected distillates was 32
ml. (R)-[1-1H,1- 2H]-Ethanol , (S)-[1-1H,1- 2 H]-butanol, (R)-[1-1H,1- 2 H]-butanol,
(S)-2-butanol, and (R)-2-butanol were synthesized by using similar
procedures.
The purity of these materials was >95% as judged by 1 H NMR
spectroscopy.
(S)-[1-1H,1-2 H]-Ethyl Tosylate. (S)-[1-1 H,1- 2 H]-Ethanol (16 mmole in
THF) was added to a stirred solution of tosyl chloride (3.8 g, 20 mmole, freshly
crystallized from hexane) in dry pyridine (15 ml, distilled from KOH under
argon) under argon at room temperature.
almost immediately.
Pyridine HCl began to precipitate
The reaction was allowed to continue for 5 h, after
which time it was terminated by the addition of 100 ml of diethyl ether and
100 ml of a saturated aqueous solution of CuSO4. The organic phase was
collected and washed with two 50 ml portions of saturated CuSO4 , 30 ml of a 1
N HCl solution, and 50 ml of a saturated aqueous solution of NaHCO 3. The
organic phase was dried over Na2 SO4 and filtered to remove the solid. The
cdesired product was isolated by silica gel column chromatography and eluted
with 7 % diethyl ether in hexane. (R)-[l- 1H,1- 2H]-Ethyl tosylate, (S)-[1-1 H,-
217
2 H]-butyl
tosylate, (R)-[1-1H,1-2 H]-butyl tosylate, (S)-2--butanyl tosylate, and
(R)-2-butyl tosylate were synthesized following analogous methods. The
purity of were judged by NMR spctroscopy >95%.
(S)-[1- 2 H,1-3 H]-Ethane. A diagram of the reaction apparatus is shown
in Figure 2. The reagent LiEt3B3 H was prepared in the following manner.2 8
3H
2
( 100 Ci) was first transferred from a uranium tritide bed thermostatted
at 350 C to a 10 ml round bottom flask fitted with a stopcock (A in Figure 2)
and septum-capped sidearm at a final pressure of 625 mm Hg. By means of a
gas-tight syringe, n-BuLi (166 gl dissolved in hexanes, 0.027 mmole) and
TMEDA2 9 (34 pl, 0.23 mmole) were added to the flask through the septum.
The resulting mixture was stirred at room temperature for 45 min, during
which time a white, creamy precipitate (Li3H) formed.
The round bottom
flask was evacuated to dryness, reaching a pressure of = 30 igm Hg. THF (100
gl) and dinitrogen were added to a final pressure of 150 jgm Hg. The resulting
Li3 H slurry was treated with Et 3B (200 gl in a THF solution, 0.2 mmol) to
dissolve the Li3 H. The alkyl tosylate (= 0.3 mmole), dissolved in 100 1 of
THF, was added to the flask and vigorous gas evolution occurred for 15 to 20
s. After a 1 h incubation period, the flask was cooled to -78 C. Stopcock C was
opened and the gas was transferred to a coconut charcoal bed (200 mg) in a test
tube immersed in liquid nitrogen. The transfer was allowed to occur for 60 s.
Identical procedures were used to synthesize (R)-[1-2 H,1- 3 H]-ethane, (S)-[12 H,1- 3H]-butane,
(R)-1-[2 Hl, 1 -3H 1]-butane, (S)-[2-3 H]-butane, and (R)-[2-3 H]-
butane from their corresponding alkyl tosylates. Purity levels >95% were
estimated by
1H
NMR.
Enzymatic Reactions.
A second round bottom flask equipped with a stopcock, sidearm, and
septum was fitted to the reaction apparatus (B in Figure 2) and evacuated.
218
Buffer (25 mM MOPS, pH 7.0, 100 to 500 gl) was added through the septum
and cooled to -78 C. Substrate gas was transferred from the charcoal bed by
opening stopcock C and stopcock B on the reaction flask. The bed was brought
to room temperature by warming with hot air during this procedure. After 60
s the stopcocks were closed and the buffer was brought to room temperature.
The protein solution was next added through the septum. A quantity of
NADH (50 or 100 gl, see Table 1) was also added. Pure 02 gas (1 ml) was added
by syringe through the septum and the mixture was incubated at 45 + 2 °C
with constant stirring for 30 min. Cooling the flask to - 78 C terminated the
reaction. Excess substrate gas was transferred from the reaction flask to the
charcoal bed by first cooling the bed to = 77 K and then opening stopcocks B
and C. Transfer was allowed to proceed for 60 s and then the stopcocks were
closed. A small amount (10 l) of the expected product alcohol (unlabeled)
was added to the reaction flask to act as a carrier in subsequent manipulations.
This flask was removed from the apparatus and transferred to a vacuum
where the volatile reaction products were collected by lyophilization.
Derivitization. 30
(R)-O-Acetyl mandelic acid (47 mg, 0.24 mmol) and DMAP (2 mg) were
mixed in 2 ml of methylene chloride at -40 C in a 100 ml round bottom flask.
Over a 5 min period, a solution of DCCI (54 mg, 0.24 mmole) in 0.5 ml
methylene chloride was added dropwise to the flask. A creamy precipitate
resulted after 10 min. The enzymatic lypholysate was added dropwise over a
5 min period.
The resulting mixture was allowed to warm to room
temperature over 3 to 6 h and then stirred overnight.
The suspension was
filtered through a silica plug ( 2 g) and washed with 35 to 40 ml of methylene
chloride.
219
The filtrate was next evaporated to near dryness under a dinitrogen
stream and then lypholized to remove any methylene chloride. The residue
was suspended in d6-benzene and filtered through a glass wool plug directly
into an NMR tube.
Product Analysis.
To measure the radioactivity of the samples, a 1 gl aliquot of the
product solution was diluted in 200 l1of methanol before being placed in a
liquid scintillation counter. NMR analysis was carried out with a 300 MHz
Bruker spectrometer at NTLF. All 3 H spectra were 1 H decoupled.
Results.
Analysis of Lypholysates. The possible alcohol products of the reaction
of (S)-[1-2H,1-3 H]-ethane with MMO are depicted in Figure 3. Their actual 1H
and
3H
NMR spectra are given in Figures 4 and 5, respectively.
spectrum shows a large water peak at 4.82 ppm.
The 1 H
Since only = 1 mole of
product was formed (mM ethanol expected in 1 ml total volume), the carrier
alcohol dominates the 1 H spectrum. Accordingly, the signals at 6 = 1.20 ppm
(t, J = 6 Hz, 3H) and 6 = 3.67 ppm (q, J = 6 Hz, 2H) were observed for ethanol. A
minor signal at 6 = 3.38 ppm is currently unassigned.
The 3 H NMR spectrum of the reaction products in Figure 5 contains
four major
resonances
ppm. The resonance at
at 6 = 1.19 (s), 6 = 3.65 (s), 6 = 3.67 (s), and 6 = 4.84 (s)
= 4.84 ppm was assigned to H 3HO, and the singlet at
= 1.19 ppm to the product of oxidation occurring at the unlabeled carbon
atom (5 in Figure 3). The inset in Figure 5 shows an expansion of the two
signals at
= 3.65 and 3.67 ppm. The narrower, more downfield resonance is
attributed to 1 and 4 (H,T products), whereas the broader, more upfield signal
is assigned as arising from 2 and 3 (D,T products).
The broadening of the
latter signal is due to unresolved coupling to the deuterium atom present in
220
the molecule.
A kinetic isotope effect can be calculated from the ratio of
alcohol products for hydroxylation of the C-H bond (2 and 3) versus products
hydroxylated at the C-D bond (1 and 4). Taking the ratio of integrals at 3.65
and 3.67 ppm gives a kH/kD value of 0.7, which should be considered
approximate since the peaks are not baseline resolved. The sum of the areas
of these two peaks can be compared to the area at
= 1.19 ppm to obtain
information about the relative hydroxylation rates at the two different carbon
atoms. The observed integrations indicate that oxidation at the unlabelled
carbon atom was favored by a factor of 3 over oxidation at both the proton and
deuteron of the labeled carbon. This value may reflect the relative reactivities
of the C-H versus C-D bonds under the assumption that once bound to the
active site, ethane is free to rotate and present either of its carbon atoms to the
active hydroxylating species. Reaction at the tritium atom will not be
revealed in the 3 H NMR spectrum, but is expected to be very slow. These
experiments were carried out under Vmax conditions, in which a saturating
amount of substrate was added to the enzyme, which could lead to an
artificially low intermolecular isotope effects, however.
Analysis of Derivatives.
The derivitization
reaction, illustrated
in
Figure 6,30 can afford five 3 H NMR resonances, two sets arising from
diastereomers and one from another product. These peaks are shown in
Figures 7 (whole spectrum) and 8 (expanded version). The products arising
from stereochemical inversion are labeled A and D, those with retention are
labeled B and C. The loss of stereochemistry, calculated by adding the areas of
peaks A and D and dividing by the combined areas of all four peaks, was 14.5
/o (85.5% retention).
The values corresponding
to products obtained by
hydroxylation at the C-H bonds versus hydroxylation of the C-D bonds were
221
39%/0
(69 % retention) and 7% (93 % retention), respectively. The reasons for
separating the products in this manner will become clear shortly.
The
3H
NMR spectrum of the alcohol products
formed by
hydroxylation of (S)-[1-2 H,1- 3 H]-butane with MMO is given in Figure 9. The
spectra of corresponding mandelate derivatives are given in Figures 10
(whole spectrum) and 11 (expanded version). The 3 H spectrum of the
mandelate derivatives of alcohols formed with (S)-[2-3 H]-butane is given in
Figure 12, and the spectrum of derivatized products formed by reactions with
a racemic mixture of this substrate is shown in Figure 13. All assignments are
indicated on the spectra.27 Table 2 summarizes the results for two trials of the
ethane hydroxylations, Table 3 reports results for reactions with (R) and (S)-[12 H,1- 3 H]-butane,
and Table 4 lists the findings for butane substrates labeled at
the C2 position.
When the results in Tables 2 and 3 are analyzed according to the
several categories of C-H bond hydroxylation, several interesting clues about
the hydroxylation mechanism of MMO emerge. For each ethane enantiomer
hydroxylation can take place either at the hydrogen atom (D,T products) or
with the deuterium atom (H,T products) of the chiral carbon. Products that
arise from hydroxylation of the C-T bond are not detected since 3 H NMR is
the analytical method.
From the relative amounts of C-H versus C-D
hydroxylation products were calculated the intramolecular kinetic isotope
effects, kH/kD (Tables 2 and 3). The 3 H NMR data also afford information
about the distribution of hydroxylation at each carbon atom of the substrate.
These ratios of products, C2/C1 and C4/C1 for ethane and butane, respectively
are also included in the Tables.
As indicated in Table 2, hydroxylation
of (R)-[1-2 H,1- 3 H]-ethane and (S)-
[1-2 H,1- 3 H]-ethane gave very different results. With (R)-[1-2 H,1- 3 H]-ethane,
222
the kinetic isotope effect calculated from the 3 H NMR of the alcohol products
was kH/kD = 1.1. This value agrees with that obtained (0.9) from adding the
integrals of peaks B and D and dividing by the sum of A and C in the
spectrum of the mandelate derivative, shown in Figure 8. The corresponding
numbers for (S)-[1-2 H,1-3 H]-ethane hydroxylation differ, however. From the
alcohol products, kH/kD = 0.2 and 0.7, in two runs, whereas from the
derivatives, values of 0.5 and 0.1 were obtained.
Large ratios of products
formed by hydroxylation at C2 vs C1 ranging from 3.5 to 4.7, were observed
with both enantiomers.
The amount of inversion for the (R) and (S)-[1-2 H,1-3 H]-ethanes also
varied.
For (R)-[1-2 H,1-3 H]-ethane, almost racemic and even inverted
mixtures of the derivatives were observed, as indicated in Table 2. Products
arising from hydroxylation of the C-H bond (D,T products), however, showed
83 and 77 % retention. Reaction at the C-D bond (H,T products) of (R)-[1-2 H,13 H]-ethane
gave predominantly inverted products, 80 and 93% inversion.
These latter values skew the summed (D,T and H,T) distribution of products
such that they appear to be almost racemic. For (S)-[1-2 H,1- 3 H]-ethane,
different behavior is again observed (Table 2). Analysis of the D,T products
reveals 69 to 100 % retention, and the H,T products show 68 to 93% retention.
This strange behavior
is also reflected in the results for MMO
hydroxylation of (R)-[1-2 H,1-3 H]-butane and (S)-[1-2 H,1-3 H]-butane (Table 3).
A kinetic isotope effect of kH/kD = 1 was computed for (S)-[1- 2 H,1- 3 H]-butane
hydroxylation, whereas a value of < 0.2 was calculated from the alcohol and
derivative spectra for hydroxylation with (R)-[1-2 H,1-3 H]-butane.
The
stereochemical results for the D,T products showed 66 % retention with (R)[1-2 H,1- 3 H]-butane and 51 % retention with (S)-[1-2 H,1-3 H]-butane. The H,T
223
alcohols gave 93 % inversion with (R)-[1-2 H,1- 3H]-butane and 51 % retention
with (S)-[1-2H,1-3 H]-butane.
Retention of stereochemistry was primarily seen with both (R)-[2-3 H]butane and (S)-[2-3 H]-butane (Table 4). Products obtained from a racemic
mixture of (R)-[2-3 H]-butane and (S)-[2- 3 H]-butane showed only a slight
preference for hydroxylation of one enantiomer over the other (Figure 13 and
Table 4). The product ratios were 1: 1.4. More interestingly, oxidation at the
C-3 position, which leads to diastereotopic products, proceeded in a
stereoselective manner. One product was formed in 2.5 fold excess over the
other, as shown in Figure 14.
Discussion.
Analysis of ProductDistributions.
Alkanes Are Not Highly Constricted in the Active Site Prior to
Hydroxylation. The reaction products obtained with racemic [2-3 H]-butane
indicate clearly that the active site does not rigorously constrain these
substrates, since there was only a slight discrimination between the two
enantiomers of this probe.
An extremely crowded active site would most
likely readily distinguish an ethyl and a methyl substituent at the reacting
carbon atom.
The products of oxidation at the C3 position using racemic
substrate, however, show some stereoselectivity. Two diastereotopic products
were observed, the ratio of one to the other as detected by
3H
NMR
spectroscopy being 2.5: 1. This behavior indicates that there is some degree of
chiral discrimination conferred by the active site, but at the 3- rather than the
a-carbon.
One possible
explanation
for the different product
alcohol
stereochemistries observed for hydroxylation of the two enantiomers of [12 H, 3 H]-ethane
or [1-2 H,3 H]-butane is that the substrates are stereochemically
224
cdistinguished in binding to the active site of the hydroxylase.
For example,
the (R) enantiomer of [1-2 H,3 H]-ethane might be specifically oriented in such
a manner that chiral discrimination between the tritium and deuterium
groups could take place prior to C-H bond hydroxylation.
Such an effect
would require restricted orientation of the methyl group with respect to the
diiron center. Similar chiral discrimination was seen with the much bulkier
radical clock substrate probe trans-2-phenylmethylcyclopropane,
25
as described
in Chapter 4.
A example of how such substrate binding prior to hydroxylation by
MMO might be manifest is given in Figure 15. If such a chiral discrimination
were taking place to account for the differences in the product distribution
between the (R) and (S) enantiomers of [1-2 H,3 H]-ethane, one would expect
the butane enantiomers to behave similarly, as illustrated for (R)-[1-2 H,1-3 H]butane in Figure 16. The methyl, deuterium, and tritium substituents of the
reacting carbon for chiral ethane would be expected to bind in the same
positions as the analogous groups of the reacting carbon atom of butane in
such a tightly constrained site.
Accordingly, similar alcohol product
stereochemistries would be expected for chiral butane as for chiral ethane.
The total (R) and (S) butane products do not, however, follow similar
stereochemical patterns, (cf. Tables 2 and 3) suggesting that this explanation is
not viable, in agreement with our interpretation of the [2-3 H]-butane results.
H,T vs D,T Product Stereochemistries and Hydrogen Exchange at the aCarbon of the Product Alcohols. Although the total amounts of retention
and inversion for the two enantiomers of chiral ethane versus chiral butane
are very different, further analysis reveals interesting trends in the data. The
(R) probes lead predominantly to retention when the reaction yields D,T
products and to inversion for most of the alcohols in the H,T pool. With the
225
(S) ethane probes, predominant retention occurs with both D,T and H,T
products. For (S) butane products, however, the products are nearly racemic
in both the D,T and H,T pools.
Although the extent of retention in the H,T and D,T product pools for
each substrate enantiomer was initially confusing, it can be rationalized by
postulating the occurrence of an exchange process in the product alcohols
which masks the true stereochemistry of the products formed in the
hydroxylation reactions. In this model, hydroxylation occurs leaving the
product molecule bound to one or both (Figure 17) iron atoms in the active
site. Alternatively, this same species could arise from binding of the released
product alcohol to a different hydroxylase molecule. A reaction in which a
protein residue exchanges hydrogen with the bound product in a
stereospecific manner might then occur. This exchange reaction must take
place stereospecifically in order to account for the observed results. Both the
extent of inversion for each enantiomer, and the composition of the H,T and
D,T product pools, can be rationalized by the extent to which such an
exchange process occurs.
As indicated below, the observed product
distributions can be accounted for by the relative binding affinities of the
product alcohols (vide infra). Product binding to the oxidized diiron site has
been demonstrated for phenol with hydroxylase from M.
trichosporium
OB3b.31 An exchange process has been suggested following hydroxylations by
cytochrome P-450.3 2
Stereochemical scrambling of product alcohols which have fully
retained or fully inverted their stereochemistry can be explained by such an
exchange mechanism.
Figures 18 and 19 show the effects of exchange on
products formed with retention and inversion for the substrate (R)-[1-2 H,13 H]-ethane.
Exchange of hydrogen in a stereospecific manner would lead to
226
the following results.
Consider the H,T (Figure 18) and D,T (Figure 19)
products formed initially. For one H,T product, exchange leaves the product
stereochemistry unmodified.
For the other, tritium is lost in the exchange
and the product is therefore not observed.
Exchange of one D,T product
similarly leads to an unobserved alcohol. The second D,T product, however,
is converted to an H,T product of the same configuration upon exchange!
This H,T product signifies an apparent inversion of configuration, even
though the hydroxylation step occurred with retention.
This last case
demonstrates how a stereospecific exchange process can lead to erroneous
conclusions about the nature of the hydroxylation step.
The quantitative
effects of such an exchange mechanism on the
product stereochemistries as well as on both the intramolecular kinetic
isotope effects (kH/kD, at C1) and C2/C1 product ratios can be computed. In
Table 5, four sets of such calculations are presented. In case A, hydroxylation
of (R)-[1-2 H,1- 3 H]-ethane is considered to occur with complete racemization
and an intramolecular kinetic isotope effect of kH/kD = 4. The corresponding
amounts of species 1 through 4 are listed for no subsequent exchange, with
50% exchange, and with 75% exchange of the product alcohols. The effects of
these exchange reactions on the various categories of H,T products, D,T
products,
C2/C1 ratio, and kH/kD are also tabulated.
A second set of
calculations was performed (case B) for the same substrate and again taking an
intial kH/kD value of 4, but assuming that the hydroxylation step occurs with
33% inversion.. The last two sets of calculations were made with (S)-[1-2 H,13 H]-ethane
and the same initial assumtions (cases C and D).
These
calculations illustrate how product exchange can skew the distribution of
products and lead to the observed stereochemistry. The effects on the kinetic
isotope effects are discussed below.
227
The low specific activity ( 40 mU/mg) observed for the hydroxylase
under conditions required to produce the large quantities of product needed
for 3 H NMR studies, = 1 mmole/1 ml, compared to the = 0.1 mmole/1 ml,
amount used under normal assay conditions, where SPA = 250 mU/mg,
suggests strong inhibition of the enzyme by the product alcohols.
These
species have a significant binding affinity for the active site which enables
them to compete with substrate.
Since the product alcohol binds well, it is
reasonable to suggest that it may bind with a long enough lifetime for an
exchange reaction to take place. No evidence for exchange was suggested in
the results for the radical clock product alcohols, which probably reflects their
relatively poor binding affinity at the active site. On the other hand, very
recent results from our laboratory3 3 have revealed that, when C2D6 is used as
substrate, a substantial amount of C2 D 4 HOH forms. Control reactions in
which hydroxylase was omitted from the reaction mixture failed to show
exchange of CH 3CD2OH. Thus an exchange mechanism such as that depicted
in Figures 18 and 19 appears to occur, and is dependent on the presence of the
hydroxylase component.
Kinetic Isotope Effects.
Figure 20 illustrates how the C2/C1 hydroxylation can depend upon the
magnitude of the intramolecular kinetic isotope effect at C1, referred to as
kH/kD.
The ability of product exchange to influence both the C2/C1 and
kH/kD values from a reaction with (S)-[1-2 H,1- 3 H]-ethane is presented in
Figure 21. Products arising from hydroxylation at the C-H bond are
designated as "kH" alcohols, whereas reactions at the C-D bond are labeled
"kD" alcohols. For both "kH" and "kD" alcohols, the tritium label can be lost,
affording undetectable products. Exchange of "kD" alcohols at their C-H bond
will have no effect since they are replaced with C-H bonds. Exchange of "kH"
228
products, however, will convert C-D bonds into C-H bonds, resulting in
species scored as "kD" products. Exchange thus depletes "kH" products while
increasing the pool of "kD" alcohols, resulting in a decrease in the apparent
kH/kD value. The exchange process will also decrease C2/C1 in the following
manner.
The sum of "kH" and "kD" alcohols represents C1 products.
Both
sets of these alcohols lose tritium upon exchange (Figure 21) depleting the
pool of C1 products detected. Exchange will have no effect on C2 products,
since this carbon only has C-H bonds and the label at the second carbon atom
will remain intact. Since C1 is diminished and C2 was left unaffected, the
value of C2/C1 will increase after exchange. The calculations presented in
Table 5 quantitatively show how the exchange process can artificially decrease
the measured kH/kD, while simultaneously increasing C2/C1 from predicted
values with a given kH/kD for the primary hydroxylation step.
Another factor that may contribute to the high C2/C1 value should be
considered, however. It is possible that the hydroxylating species in MMOH
has equal access to hydrogen atoms on both the C1 and C2 positions of chiral
ethane. In this case, the ratio of C2/C1 products would reflect the relative
ability to hydroxylate C-H versus C-D bonds, as for an intramolecular isotope
effect. Previous studies comparing R-CH3 with R-CD3 alkanes have yielded
intermolecular kinetic isotope effects of = 1.25,34 The substrates employed in
these studies were significantly larger than ethane or butane, and contained
either a tertiary or quaternary carbon atom
to the reacting methyl group.
Such highly substituted carbons react much more slowly with MMO than
primary C-H groups, and products corresponding to reaction at the a position
were not detected. It is possible that, with chiral ethane, rotation occurs in the
active site rendering both ends of the molecule accessible to the active
hydroxylating species. With chiral butane, reaction at C4 instead of C1 would
229
occur less often than the reaction of C2 versus C1 in ethane, accounting for
the lower values of C4/C1, = 3, compared to C2/C1 values = 3.5 to 4.4 for the
ethanes.
Preorientation
of the substrate at the reactive iron center may also
affect kH/kD. For example, the C-H bond of the reacting carbon atom might
interact with the reactive hydroxylating species preferentially over a C-D or a
C-T bond.
Mechanism A in Figure 22 suggests one such alignment for an
agostic interaction with the Fe atom adjacent to a ferryl species. Orientation
of the C-H bond in this manner exposes the C-D or C-T bonds preferentially
for the hydroxylation reaction. Such behavior could account for the inverse
isotope effect observed in some cases, but does not account for the different
D,T and H,T products stereochemistries.
Secondary kinetic isotope effects
from a substituent adjacent to the reacting atom in the hydroxylation step
could also play some role in determining the stereochemistry of the product,
but these effects are expected to be very small.
Mechanistic Implications.
We now inquire whether, given the foregoing analysis, the
experimental results can be accounted for if the hydroxylation step were to
occur with complete retention or complete inversion of stereochemistry
followed by exchange of protium for protium, deuterium,
or tritium, as
indicated in Figures 18 and 19. Even for cases where retention or inversion is
extremely dominant, such as > 90 % retention, an exchange mechanism alone
is not sufficient to explain all of the measured product distributions.
23 shows the four tritiated alcohols that can be observed by
3H
Figure
NMR
spectroscopy. Products C and D represent inverted species. If the alcohol were
formed initially with 100 % retention, meaning that only A and B are
produced, then conversion of these products to C and D must occur. Given
230
the restrictions on the exchange process outlined above, no pathway for
conversion of A or B into D can be found. Since this product is detected in
the 3 H NMR spectrum of the derivative made from the product of (S)-[1-2 H,13 H]-ethane
hydroxylation, it must be present in the product alcohols. All
manipulations, including the derivitization procedures, are not expected
scramble the stereochemistry. 2 7
Given the above analysis, it seems likely that there exists some radical
or possibly carbocation character in the hydroxylation step to account for all
the observed products.
The extent of such involvement seems to vary
between enantiomers of a given substrate and even between the substrates
themselves, as reflected in the different stereochemical results.
observation has several implications.
This
There could be two parallel reaction
pathways, one involving a substrate radical and one not.
Two possible
pathways are presented in Figure 22. The degree to which a radical vs. non-
radical mechanism is followed may be determined by how snugly the
substrate fits into the active site.
Parallel mechanistic pathways have
similarly been invoked in porphyrin model systems to explain the
stereochemical consequences of hydroxylation chemistry depending on steric
interactions of the substrate with the catalyst.3 5 -3 7 Alternatively, abstraction
of a hydrogen atom to form a substrate radical may occur in every reaction.
The rate constant for a rebound reaction, such as that presented in
mechanism B of Figure 22, may be so large that only a small quantity of the
substrate radical has a sufficient lifetime for rotation about the C-C bond to
occur before recombination with the bound hydroxyl group. Results with
radical clock substrate probes were consistent with either no substrate radical
formation or an extremely fast rebound step having a rate constant on the
order of 1013 s1. A rebound rate constant in the experiments reported here
231
cannot be accurately estimated since the occurrence of exchange distorts the
distribution of products formed in the actual hydroxylation step.
Comparisons with Literature.
Very different stereochemical behavior was observed in the present
experiments with MMO from M. capsulatus (Bath) compared to that found
with enzyme from M. trichosporium OB3b.2 2
With the latter MMO the
products were consistently found to have = 2:1 retention vs inversion of
sterechemistry observed for both substrate enantiomers and for both ethane
and butane substrates. Our work with the radical clock substrate probes
implied a rebound rate constant of = 1013s-1 for MMO from M. capsulatus
(Bath) and 3 - 6 1012 s- 1 for MMO from M. trichosporium OB3b.2 5 It is possible
that hydroxylation
with the M. capsulatus (Bath) MMO system has more
concerted than radical character.
Alternatively, it may display a faster
rebound rate compared to that in M. trichosporium OB3b system.
Both
explanations could account for the stereoselectivity of = 90 % retention found
in some cases with the M. capsulatus (Bath) enzyme.
Because some of the
experiments with the M. trichosporium OB3b enzyme were carried out with
identical batches of substrates reported here, it is unlikely that the different
results are due to variations in the substrates or analytical procedures.
Intramolecular kinetic isotope effects of = 4 with these substrates are
consistently observed with MMO M. trichosporium OB3b.2 7
With MMO
from M. capsulatus (Bath), low kH/kD values arise due to exchange. It is
possible that an analogous reaction does not occur with the M. trichosporium
OB3b enzyme. The intramolecular kH/kD values of = 4 agree with the figures
obtained from the radical clock substrate work with M. capsulatus (Bath).2 5 In
those experiments, an intramolecular isotope effect of = 5 was interpreted as
indicating
that substantial
C-H bond breakage is involved in the
232
hydroxylation reaction step. With cytochrome P-450 hydroxylations, kH/kD
values of 7 to 10 have been reported. 15 The isotope effect is expected to be 2 1,
the precise value depending on the linearity of the 0 ...H-C unit in the
transition state. 38 Alternative explanations for an inverse intramolecular
kinetic isotope effect in the hydroxylation reaction are not obvious. The large
C2/C1 regioselective effects observed here, ranging from 3 to 4, relative to
intramolecular isotope effect literature values, = 1, are also explained by the
exchange reaction and the access of the cacarbon to the hydroxylating species.
Concluding Remarks.
Although the foregoing interpretations are reasonably satisfying, some
mysteries remain. The relatively large quantity of H 3HO that is produced
cannot be explained by exchange alone. In some instances, this species is the
largest peak detected in the 3 H NMR spectrum of the reaction products.
Some H 3 HO could be generated during substrate synthesis, since relatively
large quantities of it have been detected in other work at the NTLF.3 9 The
C2/C1 regioselective effects of = 3 to 4 were absent in work with the radical
clock substrate probes.
This large value suggests that the apparent
regioselectivity of the reaction is not a good indicator of the intermolecular
isotope effect. Instead, the substrate may have the ability to reorient itself
preferentially in the active site prior to hydroxylation. Experiments to test the
exchange reaction, by mass spectral analysis of the products with CD3CD 3
hydroxylation, and by repeating some of the reactions with the tritiated
substrates are in progress by coworkers in our laboratory.
The results of the work described in this chapter have raised some
interesting questions about the hydroxylation mechanism. One very useful
conclusion, however, is that MMOs from M. capsulatus (Bath) and M.
trichosporium OB3b do exhibit different behavior, confirming variations seen
233
in experiments with radical clock substrate probes reported in Chapter 4 and
redox titrations in Chapter 3.
Determining the extent to which, and
understanding why, the enzyme systems vary are topics for future work.
234
References
(1)
Anthony, C. The Biochemistry of Methylotrophs; Academic Press: New
York, 1982, pp. 296-379.
(2)
Nguyen, H.-H. T.; Shiemke, A. K.; Jacobs, S. J.; Hales, B. J.; Lidstrom, M.
E.; Chan, S. I. J.. Biol. Chem. 1994, 269, 14995-15005.
(3)
Dalton,
H.; Leak, D. J. In Gas Enzymol.; al., H. D. e., Eds.; D. Reidel
Publishing Co.: London, 1985; pp. 169-186.
(4)
Colby, J.; Dalton, H. Biochem. J. 1978, 171, 461-468.
(5)
Fox, B. (G.; Lipscomb, J. D. Biochem. Biophys. Res. Comm. 1988, 154, 165-
170.
(6)
Fox, B. G(.;Froland, W. A.; Dege, J. E.; Lipscomb, J. D. J. Biol. Chem. 1989,
264, 10023-10033.
(7)
Colby, J.; Dalton, H. Biochem. J. 1979, 177, 903-908.
(8)
Green, J.; Dalton, H. J. Biol. Chem. 1985, 260, 15795-15801.
(9)
Green, J.; Dalton, H. J. Biol. Chem. 1989, 264, 17698-17703.
(10)
Dalton, H. In Adv. Appl. Microbiol.Academic Press: 1980; Vol. 26; pp
'71-87.
(11)
Colby, J.; Stirling, D. I.; Dalton, H. Biochem. J. 1977, 165, 395-402.
(12)
Rosenzweig, A. C.; Frederick, C. A.; Lippard, S. J.; Nordlund,
P. Nature
1993, 366, 537-543.
(13)
DeRose, V. J.; Liu, K. E.; Kurtz, D. M., Jr.; Hoffman,
B. M.; Lippard,
S. J. J.
Am. Chem. Soc. 1993, 115, 6440-6441.
(14)
DeWitt, J. G.; Bentsen, J. G.; Rosenzweig, A. C.; Hedman, B.; Green, J.;
Pilkington, S.; Papaefthymiou,
G. C.; Dalton, H.; Hodgson, K. O.; Lippard, S. J.
[. Am. Chem. Soc. 1991, 113, 9219-9235.
235
(15)
Ortiz de Montellano, P. R. In Cytochrome P-450 Structure, Mechanism,
and Biochemistry; Ortiz de Montellano, P. R., Ed.; Plenum Publishing Corp.:
New York, 1986; pp 217-271.
(16)
McMurry, T. J.; Groves, J. T. In Cytochrome
P-450
Structure,
Mechanism, and Biochemistry; Ortiz de Montellano, P. R., Ed.; Plenum
Publishing Corp.: New York, 1986;pp 1-28.
('17) Mansuy, D.; Battioni, P. In Activation and Functionalization of
Alkanes; Hill, C. L., Eds.; Wiley: New York, 1989; chapter VI.
(18)
Guengerich,
F. P. In Biological Oxidation Systems, Vol 1; Reddy, C. C.,
Hamilton, G. A. and Madyastha,
K. M., Eds.; Academic Press: San Diego, 1990;
pp 51-67.
(19)
Ortiz de Montellano,
P. R.; Stearns, R. A. J. Am. Chem. Soc. 1987, 109,
3415-3420.
(20)
Bowry, V. W.; Ingold, K. U. J. Am. Chem. Soc. 1991, 113, 5699-5707.
(21)
Bowry, V. W.; Lusztyk, J.; Ingold, K. U. J. Am. Chem. Soc. 1991, 113,
5687-5698.
(22)
Priestley, N. D.; Floss, H. G.; Froland, W. A.; Lipscomb, J. D.; Williams,
P. G.; Morimoto, H. J. Am. Chem. Soc. 1992, 114, 7561-7562.
(23)
Ruzicka, F.; Huang, D.-S.; Donnelly, M. I.; Frey, P. A. Biochem. 1990, 29,
1696-1700.
(24)
Rataj, M. J.; Kauth, J. E.; Donnelly, M. I. J. Biol. Chem. 1991, 266, 18684-
1.8690.
(25)
Liu, K. E.; Johnson, C. C.; Newcomb, M.; Lippard, S. J. J. Am. Chem. Soc.
1993, 115, 939-947.
('26) Pilkington, S. J.; Dalton, H. In Methods In Enzymology Academic Press:
New York, 1990; Vol. 188; pp 181-190.
('27) Priestley, N.; Wilkinson, B.; Floss, H., unpublished results.
236
(28)
Andres,
H.; Morimoto,
H.; Williams,
P. G. J. Chem. Soc., Chem.
Commun. 1990, 627-628.
(29)
Abbreviations:
hydrofuran;
TMEDA = tetramethylethylenediamine;
MOPS = N-morpholinopropane
dimethylaminopyridine;
THF = tetra-
sulfonic acid; DMAP = 4-
DCCI = dicyclohexylcarbodiimide, SPA = specific
activity.
(30)
Parker, I:). J. Chem. Soc. Perkin Trans. 2 1983, 83-88.
(31)
Andersson, K. K.; Elgren, T. E.; Que, L., Jr.; Lipscomb, J. D. J. Am. Chem.
Soc. 1992, 114, 8711-8713.
(32)
Newcomb, M., personal communication.
(33)
Valentine, A. M.; Lippard, S. J., unpublished
(34)
Choi,
S.-Y.; Eaton,
P. E.; Hollenberg,
Newcomb, M.; :Put, D. A.; Upadhyaya,
results.
P. F.; Liu, K. E.; Lippard,
S. P., manuscript in preparation.
(35)
Groves, J. T.; Viski, P. J. Org. Chem. 1990, 55, 3628-3634.
(36)
Groves, J. T.; Viski, P. J. Am. Chem. Soc. 1989, 111, 8537-8538.
(37)
Groves, J. T.; Stern, M. K. J. Am. Chem. Soc. 1987, 109, 3812-3814.
(38)
O'Ferrall, R. A. M. J. Chem. Soc. (B) 1970, 785-790.
(39)
Morimoto, H.; Williams, P. G., personal communication.
S. J.;
237
Table 1. Component Amounts Used In Enzymatic Reactions at NTLF. Total
reaction volumes were 1 ml.
25 mM MOPS, NADH (gl,
Protein B/
Substrate
(IS)-[1-2 H,1- 3 H] -
Hydroxylase a
Reductase
pH 7 buffer
(mg)
(gl)
(l)
13 (batch 1)
280 (batch 1)
500
50
13 (batch 1)
280 (batch 1)
500
50
1)
280 (batch 1)
500
100
15 (batch 2)
300 (batch 2)
400
100
25 (batch 1)
560 (batch 1)
200
100
25 (batch 1)
560 (batch 1)
200
100
24 (batch 2)
450 (batch 2)
200
100
24 (batch 2)
450 (batch 2)
200
100
24 (batch 2)
450 (batch 2)
200
100
24 (batch 2)
450 (batch 2)
200
100
0.1 M stock)
ethane
ethane
(R)-[1-21,1-3H]
-
13 (batch
ethane
(S)-[1-2 H,1-3 H] -
ethane
(R)-[1-2 H,1-3 H]--
butane
butane
(R-[2-3 H]-
butane
butane
butane
[2-3 H]-butane,
racemic mixture
aSPA (propylene) = 200 mU/mg batch 1; 224 mU/mg batch 2 under usual
conditions. SPA's of both batches under conditions to produce 1 to 2 mM
product/ml was 40 mU/mg. Fe content ranged from 2.5 to 2.7 Fe/mol
hydroxylase.
238
C:
C)C)
co)
C)
xD
O
Ccr
O
C
Cr) N
0o N.
C
C)
0
N-
m
CN
-
crUCo
-
cc
__
(i
O
o,,
In'
0Lt)
Cl
05
_
u.a _
N
CD
00
L)
N
C
CD
O
Cr)
v
6
-J
SZ z
Q'x
cU
L,
O
a
C
o
ci
C
[Cc
Cl
N
C>
"t
)
a
8>
ci
c
(:C
;
EU
O
r'
-4V
C)
cn
C)
cc
-c
C)
cn
cnn
I
I
.
: '"
I
239
u-so
LC
Lf)
o
.4 sC
c
J
L)
CIO
LQ
L"
Lr)
00
O
O
C)
&
-
Uj
K
K
K
O
-
C
-
V
dq
CI
C
t3z
UrI
r
O~
Lr)
L'x
C
cr.
,C)
L.
cro
-0
cr,
C, 72
::l cs
d:,
240
U
uLn
\o
4
oC
4t
\
CO
cl
cl
CN
C
U
N
cn
U
ru
-
c.
,
. E
C
u
_m
,
r;
.-
cn/
CZ
v C
-
Ot
uCd
Ol CO
Lr)
-E
cl,
c
v
c
3 _
aO 'd
°u
L-
Ln
C4
0
r-
cn
m
m
r
;.o
U
a
*
,
Cl
Cl
m~
C1
=j
_C
:;
<
--
wD
cci
I
Cc,
1
1:4
U)
C1
"a
241
U
U
C
C)
U
U
C~
C,4
-z C+
~z
Cos
CO
cC
+
c,
Co
O
C)
\c
C
Nd
\.6
C4
Co;
cl~
LC
Cl
Co
\D
LN
Co4
.C
co
M
N
Co
c
Co,
Cc" Co
nj
CD
d
Cl
+
-L
Cl4
N
o
ka)
C)
0
o
.4,
LC)
-
F~4
C)
o
Z
C)
-4
o
o
O
o
0--r
CL)
F-
n
'
K
;4
L cl
a0
t
C
C)
4..
C)
cn
u
U
z
3:
K
0
C'I
2
Cl
II
x
11
o
8
4._o
b
C)
U c.
UO
I3-
C)
.
U
CC
0;3-4
u
4T
C)
~4
E
0
-J
u
Co
n
bcaN
U
He
a
N
=5
Ct
C:
Co
CC~
a
0
C)
C)
K0C)
L4
o
CC
0d
CC
-4
Krf
oC)
u
Cl4
a
0II
C7
o
Cn
Lf)
Lr)
Ca
a)
4-4
to
,
cc
2
C"
"'
711
r-I
CC
X
U
xC)
0
Hr_
X
bo
C
C)
CJ
a)
bZ
CC
Cl
t-
U
.
<
7-'
x
x
x
C)
.
rl4
o
iC
n
242
c,.)
U
,
CI
u
C)
C)
c
U
n
c)
-a
C)
-
IC)
N
Cl
CD
-
-a
,I
v
-c
+
cl
Cl
0
to
C
CL
C
O
"ZI,
O
CIA
O
C)
O-1CD -CD
O
-C
"ZI
'cn
v
C,
-
,.
V4
II
C4
c~
a
O-
b-.
;H
C
CD CD
u
Lt-
cG
04
CJ
c
L
,..
;-
co
00
c~l
~C0
U
c5
C)
Ln
Cr,)
I
Ar
075
mP
NC Cr
C
CIc
clcm
,I-,
N.
0
z5
3 H
©
-z
U
rc
N
0
0
Cl
o
o
II
0
Cl
Cn
CC)
1C
Crc
Cl
li
-zfE
C)
C-
0
-'
LnCD
o
0
m
)
Cl
CC)
C
C
C)
C
Cr,
C)
'-4
cr
U
4
,I:
C)
0EcS
O,-
.)
°o
-J
LnC U]~
,<
.
a0ok
r:
Cr)
Cr)
a
m
Cr)
C
0
a)
H
Es
-u
V1
E
CI
©
C)
U
Ul)
cn
o o o
I-
C
Cr)
C)
C)
C)
0n
0
C;
r,
u
x
C)
0
4
C)
X
0
U
C)
X
a
o
UL)
L)
N
UZ
c
c
243
E
cn
E
5Cu'
E
M
o
C) C.
T~
Z
O
U
E
._
ct
o
a) a
oC(a
C
o
0
0
O
r-
r0
.
"
0
CZ
0
d .,-
o
244
U
/0
x
\
en
U
0
z0
\
P-
)I
/
W-4
A)
ni
0
01
.0
w
0
14
o.
.2
00
,a,
U
e
uen
XZ
m
u
W-
.:Nm
CI
Ui
x/
o0=/V5
H
Crz
245
5
"_
0
u
E
O
d
-z2
EZ
En
c:'
C
0
u
4,
"
£
O
U
o
&0
o
CZ
5H
.,
-,
50
.-
0
-,
5
bC7
-o
-0
r3
O
Cd
cd
k)
D
V
G
u
5
C.
0
C)
©
UD
5
X
U
c:;
0JC)
-,
C)
.,
Crd'
Z
.a
4-,
-z
Z;
c3
u
._
C)
Q
0
U)
C._
0
-z
CU)
,
c5
C3
C
ed
a0
-
.-C)
C)
r-
Cd
._
E
._
.
C3)
CI
0.
;
cs4
246
247
as
o,
is
cu
UM
C)T1
._
0
I
Cd3
o
o
Q
u
.°
sq
a,
k
cU
o
o0.
td
~3
248
W-
0
0
0
0
,
- 0
?^, *
o
es,
V-'
c
C
I
0
O
enr
-mI
m
r*
4!
t-m
o,,.,~
Nre -
'
4
-
94
un3
U)
0
U
Ue.:
X=
F
(z!
I
r-1
z
cen
U
\...
N
I
r 1
6-r'J
m
249
a)
H
CA
,cr
Ct
!
C)
o._
CO
0
CO
w"
©
©
ct
Ca
c,
C)
c1
a)
o
C)
U)
otli0o
-o
0
S
cn
U
©
C)
Cz
C
cCd
,
,-a
k)
c,0
Ik
as
U
C)
250
t
I
L
II
r
cl
L
r
Co
33~1 ~
0A
tM
=
O
CIA
L)
0r
J
Ve
251
,
-O
U
Q
0
o
,
CO
CZI
0
O
o
c
.C
~c
UO
-a
-a
CC
>, 4
C
252
tO
-
0
0
U2:
U-
I
0
1)
253
o
N
6
0
C)
cr
0
C.)
C7
co
4
254
W
-W
m
... 4
W
"O
9
m
Qj
.1-14
-W
m
7-4
5 .;*
0
"..4
>Iz
w
w
1I
U
o
NO
u
t14
'
O
u
OI
+
-4
o
!!z
0
l
I
r,
cH
C-i
W
'IO
255
C:
CO
0C"
CN
cn
.
C
U
ca
C)
-4
c)
O0
C-)
C
ci3
C)
-3
.)
to
U
._
256
0
c
OO
E3il:
I~J/
O
ru~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
o
0
o
Cm
= Cd
*
F
257
0
0
0
Cr)1
co
o
I
Cd
S
o
-d
C)
a)
c,
aD
o
C)
-z
a
N
._
._
t.
or
C)
c)
r
©
UC
CI
r,c
o
75
NI
C)
c
o...
;E
ca,
©
o,.
o,-
"U
rd
CZ
t'l
Cf)
rl
C)
C)
C
-0
cu
258
VH
CI,
C
U
.IL
no
O
0
0
0
Uo
0
U
0
+
d
+
L,
*
*
r
259
&
Or
0
0cl
I
Q
t3
©
;
0
r(
4,
3=
©
-O
en
a:
C
0
©
4..a
c
m,
._c:
E*G
G
Q
-W
aC
C
4-,
bC
:
o1.
.
CT-T
260
(U4
N
U
Z:
0
0
N
01
0.
rr
0
H
261
U
ce
CA
$-4
Ct
O
rO
a
4.
-c
N
O
o,
Q
3.
(._
6.
cJ3
.d
c
Q)
262
u
1
.-
.00
-
;;7.w
a
'N
--
0
¶4
_
0 -
u,
A
0o
-
E
-t
0
0
.-
00
'Y
©5
Lo
.
+
--W
3)
I
263
C',
CZ
4..
0
-
o
-C
I,
N c)
._
.
C
X
zc
) ¢c°
._
Z
264
C
e
a
lVe~stNI
265
Cl
cn
a
CZ
a
U
C)
aC)
cn
Ca
O
N
.a
..
C)
co
-z
or)
._
~3
· ,,,,
a,
C)
266
ki--
L
©
O
0
o
L
267
cu
k.
U
U
a
°,
°,
._
C)
r3
r_
N
U
azk
o
._
a)
o
5k
as~
C)
Cd
268
Ln
dot
i
0
o2~1
0
©g
0
269
o
0
U
of~
.(0o
cj;
o
E-e
C.
0oa,
C.
c;
._
a
oC)
Cr
,
0
c0
U
C;I
v
©
Cl
z
U:
c,
u
J
oe,
Cl
v
0
-
CI
o
oU©
©
o
t~
o;
c
CS
0
0
u
U) -..u
._
,,.o
Wo
O
"C.)
*a
Cl
o
r3
-o
©
.4
;>~
.-
270
\
\
0e
0
)t
,, A
+
+
'
n
i=\
H1
0\
%e
w
Ems/m
I
teI
,,
A.0
P-6
271
a)
U
¢eCZ
-.75
bC.
C)
C C
v
e
C))
.-,C:
f;-
-
U
._
.- ,
.4)
)
._
._
-,
0
C'
-o
©
-J
.
aC)
ti
o
0
g3
A
V
©
C)
o
o
o
a)
._ua
tO .
-
.-
cr3
-
o
a
FP
O
¢3
0
272
I'w
9!
,
1
eq,!
cu
I
Ct z
-
!I
~r-I
40(U6
!
I
273
Lt
bO
c
_ Q ._C
._
E
4·
4
.-
cr-
Z
a,
._X-
C.
_
*0a,
)
cr3c
0) c,
-~
CC
0)
t3
0,- ©
::I ,~ >,
4-
c0) tcri0)Q
.. _.
.,-
-'3
Ut
..
4...a
0)
o
gZ
._
Z:(
e
~
._
._
5
ccn
-
274
,
In!
I
-
!
tI
r-:
P4
P4
A,w
!
CZ
-4
!
In
en
r Ii
275
C
o
X
~
a)
u
:
z
o
C.
C
o
r
o
,_Z
.~
oU
U
U
o
,
,_:
-J
C
"U
U
-
U
I
:O
Q
ou
U
t.)
CL
OZ
Uz
u
Q)
z
U
.
o
:
C)
av
oCL
O
r4
W
*
uUC
z
.
E
OJ
UC
276
0
k
0uen
!
,
ee~
,9
I
-
en
zI
z
+
0O3
tO
0
_U
P-0
!
Q)
1:1
I
tI
I
I
A.0
W*
277
Figure 18. Exchange reaction interconverting the H,T products to the alcohols
as shown.
278
-T
+H
(S)-[1-3 H]-ethanol precursor
retention (3 in Table 5)
(R)-[1-3 H]-ethanol precursor
not observed
-H
----
il.
(R)-[1-3 H]-ethanol precursor
(R)-[1-3H]-ethanol precursor
inversion (1 in Table 5)
inversion
279
Figure 19. Exchange reaction interconverting the D,T products to the alcohols
as shown.
280
-T
+H
(S)-[1-2 H,1-3 H]-ethanol precursor
(R)-[1-2 H]-ethanol precursor
inversion (2 in Table 5)
not observed
-D
-~~~-~~--'
~ 0
(R)-[1-2H,1-3H]-ethanol precursor
(R)-[1-3 H]-ethanol precursor
retention (4 in Table 5)
inversion
281
)
a)
-a
u)
Cl
X
Cl
rz
0
75
a*
4._
4I
cU
k
(-
a
.-'
.- r3
C)
c
0o
0
0
45
u
c6
o
©
OU
cu
C)
C
o
o.,.a
©
-
)
U
C)
.,c3
-
o
c:
0
,ca3
0)
0)
0 ) ci
0)
r- ~
(d
4-~
\
**
282
I
/
M
ItI
II
up
N
II
Y
u
N
I
o
in
r-
U
-'?,
No
0r
i
-Ils
21
r-I
'-
.
.0".
v
/".X
.
r
I
uc
I>
lI
N
U
em
w
~ff
Cf)
l/
M1
104"
z:
114
C"4
to
-.d
· rW
.- 4
%4-4
0
114
1-1
0
:z
0
1I
0
0
11
.14
141
x
!V
O...4
,-i
p
u,
uC.
eel
:4
I
V!
! !
Il"l
A
A3
283
O
u]
o 6
-c
C)
a)
o
E;
_
rO
oC
4
~L
C
C
aJ
o x .. Ca
uQ
-e
o
4
H
U
U
-
0
C
C)
o
(
C)
U
at-
o
~
41
X
C)
i
Q)
X
-i
bX
C
-
284
"0
0
ff
"410
u
"0
u
u -o
o
0
U
-
ob=;
(A
m0
-,
o
P.*
7o
0
"O
0
"4
u
u
X
w
"a
x
W4
9CZ
0
0
Ed
m
"14
ji
"1
A
-Q v
I
L-
"0Z
_
el
o4
u
Us
u
I
To
6-.
I
C
3:4
w
U,
"0
0
v)
I--
60
cU
x1
or
us -W
;1;
cr
o
i*
CL
u
u
x
w
x
w
0
a
*;
Ed1
0
F-O
m
P
LI
285
o
o
4..
0
sv
t
U
c/
Q)
c
¢
c
O
C
4--
4.)
o .. 04X
CO
co ,
d
°,--4
_:
o
.U
-
cn
l
*-
(
U
._) L O
-
286
m
0
U-ill
cWI
U - 0
~-{II~-O0
;T
0-/
Fe
· I.....
~C~
=
e
O
,,
U - -1-1
mi
U-
U.,,
v?~
V-,,,
U--11
\
e)-
287
U
C
0.
o
U
c.
CX
a...
o0
0u
U
r;
o
r.
2u,)
A
Cr
c-
U
CO
C.)
C
ao
->
o
._(3
C
.C.
aCz
©
CJ
to'
-z
._)
a)
59
c)
o
;>
cO
4
.a
w.4
.
-
-r
Ubt
cU2
4-
;.
C
-
-o
Lo
o
·,,,
C)
C)
288
oo
1.4
0
10
~~~~~~~~0
;.4
1.4~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.
uTj
1.4
C
//
1
0
c orlu
0.v
rl
I"
enI
t(U)
r-
-4
-
vm
-
0
1.
11
w
;.4
u
0
-1
o
1-
o
M
0
e
r* 0
C~
13
I!
-
-
r i
1
ud
289
CHAPTER 6.
Intermediates in, and Kinetic Studies
of the Reaction of Reduced Hydroxylase of
MMO from Methylococcus capsulatus (Bath) with Dioxygen
290
Introduction
Methanotrophic
bacteria play an important role in the balance of
atmospheric methane and in the global carbon cycle.1 Methane is the sole
source of carbon and energy for these organisms, which rely on the enzymatic
system methane monooxygenase (MMO) to convert methane into methanol.
The first step in the metabolic pathway is shown in equation (1).1 MMO exists
in either in a particulate, membrane-bound form containing copper, or in a
soluble form containing iron at the active site.2 Soluble MMO isolated from
CH4 +
+H+
+
CH3 OH + H20 + NAD+
NADH
(1)
Methylococcus capsulatus (Bath)3 or Methylosinus trichosporium OB3b,4 two
of the better characterized systems, comprises three separate proteins, the
hydroxylase, protein B, and the reductase.
The hydroxylase component of MMO is a member of the non-heme
carboxylate protein family which contain dinuclear iron units at their active
sites. 5 Included are hemerythrin, the R2 subunit of ribonucleotide reductase,
and purple acid phosphatase.6-8
We wish to understand the factors
responsible for tuning the diiron center in each protein to perform its specific
function, which ranges from the reversible binding of dioxygen in
hemerythrin to its activation for the hydroxylation of methane in MMO. In
pursuit of this objective, we are investigating the proteins of the MMO system
and exploring the fundamental chemistry of the diiron center in the
hydroxylase.
The hydroxylase, the site for substrate binding as well as hydrocarbon
activation reaction in MMO,4 contains up to two dinuclear iron centers and is
isolated in its oxidized FeIIIFeIII (Hox) form.
The mixed-valent FeIIFeIII
291
II
(Hmv),9 and fully reduced FelIFe
(Hred) redox states are readily accessible
through chemical reduction of Hox. The crystal structure of Hox1 0 revealed
that the Fe atoms are bridged by a hydroxide and a bidentate, semibridging
glutamate ligand. Each metal is also coordinated to a histidine ligand. Two
additional glutamate residues ligate one Fe atom in a monodentate fashion,
whereas the second iron atom has a monodentate glutamate and a water
molecule in its coordination sphere.
In the sample studied by x-ray
crystallography, the iron atoms are additionally bridged by an exogenous
acetate ligand from the crystallization conditions.
Other physical
techniques, 11-14 such as EXAFS, EPR, ENDOR, and M6ssbauer spectroscopy,
support this assignment for the coordination environment of the metal
atoms.
The reductase, containing one FAD and one 2Fe-2S cluster, accepts
electrons from NADH and in turn reduces the dinuclear iron center in the
hydroxylase.
1 5 47
Protein B with MMO from M. capsulatus (Bath) regulates
electron transfer between the hydroxylase and the reductase. 18' 1 9 The effects
of protein B in the M. trichosporium OB3b MMO system include its ability to
lower the reduction potential of the hydroxylase, 20 perturb the EPR signal of
Hmv, 21 alter product distributions of the hydroxylation reaction,2 2 '2 3 and
change the initial velocity of the completely reconstituted system.2 3 Little is
known about the influence of the reductase and protein B on the reaction
chemistry of the hydroxylase. Further examination of interactions among the
three components, which influence their collective behavior in the MMO
system, is necessary to elucidate fully the role of each protein in catalysis.
Hred, the reactive species toward dioxygen, can be produced by chemical
reduction of Hox with sodium dithionite in the presence of mediators.
By
using this approach, Hred was generated in the absence of protein B and
292
reductase, and its ability to hydroxylate hydrocarbons was explored.
The
substrate nitrobenzene produces highly colored products upon oxidation.
Since neither Hred nor Hox have any optical absorption bands above 300 nm,
this reaction was easily studied by optical methods. The effects of the
reductase and protein B components on the kinetic constants, regioselectivity,
and product yields of the reaction of Hred with nitrobenzene and dioxygen
were determined. In the course of these studies, optical changes were detected
in a control reaction of Hred with dioxygen even in the absence of substrate.
Kinetic constants of these colored species were therefore determined as well,
leading to their assignment as intermediates along the reaction pathway. 2 4
Parallel studies with MMO from M. trichosporiumOB3b were reported while
our work was in progress.2 5 ' 2 6 Spectroscopic and kinetic characterization of
these intermediates was performed by using stopped-flow spectrophotometry
and rapid freeze-quench EPR, Mbssbauer, and resonance Raman spectroscopy.
We describe our findings in the above areas in the present article and suggest
assignments for the intermediates in a proposed catalytic reaction cycle.
Experimental
Bacterial Growth and Protein Purification.
Growth of the native M. capsulatus (Bath) organism and purification of
the hydroxylase, protein B, and reductase components of its sMMO were
performed as described elsewhere. 14
were in the ranges reported. 2 7
with
5 7 Fe
27
Specific activities and iron contents
For Mdssbauer work, hydroxylase enriched
was obtained from cells grown on usual medium 2 8 except that
Na 5 7FeEDTA was added as the sole source of iron.
prepared by dissolving
57Fe
Na 57 FeEDTA was
foil (52 mg, 0.91 mmole, 95% enriched, Cambridge
Isotope Labs) in 3.6 ml of 1 M ultrapure HNO 3. Next, 3.6 ml of H 2 0 and 0.58 g
(2 mmole) of Na 4 EDTA.2H 2 0 were added. After stirring for 5 min, 30 mg
293
(0).75mmole) of NaOH was added to the iron solution. The resulting mixture
was allowed to stir for 6 hours, after which time it was centrifuged to pellet
the precipitate.
The dark green supernatant was added to 100 ml 20 mM
sodium phosphate buffer (pH 7.0). The resulting solution was autoclaved and
added to the normal growth medium to a final iron concentration of 36 [tM.
Specific activities with propylene and iron content of
57 Fe
enriched
hydroxylase were in the reported ranges for native hydroxylase.
Generation of Hred.
To reduce the hydroxylase, a stoichiometric amount of a 10 mM
solution of methyl viologen was added to Hox,,as an electron transfer
mediator. The protein and mediator solution was then made anaerobic by
evacuation and back-filling with argon for at least 5 cycles. Next, a 0.1 M
solution of sodium dithionite was prepared by first degassing 2 ml of buffer
(25 mM MOPS, pH 8.5) for 15 min with dioxygen-free argon. Degassed buffer
was transferred with a gastight syringe to a vial which contained 41 mg (0.23
mmole) of sodium dithionite, previously made anaerobic by evacuation and
back-filling with argon for 3 cycles. An aliquot of the sodium dithionite
solution was then added to the protein solution by means of a gastight syringe
to achieve a 2-fold molar excess concentration over that of the hydroxylase.
Upon addition of the reductant, the reaction mixture immediately turned
blue due to formation of the methyl viologen radical. Reduction was allowed
to proceed for at least 45 min, during which time the reaction mixture was
transferred to an anaerobic chamber (Vacuum Atmospheres).
Inside this
glovebox the protein solution was pipetted into dialysis tubing having a
10,000 molecular weight cutoff, and then dialyzed against 400 ml 25 mM
MOPS, pH 7.0 buffer to remove methyl viologen and excess sodium
294
dithionite. After a 2 hour period, the protein was removed from the dialysis
bag.
Experiments with protein B and reductase were performed in two
different ways. Reduction of Hox to Hred was first carried out, after which the
other component(s) was added to the dialyzed protein. Alternatively, the
additional component(s) was combined with Hox prior to reduction, with
subsequent treatments as described.
Single Turnover Reactions.
Kinetic Studies. The reduced, dialyzed protein solution was diluted
with 25 mM MOPS, pH 7.0 buffer to achieve a 50 gM concentration of Hred.
Protein
B concentrations
ranged
from 0 to 200 gM.
The substrate
nitrobenzene was added as the neat liquid, 10 gil per 1 ml of protein solution,
to obtain saturation at approximately 5 mM. The resulting mixture was
incubated for 5 min, after which time undissolved nitrobenzene was
removed by pipette.
For stopped-flow kinetic analysis, the protein mixture
was loaded into a 10 ml Hamilton gastight syringe and then removed from
the glovebox (vide infra).
dioxygen-saturated
The reaction was initiated by mixing with
buffer, and optical changes at 404 nm were recorded to
monitor the formation of 4-nitrophenol.
Product Distribution and Yields.
To determine the distribution of
products from nitrobenzene hydroxylation, enzymatic reactions were typically
performed with 8 mg (3.2 x 10-8mole) of hydroxylase and variable quantities
of protein B and reductase. Up to 2.2 mg (1.3 x 10-7 mole) of protein B and 5
mg of (1.3 x 10-7 mole) reductase were added to achieve ratios from 0 to 4
equivalents for a given component relative to the hydroxylase.
The reaction
mixture was reduced, dialyzed and incubated with substrate as outlined
above. After dialysis, the mixture (=1 ml) was transferred to a 10 ml glass vial
295
which was then sealed with a septum. The vial was brought out of the box,
and 0.5 ml of dioxygen-saturated buffer was injected by syringe through the
septum.
The resulting solution was placed in a shaking incubator
maintained at 45 C and allowed to react for at least 15 min to insure
completion of the reaction. To compare the regioselectivity of the catalytic
system to that obtained in single turnover reactions, 1.5 mg (6 x 10-9 mole) of
hydroxylase,
02 mg (1.2 x 10-8 mole) of protein B, 0.4 mg (1.2 x 10-8 mole) of
reductase, and 1 gl (8 x 10-6 mole) of nitrobenzene were combined in a total
volume of 0.5 ml. This mixture was incubated for 30 s at 45 C in a capped
vial. By means of a syringe, 25 tglof a 0.1 M ethanol-free
solution of NADH
was then added through the septum and the mixture was allowed to react for
a 15 min period
at 45 °C.
The distribution of 2-, 3-, and 4-nitrophenol products formed in
enzymatic reactions with nitrobenzene was investigated through HPLC
analysis.
The reaction
mixtures
were treated
with 200
trichloroacetic acid solution to precipitate the protein.
l of a 3%
The resulting
suspension was centrifuged for 2 min in a table top centrifuge (Hill Scientific)
to pellet the protein. An aliquot from the supernatant was then injected into
the HPLC, and the identities of the reaction components were determined by
comparison of their retention times with those of authentic standards of the
expected products.
Conditions for HPLC analysis were as follows. Stock solutions of 1 mM
of 2-, 3-, and 4-nitrophenol and nitrobenzene were made in 25 mM MOPS
buffer, pH 7.0 buffer. A 100 gl aliquot from each solution was injected into a
Perkin-Elmer Series 4 high performance liquid chromatograph equipped with
a reverse phase C18 radial compression column (-Bondapak,
Waters).
Detection was carried out at 300 nm, a wavelength at which all four
296
components had significant absorption. The solvent system used to elute the
components was a methanol and water gradient with a flow rate of lml/min.
The solvent composition was initially 60% water and 40% methanol, and was
ramped to 100% methanol over a 20 min time period. The amount of each
component present was calculated from the peak areas after normalization
with the individual extinction coefficients at 300 nm.
The extinction coefficients for the reaction components at 300 nm were
determined in order to quantitate the HPLC peaks. Stock solutions, 10 mM in
each phenol, were carefully diluted to approximately 5 x 10-5 M in 50 % buffer
(25 mM MOPS, pH 7.0) / 50 % CH 3 0H.
The UV/vis spectra were then
recorded (Perkin-Elmer Lambda 7 spectrophotometer).
The extinction
coefficients were calculated from the absorbance at 300 nm at several
concentrations and found to obey Beer's law. The relative yields of each
product in the enzymatic reactions were then calculated from the HPLC traces
by dividing the peak area by the extinction coefficient.
The overall yields of products formed in a single turnover reaction of
Hred with nitrobenzene were calculated by first determining the amount of 4-
nitrophenol that was generated. The corresponding amounts of 2- and 3nitrophenol for the appropriate ratio of protein B to hydroxylase as
determined by the above HPLC analysis was next added to this value to
determine the total amount of product formed. Finally, the amount of
product was divided by the amount of dinuclear iron centers present in the
reaction solution to determine the percent yield of the reaction.
To optimize the yield of 4-nitrophenol in single turnover reactions,
several conditions were investigated and their effects were compared.
The
incubation time to reduce Hox was varied from 30 min to 2 hours. Reactions
with variable quantities of protein B were performed. Reductase alone and a
297
mixture of both protein B and reductase were added to the hydroxylase in
parallel reactions.
Nitrobenzene was also added prior to the addition of
sodium dithionite to determine whether the order of addition of reductant
and substrate affected the yield of this product.
KM of Nitrobenzene.
The KM constant for nitrobenzene in reactions
with MMO was measured by monitoring the optical change at 404 nm.
Reactions contained 3.3 mg (1.3 x 10-8 mole) of hydroxylase, 100 gl of
B/reductase mix (= 2.6 x 10-8mole in each protein), and 650 gl of a mixture of
buffer (25 mM MOPS, pH 7.0) and a 10 mM stock nitrobenzene solution, the
ratio of which depended on the amount of nitrobenzene desired in the
reaction. A 50 gl portion of a 0.1 M NADH stock solution was added to
initiate the reaction in a total volume of 1.2 ml. The mixture was then placed
in a thermostatted cuvette at 45 C. The absorbance change at 404 nm over a
400 s time period was used to quantitate the amount of 4-nitrophenol
produced.
From the linear region of the change in product concentration vs
time, the velocity of the reaction was determined.
Velocities at several
nitrobenzene concentrations were then plotted as a function of nitrobenzene
concentration.
From double reciprocal plots of these data, KM was
cdetermined.
Stopped-flow Spectrophotometry.
Stopped-flow experiments were performed to monitor the reaction of
in the presence of 60-200 gM protein B with dioxygen.
This ratio of
components was chosen because it gave optimal rate constants and yields
from single turnover reactions with nitrobenzene (vide infra). Dialyzed
solutions of Hred and protein B were loaded into a 10 ml Hamilton gastight
syringe fitted with a KelF threaded adaptor for connection to a anaerobic line
of a stopped-flow assembly (HiTech, Canterbury, England). The contents of
298
the 10 ml syringe were used to fill the drive syringe of the stopped-flow
apparatus. By using a Matheson gas mixer, buffer (25 mM MOPS, pH 7.0) with
variable concentrations of dioxygen (0 to 1 mM) was prepared, and then
loaded into the second drive syringe of the stopped-flow assembly. Solutions
were equilibrated for 30 min at a given temperature prior to mixing.
To
determine the pH dependence of the reaction of Hred with dioxygen, the
anaerobic dialysis was performed with 400 ml of 10 mM MOPS, pH 7.0. The
oxygenated buffer solution contained 100 mM MOPS buffer, with pH values
ranging from 6.6 to 8.6. To test for the presence of a solvent deuterium
isotope effect of the reaction velocities, hydroxylase was first exchanged into
D)20 by using a procedure described elsewhere. 12 A small volume ( 175 tgl)of
a 5 mM solution of protein B was added to the exchanged Hox (vol = = 2 ml)
prior to reduction. A stock of 25 mM MOPS in D20 was used to make the 0.1
M sodium dithionite solution (pD = 8.5, uncorrected for the deuterium
isotope effect) and for the dialysis and oxygenated buffer in the second drive
syringe (pD = 7.0, uncorrected).
For reactions with the substrates hexane and methane, the oxygenated
buffer was further treated as follows.
With hexane, 100 ml of dioxygen-
saturated 25 mM MOPS, pH = 7.0 buffer was mixed with 50 ml of hexane for
20 min to obtain a saturated solution of this substrate. The aqueous layer was
removed and used to fill the aerobic drive syringe. With methane, a similar
procedure was used, except that 25 ml of dioxygen-saturated buffer was
incubated with 25 ml of methane gas at approximately atmospheric pressure.
Spectrophotometric analysis was performed with one of two detection
systems.
A photomultiplier tube was used to collect data at single
wavelengths between 330 and 800 nm. Alternatively, a Hewlett Packard
299
UV/vis diode array spectrophotometer (model 8452A) was used to collect data
at multiple wavelengths.
Rapid Freeze-Quench Studies.
M6ssbauer samples were made with
5 7 Fe-enriched
hydroxylase.
Both
freeze-quench EPR and M6ssbauer samples were prepared by using an Update
Instruments rapid freeze-quench apparatus following methods outlined
elsewhere.
29
Solutions of 600 - 700 jtM Hred and 1.2 - 1.4 mM protein B were
mixed rapidly with a dioxygen-saturated
( 1 mM) 25 mM MOPS, pH 7.0
buffer solution at 4 C. Protein was allowed to react with dioxygen for fixed
time periods (0.025 - 60 s) before being sprayed into isopentane at -140 C. A
schematic diagram of the apparatus is illustrated in Figure 1.
EPR Spectroscopy.
EPR spectra were recorded at X-band on a Bruker Model ESP 300
spectrometer with an Oxford Instruments EPR 900 liquid He cryostat set at 9
K. The Hmv signal at gay = 1.83 was quantitated
under non-saturating
conditions by double integration of the first derivative spectrum for
comparison to a frozen solution of copper perchlorate (lmM CuSO4, 2 M
NaC10 4 , 0.01 M HCl). Transition probabilities were corrected for g-value
anisotropy.
3 0
The percentage of Hmv species was calculated from the
concentration of spin relative to the concentration of iron present in the
sample.
Mossbauer Spectroscopy.
M6ssbauer spectra were collected by using either a weak-field
M6ssbauer spectrometer equipped with a Janis 8DT variable-temperature
cryostat, or a strong field spectrometer outfitted with a Janis 12 CNDT/SC
SuperVaritemp cryostat encasing an 8 T superconducting magnet. A constant
acceleration mode in a transmission geometry was used to operate both
300
spectrometers.
The centroid of the room temperature spectrum of a metallic
iron foil corresponds to the zero velocity of the Mbssbauer spectra.
Resonance Raman Spectroscopy.
Resonance Raman spectra of stopped-flow solutions were collected
with excitation at several wavelengths
concentrations
(266, 350, and 406 nm).
Protein
as high as 150 JiM in Hred and 300 gM in protein B were
employed these experiments.
In addition, Raman spectra were collected on
the freeze-quench EPR samples, made with either 1602 or 1802 (95% enriched,
Cambridge Isotope Labs), by using Kr+ ion laser lines at 413 and 647 nm and a
dewar and at assembly in the Spiro laboratory at Princeton University.
Results
Studies of Nitrobenzene Hydroxylation.
Product Distribution and Yields in Single Turnover Experiments.
HPLC traces of extracts from single turnover reactions of Hred and various
ratios of protein B are illustrated in Figure 2. In all cases, the predominant
product was 4-nitrophenol. By contrast, under catalytic conditions, in which
all components were present and NADH sustains turnover, a mixture of 57%
4-nitrophenol and 43% 3-nitrophenol was produced. Product distributions
under all conditions tested are summarized in Table 1.
Optimal product yields were obtained with a 45 min incubation time of
reductant with Hox and a subsequent 2 h dialysis period. In all cases where an
additional component(s) was incubated with the hydroxylase, the results were
the same irrespective of whether the protein(s) was added to either Hox or
Hred. Yields from reactions of Hred with nitrobenzene in the presence of
reductase were similar to yields observed with Hred alone. Incubation with
both protein B and reductase resulted in low yields, approximately 75 % less
than with the Hred alone. The addition of protein B had a dramatic effect on
301
the yield, tripling the absorbance change at 404 nm compared to a reaction
with only Hred present.
A plot of the percent yield of product versus the
number of equivalents of protein B present in the reaction mixture is given
in Figure 3. Greater amounts of 4-nitrophenol were detected at higher ratios
of protein B relative to Hred. Addition of nitrobenzene prior to reduction of
Ho,, had no effect on the yield of 4-nitrophenol.
Kinetic Analysis of the Catalytic Hydroxylation of Nitrobenzene.
Reactions with the catalytic MMO system and nitrobenzene were carried out
to determine kcat and KM for this substrate. Traces of absorbance versus time
for various substrate concentrations, a Michaelis-Menten plot, and the
corresponding double reciprocal plot are presented in Figure 4. From these
data, a value - 5 mM was calculated for KM. Kinetic traces from single
turnover reactions of Hred and Hred plus 2 equivalents of protein B under
pseudo-first order conditions in dioxygen and nitrobenzene are given in
Figure 5. A rate constant was obtained from each trace after fitting to the first-
order growth equation indicated in equation (2). A plot of kobs values
obtained
in
this
manner
as
a
function
y=A+B(1-e-kobs*t)
of
the
ratio
(2)
of protein B to hydroxylase used in the reaction is given in Figure 6. From
this plot it is apparent that the addition of protein B increases kobs from 0.011
s- 1 to a maximum of 0.40 s-1 with the addition of between
1.5 and 2.0 molar
equivalents of protein B to hydroxylase. Further addition of protein B above
this ratio causes kobs to decrease slightly.
constant levels off to
increase of
-
0.25 s-1 .
33-fold in kobs.
Above 2.5 equivalents the rate
Protein B therefore induces a maximal
302
Intermediates Formed in the Reaction of Hred with Dioxygen.
Spectroscopic Characterization of Intermediates. Mdssbauer data were
collected from samples quenched between 0.025 and 60 s after rapid mixing.
Figures 7A, 8A, and 9A show representative spectra taken at three different
time points. Over the time range investigated, five species in the form of
quadrupole doublets were present at varying concentrations, M6ssbauer
parameters of which are summarized in Table 2. The five species correspond
to four different chemical forms of the dinuclear iron center, Hox, Hred, and
two intermediates designated L and Q. The spectrum of Ho, shown in Figure
10A, has Mbssbauer parameters indicative of the oxidized, Fe(III)Fe(III)
oxidation state of the diiron center ( = 0.51 + 0.02 mm/s and AEQ= 1.15 ± 0.03
mm/s for site 1, 6 = 0.49 ± 0.02 mm/s and AEQ = 0.85 + 0.03 mm/s for site 2).
Two diferrous signals, referred to as Hred(l) and Hred(2 ), are attributed to Hred
and both display
parameters
of 6 = 1.3
0.02 mm/s
with AEQ = 2.8 + 0.03
mm/s. The weak and strong field Mdssbauer spectra of a mixture of these two
signals are given in Figures 10B and 11A, respectively.
In accord with our earlier report,2 4 the extracted M6ssbauer spectrum of
intermediate L is a sharp symmetrical quadrupole doublet with parameters 6
= 0.66 + 0.02 mm/s and AEQ= 1.51 + 0.03 mm/s, as shown in Figure 7B. From
subsequent stopped-flow optical spectrophotometric studies, carried out at
single wavelengths by using fiber optics detection, an absorption band at = 625
nm with
= 1000 M-1cm- 1 was estimated for L. Figure 12 shows the growth of
this feature with time, measured at 625 nm, an analysis of which is given
below.
The optical spectrum of intermediate Q, shown for various time points
in Figure 13, has Xmax values of 350 (£ = 3600 M-lcm-1l),420 (£ = 7200 M-1cm-1),
and 520 nm ( = 14,000 M-1cm-1). Figure 9B shows the extracted M6ssbauer
303
spectrum of Q, which comprises two distinct doublets, having parameters
=
0.21 + 0.02 mm/s and AEQ = 0.68 + 0.03 mm/s for doublet 1 and 6 = 0.14 ± 0.02
mm/s and AEQ = 0.55 + 0.03 mm/s for doublet 2. High field Mdssbauer
spectra, shown in Figures 11B and 11C, reveal both intermediates L and Q to
be diamagnetic.
This observation agrees with low-temperature EPR spectral
studies of L and Q presented in Figures 14 and 15, respectively.
spectra, only signals corresponding
In these
to Hred (g = 15), Hmv (gav = 1.83, = 5% based
on total Fe) and a small amount of a protein-associated
free radical (g = 2.0,
5% based on total Fe), present in all oxidation states, were detected.
Resonance Raman spectra of freeze-quench samples measured at
reaction times of 10 ms, 155 ms, 8 s, and 60 s with excitation at 413 nm are
given in Figure 16. Many peaks are apparent.
Spectra for samples reacted
with either 160, or 1802 appeared very similar. Excitation at 647 nm with the
155 ms sample revealed a peak at 902 cm- 1 which shifted by 23 cm-1 upon
substitution, as summarized in Figure 17. These studies remain in progress.
Kinetic Analysis. Kinetic analysis of the time dependent changes in the
freeze-quench M6ssbauer signal revealed several important aspects of the
reaction of Hredi with dioxygen. The reaction proceeds as given in equation
(3), and plots quantitating the variation in concentration of each of these
species over time are given in Figure 18. The data were fit to equation (4),
which describes a system containing three consecutive first order processes.
k,
k2
L
Hre d
klk2k3
Y=I
e-kl*t
k2-kl Lkl(k3-kl)
k3
Q
~
e-k2*t
e-k3*t
1
k2(k3-k2)
k3
k3-k2
H(3)
1
k3-kl
1
304
The rate constant k1, = 25 s-1, corresponds both to the decay of Hred and to the
growth of intermediate L. The decay of L occurs with a rate constant of k 2
0.3 s- 1, which matches the rate constant extracted for the growth of
intermediate Q. Hox then forms with a rate constant k 3 0.03 s- 1, corresponding
to the rate constant for the decay of Q. A significant amount of Hred remained
after 30 s, and M6ssbauer analysis revealed that this species is converted to
Hox with a much smaller first order rate constant. The contributions of the
species to the Md6ssbauer spectra were subtracted during the above analysis
(vide infra).
Stopped-flow spectrophotometry was also used to analyze the kinetics
of formation and decay of both intermediates L and Q. For L, detection at 625
nm resulted in the kinetic trace shown in Figure 12. These data were fit to a
first order growth equation, giving k1
22 sl. Studies of Q at 420 nm resulted
in the kinetic trace shown in Figure 19A. Trace A was fit to the expression
given in equation (5) which contains two consecutive processes, with
corresponding
respectively.
rate constants
of k 2 = 0.5 + 0.15 s- 1 and k3
=
0.07 + 0.02 s- 1,
Trace B in Figure 19 was obtained under identical conditions
except that protein B was omitted.
k2*A
y = k2
k2- k3
[B*(1- e- k2*t)+ C*e-k3*t]
The kinetic constants k2 and k3 measured
(5)
by stopped-flow
spectrophotometry under a variety of conditions are listed in Table 3. These
rate constants were unaffected by the dioxygen concentration, as shown in
Figure 20. Similarly, the rate constants were pH independent in the range 6.6
<;pH
8.6, as indicated in Figure 21. When the reaction was run in D 2 0,
305
there was little effect on the rate constants. A kinetic trace for the reaction
run under these conditions is illustrated in Figure 22, and fitting to the
equation yielded rate constants of k2 = 0.35 s- 1 and k 3 = 0.15 s- 1
biexponential
at 6 °C. The temperature dependence of the rate constants for the formation
and decay of compound Q was also measured. Arrhenius and Eyring plots of
these data are given in Figures 23A and 23B, respectively. The parameters
determined from these plots are listed in Table 4.
The optical spectrum of Q was also recorded in the presence of
substrate.
With methane, high protein concentrations (> 150 M) were
necessary to detect this intermediate.
Fits the observed kinetic traces, an
example of which is shown in Figure 24, revealed both k 2 and k3 to increase
slightly, to values of = 0.95 s-1 and 0.17 s- 1, respectively. Analogous data were
collected with the substrate hexane. The corresponding kinetic trace was fit
with k 3
=
0.13 s -1, but k2 was unchanged. The optical spectrum of Q, taken in
the presence of' either methane or hexane, revealed no new chromophoric
species present.
Data with nitrobenzene differ from those collected with
methane and hexane since the product of the hydroxylation
reaction, 4-
nitrophenol, absorbs significantly at 400 nm. When monitored at 420 nm, k 2
was determined to be 0.26 s-1 and, k3 0.11 s- 1. No other new optical bands
were detected.
Discussion
MMO Component Interactions as Assessed by Hydroxylation of Nitrobenzene.
The interactions of protein B and reductase with the hydroxylase were
evaluated
through the influence of these two components on the
hydroxylation of nitrobenzene by Hred. Three aspects of this reaction were
investigated, the regioselectivity of hydroxylation, the overall product yield,
306
and the rate constant for product formation.
Examination of the product
alcohols indicated that hydroxylation of nitrobenzene proceeds in a
regioselective manner depending on the reaction conditions.
With
chemically reduced hydroxylase (Hred), 4-nitrophenol is the predominant
product irrespective of the amount of added ratios of protein B up to a
fourfold excess. Under catalytic conditions with all components (Hox, protein
B, reductase, N'ADH) present, the distribution of products changed to an
approximately equimolar mixture of 4-nitrophenol and 3-nitrophenol.
The
presence of reductase and NADH, therefore, alters the regioselectivity of
hydroxylation to afford 3-nitrophenol, which was not produced by protein B
alone.
These findings are compared in Table 1 to the results of analogous
studies of catalytic reactions with MMO from M. trichosporium OB3b.2 3 In
this MMO system, unlike MMO from M. capsulatus (Bath), protein B is
apparently not required to sustain turnover under catalytic conditions,4 and
H202 can support high levels of turnover in a shunt pathway with Hox
alone. 2 2'
2 3 31
Investigation of the regioselectivity of nitrobenzene
hydroxylation with the M. trichosporium OB3b system2 3 revealed that 2 90%
of the product is 3-nitrophenol under catalytic conditions without protein B.
Adding 0.1 to 2 equivalents of protein B to hydroxylase, reductase, and NADH
afforded 83 to 89% of 4-nitrophenol.
In this case, therefore, protein B shifts
the product distribution toward 4-nitrophenol. In the H 2 0 2 shunt pathway of
Hox from M. trichosporium OB3b,2 2 almost equal amounts of 4-nitrophenol
and 3-nitrophenol were produced, which matches behavior seen with the
catalytic system from M. capsulatus (Bath). It is also noteworthy that the KM
(-::5 mM) for nitrobenzene with the catalytic MMO system from M. capsulatus
(Bath) matches the KM value of Hox and H2 0 2 , but not the KM value ( 100
307
pM) for the catalytic system from M. trichosporium OB3b. From these results
and the data in Table 1 it is obvious that the component interactions in the
two systems are complex and alter the hydroxylation chemistry in different
manners.
Significant formation of 4-nitrophenol is somewhat surprising if the
hydroxylation mechanism involves electrophilic attack on the aromatic ring.
The nitro substituent is a powerful deactivating group and a strong meta
director, and if substrate reactivity were the determining factor, 3-nitrophenol
might be expected to be the major product.
We therefore suggest that the
stereochemistry of the substrate binding in the active site determines the
regioselectivity of the reaction. This property will be affected by the presence
of the other protein components, since their interactions may cause
perturbations in the active site structure.3 2 An alternative possibility is that
the hydroxylation mechanism does not involve electrophilic aromatic
substitution.
The product yield in single turnover reactions with Hred is also affected
by the presence of the other components. Since HPLC analysis identified 4nitrophenol as the predominant product under all conditions with Hred, the
increased yields are not due to a shift in product distribution.
Protein B
markedly increased the yields, as illustrated in Figure 3. Addition of protein
B and reductase to Hred resulted in diminished yields, and when added to Hox
prior to reduction, extremely low levels of hydroxylation were observed. The
latter finding is consistent with the difficulty reducing Hox to Hred with
protein B and reductase present, as monitored by EPR spectroscopy.l 9 Adding
reductase alone to Hox or Hred had very little effect on the yields relative to
reactions with Hred alone. These results imply that protein B, and not the
reductase, increases the yield of the hydroxylation reaction of nitrobenzene.
308
It is possible that reduction of the complex formed by Hox,,with protein
B and reductase might produce Hred that would be a good hydroxylation agent
when dioxygen is added.
This alternative is consistent with increased
reduction potentials of the substrate/reductase/hydroxylase
to Hox alone.
complex relative
Such a reactive Hred species formed in this manner cannot be
investigated under single turnover conditions, however, since reduction of
Hox to Hred is inhibited by protein B and reductase in the absence of substrate,
and nitrobenzene, the substrate in these experiments, prevented the mediated
reduction of Hox to Hred. Addition of protein B and reductase to Hred may not
produce the same reactive species because of likely structural alterations
produced along with the change in oxidation state of the hydroxylase.
Protein B also increases the rate constant for single turnover reactions
of Hred with nitrobenzene.
Neither addition of reductase to Hox or Hred, nor
of both protein B and reductase to Hred, increased the rate constant relative to
reactions of Hred alone. These observations again imply that protein B alone
activates the hydroxylase and that protein B and reductase do not significantly
affect Hred. The pseudo-first order rate constant for hydroxylation increased
by up to 33-fold when Hred was treated with protein B. This behavior is
reminiscent of that reported for protein B from M. trichosporium OB3b for
titrations performed with Hox,,and reductase under catalytic turnover
conditions.
21
In contrast to the present single turnover reactions of
nitrobenzene with Hred from M. capsulatus (Bath), protein B had little effect
on the rates of analogous reactions of Hred from M. trichosporium OB3b.2 3
Instead, protein B from M. trichosporium OB3b increased the initial velocity
of the complete reconstituted system (Ho, reductase, NADH, catalytic
conditions), but decreased initial velocities in reactions of Hox with H 20 2 . As
seen in the product distribution studies, component interactions affect the
309
kinetic behavior in a complex manner which differs for the two MMO
systems.
Complex formation among the three protein components of M.
trichosporium OB3b MMO has been demonstrated by use of chemical crosslinking agents and other methods.2 1 In this study, it was hypothesized that
protein B bind s tightly to Hox to form an activated complex, but then
dissociates from Hred after electron transfer occurs. The present results are
consistent with initial binding of protein B to Hox to form an activated
complex. Because identical behavior is seen when protein B is added to Hred,
we conclude that protein B can bind to Hred to form an activated complex as
well. If the Hred-protein B complex is the species responsible for the increased
rate constants, and Hred and protein B are in equilibrium with one another,
then the addition of more protein B would ensure that Hred is in the protein
B-bound complex.
equivalents
Elevated concentrations of protein B beyond two
per Hred may diminish the rate constant of the reaction by
sterically hindering substrate binding to Hred. The total yield of the reaction
could still remain high, since nearly all of the Hred is in the active, protein Bbound form. If there are adventitious binding sites for additional molecules
of protein B beyond two equivalents, excess protein B would have little effect
once all such sites were fully occupied. This model is consistent with the B
dependence of kobs shown in Figure 6 and with the yields reported in Figure
3. Close examination of Figure 3 indicates that some cooperative behavior
may be taking place. It is possible that binding of protein B to one dinuclear
iron site in Hred increases the binding affinity for the second site. The errors
of these data points are large, however, which may account for the apparent
curvature in the data.
310
The M. capsulatus (Bath) reductase alters the product distributions of
Hred with nitrobenzene, but protein B plays many more roles in the MMO
system.
As mentioned earlier, it regulates electron transfer from the
reductase to the hydroxylase so as to occur only in the presence of substrate; 17
it shifts the reduction potentials of the substrate/reductase/hydroxylase
complex;
19
and it affects kobs in single turnover reactions of Hred. Protein B
from M. trichosporium OB3b alters the reduction potential of Hox,20 perturbs
the EPR signal of Hmv,21 changes product distributions with Hred,23 and affects
the initial velocity of the complete reconstituted system and the H 2 0 2 shunt
system. 2 3 A molecular explanation for these observed functions remains a
challenge for the future.
More structural work on component interactions
with both MMO systems is essential to understand fully the fundamental
reasons for the observed effects on the rate constants and yields of the
reactions.
Spectroscopic Studies of Intermediates in the Reaction Cycle.
Because optimal yields and rate constants under single turnover
conditions occurred in the presence of 2 equivalents of protein B, these
conditions were adopted for investigating the reaction of Hred with dioxygen.
Reactions were carried out in the absence of substrate in order to allow
detectable concentrations of intermediates to accumulate, since these
activated species were presumed to react quickly with substrate.
The
temperature was also lowered to 4 C to slow down the reaction and facilitate
the detection of short-lived species.
In earlier communications we reported the optical spectroscopic and
M6ssbauer parameters of intermediates L and Q and the resonance Raman
features of L formed in the reaction of Hred with MMO from M. capsulatus
(Bath) with dioxygen.2 4
33
For intermediate L, a doublet with parameters of 6
311
= 0.66 + 0.02 mm/s and AEQ = 1.51 + 0.03 mm/s was observed. Carboxylatebridged high spin diiron(III) complexes generally have isomer shift values
between 0.45 and 0.55 mm/s,6 -8 and isomer shift values for diiron(II)
compounds fall in the range of 1.1 - 1.3 mm/s. 6 -8
Although parameters
matching the values of L have not been observed in model compounds, the
magnitude
of the isomer shift is close to parameters seen for diiron(III)
clusters.
Since Hred decays with concomitant production of intermediate L, the
minimal mechanism dictates that L is an early species formed in the reaction
of Hred with dioxygen. Diiron(III) peroxide complexes, suggested as products
in reactions of several diiron(II) model complexes with dioxygen, result from
two-electron transfer from the Fe atoms to the dioxygen ligand.5 Based on the
kinetic data, Raman spectra (vide infra), and M6ssbauer parameters, we
propose L to be a diiron(III) peroxide intermediate in the catalytic reaction
cycle of MMO. Isomer shift values of 6 = 0.52 - 0.54 mm/s have been reported
for a few diiron(III) peroxide complexes.34-36 In general, isomer shift values
increase according to increasing coordination number and negative charge on
the complex. 3 7 The high 6 value of 0.66 mm/s for L could indicate the
presence of six-coordinate iron atoms and considerable peroxide-to-iron
charge transfer character. A coordination number of seven for one or both of
the iron atoms could also account for the high isomer shift. Among other
1 peroxide with iron coordination numbers of six, or an a
possibilities, an ii1,T
T2,T2 binding mode with seven-coordinate iron atoms are both consistent
with this proposal.
The sharp, symmetrical shape of the M6ssbauer signal
strongly implies that the iron atoms are in nearly identical coordination
environments.
The optical band at 625 nm ( = 1000 Ml1cml)
for L is close to
312
that reported (ma,
complexes.
= 600-610 nm) for several diiron(III) peroxide
3 4 ,3 8 ,39
The Raman spectrum, 3 3 obtained at an excitation wavelength of 647
nm, of a rapid freeze-quench sample, frozen 155 ms after mixing to optimize
the quantity of L formed, exhibits a feature at 902 cm- 1 which shifts with 180
substitution.
This frequency is in the reported range for u(O-O) of several
iron(III) peroxide model complexes, and provides strong evidence for the
existence of such a unit in L. The 902 cm-1 feature was much less enhanced in
spectra obtained with excitation at 413 nm, implying that the 625 nm optical
absorption
is a peroxide-to-iron(III)
charge transfer band.
The Au of 23 cm - 1
observed with the use of 1802 is less than the 42 cm-1 shift predicted for a
diatomic harmonic oscillator. Coupling of the 0-0 stretching mode to
another mode in the diiron unit may lead to a lower value for Atu. A second
possible explanation is that the bound peroxide exchanges with the oxygen
atom of the bridging hydroxide ligand, since the Au value matches that
expected for 160-180 substitution.
A vibration characteristic of v(Fe-0 2)
between 400 and 500 cm-1 was not observed. The geometry of excited state
strongly affects the intensity of this stretch, which was not observed in several
peroxide complexes, including including oxyhemacyanin.4041
Figure 25A
proposes structures for L that are compatible with the spectroscopic data.
The optical and Raman spectroscopic properties of L rule out other
models proposed
previously. 2 4
For example, a delocalized mixed valent
system with a coordinated superoxide anion radical arising from single
electron transfer from the iron(II) atoms was suggested to account for the high
6 value.
Another alternative was that L might be a dioxygen adduct of a
diiron(II) unit which is antiferromagnetically coupled with intermediate spin
(S = 1).
313
The decay of L proceeds with concomitant formation of Q. This second
intermediate has been observed in both the M. capsulatus (Bath)24 and M.
trichosporiunm OB3b MMO systems. 2 6 The M6ssbauer spectrum of Q from M.
trichosporium OB3b is a doublet with 6 = 0.17 mm/s
and AEQ = 0.53 mm/s.
Since this doublet appeared to be symmetrical, the iron atoms were assumed
to reside in equivalent coordination environments. In the M. capsulatus
(Bath) system, two unresolved doublets of equal intensity were readily
distinguished and fit to the parameters given in Table 2. The average isomer
shift and quadrupole splitting of the two doublets, 0.18 mm/s and 0.62 mm/s,
respectively, agree with parameters obtained from the corresponding
spectrum
of Q from the M. trichosporium OB3b organism.
In the M.
capsulatus (Bath) system, however, the presence of two distinct signals for Q
indicated that the iron atoms have in inequivalent coordination spheres.
The average isomer shift value of 0.18 mm/s is outside the range for
carboxylate-bridged, octahedral high-spin diiron(III) clusters.6 ' 7 Instead, the
value is near the range reported for the few characterized heme and nonheme iron(IV) complexes,4
species.
2 45
suggesting that Q might be an iron(IV) oxo
Figure 25B illustrates such possible structures for Q.
The
inequivalence of the Fe sites could reflect significant rearrangement of the
protein ligands in Q as compared to L and Hox. Alternatively, the oxygen
atom of the proposed Fe(IV) oxo species may not be bound symmetrically
between the Fe atoms. A species having a low spin diiron(III) unit with a
coordinated oxyl ligand, which is antiferromagnetically coupled to a cysteinyl
radical, has also been suggested for Q.5
The widely accepted hydroxylation mechanism for the heme protein
cytochrome P-450 similarly invokes an Fe(IV) oxo species which abstracts a
hydrogen atom from the substrate molecule.4 6 Cytochrome P-450 does not
314
hydroxylate methane, however. Very recently, Fe(IV) intermediates have
been ruled out in the mechanism for the generation of the tyrosyl radical in
the R2 protein of ribonucleotide reductase, which has an active site structure
very similar to that of the MMO hydroxylase.2 9 4 7 4 8
Instead, a diiron(III)
oxygen radical species was implicated in the oxidation of the tyrosine residue.
A similar species, perhaps the oxyl radical suggested above, which is reactive
toward CH4 may be present in MMO.
The time course freeze-quench M6ssbauer study revealed that two
types of iron centers exist in Hred. Of the = 0.7 mM concentration Hred used
in this study, only = 45% was consumed after 8 s, at which time L had been
converted
to Q.
Lower Hred concentrations,
420
M, resulted
in
approximately the same percent consumption at 8 s, thus ruling out the
possibility that dioxygen was a limiting reagent at the higher protein
concentrations.
Close examination of the M6ssbauer spectrum of Hred in
Figure 10B reveals the existence of two types of iron signals, designated
Hred(1) and Hred(2).
The high-field Mdssbauer spectrum in Figure 11A
confirms this finding. The time dependence of one of these signals, attributed
to Hred(2 ), revealed that it reacts with dioxygen with a slower rate constant of
= 0.1 s- 1. The EPR spectrum
of a freeze-quench
sample made after 8 s of
reaction time shows a distinguishable signal at g = 15, further confirming the
persistence of Hred(2 ), although accurate quantitation of this resonance was
not possible. M6ssbauer analysis of Hred revealed that Hred(l) and Hred(2) are
present in roughly equal amounts in the sample. Both forms of Hred react
with dioxygen. Hred(l) reacts rapidly and produces L, while Hred(2 ) reacts
more slowly to produce Hox. The latter reaction does not produce any
distinguishable intermediates, however. This finding is consistent with the
fact that maximal yields of only = 40% are observed in single turnover
315
reactions of nitrobenzene and Hred and dioxygen. Presumably, conversion
results only from Hred(l).
Similar heterogeneity in Hred from M.
trichosporium OB3b may not be as pronounced, since the g = 15 signal of Hred
was reported to disappear completely with a rate constant of = 22 s-1.26 On the
other hand, single turnover reactions with this system do not fully convert
nitrobenzene either.
The extinction coefficents previously reported for Q,2 4 £ = 1800 M-lcm - 1
(350 nm),
= 3600 M-lcm
-1
(420 nm), and e = 7000 M-lcm
-1
(520 nm) are
adjusted here to values of 3600, 7200, and 1400 M-lcm-1 , respectively, to reflect
the presence of the inactive Hred clusters. For L, this adjustment was already
considered in the calculation of £ : 1000 M-lcm -1 at 625 nm.
Stopped-flow kinetic studies of Q with optical detection at 420 nm
yielded rate constants for the growth and decay of this intermediate that were
fit cleanly to a biexponential equation with a single growth and a single decay
phase. These rate constants matched within error those determined from the
kinetic M6ssbauer data (Table 3). In studies with M. trichosporium OB3b, the
growth and decay of Q were best fit to a triexponential equation, which
included
two separate growth curve components.
26
The origin of these
separate processes was attributed to discrete Hred/B complexes that react with
cdioxygen with different rate constants to form Q, behavior which was not
observed in the M. capsulatus (Bath) system. The average rate constant for
growth, 1
0.1 s-l1,was about twice the value reported here for M. capsulatus
(Bath), but both MMO systems had similar decay constants of = 0.05 to 0.065
s-1. The optical spectrum of an unexplained species that formed but did not
further react was observed in the reaction with M. trichosporium OB3b, and
was subtracted from the observed spectrum to reveal that of Q.26 This species
may correspond to Hred(2 ) observed with M. capsulatus (Bath), but may be
316
present in smaller quantities, rendering its EPR signal undetectable.
Alternatively, it could be EPR silent in this system.
Kinetic traces for the formation and decay of Q were recorded under a
variety of conditions. Without protein B, very little Q was detected, as shown
in Figure 18B. Addition of protein B to Hred yielded the kinetic trace in
Figure 18A. These observations are consistent with behavior seen with
reactions carried out under single turnover conditions of Hred with
nitrobenzene, and indicate that protein B activates Hred.
Conversion of L, a diiron(III) peroxide complex, into any of the two
species proposed above for Q requires cleavage of the 0-0 bond followed by
the release of a water molecule. From the temperature dependence of the rate
constant for this process, a large ASt value of 147 + 5 Jmole-1 K-
1
calculated, which is consistent with release of a water molecule.
corresponding
was
The
ASt value of = 8 + 0.5 Jmole- 1 K- 1l for the decay of Q is low, as
expected for reaction of Q with substrate, to yield Hox and the alcohol product.
The significant
formation
AHt values, = 113 ± 4 kJmole -1 (Ea = 113 ± 4 kJmole - 1) for the
and = 75
3 kJmole- 1 (Ea = 75 + 3 kJmole - 1) for the decay, are
consistent with significant bond rearrangement occurring in these steps.
Rate constants for both the formation and decay of Q were independent
of the concentration of dioxygen. This observation is consistent with L being
the intermediate formed from the reaction of Hred with dioxygen. The rate
constant for L formation is expected to have a first-order dependence on the
concentration
of dioxygen in the system.
Study of L by stopped-flow
techniques has proved to be more difficult, owing to its lower extinction
coefficient and shorter lifetime and remain in progress. Because L formation
is most likely irreversible and proceeds with a rate constant ( 25 s- 1) which is
nearly two orders of magnitude greater than the rate constant for Q formation
317
(- 0.4 s-l), the latter should only be dioxygen-dependent
at very low 02
concentrations, which are experimentally impractical to attain.
Lack of
dioxygen dependence as measured through intermediate Q is therefore
consistent with a mechanism in which Hred reacts with 02 to produce L
which then decays to form Q.
Since the reaction catalyzed by the MMO system requires two protons,
the effects of pH and a D 20 medium on the kinetic behavior of Q were
investigated.
Except for an increase in k 3 with D2 0 to values
differences were observed.
0.145 s- 1, no
This finding does not address whether or not
proton transfer occurs in either of these steps, but rather indicates that proton
transfer is not rate determining. If proton transfer were rate determining in
either the conversion of L to Q or in the decay of Q, a significant decrease in k 2
or k 3 would be expected. Another process, such as cleavage of the 0-0 bond,
may be rate determining in Q formation.
The presence of the substrates methane, hexane, and nitrobenzene
caused much less Q to accumulate when compared to reactions run in the
absence of a hydrocarbon substrate. With methane, Q was detected only at
h:igh concentrations of Hred, and both k2 and k 3 increased. With both hexane
and nitrobenzene,
k2 was not affected significantly, as shown in Table 3,
which implies that these substrates had little effect on the conversion of L to
Q. Since k 3 increased and much less of intermediate Q was detected in all
cases, the substrates are assumed to react with Q. In no case were additional
optical bands were detected following the decay of Q. This result contrasts
with hydroxylation
of nitrobenzene in the M. trichosporium OB3b system
where an intermediate,
termed T, was detected. 2 6
Intermediate T was
observed only with nitrobenzene, and was attributed to an optical band
a:rising from the bound product alcohol, the majority of the absorbance at 430
318
nm arising from 4-nitrophenol.
This difference may reflect the different
binding affinities of the product to the active site in the two systems.
Conclusions.
Single turnover reactions of Hred with nitrobenzene are highly
dependent on the concentration of protein B in the system. Both the yield
and the rate constant of the reaction are optimized with two equivalents
present.
The regioselectivity of the reaction under these conditions differs
substantially from reactions performed under catalytic conditions with all
components present, where the reductase plays a key role. By lowering the
temperature
and omitting substrate from the above reactions, two
intermediates,
L and Q, along the reaction pathway were identified and
kinetically characterized. Structures consistent with the Mbssbauer, EPR, and
optical spectroscopic characterization of these intermediates were proposed
and related to one another in a catalytic cycle. Unusual M6ssbauer
parameters for L and Q were observed, which will require the synthesis of
appropriate model complexes for definitive structural characterization.
From the foregoing findings, we propose the catalytic hydroxylation
cycle for MMO from M. capsulatus (Bath) illustrated in Figure 26. The
reaction begins with the resting state of the enzyme, Ho. Binding of substrate
and subsequent electron transfer through the reductase yields the Hredsubstrate complex. Dioxygen next binds to Hred and forms intermediate L, a
diiron(III) peroxide.
Rearrangement of L occurs with release of a water
molecule and to produce intermediate Q, of uncertain composition and
structure, which then reacts with substrate. Product dissociation leaves Hox,
and the cycle is ready to begin again. Rate constants for each step are as
indicated in the figure, and activation parameters are consistent with this
proposal.
319
Acknowledgments.
Rapid freeze-quench samples were prepared with the
assistance of Professors B. H. Huynh and D. E. Edmondson at Emory
University.
Mssbauer
samples were analyzed by Dr. D. Wang and Prof.
Hluynh. Resonance Raman data were collected by D. Qiu and Professor T. G.
Spiro.
I thank A. M. Valentine for experimental assistance and helpful
discussions, and Dr. A. Salifogolou for providing cell paste. I am also grateful
to Dr. Y. Wang for driving to Princeton to initiate the Raman collaboration,
even though I slept almost the entire way!
320
References
(1)
Anthony, C. The Biochemistry of Methylotrophs; Academic Press: New
York, 1982, p 296-379.
(2)
Scott, D.; Brannan, J.; Higgins, I. J. J. Gen. Microbiol. 1981, 125, 63-72.
(3)
Colby, J.; Dalton, H. Biochem. J. 1978, 171, 461-468.
(4)
Fox, B. G.; Lipscomb, J. D. Biochem. Biophys. Res. Comm. 1988, 154, 165-
17'0.
(5)
Feig, A. L.; Lippard, S. J. Chem. Rev. 1994, 94, 759-805.
(6)
Que, L., Jr.; True, A. E. In Prog. Inorg. Chem.; Lippard, S. J., Eds.; John
Wiley & Sons: New York, 1990; Vol. 38; pp 97-200.
(7)
Kurtz, D. M. Chem. Rev. 1990, 90, 585-606.
(8)
Wilkins, R. G. Chem. Soc. Rev. 1992, 171-178.
(9)
Woodland,
M. P.; Patil, D. S.; Cammack,
R.; Dalton,
H. Biochim.
Biophys. Acta 1986, 873, 237-242.
(10)
Rosenzweig, A. C.; Frederick, C. A.; Lippard, S. J.; Nordlund,
P. Nature
1993, 366, 537-543.
(11)
Fox, B. G.; Hendrich,
M. P.; Surerus,
K. K.; Andersson,
K. K.; Froland,
AW.A.; Lipscomb, J. D.; Minck, E. J. Am. Chem. Soc. 1993, 115, 3688-3701.
(12)
DeRose, V. J.; Liu, K. E.; Kurtz, D. M., Jr.; Hoffman,
B. M.; Lippard,
S. J. J.
Am. Chem. Soc. 1993, 115, 6440-6441.
(13)
Hendrich,
M. P.; Fox, B. G.; Andersson,
K. K.; Debrunner,
P. G.;
Lipscomb, J. D. . Biol. Chem. 1992, 267, 261-269.
(14)
DeWitt, J. G.; Bentsen, J. G.; Rosenzweig, A. C.; Hedman, B.; Green, J.;
Pilkington, S.; Papaefthymiou,
G. C.; Dalton, H.; Hodgson, K. O.; Lippard, S. J.
J. Am. Chem. Soc. 1991, 113, 9219-9235.
(15)
Lund, J.; Dalton, H. Eur. J. Biochemistry 1985, 147, 291-296.
321
(1.6)
Lund, J.; Woodland,
M. P.; Dalton, H. Eur. J. Biochem. 1985, 147, 297-
305.
(1.7)
Green, J.; Dalton, H. Biochem. J. 1989, 259, 167-172.
(1.8)
Green, J.; Dalton, H. J. Biol. Chem. 1985, 260, 15795-15801.
(19)
Liu, K. E.; Lippard, S. J. J. Biol. Chem. 1991, 266, 12836-12839.
(20)
Paulsen,
K. E.; Liu, Y.; Fox, B. G.; Lipscomb,
J. D.; Miinck,
E.;
Stankovich, M. T. Biochemistry 1994, 33, 713-722.
(21)
Fox, B. G.; Liu, Y.; Dege, J. E.; Lipscomb, J. D. J. Biol. Chem. 1991, 266,
540-550.
(22)
Andersson,
K. K.; Froland, W. A.; Lee, S.-K.; Lipscomb, J. D. New J.
Chem. 1991, 15, 411-415.
(23)
Froland, W. A.; Andersson, K. K.; Lee, S.-K.; Liu, Y.; Lipscomb, J. D. J.
Biol. Chem. 1992, 267, 17588-17597.
(24)
Liu, K. E.; Wang, D.; Huynh, B. H.; Edmondson,
D. E.; Salifoglou, A.;
Lippard, S. J. J. Am. Chem. Soc. 1994, 116, 7465-7466.
(25)
Lee, S.-K.; Fox, B. G.; Froland, W. A.; Lipscomb, J. D.; Miinck, E. J. Am.
Chem. Soc. 1993, 115, 6450-6451.
(26)
Lee, S.-K.; Nesheim, J. C.; Lipscomb, J. D. J. Biol. Chem. 1993, 268, 21569-
21577.
(27)
Liu, K. E.; Johnson, C. C.; Newcomb, M.; Lippard, S. J. J. Am. Chem. Soc.
1992, 115, 939-947.
(28)
Pilkington, S. J.; Dalton, H. In Methods In Enzymology Academic Press:
New York, 1990; Vol. 188; pp 181-190.
(29)
Ravi, N.; Bollinger, J. M., Jr.; Huynh, B. H.; Edmondson,
Am. Chem. Soc. 1994, 116, 8007-8014.
(30)
Asa, R.; Vnngard,
T. J. Mag. Res. 1975, 19, 308-315.
D.; Stubbe, J. J.
322
(31)
Froland, W. A.; Andersson, K. K.; Lee, S.-K.; Liu, Y.; Lipscomb, J. D. In
IUCCP Symposium on Applications of Enzyme Biotechnology; Plenum Press:
College Station, TX, 1991; pp 39-54.
(32)
See discussions in refs. 9 and 31.
(33)
Liu, K. E.; Valentine, A. M.; Qiu, D.; Edmondson,
D. E.; Huynh, B. H.;
Spiro, T. G.; Lippard, S. J., to be submitted.
(34)
Brennan, B. A.; Chen, Q.; Juarez-Garcia, C.; True, A. E.; O'Connor, C. J.;
Que, L., Jr. Inorg. Chem. 1991, 30, 1937-1943.
(35)
Menage, S.; Brennan, B. A.; Juarez-Garcia,
C.; Miinck, E.; Que, L., Jr. J.
Am. Chem. Soc. 1990, 112, 6423-6425.
(36)
Micklitz, W.; Bott, S. G.; Bentsen, J. G.; Lippard, S. J. J. Am. Chem. Soc.
1989, 111, 372-374.
(37)
Greenwood, N. N.; Gibbs, T. C. In Mdssbauer SpectroscopyChapman
and Hall, Ltd.: London, 1971;pp 239-303.
(38)
Nishida, Y.; Takeuchi, M.; Shimo, H.; Kida, S. Inorg. Chim. Acta 1984,
96, 115-119.
(39)
Nishida, Y.; Takeuchi, M. Z. Naturforsch. 1987, 42B, 52-54.
(40)
Ling, J.; Nestor, L. P.; Czernuszewicz,
R. S.; Spiro, T. G.; Fraczkiewicz,
R.; Sharma, K. D.; Loehr, T. M.; Sanders-Loehr,
J. J. Am. Chem. Soc. 1994, 116,
7682-7691.
(41)
Kitajima, N.; Fujisawa, K.; Fujimoto, C.; Moro-oka, Y.; Hashimoto,
KIitagawa, T.; Toriumi, K.; Tatsumi, K.; Nakamura,
S.;
A. J. Am. Chem. Soc. 1992,
114, 1277-1291.
(42)
Iron Porphyrins; Debrunner, P. G., Ed.; VCH Publishers: New York,
1990, pp. 139-234.
(43)
Collins,
T. J.; Kostka, K. L.; Miinck, E.; Uffelman,
1.990, 112, 5637-5639.
E. S. J. Am. Chem. Soc.
323
(4.4)
Kostka, K. L.; Fox, B. G.; Hendrich,
M. P.; Collins, T. J.; Rickard,
C. E. F.;
Wright, L. J.; Miinck, E. J. Am. Chem. Soc. 1993, 115, 6746-6757.
(45)
Vogel, E.; Will, S.; Tilling,
Trautwein,
(46)
A. S.; Neumann,
L.; Lex, J.; Bill, E.;
A. X.; Wieghardt, K. Angew. Chem., Int. Ed. Engl. 1994, 33, 731-735.
Ortiz de Montellano, P. R. In Cytochrome P-450 Structure, Mechanism,
and Biochemistry; Ortiz de Montellano, P. R., Eds.; Plenum Publishing Corp.:
New York, 1986; pp 217-271.
(47)
Bollinger, J. M., Jr.; Tong, W. H.; Ravi, N.; Huynh, B. H.; Edmondson,
D. E.; Stubbe, J. J. Am. Chem. Soc. 1994, 116, 8015-8023.
(48)
Bollinger, J. M., Jr.; Tong, W. H.; Ravi, N.; Huynh, B. H.; Edmondson,
D. E.; Stubbe, J. J. Am. Chem. Soc. 1994, 116, 8024-8032.
324
0
ON
QN
I
,-
I-
r.
C)
zo
4.-
75
I-
V
A
C)
N
C
)
0
4-
o
I
-
00
n
O
Ln
o00
.
-4
a
all
cr
U
0.
,u
0>o
o
c,
U
a
r~
0)
:t:
0)
4k
0)
o
rr
0
z
-4
0
C
C)
4o
cU
~C)
OC)
m(
E
o
X.
o
cE
+1
xo
~
C)
0><
x
C)o
o
c©
o
-z
a
o.
E
o
O
~C C
~~63:
Ur4
-4
75
O
=j~
.C
C)
k
C~l
CU
:
ct
cO
fEC
cra
325
Table 2. M6ssbauer Parameters of Species Detected in Rapid Freeze-Quench
Samples from the Reaction of Hred with Dioxygen.
Species
8 + 0.02 mm/s
AEQ ± 0.03 mm/s
Ho, site 1
0.51
1.15
Ho, site 2
0.49
0.85
Hred(1)
1.3
2.8
Hred(2)
1.3
2.8
L
0.66
1.51
Q (site 1)
0.21
0.68
0.14
0.55
.~ (st)0.405
Q (site 2)
326
Table 3. Rate Constants for Formation (k2 ) and Decay (k3 ) of Intermediate Q
Under a Variety of Conditions. Unless indicated otherwise, experiments were
carried out at 4 °C.
k 2 (s- 1)
k 3 (s-1)
Mbssbauer Time Course
0.3 + 0.15
0.03 + 0.02
Stopped-Flow
0.5 + 0.15
0.07 + 0.02
Varying [02]
0.35+ 0.15
0.09 + 0.02
Varying pH
0.4 ± 0.15
0.074 + 0.02
0.35+ 0.15
0.15 + 0.02
CH4
0.95+ 0.22
0.17 + 0.02
Hexane
0.5 ± 0.15
0.13 + 0.02
Nitrobenzene
0.26± 0.15
0.11 ± 0.02
274.5 K
0.16 + 0.05
0.06 + 0.01
277 K
0.28 + 0.15
0.09 + 0.02
283 K
0.96 + 0.10
0.30 ± 0.02
288 K.
2.6 + 0.3
0.40 ± 0.02
293 K.
6.4 ± 0.4
0.48 + 0.05
298 K
19.7 ± 1.0
1.0 ± 0.1
303 K
53+4
1.8 + 0.3
318 K
132 + 10
8.35 + 1.0
Experiment
D 2 0 Exchange
(6 °C)
327
-
Table 4. Activation Parameters from Arrhenius and Eyring Plots of the
Reaction of Hred with Dioxygen.
Rate Constant
Ea
)
AH (kJ)
ASt (Jmole 1 K-1 )
k2
113
111
147
k3
75
75
8
328
CA
4-
CQ
a.
N
C.
Ca
co
~L4
329
1-
4
=
N W,
::
r 01
o
X
w
V U
o =
._
.-
;·
e
E
;rr
u
6 :;,.
.:;tl
C)
_
m
oZ
330
Figure 2. HPLC traces of single turnover reactions of Hred with nitrobenzene
and dioxygen. A control reaction carried out with no NADH is given in trace A.
Reactions with one and two equivalents of protein B present are given in traces B
and C, respectively. 2-Nitrophenol and 3-nitrophenol would elute where
indicated.
331
A.
I..
__
-__
PhNO2
B.
4-nitrophenol
-4<
"K
2-nitrophenol
3-nitrophenol
PhNO2
--C.
_.
-_..
4-nitrophenol
PhNO2
332
CZ
o
CZ
o
u
en
t~
a)
N
-e
,._
4.a
cu
xo
-3
X:
._
u
0
o
O
o .
0
"
qc;
,..
or
333
.
\.
ci
~~~pI
~ ~
0 t-~
~~
---
0
P--
ul)
Tt
4-
00
c1·
r_
a)T
. -.
Ip,~~~~~~~~·
1
0:I
I1
I
In
mn
P
N
!I
A °/0
pTaOT
I
%
0
H
o
r-.
-
334
Figure 4. Plots generated from the catalytic reaction of the complete MMO
system with nitrobenzene. The kinetic traces in A were obtained at 404 nm, and
the linear portions were used to generate plots B and C. The lag present in A is
most likely due inadequate thermal equilibration of the cuvette at 45 °C prior to
initiation of the reaction.
335
A.
1
0.8
c)
c
0.6
°
0.4
0.2
0
0
50
100
150
200
250
300
5
6
Time (sec)
B.
0.005
0.004
M 0.003
0.002
3
0.001
0
0
1
2
3
4
[Substrate] mM
C.
y = 92.785 + 486.29x R= 0.98811
I
1
c, 3440
v,
<
2580
;:
1720
u
?
860
0 !-
0
I
2
I
I
4
6
1/[Substrate] (mM')
I
8
336
Figure 5. Kinetic traces at 404 nm of the reaction of Hred with dioxygen in the
absence (trace A) and presence (trace B) of two equivalents of protein B. First
order fits are superimposed and yielded rate constants of 0.011 s - l and 0.332 s-l
for A and B, respectively.
337
A.
0.024
0.021
..
rn
0.018
u
.
0.015
©U
CL4
O
0.01.2
0
60
120
180
240
300
9.6
12
Time (sec)
B.
0.2
0.18
,.6.a
*_-4
en
0.16
©d
Cd
._u
O
0.14
0.12
0.1
0
2.4
4.8
7.2
Time (sec)
338
Figure 6. Plot of pseudo-first order rate constant of reactions of Hred with
nitrobenzene and dioxygen as a function of the amount of protein B present. The
rate constant
kobs increases from = 0.011 s - 1 to = 0.36 s - 1 .
339
,% 1-
U.Z
0.4
0.3
'-C'
CA
-0
0.2
0.1
n
0
0.5
1
1.5
2
2.5
3
3.5
Equivalents Protein B
4
340
Figure 7. (A) Missbauer
spectrum in a weak magnetic field (50 mT) of a 155 ms
rapid freeze-quench sample.
(B) M6ssbauer spectrum for L obtained after
subtracting signals for Ho,, Hred, and Q from A.
341
_
·
·
I
r
I
j
0.0
A.
I
I
II
II
1.0
II
I
Y
I
C::
I--,
.C_
II
0
II
0
-(A
u
0.0
B.
0.5
I
I
....
B
I
I
-4
-4
I
II
i
I
I
B
0
Velocity (mm/s)
I
I
4
-
I
-
342
Figure 8. Mbssbauer spectrum in a weak magnetic field (50 mT) of a 3 s rapid
f:reeze-quench sample.
343
I
·
I
·
I
r
I
·
·
I
I
I
I
I
I
I
I
I
I~~~~~~~~~~~~~~~~~~~~~~~~~~~~
0.0
J1
I0I0poioll
0.2
I
0.4
0.6
z0
0
_
0.8
I
1.0
_
1.2
_
I
II
1.4
1.6
-
I
I
=
I
-4
I
I
-2
L
I
I
0
VELOCITY (mm/s)
I
2
I
4
344
Figure 9. (A) Mdssbauer spectrum in a weak magnetic field (50 mT) of a 8 s
rapid freeze-quench sample. (B) Mbssbauer spectrum for Q obtained after
subtracting signals for Hox, Hred, and L from A.
345
_
0.0
A.
1.0
I
II
I:
v.,
v:
0.0
0.5
I
II
I
-4
I
I
0
Velocity (mm/s)
I
4
346
Figure 10. M/ssbauer
Hred (B) at 4 K.
spectrum in a weak magnetic field (50 mT) of Hox (A) and
347
A.
i
I
I
I
I
I
I
I
I
I
0.0 -
0.5
I
I
1.0
z
o 1.5
0
=2.0
I
I
2.5
I
I
I I
I I
I
I
II
I I
3.0
-4
-2
0
2
VELOCITY (mm/s)
4
348
B.
I
I
I
I
I
I
I
I
I
I
I
0.0
00010411111
[P
0.5
l
I
I
II
I
I
z 1.0
I
I
I-.
I
0
I
I
I
I
I
C
<
1.5
I
I I
I
I
2.0
I
I
I
-4
II
,
-2
I
I
0
I
I
I
2
VELOCITY (mm/s)
I
4
I
349
Figure 11. High field M6ssbauer spectra (8 T, 4 K) of Hred (A), L (B), and Q (C).
Two species, Hred(l) and Hred(2 ), were found in spectrum A, and a mixture of
these gives rise to the spectrum.
remains in progress.
Simulation of the individual components
350
I
I
I
I
I
I I
I
I
I
I
I
I
I
0.0
0.2
0.4
0.6
0.8
,
0.0
0.2
z
0
0
<
0.0
0.2
11
_
=
-
I
-6
I
I
-4
I
I
-2
I
0
2
VELOCITY (mm/s)
I
I I
4
6
I
351
._
to
--
o,-
I
oC
aD
v
m
©
'U
04
0
a
352
Ln
N
0o
U
o
an
05
C;
(l6suaa Ie;6do
XpIsua~j iumid0
353
cr.
c
5
C
4-
cn
C)
C
a
'"L
._
_
r
C)
C)
U)
t
o
C.
-)
C-o
*~
354
oo
0
U.)
0
o
b.
o
LIo
0
m, E~
0W
be
C,
3
Ln
o
CD
CD
O
o
to
6
6
o
n
0'
c
6
in
0
r
6
aueqlosqv
on
0
0
X
355
.r
a)
E
N
c)
11
II
tC
cu
cn
ocZ,
co
Q
C)
'a
cn
C.
.as
a3
cn
u
C
co
a)
CZ
-C
,.cn
C0
*v--
c~
0
._
CD
C1
356
L
C
O
*
Cfo
CD
II
'-4
V)
V)
N
CI
o
CD
ti
In)
'-.
rn
IC
In
357
as
,o
C)
aC)
cnl
,.)
0
L.)
C)
Cl
C.
O
N
aC)
C)
CD
N.
C)
C)
C.)
C.)
To
C)
Cl
X
O4
._
X
C14
358
o
Lo
ef1
0
Ori
m
359
a
X
a
0
r.
0
a)
c4
a)
U
cD
u
o,
._
0
0
cc
a)
a.
c-,
U
X
r.
360
0
0
ro0
C)
O
O
rl
o .
LfC
U)
E
0c,
X1TsuaauJ
0
I
361
Figure 17. Resonance Raman spectra of a rapid freeze-quench sample of L from
the reaction of Hred with dioxygen (from ref. 33). Excitation was carried out at
647 nm. The inset shows the 860 to 940 cmql region of the spectra obtained from
samples made with 1602 and 1802, as well as the difference spectrum.
362
MMO/L intermediate
Xo= 647.1 nm
CoDu
ag
l
l
LO
0
CO o
I
.
C
f
1
I
I
1000
80)0
600
A
LnA/
/
-1
cm
400
363
Figure 18. Plots of % M6ssbauer area of the species L (squares), Q (diamonds),
and Hox (triangles) over time. Lines represent the fit to equation (4) with k = 28
s1, k 2 = 0.46 s-, k3 = 0.026 s-.
364
A.
27
< 20
u
c
14
3
:0
. 6.8
0
0
12.8
25.6
38.4
51.2
Time (sec)
B.
"
26
<
19
C.0
c
:
13
6.4
6.4
0
42
8.4
12.6
16.S
Time (sec)
C.
'
26
-
19
t,
13
..
n
:0
6.4
0
I
I
-1.8
I
-0.9
I
0
I
0.9
log Time (sec)
I
1.8
365
;
aQO
-.
EQ
U
II
cJC;
'T,
·c~
a
yCUcJ,
u
·=aJ
.
O
aQ
°
o
,.
c
-
_O- , o
V
0)
U
¢
,O
o
O,
uO
0>_
.°
5
;Z
t
._
U
Gt
r
._,
bD
5
-
366
o
0
CD
It
u
v
0r)F
c,
0
u
e ".
*_
N
Co
r-
I
e
co
N
rq
N
i!suaa le'ldo
c
t"
o0
367
Cr4
"1
od
c1
II
Cr
II
C)
.)
.4
>-.
-
ct
-,
c.
'-,
0
u
CQ,)
Ce
U~
-rr)
0
a
4::
rd
U3-
(d
.%
.i
C.
rJ
C.
3-
3-
368
I
Ift
I
,i
N
crl
H-@-
H-U
6
d
I .'
,
c
J
l
oe
~
=
-- ~---
*
0
;
X
o z
en
X
I
O
*1
-1
I
.,
O
0
w
I
I
.o
-w
III
oNN
(I S) lUqlSUOD ao
0
369
rn
a,
0.C
o
ca)
a.
C)
-
>
o,
,-
CL)
:
C)
X
0
_
Cab
C
o
370
,
i
lw
-w
I
I
i
i
0
-I
i
__i:
:
0
-w
I
i
LO
tI
i
[~~
Ii
o0
(ITS)
*zt
D
I
of)
0
II
E~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
l~~~~~
a
6H
:UIuisuoJ caw}J
C
371
0
0
o
II
M
.11
-e
c-I
0
C'4
C/2
o4
CZ
CZ
.S
Cl
c
c.
C)
-,
0X
-u
1:
IC)
75
xcur
0,
-4
S
C14
Cl
I-
C)
._
CD
u
11:d
.4~J
C)
.5
C)
0
Cl
- as
0
oo
o
C
-,
. 4
s
H
II
CI)
372
rll
om0
oo
0
C
cm
of)
0
t
6
6° ° ~ O6 r
Xp!sula leI:ldo
373
Figure 23. An Arrhenius plot (A) and an Eyring plot (B) for the temperature
dependence of the rate constants k 2 and k3.
374
A.
y = 48.152 + -13648x R= 0.99494
I
In k2
y = 30.398 + -9060.9x R= 0.98746
-
In k3
U
4
2
0
-2
-4
I
0.0031
I
0.0032
I
0.0034
I
0.0035
0.0036
1/T
B.
-B.y = 17.657 + -13338x R= 0.99465
y = 0.91047 + -9039.3x R= 0.98752
k3
-
-24
I-,
*
__
-26
-28
-30
-30
-32
II
0.0031
I
0.0032
I
0.0034
I
0.0035
1/T
I
0.0036
I
375
oUCa)
7
uU
;n
o
c-I;
,
C
c5
1
::
O
,0
o
ta3
C.)
.,
a
C)
En
oz
c
.qq
.-,as
.
. ..
376
en
es
o
6
0
o
o
o
o
c;
o
c;~~
AIsuaa lu:)Tldo
0o
0
6
0
6
377
n
4--
u
uC
"
c
,
.
U4
0
rO
C
a)
I
1
-
o
378
0
0
I
z
I
z
n
_
T
0
0O
0
owl
P-
=0
\
I=
0
_ /0
/
1 /
II:
:L
Z
Z
0
0
Z
F,
,IO
9
O0
I
-
/
-
o
0
0
I
/z
NP7
0
o
0
I
Z
o
0
z
I
o - W_
1.4
"10
O)
(U
.4.0
N
=0 0-0
ed
F
::L
rd
wr
-W
ti
0-4
C41
\o
-
I
/
Z
o
379
Figure 26. Proposed catalytic cycle for the oxidation of hydrocarbons by MMO.
See Figure 25 for proposed structures of intermediates L and Q. Substrate first
binds to Ho, followed by the reduction of Hox to Hred. Dioxygen binds to Hred
and produces intermediate L. Rearrangement, which may include proton
transfer, yields intermediate Q. Substrate then reacts with Q to produce alcohol,
1H20, and Hox.
380
R-H
--
ROH + H2 0
FeI',O
Fe"'
H
-1
R-H
H2 0 2
R-H
Fe"Io"
shunt
pathway
H
NADH
0.31 s q1
j +
2H
R-H
Fe2 III(2 -2)
w"L' 2
Fe"'
..
I eIIo
25 Sl
R-H
02
FeII
H
-.
w,
J.
NAD'T+H'
381
Biography
Katherine Liu was born in Dayton, Ohio, on September 28, 1967. She
spent her childhood mainly in suburban Detroit, but moved with her family
to South Bend, Indiana for a brief period. Kathy attended Cornell University
in Ithaca, New York, where she obtained her B. A., majoring in chemistry. At
Cornell, Kathy was introduced to chemical research in the laboratory of Dr.
1H-ctor D. Abrufia. After graduating in 1989, she enrolled in the graduate
program in chemistry at the Massachusetts Institute of Technology in
Cambridge, Massachusetts. Her thesis work at MIT was carried out in the
laboratory of Dr. Stephen J. Lippard, and focused on the mechanistic aspects of
the enzymatic system methane monooxygenase. Kathy completed her Ph.D.
in February, 1995.