1 COENZYME A DEPENDENT AEROBIC METABOLISM OF BENZOATE VIA EPOXIDE ...

22 Pages·2010·1.19 MB·English

JBC Papers in Press. Published on May 7, 2010 as Manuscript M110.124156
The latest version is at http://www.jbc.org/cgi/doi/10.1074/jbc.M110.124156
  1
COENZYME A DEPENDENT AEROBIC METABOLISM OF BENZOATE VIA EPOXIDE
FORMATION*

Liv J. Rather1, Bettina Knapp2, Wolfgang Haehnel2, and Georg Fuchs1

From 1Lehrstuhl Mikrobiologie and 2Lehrstuhl Biochemie, Fakultät Biologie,
Schänzlestrasse 1, Universität Freiburg, D-79104 Freiburg, Germany.

Running title: CoA dependent aerobic metabolism of benzoate via an epoxide

Corresponding author. Mailing address: Mikrobiologie, Fakultät Biologie, Schänzlestr. 1, D-79104
Freiburg, Germany; Phone: [phone]; Fax: [phone]; E-mail: [email protected]
freiburg.de.

In the aerobic metabolism of aromatic Aromatic compounds like benzoate belong
substrates oxygenases use molecular oxygen to to the second most abundant class of organic
D
hydroxylate and finally cleave the aromatic growth substrates, next to carbohydrates, which ow
n
ring. In the case of the common intermediate nature provides as feedstock to animals (with lo
a
d
benzoate the ring cleavage substrates are either limited aromatic metabolism), fungi and bacteria. e
d
catechol (in bacteria) or 3,4-dihydroxybenzoate Normally benzoate is converted by dioxygenases fro
m
(protocatechuate, mainly in fungi). We have to catechol (bacteria) or protocatechuate (mainly h
shown before that many bacteria, e.g. Azoarcus fungi), followed by ring cleavage catalyzed by ttp
evansii the organism studied here, use a central ring cleaving dioxygenases. The well- ://w
w
completely different mechanism. This elaborate known β-ketoadipate pathway is the paradigm of w.jb
pathway requires formation of benzoyl- this metabolic achievement (1,2). A new pathway c.o
coenzyme A (CoA), followed by an oxygenase for aerobic benzoate oxidation has been postulated brg/
reaction and a non-oxygenolytic ring cleavage. for Azoarcus evansii and for a Bacillus y g
u
Benzoyl-CoA transformation is catalyzed by stearothermophilus-like strain (3,4). In the e
s
the iron-containing benzoyl-CoA oxygenase meantime, this pathway or its characteristic genes t on
D
(BoxB) in conjunction with an FAD and Fe-S were found in various other bacteria, either as the e
c
centers containing reductase (BoxA), which only pathway or as an additional benzoate em
b
donates electrons from NADPH. Here we show metabolic pathway that comes into play under er 2
that benzoyl-CoA oxygenase actually does not certain conditions (5,6). 1, 2
form the 2,3-dihydrodiol of benzoyl-CoA, as The new principle differs in many aspects 01
8
formerly postulated, but the 2,3-epoxide. An from the orthodox situation. First, all intermediates
enoyl-CoA hydratase (BoxC) uses two starting with benzoyl-CoA are processed as
molecules of water to first hydrolytically open coenzyme A thioesters rather than as free acids
the ring of 2,3-epoxybenzoyl-CoA, which may (4). Second, ring cleavage is hydrolytic rather than
proceed via its tautomeric seven-membered oxygenolytic (7). Third, none of the classical
oxepin ring form. Then ring-C2 is hydrolyzed enzymes is involved in any step, except
off as formic acid, yielding 3,4-dehydroadipyl- β-ketothiolase, which cleaves β-ketoadipyl-CoA
CoA semialdehyde. The semialdehyde is into succinyl-CoA and acetyl-CoA (8). Fifteen
oxidized by a NADP+-dependent aldehyde genes coding for the benzoate oxidation (box)
dehydrogenase (BoxD) to 3,4-dehydroadipyl- pathway are clustered on the A. evansii
CoA. Final products of the pathway are formic chromosome (8). These box genes code for the
acid, acetyl-CoA and succinyl-CoA. This following functions: a putative ATP-dependent
overlooked pathway occurs in 4-5 % of all benzoate transport system, a benzoate-CoA ligase,
bacteria whose genome has been sequenced and a benzoyl-CoA oxygenase and reductase, a ring-
represents an elegant strategy to cope with the opening enzyme, enzymes for β-oxidation of CoA-
high resonance energy of aromatic substrates activated intermediates, a thioesterase, and a
by forming a non-aromatic epoxide.
Copyright 2010 by The American Society for Biochemistry and Molecular Biology, Inc.
2
lactone hydrolase, as well as completely unknown dioxygenase/reductase forming the 2,3-
enzymes belonging to new protein families (8). dihydrodiol of benzoyl-CoA. BoxC acts on this
Benzoate is activated by benzoate-CoA epoxide or its oxepin tautomer by adding two
ligase forming benzoyl-CoA (8-10). Benzoyl-CoA molecules of water and thus eliminating ring-C2
conversion requires NADPH, O and two protein as formic acid. This new principle is widely
2
components, BoxA and BoxB (3). BoxA is a distributed and has a counterpart in a novel
homodimeric iron-sulfur-flavoprotein (46 kDa pathway of phenylacetyl-CoA oxidation that also
subunit), which acts as a reductase (3). In the involves coenzyme A thioesters, epoxide and
absence of BoxB, BoxA catalyzes the benzoyl- oxepin intermediates (§).
CoA stimulated artificial electron transfer from
NADPH to O to produce H O (3). EXPERIMENTAL PROCEDURES
2 2 2
Physiologically, BoxA uses NADPH to reduce Materials. Oxygen-18O (normalized, > 97 atom%)
BoxB, a monomeric 55 kDa iron-protein that acts and water-18O (normalized with respect to
as benzoyl-CoA oxygenase (11). The product of hydrogen, > 97 atom%) were obtained from
benzoyl-CoA oxidation was tentatively identified Campro Scientific (Berlin, Germany). Glucose
by NMR spectroscopy as its dihydrodiol 6-phosphate dehydrogenase from baker´s yeast
derivative, 2,3-dihydro-2,3-dihydroxybenzoyl- (268 U mg-1 protein, 0.91 mg mL-1) was obtained
D
CoA (11). This suggested that BoxAB act as a from Fluka Analytical (Buchs, Switzerland). ow
n
benzoyl-CoA dioxygenase/reductase. The Vector pASK-IBA43plus, anhydrotetracycline, lo
a
d
benzoyl-CoA oxygenase system has very low and desthiobiotin were obtained from IBA GmbH e
d
similarity to known oxygenase systems (11). (Göttingen, Germany). Restriction enzymes KpnI fro
m
Unexpectedly, benzoyl-CoA and BamHI and T4 DNA ligase were obtained h
transformation by BoxAB was greatly stimulated from Fermentas GmbH (St. Leon-Rot, Germany). ttp
when an enoyl-CoA hydratase/isomerase-like N,N-Diethyldithiocarbamate was obtained from ://w
w
protein, BoxC, was added (11). BoxC, a Sigma Aldrich (Steinheim, Germany). w.jb
homodimeric enzyme (61 kDa subunits), catalyzes Synthesis of CoA esters. Benzoyl-CoA was c.o
the hydrolytic conversion of the BoxAB product, prepared according to published procedures (4). brg/
which inactivates BoxAB, to 3,4-dehydroadipyl- Bacterial cultures. A mutant of Azoarcus evansii y g
u
CoA semialdehyde and formic acid (7). It contains KB740 (DSMZ6869) (13) with a modified e
s
domains characteristic for enoyl-CoA chromosomal boxB gene coding for BoxB with a t on
D
hydratases/isomerases, besides a large central C-terminal Strep-Tag (BoxBStrep, courtesy of J. ec
domain with no significant similarity to sequences Gescher*) (11) was grown aerobically at 37 °C em
b
in the database (7). with benzoate as the sole source of cell carbon and er 2
BoxD, a homodimer composed of 54 kDa energy (14) in a 200 L fermentor (air flow 1, 2
subunits, is a NADP+-specific aldehyde 50 L min-1; 200 rpm). Benzoate was added 01
8
dehydrogenase, which oxidizes the product of continuously when the initially added substrate
BoxC, 3,4-dehydroadipyl-CoA semialdehyde, to (5 mM) was nearly consumed. Cells were
the corresponding acid, 3,4-dehydroadipyl-CoA harvested in the exponential growth phase at an
(12). The further metabolism probably requires a optical density at 578 nm of 6, which
kind of β-oxidation leading to β-ketoadipyl-CoA, corresponded to 1.6 g of cells (dry mass) L-1 (4).
the last intermediate at which the conventional The culture was cooled to 8 °C, and cells were
β-ketoadipate pathway and the unorthodox new harvested by continuous flow centrifugation. The
pathway merge (4,7). yield was 200 g of cells (wet mass) mol-1
To elucidate the exciting mechanism of benzoate. Cells were frozen in liquid N and stored
2
benzoyl-CoA oxidation catalyzed by BoxAB we at -70 °C.
applied 18O-labeling studies and analyzed the Cloning and overexpression of the box C
His Strep
products by mass spectrometry. It turned out that gene. The gene was amplified from chromosomal
the product contained only one additional oxygen DNA of A. evansii by colony polymerase chain
and could be derivatized by an epoxide trapping reaction (PCR) with Pfu-polymerase, using
agent. In summary, the results indicate that forward primer KpnIBoxC_fwd and reverse
BoxAB is a benzoyl-CoA epoxidase forming 2,3- primer rev_BoxCBamHI: KpnIBoxC_fwd: 5’-
epoxybenzoyl-CoA, rather than a benzoyl-CoA GATCCTGGTACCCAAGCAGTCGCGAACA
3
AGC-3’; rev_BoxCBamHI: 3’-CAAGCTGACC 1-5 mL (Amicon, 30 kDa). Enzymes were stored
TTGGCGCACCCTAGGATGAAG-5’. Bold: at -70 °C with 10 % (v/v) glycerol. Box C
His Strep
boxC gene, underlined: restriction sites. The was purified similar to BoxB ; the column was
Strep
forward primer contained a KpnI restriction site washed with 60 mL buffer A and 180 mL buffer A
instead of the start codon, the reverse primer a containing 100 mM KCl. BoxD was purified as
Mal
BamHI restriction site instead of the stop codon. described (12).
The amplified gene was purified using the PCR Protein analysis. Protein concentration was
purification kit as described in the instruction determined by the Bradford method (15) and the
manual (Qiagen, Hilden, Germany). The amplified BC assay kit as described in the instruction manual
gene and the vector pASK-IBA43plus were cut (Uptima, Interchim, Montlucon Cedex, France)
with KpnI and BamHI, purified using the PCR using bovine serum albumin as standard. Sodium
purification kit, and ligated with T4 DNA ligase. dodecyl sulfate-polyacrylamide (11.5 %) gel
The vector pASK-IBA43plus contains a electrophoresis used the Laemmli method (15) and
N-terminal start codon followed by codons for a Coomassie blue staining (16).
6xHis-Tag and C-terminal codons for a Strep-Tag Stoichiometry of BoxAB reaction. NADPH
in front of the stop codon. The construct pASK- oxidation at 30 °C was measured
IBA43plus_Box C was transformed into spectrophotometrically at 377 nm
His Strep D
E. coli DH5α by electroporation and the correct (ε[NADPH]377 nm extrapolated 1,620 M-1 cm-1). ow
n
sequence was verified by restriction and sequence Either limiting concentrations of benzoyl-CoA or lo
a
d
analysis. E. coli/pASK-IBA43plus_Box C oxygen were applied. Air saturated (30 °C) assay e
His Strep d
was grown aerobically at 37 °C in lysogeny broth mixtures (500 µL) with 100 mM Tris/HCl buffer, fro
m
medium with 50 µg of ampicillin mL-1 in a 10 L pH 8.0, contained either 0.4 mM, 0.1 mM, or h
fermentor (300 rpm). At an optical density (light 0.05 mM benzoyl-CoA. They were mixed with ttp
path 1 cm) at 578 nm (OD ) of 0.5, 200 µg of 0.04 mg BoxA mL-1, 1.02 mg BoxB mL-1, and ://w
578 nm Strep w
anhydrotetracycline L-1 was added for induction. 0.96 mg BoxHisCStrep mL-1. After addition of w.jb
Cells were harvested in the exponential growth 0.6 mM NADPH the assay mixture was covered c.o
phase at an OD578 nm of 1.6. The culture was cooled immediately with 200 µL of paraffin oil to avoid brg/
down to 4 °C, cells were harvested by further oxygen uptake. NADPH oxidation y g
u
centrifugation and stored at -70 °C. The yield was observed after the BoxABC reaction caused by e
s
20 g of cells (wet mass). BoxA was subtracted. The slow endogenous, t on
D
Preparation of cell extracts. All steps were benzoyl-CoA independent NADPH oxidation with e
c
performed at 4 °C under anaerobic conditions. O , which is catalyzed by BoxA, was monitored as em
2 b
Frozen cells were suspended in an equal volume of constant slow absorption decrease at 377 nm, er 2
20 % (v/v) glycerol containing 0.05 mg of when benzoyl-CoA was consumed. This blank 1, 2
DNaseI mL-1. The suspension was passed through reaction was extrapolated back to the time point 01
8
a French pressure cell at 137 MPa and centrifuged zero. This extrapolated A value was
377 nm
(1 h, 100,000 x g). subtracted from the initial A value.
377 nm
Enzyme purification. All steps were performed at Enzyme assays for BoxAB, BoxABC, and
4 °C under anaerobic conditions. BoxA was BoxABCD. Standard assay mixtures (0.04-1 mL)
purified and assayed according to Mohamed et al. containing 0.6 mM NADPH and 0.2 mM benzoyl-
(3). BoxB was purified by affinity CoA in 10 mM Tris/HCl buffer, pH 8.0, were
Strep
chromatography. Cell extract (20-150 mL, mixed at 4 °C with 0.08 mg BoxA mL-1.
70 mg protein mL-1, 100,000 x g supernatant) was Routinely, a NADPH regenerating system was
applied to a column of Strep-Tactin Superflow included consisting of 3.3 mM MgCl , 3.3 mM
2
(25 mL, IBA GmbH, Göttingen, Germany), which D-glucose 6-phosphate, and 2 U glucose
was equilibrated with 200 mL 10 mM Tris/HCl, 6-phosphate dehydrogenase mL-1. The reaction
pH 8.0, (buffer A) at a flow rate of 3 mL min-1. was started by addition of 0.64 mg BoxB mL-1.
Strep
The column was washed with 60 mL buffer A, In some experiments, 0.16 mg Box C mL-1
His Strep
180 mL buffer A containing 250 mM KCl, and was added, and in some cases additionally 0.08 mg
60 mL buffer A. Protein was eluted with 60 mL BoxD mL-1. Labeling assays were performed in
Mal
buffer A containing 2.5 mM desthiobiotin. Eluted a closed tube (7.5 mL) with 50 % 18O / 50 % 16O
2
protein (30 mL, 20 mg) was concentrated to (v/v) gas phase or with 50 % H 18O / 50 % H 16O
2 2
4
(v/v). Assay mixtures were stirred at 24 °C for enzymatic reaction were injected with a FAMOS
30 min. The enzymatic reaction was stopped by autosampler (Dionex) and desalted by transfer
adding a fivefold volume of ethanol (-20 °C). with an Agilent HPLC 1100 to a RP-trap column
After incubation for 20 min at -20 °C denatured (Zorbax Eclipse XDB-C18, 5 µm; 0.1 x 15 mm).
protein was removed by centrifugation. The The sample was eluted at 200 nL min-1 with an
supernatant was evaporated under reduced ACN gradient from a quaternary HPLC pump
pressure at 30 °C. The residue was resolved in (Ultimate, Dionex) and separated on a fused silica
100 µL H O and the products were purified by emitter of 0.075 x 105 mm (Proxeon) packed with
2
reverse-phase HPLC. Pro C18, 3 µm (YMC). Elution started with 100 %
BoxAB reaction in the presence of A (H O / 3 % ACN / 0.1 % formic acid) and 0 %
2
N,N-diethyldithiocarbamate (DTC). Standard B (H O / 80 % ACN / 0.1 % formic acid) for
2
assay mixtures (500 µL) included 10 mM DTC. 11 min, followed by successive linear gradients in
Aliquots (40 µL) were taken at different points 4 min to 5 % B, 20 min to 30 % B, 21 min to 70 %
(t = 0, 2, 4, 6, 8, 10, 15, 20, 40 and 60 min) and the B and terminated by 2 min at 70 % B. The fused
samples were analyzed. silica emitter connected to -2 kV was mounted in
Purification of products by reverse-phase HPLC. the nano electrospray interface of the LTQ-FT.
Aliquots were applied to a column of The mass spectrometer was operated in the data
D
Lichrospher 100 RP 18E, 5.0 µm, 125 x 4 mm dependent mode to automatically switch between ow
n
(Wicom, Heppenheim, Germany), equilibrated MS and MS/MS acquisition. Survey MS spectra lo
a
d
with 40 mM ammonium acetate (NH Ac), pH 6.8, (from m/z 250 – 1800) were acquired in the FT- e
4 d
containing 5 % (v/v) acetonitrile (ACN) at a flow ICR with a resolution of 25,000. The most intense fro
m
rate of 1 mL min-1. An ACN gradient in the same ion was isolated for high resolution (50,000) h
buffer was used: 2 min 5 %, 1 min to 10 %, measurement in the FT-ICR with a 10 Da mass ttp
11 min to 30 %, 1 min change from NH Ac at range. These ions were then fragmented in the ://w
4 w
30 % to water at 30 %, 3 min to 50 %, 3 min back linear ion trap using collision induced dissociation w.jb
to 5 %, 3 min change from water plus 5 % to (CID) and recorded at low resolution (MS/MS c.o
NH4Ac plus 5 %, 6 min equilibration 5 % in scan). The latter ions were dynamically excluded brg/
NH4Ac. Elution was monitored with an UV-diode for the following 30 s. The total cycle time was y g
u
array detector routinely at 260 nm. The amount of approximately 0.3 s. e
s
the CoA thioesters was estimated based on the Analysis of oxygen exchange by mass t on
D
assumption that they exhibited identical absorption spectrometry. 3,4-Dehydroadipyl-CoA e
c
coefficients at 260 nm as benzoyl-CoA (ε = semialdehyde was enzymatically synthesized em
260 nm b
21,100 M-1 cm-1). Retention times were as follows: without 18O and purified by reverse-phase HPLC. er 2
1.0 min polar products; 3.5 min product of BoxA, 100 µL of unlabeled semialdehyde was mixed 1, 2
BoxBStrep, BoxHisCStrep, and BoxDMal; 6.4 min with an equal amount of H218O (resulting in 50 % 018
product of BoxA, BoxB , and Box C ; H 18O / 50 % H 16O) and analyzed by MS over a
Strep His Strep 2 2
7.0 min product of BoxA and BoxB ; 9.5 min period of 14 minutes after different incubation
Strep
benzoyl-CoA; 12.0 min derivative product of times (t = 1 min, 15 min, 45 min, 17 h). As control
BoxA and BoxB with DTC. Fractions of 100 µL of unlabeled semialdehyde was mixed
Strep
0.5 mL were collected and frozen at -20 °C. with 100 µL of H 16O and analyzed.
2
Analysis by mass spectrometry (MS). Samples of
HPLC fractions were transferred by a syringe RESULTS
pump or via nano-reverse-phase (RP) HPLC into Conversion of benzoyl-CoA with
the nano electrospray ionization (ESI) source of a NADPH and oxygen and characterization of the
Finnigan LTQ-FT mass spectrometer (Thermo product. Benzoyl-CoA was transformed by the
Electron Corporation, Waltham, Mass.) for online recombinant enzyme BoxB in conjunction with
Strep
mass detection assembled from a linear ion trap the native enzyme BoxA (here referred to as
and an ion cyclotron (7 Tesla magnet) with BoxAB) with O and NADPH as electron donor.
2
Fourier-transform ion cyclotron resonance mass The stoichiometry of the oxygen dependent
spectrometry. reaction was 1 NADPH oxidized and 1 O
2
Analysis by direct coupling of HPLC, MS, and consumed per 1 benzoyl-CoA transformed
MS/MS. Benzoyl-CoA and the product of the (Table 1). This ratio could be interpreted as the
5
result of a dioxygenase/reductase reaction leading additional oxygen atom. A monohydroxylated,
to the non-aromatic cis-2,3-dihydrodiol of aromatic derivative like 2-hydroxybenzoyl-CoA or
benzoyl-CoA. The product was purified by HPLC 3-hydroxybenzoyl-CoA is an unlikely candidate
and analyzed by ESI mass spectrometry (Fig. 1). for the following non-oxygenolytic cleavage of the
However, no mass of benzoyl-CoA plus two ring, and the benzoate oxidation gene cluster does
oxygen atoms and two hydrogen atoms (expected not code for a ring cleaving dioxygenase. In case
MH+ = 906) was found. The observed value of of an epoxide one may expect a hydratase yielding
m/z = 888.144 agrees well with the monoisotopic the trans-2,3-dihydrodiol of benzoyl-CoA, but a
mass of benzoyl-CoA plus one oxygen atom corresponding hydratase gene is neither found in
(expected MH+ = 888.1436). The peak at 910.127 the gene cluster.
represents the same product but associated with 18O-Labeling of the product after
Na+ instead of H+. Since not even traces of a transformation of benzoyl-CoA by BoxAB in
product carrying two additional oxygen atoms the presence of 18O . The free product of benzoyl-
2
could be observed, the product as isolated is not CoA oxygenase BoxAB cannot be unreasonably
the expected benzoyl-CoA dihydrodiol. It is rather labile; nevertheless, a transient formation of a
a hydroxylated benzoyl-CoA or an epoxide of labile dihydrodiol might be possible. If a
benzoyl-CoA. dihydrodiol transiently emerges in the course of
D
Still, the possibility exists that the actual the reaction, this should be a cis-dihydrodiol in ow
n
product of the reaction was a labile dihydrodiol, case of a dioxygenase/reductase, and a trans- lo
a
d
which during isolation becomes stabilized by dihydrodiol in case of epoxide formation followed e
d
rearomatization through water elimination, thus by water addition. To test the different options fro
m
fporromduincgt. 2W- oer 3th-heyredfroorxey buesnezdo ydli-rCeoctA caosu dpeliandg- enodf pbreenszeonycle- CoofA 5 0w %as 1t8rOans /f o5rm0 %ed 1i6nO H, 2i1s6oOla tiend tbhye http
HPLC, MS, and MS/MS in a fast experiment to reverse-phase HPLC 2and analyz2ed by mass ://w
w
minimize possible side reactions and to detect spectrometry. In the first case, half of the diol w.jb
possible dihydroxylated intermediates. Direct should be unlabeled and the other half should c.o
injection is gentle compared to freeze-drying of carry two 18O; water elimination should yield brg/
the HPLC sample and dissolving it again. The 50 % unlabeled and 50 % 18O-labeled benzoyl- y g
u
fragments of CoA served as indicator that the CoA derivative carrying one oxygen atom. In the e
s
detected parent mass even of trace unknown second case, 50 % should be unlabeled and 50 % t on
products is indeed a derivative of CoA. carrying one 18O atom; random water elimination De
c
Fragmentation of CoA in a mass spectrometer has would yield 75 % unlabeled and 25 % benzoyl- em
b
been reported, but only the mass of 428 was CoA carrying one 18O. Two molecule species with er 2
attributed to adenosine 3’,5’-bisphosphate (17). equal concentrations were observed (888.147 and 1, 2
We observed this mass and additional masses 890.151), one corresponding to a benzoyl-CoA 01
8
representing fragments of acyl-CoA thioesters. derivative carrying one additional 16O (888.1436)
The fragments contain either the aromatic ring or and the other benzoyl-CoA carrying one additional
the CoA adenylate moiety (Fig. 2). The method 18O (890.1480) (Fig. 3). Again, no molecule
proved valid when tested with benzoyl-CoA species carrying two oxygen atoms were detected.
(Fig. 2A). We observed in addition a fragment of 18O-Labeling of products after
the mass 365.2, which is likely caused by an transformation of benzoyl-CoA by BoxAB in
additional loss of phosphoric acid (97.97) the presence of H 18O or 18O and subsequent
2 2
(463.13 - 97.97 = 365.15). A related elimination of transformation by enoyl-CoA hydratase BoxC
phosphate is known for phosphoserine leading to and aldehyde dehydrogenase BoxD. The
dehydroalanine (18). The product of BoxAB products of the transformation of benzoyl-CoA
showed a minor 479.1 fragment and a major with the three enzymes BoxABC are considered
fragment of 381.2, which is likely formed from the established, 3,4-dehydroadipyl-CoA semialdehyde
minor fragment by an additional loss of and formic acid derived from C2 of the aromatic
phosphoric acid as in case of benzoate (Fig. 2B). ring (7). Formic acid cannot be detected due to its
These results indicate that the product small size (m/z < 50 cannot be detected) nor can a
detected under these gentle conditions was also a derivative of higher mass be obtained without loss
benzoyl-CoA derivative carrying only one of one oxygen atom of formate. BoxD catalyzes
6
the oxidation of the CoA linked C semialdehyde reaction was analyzed with enzymatically
6
to the dicarboxylic acid under incorporation of synthesized, HPLC purified 16O-semialdehyde in
water (12). Transformations were conducted with 50 % H 18O / 50 % H 16O and analyzed by MS as
2 2
16O in 50 % H 18O / 50 % H 16O. When BoxAB or a function of time. Indeed, the formation of the
2 2 2
BoxABC were added, solely the mass peaks of 18O-semialdehyde species was observed (Fig. 4).
unlabeled products were observed (Table 2). The In our previous preparation (see above) the
mass peaks corresponded to benzoyl-CoA carrying semialdehyde was in contact with unlabeled water
one additional oxygen atom and 3,4- for several hours, allowing the complete exchange
dehydroadipyl-CoA semialdehyde, respectively. between the carbonyl 18O of the semialdehyde and
This finding seems to exclude that oxygen from 16O from water.
water is incorporated into the C product formed Evidence for an epoxide as the
6
by BoxABC. Correspondingly, the assay with actual product of BoxAB. The possibility that a
BoxABCD yielded one product exhibiting two dihydrodiol intermediate rearomatizes under
mass peaks with equal intensity corresponding to elimination of water can be excluded, as the
unlabeled and singly 18O-labeled 3,4- product of BoxAB is cleaved by BoxC non-
dehydroadipyl-CoA (Table 2). The single oxygenolytically, which would not be possible in
18O-label therefore comes from water in the course case of an aromatic ring. The remaining option
D
of the aldehyde oxidation by BoxD. No double- how benzoyl-CoA may be activated by BoxAB ow
n
labeled product was found. using 1 O2 plus 1 NADPH is the formation of an loa
d
When benzoyl-CoA was transformed in epoxide most likely between C2 and C3 of the e
d
H216O in the presence of 50 % 18O2 / 50 % 16O2 the ring. The second oxygen atom is released as H2O. from
product of BoxABC unexpectedly showed only We set out to obtain a derivative with h
one mass peak of 878 (Table 2), which N,N-diethyldithiocarbamate (DTC), which reacts ttp
corresponds to the unlabeled 3,4-dehydroadipyl- with epoxides even under gentle enzyme assay ://w
w
CoA semialdehyde. In contrast, the product of the conditions under opening of the epoxide ring w.jb
assay with BoxABCD exhibited two mass peaks (20,21). DTC did not disturb the enzymatic c.o
of equal amplitude (Table 2). Mass peak 894 reaction. The 2,3-epoxide of the benzene ring of brg/
corresponds to unlabeled 3,4-dehydroadipyl-CoA, benzoyl-CoA, 2,3-epoxybenzoyl-CoA, is expected y g
mass peak 896 to single 18O-labeled 3,4- to yield two possible DTC adducts depending on ue
s
dehydroadipyl-CoA. Since the 18O-labeled adipyl the steric hindrance of the products: a main t on
D
C6 carboxyl group is derived from C3 of the 3,4- product (little steric hindrance) and a side product e
c
dehydroadipyl-CoA semialdehyde (Fig. 6), C3 of (more steric hindrance) (Fig. 5A). When benzoyl- em
b
the semialdehyde must have been linked to 18O. CoA was transformed at 24 °C with BoxAB, the er 2
However, the observed missing labeling of the reaction did not go to completion due to 1, 2
semialdehyde seemingly is contradictory to this inactivation of the enzymes by its product 01
8
conclusion since 50 % of this product should also (Fig. 5B) (11). This alone indicates that the
contain 18O. This inconsistency can be explained if BoxAB product is chemically reactive. When
the free carbonyl oxygen of the aldehyde (product DTC was added in excess, the reaction went to
of BoxABC) rapidly exchanges 18O with 16O from completion and nearly stoichiometric amounts of a
water via the aldehyde hydrate, when the sample new product appeared (Fig. 5B), which migrated
was prepared and the product isolated in unlabeled in reversed phase HPLC at 12-14 min. The
water. This exchange reaction is well known (19). derivative apparently did not inactivate BoxAB
3,4-Dehydroadipyl-CoA semialdehyde hydrate anymore. The new product peak was collected and
was observed before in NMR studies of the analyzed by UV-vis spectroscopy (Fig. 5C) and
reaction mechanism of BoxC (7). MS (Fig. 5D).
Determination of the rate of oxygen The UV-vis spectrum of the BoxAB product
exchange between the carbonyl oxygen of 3,4- showed an absorption maximum at 260 nm
dehydroadipyl-CoA semialdehyde and oxygen (ε ≈ 14,000 M-1 cm-1 at pH 6.8) and another
260 nm
from water. To estimate the rate of oxygen characteristic maximum at 310 nm (ε ≈
260 nm
exchange between carbonyl 18O of the 7,000 M-1 cm-1 at pH 6.8) compared to benzoyl-
semialdehyde and 16O from water, the exchange CoA (ε = 21,100 M-1 cm-1 at pH 6.8) (11)
260 nm
was measured in the opposite direction. The (Fig. 5C). The DTC derivative had nearly lost the
7
absorption at 310 nm, which would be consistent epoxide forming). The suggested trivial name is
with an opening of the epoxide (Fig. 5C). The MS benzoyl-CoA 2,3-epoxidase (E. C. 1.14.13.58).
spectrum showed a main mass peak at 1037.177 In view of the new findings, the former
that agrees well with the theoretical mass of the NMR data were reevaluated (Table 3). The
DTC adduct of 2,3-epoxybenzoyl-CoA (expected product that was studied before and in this work
mass 1037.1769) (Fig. 5D). The same derivative had the same UV-vis spectrum and therefore
was obtained when the product of BoxAB was appears to be identical. The true nature of the
isolated and then treated with DTC. This indicates product may well be 2,3-epoxybenzoyl-CoA
that the epoxide is not a transiently formed (7-oxabicyclo [4.1.0] hepta-2,4-diene-2-carboxyl-
intermediate in the BoxAB catalyzed reaction, but CoA), as the comparison between data based and
it is a true product. Benzene epoxide is known to rule based chemical shifts indicates. Still, these
be in equilibrium with its tautomeric oxepin form NMR data do not prove the proposed structure.
(22). 2,3-Epoxybenzoyl-CoA might undergo an NMR spectroscopy in this instance does not allow
epoxide-oxepin valence tautomerism (Fig. 5E and discriminating unambiguously between a cis-
Fig. 6). An oxepin is an unsaturated seven- dihydrodiol and an epoxide. This is because both
membered heterocycle having six carbon atoms, compounds exhibit cis carbon-oxygen bonds at C2
one oxygen atom and three double bonds. This and C3 of the ring. The tautomeric oxepin has no
D
oxepin is not aromatic as it does not obey the carbon-carbon bond between C2 and C3, thus it ow
n
Hückel rule. We conclude that benzoyl-CoA is would have different chemical shifts and signal lo
a
d
oxidized by BoxAB to an epoxide, which is non- splitting (Table 3), which were not observed. e
d
aromatic, thus allowing a non-oxygenolytic ring Taking into account the DTC derivative and the fro
m
cleavage catalyzed by the following enzyme spectral properties of the product, we consider the h
BoxC. BoxAB therefore is not a epoxide as the true product and not only as a ttp
dioxygenase/reductase, but an epoxide forming transient intermediate. BoxB has a counterpart in ://w
w
monooxygenase, an epoxidase. phenylacetyl-CoA oxygenase, which also forms an w.jb
epoxide of an aromatic CoA thioester that is c.o
DISCUSSION converted to an oxepin form (§). Thus, this brg/
Product of BoxAB. In previous work common unprecedented epoxide formation y g
u
the product of the oxygen and NADPH dependent represents a new paradigm of aerobic aromatic e
s
transformation of benzoyl-CoA by BoxAB was metabolism. t on
D
tentatively assigned to cis-2,3-dihydro-2,3- Occurrence of the new benzoate e
c
dihydroxybenzoyl-CoA and the enzyme system pathway and relation of benzoyl-CoA em
b
BoxAB was named benzoyl-CoA, oxygenase (BoxB) to other enzymes. A BLAST er 2
NADPH:oxygen oxidoreductase (2,3- search (BLASTP 2.2.22+) (23,24) with BoxBC 1, 2
hydroxylating) (11). Although the NMR data were from A. evansii (NCB Accession # Q9AIX7, NCB 01
8
consistent with a cis-diol configuration, they did Accession # Q84HH6) revealed that 4-5 % of all
not provide unequivocal evidence (11). In the fully sequenced eubacterial genomes (mostly α-
conventional metabolism of aromatic compounds, and β-proteobacteria) harbor the two key genes of
a cis-dihydrodiol undergoes oxidation and the CoA dependent benzoate oxidation pathway,
rearomatization to a dihydroxy aromatic product, which is lacking in Archaea (Suppl. 1). For
catalyzed by a diol dehydrogenase. This comparison, 7 % of the species harbor benzoate
dihydroxylated aromatic central intermediate is 1,2-dioxygenase benABC and cis-diol
then cleaved by ring cleaving dioxygenases. Yet, a dehydrogenase benD genes characteristic for the
putative diol dehydrogenase gene could not be classical benzoate pathway involving ring cleaving
found in the benzoate oxidation gene cluster. In dioxygenases. Some species contain even both
addition, BoxB amino acid sequence shows no options (1.3 %) (Suppl. 1), which might be
similarity to known oxygenase subunits of ring required for high turnover of benzoate or if
hydroxylating or ring cleaving dioxygenases. This reduced oxygen tension is present, as has been
study revealed that the BoxAB enzyme system suggested for Burkholderia xenovorans LB400
forms an epoxide and should be renamed benzoyl- (6,25). These percentages indicate that the new
CoA, NADPH:oxygen oxidoreductase (2,3- pathway is not a minor route. The amino acid
8
sequence similarity/identity for BoxB is in the essential concerted acid/base catalysts (7,44). It is
range between 97-92 % and 72-57 %, respectively. obvious from the 18O -labeling experiment that
2
The active center of BoxB is a dinuclear BoxC opens the 18O-2,3-epoxide by adding OH-
iron center ($) resembling the active site of soluble regiospecifically at ring-C2 of the epoxide or its
methane monooxygenase (E. C. 1.14.13.25, PDB oxepin tautomer, leading to a common seven-
1MMO) (26), ribonucleoside-diphosphate membered ring (Fig. 6). This results in 18O being
reductase (E. C. 1.17.4.1, PDB 1RIB) (27), linked to ring-C3, which gives rise to the C6
multicomponent phenol hydroxylase (E. C. carbonyl group of 3,4-dehydroadipyl-CoA
1.14.13.7, PDB 2INN) (28), toluene/o-xylene semialdehyde formed by BoxC, in which 18O was
monooxygenase (PDB 1T0Q) (29), ∆9 stearoyl- retained. The electron-withdrawing effect of the
acyl carrier protein desaturase (E. C. 1.14.99.6, CoA activated carboxy group facilitates ring
PDB 1AFR) (30), p-aminobenzoate N-oxygenase opening of the epoxide by addition of OH- to form
(PDB 3CHH) (31), and alkene monooxygenase a dialdehyde. The enolate anion intermediate is
(E. C. 1.14.13.69) (32). Soluble methane stabilized by an oxyanion hole characteristic for
monooxygenase (33,34) and alkene enoyl-CoA hydratases/isomerases (7,44).
monooxygenase (32,35) are also able to form Protonation at ring-C1 prepares the intermediate
epoxides. The dinuclear iron center in general is for the next addition of OH- at ring-C2, which
D
thought to be able to perform epoxidation (36,37). leads to elimination of the C2-atom as formic acid. ow
n
It should be stressed that other epoxide forming Finally, a second protonation at ring-C1 leads to lo
a
d
enzymes exist that are not related to BoxB. the product 3,4-dehydroadipyl-CoA semialdehyde e
d
Examples are cytochrome P450 (E. C. 1.14.14.1) (7). The production of formic acid explains the fro
m
(38), vitamin-K reductase (E. C. 1.1.4.1 & 1.1.4.2) odd finding that genes for enzymes of benzoate h
(39,40), squalene monooxygenase (E. C. and formate metabolism are co-induced (5-7,45). ttp
1.14.99.7) (41,42), and zeaxanthine epoxidase N-terminal and C-terminal domains of BoxC are ://w
w
(E. C. 1.14.13.90) (43). characteristic for enoyl-CoA hydratases/ w.jb
Possible mechanism of BoxAB and isomerases (crotonase) protein family. This c.o
catalysis of BoxC revisited (Fig. 6). BoxAB mechanism is consistent with the common brg/
introduces one oxygen atom from molecular mechanism of the enoyl-CoA hydratase/isomerase y g
u
oxygen into benzoyl-CoA to form 2,3- (crotonase) protein family, which catalyzes a e
s
epoxybenzoyl-CoA. The other oxygen atom is variety of different reactions including enoyl-CoA t on
D
eliminated as water, which fuels the reaction and hydration, enoyl-CoA isomerization and C-C bond e
c
renders it irreversible. Oxygen becomes activated cleavage, all based on abstraction/addition of the em
b
at the diiron center and the enzyme structure and α-proton of the carboxylic acid and the reversible er 2
properties of the active site are topics of current syn-addition of water to enoyl-CoA-ester (46). 1, 2
studies ($). BoxC converts the product of BoxAB, Summing up the new pathway. 01
8
2,3-epoxybenzoyl-CoA, possibly via its oxepin Benzoate appears to be transported into the
form, to an aldehyde plus formic acid by bacterium by an ABC-transporter system and
integration of two water molecules (Fig. 6). BoxC immediately becomes activated by benzoate-CoA
was formerly termed benzoyl-CoA-dihydrodiol ligase forming benzoyl-CoA (Fig. 6) (8-10). In the
lyase and should be renamed 2,3-epoxybenzoyl- following reactions all intermediates are
CoA dihydrolase. The quick removal of the coenzyme A thioesters. This may be advantageous
reactive epoxide by BoxC is probably vital. In the as the CoA activated group has an electron-
absence of BoxC BoxAB become inactivated in withdrawing effect, which stabilizes negative
the course of the reaction; in vivo BoxC may even charge and thus activates the aromatic ring.
form a complex with BoxB (11). Furthermore, thioester formation allows an
The crystal structure of BoxC without efficient trapping of aromatic acids within the cell
substrate was solved and benzoyl-CoA 2,3- and CoA-intermediates - especially the epoxide -
dihydrodiol was modeled into the active center might be less toxic. As the energy-rich thioester
(44). This model needs revision in view of the true bond is retained in the products, the energy
epoxide substrate. Water addition may be initially spent is not lost. Benzoyl-CoA 2,3-
catalyzed consecutively by the same amino acids. epoxidase BoxAB catalyzes the introduction of
Mainly two glutamate residues function as one oxygen atom to form 2,3-epoxybenzoyl-CoA.
9
The 2,3-epoxybenzoyl-CoA dihydrolase BoxC aerobic benzoate degradation via CoA ligation
integrates two water molecules to form the open follows the equation:
chain intermediate 3,4-dehydroadipyl-CoA Benzoate + ATP + 2 CoA + O + 3 H O + NAD+
2 2
semialdehyde, formic acid is split off. 3,4- → Acetyl-CoA + Succinyl-CoA + Formic acid +
Dehydroadipyl-CoA semialdehyde dehydrogenase AMP + PP + NADH + H+.
i
BoxD oxidizes the semialdehyde to its Aerobic benzoate degradation via the
corresponding acid, 3,4-dehydroadipyl-CoA β-ketoadipate pathway follows the equation:
(4,12). Modified β-oxidation leads to Benzoate + CoA + 2 O + H O → Acetyl-CoA +
2 2
β-ketoadipyl-CoA, which is finally cleaved into Succinate + CO .
2
acetyl-CoA and succinyl-CoA by β-ketoadipyl- Hence, the new pathway uses less oxygen and
CoA thiolase (8). The overall stoichiometry of produces reduced products, NADPH and formic
acid.

REFERENCES
1. Harwood, C. S., and Parales, R. E. (1996) Annu. Rev. Microbiol. 50, 553-590
2. Ornston, L. N., and Stanier, R. Y. (1966) J. Biol. Chem. 241, 3776-3786
3. Mohamed, M. E., Zaar, A., Ebenau-Jehle, C., and Fuchs, G. (2001) J. Bacteriol. 183, 1899-1908
D
4. Zaar, A., Eisenreich, W., Bacher, A., and Fuchs, G. (2001) J. Biol. Chem. 276, 24997-25004 ow
n
5. Denef, V. J., Patrauchan, M. A., Florizone, C., Park, J., Tsoi, T. V., Verstraete, W., Tiedje, J. M., lo
a
d
and Eltis, L. D. (2005) J. Bacteriol. 187, 7996-8005 e
d
6. Denef, V. J., Klappenbach, J. A., Patrauchan, M. A., Florizone, C., Rodrigues, J. L., Tsoi, T. V., fro
m
Verstraete, W., Eltis, L. D., and Tiedje, J. M. (2006) Appl. Environ. Microbiol. 72, 585-595 h
7. Gescher, J., Eisenreich, W., Wörth, J., Bacher, A., and Fuchs, G. (2005) Mol. Microbiol. 56, ttp
1586-1600 ://w
w
8. Gescher, J., Zaar, A., Mohamed, M., Schägger, H., and Fuchs, G. (2002) J. Bacteriol. 184, 6301- w.jb
6315 c.o
9. Schühle, K., Gescher, J., Feil, U., Paul, M., Jahn, M., Schägger, H., and Fuchs, G. (2003) brg/
J. Bacteriol. 185, 4920-4929 y g
u
10. Kawaguchi, K., Shinoda, Y., Yurimoto, H., Sakai, Y., and Kato, N. (2006) FEMS Microbiol. Lett. e
s
257, 208-213 t on
D
11. Zaar, A., Gescher, J., Eisenreich, W., Bacher, A., and Fuchs, G. (2004) Mol. Microbiol. 54, 223- e
c
238 em
b
12. Gescher, J., Ismail, W., Ölgeschläger, E., Eisenreich, W., Wörth, J., and Fuchs, G. (2006) er 2
J. Bacteriol. 188, 2919-2927 1, 2
13. Anders, H. J., Kaetzke, A., Kämpfer, P., Ludwig, W., and Fuchs, G. (1995) Int. J. Syst. Bacteriol. 01
8
45, 327-333
14. Tschech, A., and Fuchs, G. (1987) Arch. Microbiol. 148, 213-217
15. Coligan, J. E., Dunn, B. M., Ploegh, H. L., Speicher, D. W., and Wingfield, P. T. (1995) John
Wiley & Sons, New York, N.Y.
16. Zehr, B. D., Savin, T. J., and Hall, R. E. (1989) Anal. Biochem. 182, 157-159
17. Dalluge, J. J., Gort, S., Hobson, R., Selifonova, O., Amore, F., and Gokarn, R. (2002) Anal.
Bioanal. Chem. 374, 835-840
18. Bennett, K. L., Stensballe, A., Podtelejnikov, A. V., Moniatte, M., and Jensen, O. N. (2002)
J. Mass. Spectrom. 37, 179-190
19. Rétey, J., Umani-Ronchi, A., and Arigoni, D. (1966) Experientia 22, 72-73
20. Dupard-Julien, C. L., Kandlakunta, B., and Uppu, R. M. (2007) Anal. Bioanal. Chem. 387, 1027-
1032
21. Stanior, U., and Wiessler, M. (1984) Arch. Pharm. 317, 1042-1047
22. Vogel, E., and Günther, H. (1967) Angew. Chem., Int. Ed. Engl. 6, 385-401
23. Altschul, S. F., Madden, T. L., Schäffer, A. A., Zhang, J., Zhang, Z., Miller, W., and Lipman, D.
J. (1997) Nucleic. Acids Res. 25, 3389-3402
10
24. Altschul, S. F., Wootton, J. C., Gertz, E. M., Agarwala, R., Morgulis, A., Schäffer, A. A., and Yu,
Y. K. (2005) FEBS J. 272, 5101-5109
25. Pérez-Pantoja, D., De la Iglesia, R., Pieper, D. H., and Gonzàlez, B. (2008) FEMS Microbiol.
Rev. 32, 736-794
26. Rosenzweig, A. C., Frederick, C. A., Lippard, S. J., and Nordlund, P. (1993) Nature 366, 537-543
27. Nordlund, P., Sjoberg, B. M., and Eklund, H. (1990) Nature 345, 593-598
28. Sazinsky, M. H., Dunten, P. W., McCormick, M. S., DiDonato, A., and Lippard, S. J. (2006)
Biochemistry 45, 15392-15404
29. Sazinsky, M. H., Bard, J., Di Donato, A., and Lippard, S. J. (2004) J. Biol. Chem. 279, 30600-
30610
30. Lindqvist, Y., Huang, W., Schneider, G., and Shanklin, J. (1996) EMBO J. 15, 4081-4092
31. Choi, Y. S., Zhang, H., Brunzelle, J. S., Nair, S. K., and Zhao, H. (2008) Proc. Natl. Acad. Sci.
USA 105, 6858-6863
32. Gallagher, S. C., Cammack, R., and Dalton, H. (1997) Eur. J. Biochem. 247, 635-641
33. Colby, J., Stirling, D. I., and Dalton, H. (1977) Biochem. J. 165, 395-402
34. Murray, L. J., and Lippard, S. J. (2007) Acc. Chem. Res. 40, 466-474
35. Small, F. J., and Ensign, S. A. (1997) J. Biol. Chem. 272, 24913-24920
D
36. de Visser, S. P. (2008) Chemistry 14, 4533-4541 ow
n
37. Musaev, D. G., Basch, H., and Morokuma, K. (2002) J. Am. Chem. Soc. 124, 4135-4148 lo
a
d
38. Isin, E. M., and Guengerich, F. P. (2008) Anal. Bioanal. Chem. 392, 1019-1030 e
d
39. Lee, J. J., and Fasco, M. J. (1984) Biochemistry 23, 2246-2252 fro
m
40. Mukharji, I., and Silverman, R. B. (1985) Proc. Natl. Acad. Sci. USA 82, 2713-2717 h
41. Cory, E. J., Russey, W. E., and Ortiz de Montellano, P. R. (1966) J. Am. Chem. Soc. 88, 4750- ttp
4751 ://w
w
42. Nakano, C., Motegi, A., Sato, T., Onodera, M., and Hoshino, T. (2007) Biosci. Biotechnol. w.jb
Biochem. 71, 2543-2550 c.o
43. Büch, K., Stransky, H., and Hager, A. (1995) FEBS Lett. 376, 45-48 brg/
44. Bains, J., Leon, R., and Boulanger, M. J. (2009) J. Biol. Chem. 284, 16377-16385 y g
u
45. Rabus, R., Kube, M., Heider, J., Beck, A., Heitmann, K., Widdel, F., and Reinhardt, R. (2005) e
s
Arch. Microbiol. 183, 27-36 t on
D
46. Eberhard, E. D., and Gerlt, J. A. (2004) J. Am. Chem. Soc. 126, 7188-7189 e
c
em
b
FOOTNOTES er 2
*We thank Dr. Wolfgang Eisenreich for simulation of 13C-NMR data and Dr. Johannes Gescher for the 1, 2
BoxBStrep mutant of A. evansii, which was cloned in the same manner as the BoxBHis mutant (11). This 018
work was supported by the Deutsche Forschungsgemeinschaft (FU 118/16-3 and HA 1084/9-1) and the
Graduiertenkolleg Biochemie der Enzyme.
§ R. Teufel, V. Mascaraque, W. Ismail, M. Voss, J. Perera, W. Eisenreich, W. Haehnel and G. Fuchs,
unpublished results
$ L. Rather, T. Weinert, U. Demmer, E. Bill, U. Ermler, and G. Fuchs, unpublished results

The abbreviations used are: CoA, coenzyme A; box, benzoate oxidation gene cluster; BoxAB, native
enzyme BoxA in conjunction with recombinant enzyme BoxB ; BoxABC, native enzyme BoxA in
Strep
conjunction with recombinant enzyme BoxB and recombinant enzyme Box C ; BoxABCD, native
Strep His Strep
enzyme BoxA in conjunction with recombinant enzyme BoxB , recombinant enzyme Box C and
Strep His Strep
recombinant enzyme BoxD ; A. evansii, Azoarcus evansii; DTC, N,N-diethyldithiocarbamate; HPLC,
Mal
high performance liquid chromatography; MS, mass spectrometry; MS/MS, second mass spectrometry of
the most abundant parent ion which was fragmented by collision induced dissociation; LTQ-FT, linear
ion trap with Fourier-transform ion cyclotron resonance mass spectrometry; FT-ICR, Fourier-transform
ion cyclotron resonance mass spectrometry.