The non-heme diiron alkane monooxygenase of Pseudomonas oleovorans (AlkB) hydroxylates via a substrate radical intermediate [12]

Full text of DOI:10.1021/ja001500v

J. Am. Chem. Soc. 2000, 122, 11747-11748
11747
The Non-Heme Diiron Alkane Monooxygenase of
Scheme 1. Mechanism of Norcarane Hydroxylation by AlkB
Pseudomonas oleoWorans (AlkB) Hydroxylates via a
Substrate Radical Intermediate
Rachel N. Austin, Hung-Kuang Chang,§
†,‡
§
,†
Gerben J. Zylstra, and John T. Groves*
Department of Chemistry, Princeton UniVersity
Princeton, New Jersey 08544
Biotechnology Center for Agriculture and the EnVironment
Rutgers UniVersity, New Brunswick New Jersey 08901
ReceiVed May 1, 2000
One of the most remarkable oxidations found in nature is the
insertion of oxygen into a carbon-hydrogen bond. Although the
chemical inertness of paraffins has been well-known to chemists
for centuries, the hydroxylation of these materials is a common
biological strategy. The soil organism Pseudomonas oleoVorans
1
TF4-1L (ATCC 29347) can grow on octane as its sole source
2
of carbon, a facility that has been successfully directed toward
3
the large-scale production of octanol from octane. The mono-
oxygenase AlkB from TF4-1L is a dinuclear non-heme iron
4
enzyme that catalyzes the terminal hydroxylation of simple
alkanes (C -C12). This is the initial step in the metabolic process
5
by which TF4-1L obtains energy from alkanes. Recent genetic
experiments have suggested that enzymes with high homology
to AlkB, especially in the histidine-rich region thought to be
nutrients, glucose, and ampicillin (TOP10). E. coli was induced
with IPTG. Both organisms were incubated with substrates for
23 h. After transformation, the supernatant was extracted and the
products assayed directly by GC-MS. The retention times and
fragmentation patterns of authentic products were compared to
those of the identified peaks in the GC-MS spectra.
5
essential for iron binding, are widely distributed in nature. Hence,
mechanistic insights into AlkB hydroxylation may shed light on
a significant portion of the alkane transformations that take place
in the environment and also on the differences and similarities
between this enzyme and two, more extensively studied alkane
hydroxylating enzymes, cytochrome P450 and soluble methane
monooxygenase (sMMO).
Diagnostic substrates that undergo characteristic structural
changes in response to forming specific reaction intermediates
have been very useful in elucidating key features of a number of
metalloenzymes and model compounds.7 2-Methyl-1-phenyl-
cyclopropane, which has been used previously as a probe for AlkB
reactivity,11 can undergo very rapid cyclopropyl carbinyl-
We have investigated the mechanism of alkane hydroxylation
catalyzed by AlkB using the diagnostic substrates norcarane
-10
(bicyclo[4.1.0]heptane) and 2-methyl-1-phenylcyclopropane, both
in wild-type TF4-1L and Escherichia coli expressing the cloned
alkane hydroxylase genes. The results give conclusive evidence
for a carbon-centered radical intermediate with a lifetime of
approximately 1 ns in the hydroxylation process.
1
1 -1 11
homoallyl radical rearrangement (k
r
) 2 × 10 s ). However,
this probe cannot differentiate unambiguously between cation and
radical intermediates. Norcarane can conclusively distinguish
between radical and cationic pathways since the 2-norcaranyl
radical is known to open so as to afford the 3-cyclohexenylmethyl
radical.2 By contrast, the 2-norcaranyl cation is a resonance
hybrid with the 3-cyclohepten-1-yl cation, giving products from
The oxidations of both substrates were performed using
growing cells from both TF4-1L and E. coli TOP10-(pGJZ1371).
The latter plasmid was constructed by cloning an alkBFG PCR
product encoding the alkane hydroxylase and its associated
rubredoxins along with an alkT PCR product encoding the
rubredoxin reductase into the expression vector pTrcHis2-TOPO.
The use of TF4-1L, E. coli expressing the alkane hydroxylase,
and a negative control of E. coli with the expression vector only
allow us to unequivocally attribute the chemistry reported here
to the AlkB protein.
6a
1
2
both systems, depending upon the nucleophile (Scheme 1).
Thus, norcarane allowed the determination that oxomanganese
porphyrins, as models of cytochrome P450, hydroxylated with
discreet radical intermediates.8
While diagnostic probes have been widely used, recent experi-
ments have demonstrated that subtle considerations can obscure
the information they yield.1
3-19
For this reason, we reexamined
Wild-type TF4-1L and E. coli TOP10-(pGJZ1371) were grown
6
in mineral salts basal (MSB) medium with octane (TF4-1L) or
2-methyl-1-phenylcyclopropane as a substrate with TF4-1L. We
†
Princeton University.
(6) Stanier, R. Y.; Palleroni, N. J.; Duodoroff, M. J. Gen. Microbiol. 1966,
43, 159-271.
‡
Permanent Address: Department of Chemistry, Bates College.
§
Rutgers University.
(7) Ortiz de Montellano, P. R.; Stearns, R. A. J. Am. Chem. Soc. 1987,
109, 3415-3420.
(1) Recent 16S rRNA sequencing by van Beilen and Witholt has reclassified
Pseudomonas oleoVorans TF4-1L as P. putida (GenBank accession AJ249825).
To facilitate further communication, we have elected to use the older name
and acknowledge the nomenclature issue here.
(8) Groves, J. T.; Kruper, W. J.; Haushalter, R. C. J. Am. Chem. Soc. 1980,
102, 6375-6377.
(9) Groves, J. T.; McClusky, G. A.; White, R. E.; Coon, M. J. Biochem.
Biophys. Res. Commun. 1978, 81, 154-160.
(
2) Peterson, J.; Basu, D.; Coon, M. J. Biol. Chem. 1966, 241, 5162-
164.
3) (a) Bosetti, A.; van Beilen, J. B.; Preusting, H.; Lageveen, R.; Witholt,
5
(10) Newcomb, M.; Tadic-Biadatti, M. H.; Chestney, D. L.; Roberts, E.
S.; Hollenberg, P. F. J. Am. Chem. Soc. 1995, 117, 12085-12091.
(11) Fu, H.; Newcomb, M.; Wong, C.-H. J. Am. Chem. Soc. 1991, 113,
5878-5880.
(
B. Enzyme Microb. Technol. 1992, 14, 702-708. (b) Schmid, A.; Kollmer,
A.; Mathys, R.; Witholt, B. Extremophiles 1998, 2, 249-256.
(
4) Shanklin, J.; Achim, C.; Schmidt, H.; Fox, B.; M u¨ nck, E. Proc. Natl.
(12) Friedrich, E. C.; Jassawalla, J. D. C. J. Org. Chem. 1979, 44, 4224.
(13) Choi, S. Y.; Eaton, P. E.; Hollenberg, P. F.; Liu, K. E.; Lippard, S. J.;
Newcomb, M.; Putt, D. A.; Upadhyaya, S. P.; Xiong, Y. J. Am. Chem. Soc.
1996, 118, 6457-6555.
Acad. Sci. U.S.A. 1997, 94, 2981-2986.
5) Smits, T. H. M.; Rothlisberger, M.; Witholt, B.; van Beilen, J. B.
EnViron. Microbiol. 1999, 1, 307-317.
(
1
0.1021/ja001500v CCC: $19.00 © 2000 American Chemical Society
Published on Web 11/08/2000

11748 J. Am. Chem. Soc., Vol. 122, No. 47, 2000
Communications to the Editor
as those of cytochrome P450 and sMMO, it is reasonable to
presume that oxygen binding and reduction can lead to a reactive
ferryl species, OdFe(IV)-Fe(IV),24 a two-electron oxidant with
respect to the resting ferric enzyme. Hydrogen abstraction from
the alkane will lead to the intermediate substrate radical. The rate
of the norcaranyl radical rearrangement is relatively slow on the
25
8 -1 26
scale of well-characterized radical clocks, ∼2 × 10 s . Thus,
the observation of both ring-opened and ring-closed products
provides an opportunity to estimate the lifetime of the intermediate
radical, ∼1 ns. This estimate comes from consistent results of 15
experiments involving both organisms (see Supporting Informa-
tion). This lifetime is consonant with the observation of only ring-
opened products in the case of AlkB oxidation of 2-methyl-1-
phenylcyclopropane.
Figure 1. GC-MS total ion trace in the product region for the oxidation
of norcarane by E. coli TOP10(pGJZ1371).
also used it as a substrate for the first time with a cloned organism,
E. coli TOP10(pGJZ1371). Control experiments showed 1-phenyl-
The nanosecond lifetime for the substrate intermediate observed
here is certainly long enough to qualify this species as a distinct
intermediate and supports the hypothesis that non-heme iron
3
-buten-1-ol, trans-2-phenylcyclopropylmethanol, and trans-2-
phenylcyclopropyl carboxylic acid were not further transformed.
2
1
-Methyl-1-phenylcyclopropane yielded only 1-phenyl-3-buten-
-ol, with both TF4-1L and TOP10(pGJZ1371).
Norcarane was found to be a good substrate for AlkB both in
23
enzymes can participate in the oxygen rebound mechanism. The
20
unambiguous presence of the radical ring-opened product in the
hydroxylation of norcarane enables us to conclusively identify a
radical pathway. We note that a stepwise radical process would
also explain the unusual stereochemical outcome for the epoxi-
dation of 1-octene mediated by AlkB described in the pioneering
TF4-1L and in E. coli TOP10(pGJZ1371), affording a mixture
of products. Approximately 4% of the starting material was
metabolized to products over the course of the reaction giving
strong, well-resolved peaks in the GC-MS (Figure 1). For both
the cloned and wild-type organism, of the total amount of
norcarane-derived products, approximately 85% were cis- and
trans-2-norcaranol and 15% were 3-(hydroxymethyl)-cyclohexene,
the ring-opened radical product. In the longer incubations with
TF4-1L, a small amount of norcaranone was detected. When
present, it was included in estimates of radical lifetimes. E. coli
TOP10(pGJZ1371), which does not have an overexpressed alcohol
dehydrogenase and has only an inefficient native alcohol dehy-
drogenase21 did not generate further oxidation products.22 Neither
the cationic ring-opened product cycohept-3-ene-1-ol nor its
corresponding ketone was detected. A series of control experi-
ments with both observed and non-observed products produced
only a small amount of norcaranone from incubations of TF4-1L
with norcaranol. No further oxidation products were detected with
E. coli.
27
work on P. oleoVorans by May.
It is interesting to speculate on the observed differences be-
tween the conclusions presented here and results obtained with
15,28
4
sMMO.
Such variables as the iron coordination sphere, the
strength of the scissile C-H bond, and possible spin state crossing
29
effects are in need of careful scrutiny.
Acknowledgment. We (all authors) thank the National Science
Foundation for funding through the Environmental Molecular Science
Institute, CEBIC (Center for Environmental Bioinorganic Chemistry)
NSF9810248 and CHE9814301 (J.T.G.). G.J.Z. thanks Exxon Research
and Engineering for initial funding on alkane oxidation. R.N.A. thanks
Bates College for granting and supporting a sabbatical leave. We dedicate
this paper to the 100th anniversary of the discovery of free radicals by
Moses Gomberg.
Supporting Information Available: Figure showing estimates of
radical lifetimes with norcarane as substrate (PDF). This material is
available free of charge via the Internet at http://pubs.acs.org.
Taken together, the products of these oxidations are consistent
with an oxygen rebound scenario2 for hydroxylation by AlkB as
shown in Scheme 1. Since the stoichiometry of AlkB is the same
3
JA001500V
(
23) Groves, J. T. J. Chem. Educ. 1985, 928-932. (b) McLain, J. L.; Lee,
(
14) Jin, Y.; Lipscomb, J. D. Biochemistry 1999, 38, 6178-6186.
J.; Groves, J. T. In Biomimetic Oxidations Catalyzed by Transition Metal
Complexes; Meunier, B., Ed.; Imperial College Press: London, 2000; pp 91-
169.
(
15) Choi, S.-Y.; Eaton, P. E.; Kopp, D. A.; Lippard, S. J.; Newcomb, M.;
Shem, Runnan. J. Am. Chem. Soc. 1999, 121, 12198-12199.
(
16) Toy, P.; Newcomb, M.; Hager, L. Chem. Res. Toxicol. 1998, 11, 816-
23.
17) Atkinson, Jeffrey K.; Hollenberg, Paul F.; Ingold, K. U.; Johnson, C.
(24) (a) Feig, A.; Lippard, S. J. Chem. ReV. 1994, 94, 759-805. (b) Shu,
L.; Neshein, J. C.; Kauffman, K.; M u¨ nck, E.; Lipscomb, J. D.; Que, L., Jr.
Science 1997, 275, 515-518.
(25) (a) Bowry, V. W.; Lusztyk, J.; Ingold, K. U. J. Am. Chem. Soc. 1991,
113, 5687-5698. (b) Newcomb, M.; Le Tadic-Biadatti, M.-H.; Chestney, D.
L.; Roberts, E. S.; Hollenberg, P. F. J. Am. Chem. Soc. 1995, 117, 12085-
12091.
(26) The rate of norcaranyl radical ring opening was calculated from product
ratios for the reaction of 2-chloronorcarane with tri-n-butyltin hydride. (a)
Friedrich, E. C.; Holmstead, R. L. J. Org. Chem. 1972, 37, 2550. (b) Rahm,
A.; Amardeil, R.; Degueil-Castaing, M. Organomet. Chem 1989, 371, C4-
C8. (c) Carlsson, D. J.; Ingold, K. U. J. Am. Chem. Soc. 1968, 90, 7047-
7055.
8
(
C.; Le Tadie, M.-H.; Newcomb, M.; Putt, D. A. Biochemistry 1994, 33,
1
0630-10637.
18) Toy, P. H.; Newcomb, M.; Hollenberg, P. F. J. Am. Chem. Soc. 1998,
20, 7719-7729.
19) Newcomb, M.; Shen, R.; Choi, S.-Y.; Toy, P. H.; Hollenberg, P. F.;
(
1
(
Vaz, A. D. N.; Coon, M. J. J. Am. Chem. Soc. 2000, 122, 2677-2686.
(20) Identification and quantification of 1-phenyl-3-buten-1-ol was straight-
forward. An upper limit of 1% trans-(2-phenylmethyl-cyclopropyl)methanol
was determined by measuring the intensity of m/e 117 at the retention time
of this product. Other MS peaks were uncharacteristic of this product,
indicating that little or none was formed.
(27) Katopodis, A. G.; Wimalasena, K.; Lee, J.; May, S. W. J. Am. Chem.
Soc. 1984, 106, 7928-7935.
(28) Wallar, B. J.; Lipscomb, J. D. Chem. ReV. 1996, 96, 2625-2657.
(29) Schr o¨ der, D.; Shaik, S.; Schwarz, H. Acc. Chem. Res. 2000, 35, 139-
145.
(
21) Brand, J. B.; Cruden, D. L.; Zylstra, G. J.; Gibson, D. T. Appl. EnViron.
Microbiol. 1992, 58, 3407-3409.
22) van Beilen, J. B.; Kingma, J.; Witholt, B. Enzyme Microb. Technol.
994, 16, 904-911.
(
1