Mechanistic studies on the conversion of arylamines into arylnitro compounds by aminopyrrolnitrin oxygenase: identification of intermediates and kinetic studies

Article abstract of DOI:10.1002/anie.200502903

(Chemical Equation Presented) Rieske business! The Rieske N-oxygenase, aminopyrrolnitrin oxygenase (PrnD), catalyzes the unusual oxidation of arylamines to arylnitro compounds. Kinetic analysis has confirmed that PrnD catalyzes the conversion of the substrate pABA (4-aminobenzylamine) through at least three consecutive reactions: two monooxygenation steps and one dehydrogenation step (see scheme).

Full text of DOI:10.1002/anie.200502903

Communications
Enzyme Catalysis
DOI: 10.1002/anie.200502903
Mechanistic Studies on the Conversion of
Arylamines into Arylnitro Compounds by
Aminopyrrolnitrin Oxygenase: Identification of
Intermediates and Kinetic Studies**
Jungkul Lee and Huimin Zhao*
Rieske oxygenases are widespread in nature and catalyze a
diverse set of oxidation reactions including cis-dihydroxyla-
tion, monohydroxylation, desaturation, sulfoxidation, and O-
and N-dealkylation.[1,2] Recently, we reported the character-
ization of a Rieske N-oxygenase, aminopyrrolnitrin oxygen-
ase (PrnD), which catalyzes unusual arylamine oxidation
reactions.[3] Although arylamine oxidases seem to be widely
distributed and used in a variety of metabolic reactions,[4–9]
PrnD represents one of only two known examples of aryl-
amine oxidases or N-oxygenases involved in the formation of
nitro groups, the other being AurF involved in aureothin
biosynthesis.[9] PrnD is involved in the biosynthesis of anti-
biotic pyrrolnitrin[3-chloro-4-(2’-nitro-3’-chlorophenyl)pyr-
role] (1), which is produced by many Pseudomonads and
displays broad-spectrum antifungal activity.[10] In the pro-
posed biosynthetic pathway of 1 (Scheme 1),[11] PrnD cata-
lyzes the oxidation of the amino group of aminopyrrolnitrin
(2) to a nitro group to form 1. Although PrnD has been
characterized, no experimental evidence has been available
for the mechanism of conversion of the arylamine. Therefore,
Scheme 1. Proposed biosynthetic pathway for pyrrolnitrin 1.[11]
[*]Dr. J. Lee, Prof. H. Zhao
Department of Chemical and Biomolecular Engineering
University of Illinois at Urbana-Champaign
600 S. Mathews Ave., Urbana, IL 61801 (USA)
Fax: (+1)217-333-5052
E-mail: zhao5@uiuc.edu
[**]This research was supported by a grant from the Office of Naval
Research (N00014–02-1-0725). We thank Ee Lui Ang for his help in
homology modeling. J.L. was partially supported by the Korea
Research Foundation Grant funded by the Korean Government
(Basic Research Promotion Fund, M01-2004-000-10159-0).
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
622
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 622 –625

Angewandte
Chemie
we decided to investigate the intermediates formed during the
conversion of arylamine into arylnitro compounds, as these
might provide valuable mechanistic information.
in the dioxygenation reaction, it is necessary to show that the
partial reaction does occur and that it does so at a rate equal
to or greater than the rate of the overall reaction. The
intermediacy of the hydroxylamine and nitroso compounds
was further assessed through their chemical synthesis and
kinetic characterization.
Rieske oxygenases are typically comprised of two protein
components: a terminal oxygenase and a flavin reductase.[12]
Thus, a reductase was deemed necessary for the function of
PrnD. On the basis of the nonspecificity of the flavin
reductase,[21] the unknown reductase for PrnD from the
Pseudomonas strain was substituted by flavin reductase SsuE
from E. coli, with optimization of the concentration of SsuE
for PrnD activity.[3] As compound 2, the physiological
substrate for PrnD, is not commercially available and is
difficult to obtain in large amounts either from natural
sources or by chemical synthesis, we used an alternative
substrate 4-aminobenzylamine (pABA) to investigate the
conversion mechanism of the arylamine. To observe the
intermediates, PrnD was rapidly mixed with a 1 mm solution
of pABA at 158C to slow the reaction, and the reaction
mixture was analyzed every 30 min. Rieske [2Fe-2S] cluster-
reconstituted PrnD[3] in combination with SsuE reductase
showed a significant conversion of the substrate pABA to
produce several compounds that migrated in HPLC with
retention times of 4.9 min (authentic 4-nitrobenzylamine
(pNBA)), 1.9 min (product A), and 4.5 min (product B;
Figure 1). These products were absent if heat-denaturated
The intermediates pHABA and pNOBAwere synthesized
from pNBA and pABA, respectively, as described previ-
1
ously[14,15] and were characterized by H NMR spectrometry
and high-resolution (EI) mass spectrometry.[16] Initial-velocity
studies were performed with pHABA and pNOBA as
variable substrates in the presence of a fixed concentration
of PrnD ( ꢀ 1 mm). The data presented in Table 1 demonstrate
Table 1: Kinetic constants for aminopyrrolnitrin oxygenase.[a]
Substrate
kcat [minÀ1
]
Km [mm]
kcat/Km [mmÀ1 minÀ1
]
2
6.8Æ0.59
6.5Æ0.62
7.9Æ0.98
84.5Æ9.72
191Æ20
379Æ34
404Æ56
580Æ73
0.036Æ0.004
0.017Æ0.002
0.019Æ0.003
0.146Æ0.021
pABA
pHABA
pNOBA
[a] Km =Michaelis constant.
that PrnD catalyzes the oxygenation of pHABA and pNOBA
to the common product pNBA, indicating that the rate of
oxygenation of the intermediates was high enough for them to
be intermediates in the dioxygenation of the arylamine
catalyzed by PrnD. As could be predicted from the fact that
only small amounts of nitroso intermediate can be detected in
the reaction mixture catalyzed by PrnD, the turnover rate of
the nitroso compound is approximately 13-fold greater than
the overall rate of dioxygenation. Thus, PrnD catalyzes at
least three consecutive reactions (Scheme 2), in which the
Scheme 2. Proposed mechanism for bioconversion of arylamines into
arylnitro compounds by PrnD.
Figure 1. HPLC analysis of the PrnD reaction product. Each product
was identified by LC-MS/MS analysis, and the fragmentation patterns
are indicated.
conversion of arylamine into arylnitro compounds proceeds
in two monooxygenation steps and one dehydrogenation step,
via hydroxylamine and nitroso compounds as intermediates.
The involvement of radical species in the reaction catalyzed
by PrnD was excluded in the previous studies.[3]
To investigate the kinetic competence of intermediate
pHABA, several PrnD mutants were prepared. The position
of site-directed mutagenesis was decided based on homology
modeling followed by analysis of the substrate-binding
pocket. A homology model of PrnD was built based on the
structure of the a-subunit of the naphthalene dioxygenase of
Pseudomonas sp. strain NCIB 9816-4—the Rieske oxygenase
for which a crystal structure was determined.[17] The a-subunit
contains a Rieske [2Fe-2S] center and mononuclear non-
heme-iron site, which is believed to be the site of activation of
molecular oxygen and substrate oxygenation.[18] In consider-
PrnD was used or if pABA was omitted from the assay which
indicated that the new metabolites were derived enzymati-
cally from pABA. LC-MS/MS analysis of products A and B
revealed that these compounds showed the calculated masses
([M+H]+: m/z 139 and 137) of 4-hydroxylaminobenzylamine
(pHABA) and 4-nitrosobenzylamine (pNOBA), respectively.
Furthermore, it showed the characteristic fragmentation
pattern of a hydroxylamine derivative (m/z 122 [MÀ16]+
and 105 [MÀ33]+) and
a nitroso derivative (m/z 120
[MÀ16]+), respectively. Thus, pHABA and pNOBA can be
considered as intermediates or dead-end products in the
dioxygenation of pABA by PrnD. To establish that the
hydroxylamine and nitroso compounds are the intermediates
Angew. Chem. Int. Ed. 2006, 45, 622 –625
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
623

Communications
ation of the hydrophobicity of the residues and the distance
the PrnD-catalyzed oxygenation reaction of arylamines,
substantiating for the first time the catalytic mechanism for
the conversion of arylamines into arylnitro compounds. The
intermediates are at least partially released from the active
site of the enzyme during catalysis, providing a rare example
of three consecutive chemical reactions catalyzed by one
active site. This may be the primary mechanism by which
arylamines are oxygenated to give arylnitro compounds in
biological processes. The characterization of PrnD adds a new
and interesting member to the family of Rieske nonheme-iron
oxygenases and demonstrates the pathway for the formation
of arylnitro compounds in nature.
between the substrate and the residues in the substrate-
binding pocket, three residues (D269, F312, W209) were
mutated. The catalytic activities of PrnD mutants of D269A,
F312V, and W209A toward pHABA were determined and
compared with those of pABA. Wild-type (WT) PrnD
exhibited similar activity toward pHABA and pABA, con-
sistent with the notion that formation of intermediate
pHABA is a rapid step in WT PrnD. F312V and W209A
mutants showed no conversion of either pHABA or pABA.
However, D269A mutant PrnD showed much higher activ-
ities toward the intermediate pHABA (kcat = 8.4 minÀ1, kcat
/
Km = 0.020 mmÀ1 sÀ1) than pABA (kcat = 0.3 minÀ1, kcat/Km =
0.004 mmÀ1 sÀ1). These results indicate that the hydroxylamine
intermediate is kinetically competent. In addition, the accu-
mulation of the hydroxylamine intermediate in WT PrnD
indicates that its formation is fast relative to its decay. The
lack of significant accumulation of the nitroso intermediate
suggests that it decays faster than it is formed. Thus, the rate-
limiting step is the dehydrogenation of pHABA to give
pNOBA. Of note, in the absence of NADPH flavin reductase
system, the hydroxylamine intermediate was not oxidized by
O2 to the nitroso intermediate and H2O2 was not formed.
Artificial electron acceptors such as methylene blue and
ferricyanide were less than 10% as effective as O2 under
anoxic conditions. In addition, the D269A mutation signifi-
cantly slows down the formation of hydroxylamine, suggest-
ing that Asp269 is involved in the early steps of the reaction
and likely responsible for the abstraction of a proton from the
substrate. A homology model shows that Asp269 is properly
positioned in the substrate-binding pocket to perform the
proposed function.
To confirm whether one or both oxygen atoms of O2 were
incorporated into the product, we carried out isotope-labeling
experiments with 18O2. Incorporation of the label was
analyzed by ESI/LC-MS. In the first experiment, PrnD was
incubated with pABA in an 18O2 atmosphere. Unlabeled
pNBA showed the molecular ion, [M+H]+, at m/z 153. In
contrast, most of the pNBA produced under the 18O2
atmosphere revealed the molecular ion at m/z 157, indicating
the incorporation of two 18O atoms. In addition, incorporation
of one 18O atom into pHABA was demonstrated by the signal
for the molecular ion at m/z 141, in comparison to m/z 139 for
the unlabeled pHABA. For the second labeling experiment,
unlabeled pHABA and pNOBA were incubated in an 18O2
atmosphere with PrnD. The incorporation of one 18O atom
into pNBAwas demonstrated by the molecular ion at m/z 155.
In addition, no 18O was incorporated into pNOBA in the
reaction of pHABA with PrnD, ruling out the possibility of
the involvement of the dihydroxylamine intermediate. Thus
two 18O atoms are sequentially incorporated into the product,
confirming that the dioxygenation by PrnD proceeds in a
consecutive monooxygenase-type reaction. Note that some
iron(ii)-dependent oxygenases have been reported to be
bifunctional or even trifunctional, catalyzing several consec-
utive oxidative transformations within a single biosynthetic
pathway.[19,20]
Experimental Section
Site-directed mutagenesis of pTKXb-PrnD was carried out by using
the QuikChange site-directed mutagenesis kit from Stratagene (La
Jolla, CA). The pTKXb-prnD plasmid was used as the DNA
template.[3] Trp209 and Asp269 residues were mutated to Ala
individually. Residue Phe312 was mutated to Val. The plasmids
containing the correct mutant genes were cloned into pMAL-c2x and
then used to transform E. coli BL21(DE3), and colonies selected by
ampicillin resistance were used for protein expression. The PrnD
mutants were expressed, purified, and reconstituted with Fe2+ and S2À
according to the same procedure for the wild-type enzyme, as
described elsewhere.[3]
Enzyme activity was routinely assayed by HPLC. The assay
mixture (0.5 mL final volume) contained 500 mm NADPH, 3 mm FMN,
500 mm substrate, and SsuE and PrnD (SsuE/PrnD molar ratio of 4.0)
in 20 mm TrisHCl at pH 7.8 and was stirred at 308C. Reactions were
initiated by addition of PrnD to the reaction mixture and analyzed by
HPLC. One unit of activity was defined as the amount of enzyme
forming 1 mmol of product per minute at 308C under standard assay
conditions, calculated from the rate of substrate depletion. Kinetic
parameters determined in atmospheric oxygen were obtained by
fitting the data to the Michaelis–Menten equation.
For labeling studies with 18O2, two vials, one containing substrate
and one containing holo-PrnD reaction mixture, were degassed by
application of a vacuum and flushed with argon three times. The
anaerobic holo-PrnD solution was transferred to the vial containing
the substrate. The argon was removed by application of a vacuum, and
finally 18O2 was allowed to enter into the vial. After incubation for 1 h
at 308C, the reaction was analyzed by HPLC coupled to an
electrospray ionization (EI) mass spectrometer (TSQ Quantum,
ThermoFinnigan, San Jose, CA) in positive-ion mode. A linear
gradient of MeOH (0–75%) in aqueous acetic acid (0.1%) was used.
Products of enzyme reactions were analyzed by an Agilent 1100
Series HPLC System. The sample was eluted on a ZORBAX SB-C8
Column (4.6 ” 150 mm2, Agilent). HPLC parameters were as follows:
258C; solvent A: 1% acetic acid in water; solvent B: methanol;
gradient: 5% B for 2 min; then to 100% B in 18 min, and finally
maintain at 100% B for 2 min; flow rate: 1.0 mLminÀ1; detection by
UV absorbance at 254 nm. Hydrogen peroxide was determined by
monitoring the oxidation of 2,2’-azino-bis(3-ethylbenzthiazoline-6-
sulfonic acid) in the presence of horseradish peroxidase at 725 nm, as
described previously.[21]
Received: August 16, 2005
Published online: December 12, 2005
In this study, we have obtained direct evidence for the
involvement of hydroxylamine and nitroso intermediates in
Keywords: amines · enzyme catalysis · oxygenation ·
reaction mechanisms · reactive intermediates
.
624
www.angewandte.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 622 –625

Angewandte
Chemie
[1] M. Costas, M. P. Mehn, M. P. Jensen, L. Que, Chem. Rev. 2004,
104, 939.
[2] S. M. Resnick, K. Lee, D. T. Gibson, J. Ind. Microbiol. Biotech-
nol. 1996, 17, 438.
[3] J. K. Lee, M. Simurdiak, H. Zhao, J. Biol. Chem. 2005, 280,
36719.
[4] C. Rebstock, H. J. Crooks, J. Controulis, Q. Bartz, J. Am. Chem.
Soc. 1949, 71, 2458.
[5] M. Morris, G. Pagan, H. Warmke, Science 1954, 119, 322.
[6] G. C. Lancini, D. Kluepfel, E. Lazzari, G. Sartori, Biochim.
Biophys. Acta 1966, 130, 37.
[7] R. R. King, C. H. Lawrence, L. A. Calhoun, Phytochemistry
1998, 49, 1265.
[8] B. Meunier, S. P. de Visser, S. Shaik, Chem. Rev. 2004, 104, 3947.
[9] J. He, C. Hertweck, J. Am. Chem. Soc. 2004, 126, 3694.
[10] P. E. Hammer, D. S. Hill, S. T. Lam, K. H. VanPꢀe, J. M. Ligon,
Appl. Environ. Microbiol. 1997, 63, 2147.
[11] K. H. Van Pꢀe, O. Salcher, F. Lingens, Angew. Chem. 1980, 92,
855; Angew. Chem. Int. Ed. Engl. 1980, 19, 828.
[12] J. R. Mason, R. Cammack, Annu. Rev. Microbiol. 1992, 46, 277.
[13] S. Keller, T. Wage, K. Hohaus, M. Holzer, E. Eichhorn, K. H.
Van Pꢀe, Angew. Chem. 2000, 112, 2380; Angew. Chem. Int. Ed.
2000, 39, 2300.
[14] C. Theodore, F. G. William, J. Org. Chem. 1972, 37, 741.
[15] R. H. Richard, P. B. Raymond, J. Am. Chem. Soc. 1960, 82, 3454.
[16] pHABA: 1H NMR ([D6]acetone, 500 MHz): d = 2.92 (br s, NH2),
1.91 (br s, NHOH), 4.38 (d, 2H), 6.63 (d, 2H), 7.02 (d, 1H),
7.28 ppm (d, 1H); HR-EI-MS: m/z calcd for pHABA [M+]:
138.07931; found: 138.07914. pNOBA: 1H NMR ([D6]acetone,
500 MHz): d = 3.39 (br s, NH2), 4.18 (s, 2H), 7.81 (d, 2H),
8.23 ppm (d, 2H); HR-EI-MS: m/z calcd for pNOBA [M+]:
136.06367; found: 136.06366.
[17] B. Kauppi, K. Lee, E. Carredano, R. E. Parales, D. T. Gibson, H.
Eklund, S. Ramaswamy, Structure 1998, 6, 571.
[18] C. S. Butler, J. R. Mason, Adv. Microb. Physiol. 1997, 38, 47.
[19] Z. H. Zhang, J. S. Ren, D. K. Stammers, J. E. Baldwin, K. Harlos,
C. J. Schofield, Nat. Struct. Biol. 2000, 7, 127.
[20] L. D. Thornburg, M. T. Lai, J. S. Wishnok, J. Stubbe, Biochem-
istry 1993, 32, 14023.
[21] D. E. Ward, C. J. Donnelly, M. E. Mullendore, J. van der Oost,
W. M. de Vos, E. J. Crane III, Eur. J. Biochem. 2001, 268, 5816.
Angew. Chem. Int. Ed. 2006, 45, 622 –625
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
625