Báo cáo khoa học: 2-Hydroxyisocaproyl-CoA dehydratase and its activator from Clostridium difficile pdf

2-Hydroxyisocaproyl-CoA dehydratase and its activator
from Clostridium difficile
Jihoe Kim, Daniel Darley and Wolfgang Buckel
Laboratorium fu
¨
r Mikrobiologie, Fachbereich Biologie, Philipps-Universita
¨
t, Marburg, Germany
2-Hydroxyacyl-CoA dehydratases are the key enzymes
in the fermentation pathways of 12 proteinogenous
amino acids to ammonia, CO
2
, short chain fatty acids
and in some cases molecular hydrogen [1]. In Acidami-
nococcus fermentans (Clostridiales), Clostridium sym-
biosum and Fusobacterium nucleatum glutamate is
oxidized to 2-oxoglutarate and ammonia, reduced to
(R)-2-hydroxyglutarate and transformed to (R)-2-hy-
droxyglutaryl-CoA, which is reversibly dehydrated to
(E)-glutaconyl-CoA. Subsequent decarboxylation leads
to crotonyl-CoA, which disproportionates to acetate,
butyrate and H
2
[2]. The 2-hydroxyglutaryl-CoA dehy-
dratase, also called component D, requires activation
by component A, the activator or initiator, which
transfers one electron to the dehydratase concomitant
with hydrolysis of ATP. It has been postulated that
further transfer of the electron to the substrate initiates
the syn-elimination of water via radical intermediates
[3,4]. Upon completion of the catalytic cycle the elec-

tron is thought to be recycled to the next incoming
substrate enabling many turnovers without further
ATP hydrolysis. The extremely oxygen-sensitive com-
ponent A from both, A. fermentans [5] and F. nuclea-
tum [6], are homodimeric enzymes with one [4Fe )4S]
cluster bound between the two subunits, with each
capable of binding one ATP. The dehydratases also
contain [4Fe)4S] clusters; the enzymes from A. fermen-
tans [7] and F. nucleatum [8] contain one, whereas in
the enzyme from C. symbiosum two such clusters have
been detected [9]. Components D from A. fermentans
and C. symbiosum are heterodimers and contain in
addition to the [4Fe)4S] cluster about one mole of
riboflavin-5¢-phosphate (FMN) as well as small
amounts of riboflavin and molybdenum [7,9]. In con-
trast, the dehydratase from F. nucleatum lacks FMN
and molybdenum but contains riboflavin and is
Keywords
ATP; iron–sulfur; leucine fermentation;
electron recycling; radical mechanism
Correspondence
W. Buckel, Laboratorium fu
¨
r Mikrobiologie,
Fachbereich Biologie, Philipps-Universita
¨
t,
35032 Marburg, Germany
Fax: +49 6421 28 28979
Tel: +49 6421 28 21527

E-mail:
(Received 4 August 2004, revised 12
November 2004, accepted 22 November
2004)
doi:10.1111/j.1742-4658.2004.04498.x
The hadBC and hadI genes from Clostridium difficile were functionally
expressed in Escherichia coli and shown to encode the novel 2-hydroxyiso-
caproyl-CoA dehydratase HadBC and its activator HadI. The activated
enzyme catalyses the dehydration of (R)-2-hydroxyisocaproyl-CoA to
isocaprenoyl-CoA in the pathway of leucine fermentation. The extremely
oxygen-sensitive homodimeric activator as well as the heterodimeric dehy-
dratase, contain iron and inorganic sulfur; besides varying amounts of zinc,
other metal ions, particularly molybdenum, were not detected in the dehy-
dratase. The reduced activator transfers one electron to the dehydratase
concomitant with hydrolysis of ATP, a process similar to that observed
with the unrelated nitrogenase. The thus activated dehydratase was separ-
ated from the activator and ATP; it catalyzed about 10
4
dehydration turn-
overs until the enzyme became inactive. Adding activator, ATP, MgCl
2
,
dithionite and dithioerythritol reactivated the enzyme. This is the first
demonstration with a 2-hydroxyacyl-CoA dehydratase that the catalytic
electron is recycled after each turnover. In agreement with this observation,
only substoichiometric amounts of activator (dehydratase ⁄ activator ¼ 10
mol ⁄ mol) were required to generate full activity.
Abbreviations
FldA, CoA-transferase; FldBC, phenyllactyl-CoA dehydratase; FMN, riboflavin-5¢-phosphate; HadBC, 2-hydroxyisocaproyl-CoA dehydratase;
HadI, initiator, activator or archerase of HadBC; ICP-AES, inductively coupled plasma-atomic emission spectroscopy.

550 FEBS Journal 272 (2005) 550–561 ª 2004 FEBS
composed of three subunits [8]; the extra subunit is not
related to any known protein [10].
Besides 2-hydroxyglutaryl-CoA dehydratases, enz-
ymes catalyzing the dehydration of lactyl-CoA to acry-
loyl-CoA from Clostridium propionicum [11,12] and
phenyllactyl-CoA to cinnamoyl-CoA from Clostridium
sporogenes [13] have also been purified. Whereas the
lactyl-CoA dehydratase system resembles that of 2-hy-
droxyglutaryl-CoA dehydratase from C. symbiosum,
phenyllactyl-CoA dehydratase (FldBC) forms a com-
plex with a highly specific class III CoA-transferase
(FldA). The complex FldABC catalyses the overall
dehydration of (R)-phenyllactate to cinnamate in the
presence of catalytic amounts of cinnamoyl-CoA after
activation by ATP, MgCl
2
and a reducing agent medi-
ated by FldI [13,14]. Our studies with phenyllactate
dehydratase revealed a similar arrangement of homo-
logous genes in the genome of Clostridium difficile,
designated as hadA, hadI, hadB and hadC, for hydroxy-
acyl-CoA dehydratase [13]. Upstream of hadA an open
reading frame in the opposite direction (ldhA) was detec-
ted encoding a putative d-2-hydroxy acid
dehydrogenase (Fig. 1). We speculated that these genes
could be involved in the fermentation of leucine, the pre-
ferred substrate of C. difficile [14,15]. Three moles of
leucine are fermented by this organism to a mixture of
fatty acids; two moles of leucine are reduced to

isocaproate, whereas one mole is oxidized to isovalerate
and CO
2
(Eqn 1) ([16,17]; for structures, see Fig. 2).
3 lÀLeucine þ 2H
2
O ¼ 3NH
þ
4
þ CO
2
þ isovalerate
À
þ 2 isocaproate
À
Eqnð1Þ
DG°¢ ¼ )146 kJÆreaction
)1
[18].
A proposed pathway is shown in Fig. 2. The forma-
tion of isocaproate should proceed via the dehydration
of (R)-2-hydroxyisocaproyl-CoA to 2-isocaprenoyl-
CoA. In this paper we describe the expression of hadI
and hadBC in Escherichia coli and characterize the
respective gene products as functional activator ⁄
initiator and 2-hydroxyisocaproyl-CoA dehydratase,
respectively.
Fig. 1. Gene arrangement for the 2-hydroxyisocaproyl-CoA dehydra-
tase system of C. difficile. ldhA, hydroxyisocaproate dehydroge-
nase; hadA, isocaproyl-CoA: 2-hydroxyisocaproate CoA-transferase;

hadI, activator of the dehydratase; hadBC, dehydratase, acdB; acyl-
CoA dehydrogenase; etfBA, electron transferring flavoprotein
Fig. 2. Proposed L-leucine fermentation
pathway of C. difficile. LdhA, (R)-2-hydroxy-
isocaproate dehydrogenase; HadA,
isocaproyl-CoA: 2-hydroxyisocaproate
CoA-transferase; HadI, activator of dehydra-
tase; HadBC, 2-hydroxyisocaproyl-CoA
dehydratase; Fd
–
, reduced ferredoxin.
J. Kim et al. 2-Hydroxyisocaproyl-CoA dehydratase
FEBS Journal 272 (2005) 550–561 ª 2004 FEBS 551
Results
Cloning of the genes, hadI and hadBC
The genes, hadI for activator or initiator and hadBC
for the two subunits of the dehydratase, were identified
in the gene cluster of the putative 2-hydroxyisocap-
royl-CoA dehydratase system of C. difficile as des-
cribed earlier [14]. For gene cloning, PCR primers
were designed on the basis of identified ORFs. In the
case of hadI and hadB the second in frame ATG start
codons were chosen, because their distance of nine and
seven nucleotides, respectively, from the Shine–Dal-
garno sequences [19] were similar to those in the ldhA,
hadA and hadC genes (Fig. 3). The primers for cloning
into the expression vectors (pASK-IBA7 or 3) con-
tained a cleavage site for the restriction enzyme BsaI
and provided an 8-amino acid Strep-tag II peptide
(Trp-Ser-His-Pro-Gln-Phe-Glu-Lys) on the C-terminal

ends of the proteins for one-step purification. The nuc-
leotide sequence (810 bp) of the cloned hadI showed
two nucleotide substitutions at C303fiT and
A645fiG, but the encoded 266 amino acids were
100% identical to those of the sequence from the
Sanger Center (see acknowledgement), GenBank acces-
sion number AY772815. The genes hadB and hadC
encoding the two subunits of the dehydratase were
amplified as one fragment and cloned into pASK-
IBA3 giving Np3 BC. The nucleotide sequence of the
cloned hadBC was composed of 2354 bp encoding 783
amino acids; three silent nucleotide substitutions were
found at G285fiA, T870fiC and T2003fiC, GenBank
accession number AY772816.
Activator of 2-hydroxyisocaproyl-CoA
dehydratase, HadI
HadI was identified by homology analysis with
the known activators of C. sporogenes (FldI) and
A. fermentans (HgdC) showing 55 and 51% amino
acid sequence identities, respectively. Similar to HgdC
[7] and FldI [14], HadI was produced in E. coli
through gene expression and purified by affinity
chromatography, because these activators are extre-
mely sensitive against oxygen and difficult to purify in
sufficient amounts from the original organism. In
order to produce the activator of 2-hydroxyisocaproyl-
CoA dehydratase HadI from C. difficile, E. coli cells
harbouring the Np3I plasmid were grown and induced
under anaerobic conditions. The harvested cells were
opened by a French Press to avoid heating the sensi-

tive enzyme by sonication and the produced protein
fused with a C-terminal Strep-tag II peptide was puri-
fied by one-step affinity column. The pure activator
was eluted in buffer (see Experimental procedures)
containing 1 mm ADP and 10 mm MgCl
2
to maintain
stability (Fig. 4). By chemical analysis, 4 ± 0.5 nonh-
eme iron and 2 ± 0.1 acid-labile sulfur were detected
indicating one [4Fe)4S] cluster in the isolated protein.
The low observed sulfur content presumably resulted
from loss of H
2
S during storage of the extremely labile
[4Fe)4S] protein. The UV-visible spectrum of the puri-
fied activator as isolated showed a maximum around
370 nm (Fig. 5). Reduction of the [4Fe)4S]
2+
cluster
with 11 equivalents of dithionite gave a shoulder
around 420 nm concomitant with a 20% decrease in
Fig. 3. Nucleotide sequences around the ribosome binding site and start codons of the genes. The ribosome binding sites and the start codons
are shown in bold letters. Abbreviations of the genes are as described in Fig. 1. The number of nucleotides shows the nucleotide space
between ribosome binding site and start codon. By using the italicized start codons no active proteins could be obtained.
Fig. 4. Purification of recombinant HadI, activator of dehydratase.
SDS ⁄ PAGE (15%) stained with Coomassie brilliant blue. M,
molecular mass marker; CFE, cell-free extract induced with
anhydrotetracycline, 200 lgÆL
)1
; FT, flow through from the column;

Ac, purified activator.
2-Hydroxyisocaproyl-CoA dehydratase J. Kim et al.
552 FEBS Journal 272 (2005) 550–561 ª 2004 FEBS
absorbance and the appearance of second shoulder
between 500 and 600 nm. A 10-fold excess of thionine
oxidized HadI (as isolated) with a maximum around
400 nm. Oxidized (% 1s
)1
) well as reduced HadI
(% 2s
)1
) showed low ATPase activities. But in the
presence of the dehydratase (20 lg HadBC ⁄ mL)
reduced HadI (1.0 lgÆmL
)1
) catalyzed the hydrolysis
of ATP very efficiently (50 s
)1
), almost independent of
whether the substrate was added (45 s
)1
).
2-Hydroxyisocaproyl-CoA dehydratase (HadBC)
An E. coli cell-free extract containing the recombinant
dehydratase produced from the hadBC genes showed
by SDS ⁄ PAGE thick protein bands around the
43-kDa molecular mass marker, which were not seen
in the extract of noninduced E. coli cells. A dehydra-
tase activity of 9 UÆmg
)1

, equal to that in the
C. difficile cell-free extract, was obtained. Unfortu-
nately, the produced protein could not be purified
using the affinity column, probably because the Strep-
tag II peptide at the C-terminus of the HadC subunit
was buried inside the protein and could not bind to
the column. The dehydratase was therefore purified
from C. difficile cell-free extracts by three chromatog-
raphy columns (Table 1). SDS ⁄ PAGE of the purified
enzyme showed two protein bands (calculated masses
of two subunits, HadB ¼ 46 578 Da and HadC ¼
42 350 Da) just below the 43 kDa protein molecular
mass marker (Fig. 6). On a gel filtration column, the
enzyme eluted at a size (% 90 kDa) corresponding to
the heterodimer (89 kDa). The N-terminal amino acid
sequences of two subunits determined by the Edman
degradation method revealed that the upper band was
the slightly smaller HadC (MEAILSKMKE) and the
lower band the somewhat larger HadB (SEKKE
ARVVI) confirming the correct start codon. The UV-
visible spectrum of purified 2-hydroxyisocaproyl-CoA
dehydratase showed a typical spectrum of iron–sulfur
Fig. 5. UV-visible spectra of purified activator of the dehydratase,
HadI. Solid line, as isolated (4.2 mgÆmL
)1
); dotted line, reduced
with a 10-fold excess of dithionite in 50 m
M Mops pH 7.0, 10 mM
MgCl
2

,1mM ADP and 5 mM dithiothreitol (0.5 mgÆmL
)1
; eightfold
amplified); dashed line, oxidized with a 10-fold excess of thionine in
the same buffer condition (0.5 mgÆmL
)1
; eightfold amplified).
Excess reductant or oxidant was removed by desalting through
Sephadex G-25 columns.
Table 1. Purification of 2-hydroxyisocaproyl-CoA dehydratase from C. difficile and E. coli.
Step Protein (mg) Activity (U) Specific activity (UÆmg
)1
) Enrichment (fold) Yield (%)
C. difficile cell-free extract
a
700 6300 9 1 100
DEAE Sepharose 140 3220 23 3 51
Phenyl Sepharose 50 2250 45 5 36
Q-Sepharose 17 2210 130 15 35
E. coli cell-free extract
b
153 1363 9 1 100
DEAE Sepharose 43 887 21 2 65
Phenyl Sepharose 6 593 99 11 44
a
Starting from 15 g wet cell paste;
b
6 g wet cell paste.
Fig. 6. SDS ⁄ PAGE of purified 2-hydroxyisocaproyl-CoA dehydratase
(HadBC). The gel was (8%) stained with Coomassie brilliant blue.

M, molecular mass marker; lanes 1–3, purified protein.
J. Kim et al. 2-Hydroxyisocaproyl-CoA dehydratase
FEBS Journal 272 (2005) 550–561 ª 2004 FEBS 553
cluster(s) (Fig. 7). Chemical analysis revealed
5.7 ± 0.1 nonheme iron and 6.1 ± 0.5 acid-labile sul-
fur. Metal contents were also estimated by inductively
coupled plasma atomic emission spectroscopy (ICP-
AES, model Optima 3000, PerkinElmer, Rodgau-Ju
¨
ge-
sheim, Germany). Two preparations of the dehydratase
were analyzed. The iron content was estimated as 3.8
and 4.1 mol ⁄ mol homodimer; cobalt, nickel and
molybdenum were absent (< 0.01), but surprisingly
stoichiometric amounts of zinc were found in the pre-
parations, 1.1 and 1.9 molÆmol
)1
, respectively. The
supernatant of the enzyme after treatment with anoxic
0.2 m trichloroacetic acid showed a characteristic UV-
visible spectrum of oxidized flavin (peaks at 370 nm
and 450 nm), but no significant flavin content (< 5%
of the dehydratase) was detected by HPLC comparing
with FMN, FAD and riboflavin standards. After oxi-
dation with air the UV-visible spectrum of the super-
natant showed a new peak at 300 nm and a shoulder
around 325 nm, which could not be assigned to any
known cofactor. Probably this absorption was due to
oxidized iron sulfide.
Using the same method as applied for the purifica-

tion of the 2-hydroxyisocaproyl-CoA dehydratase from
cell-free extracts of C. difficile, the recombinant
enzyme with a nonfunctional Strep-tag at the C-termi-
nus of the C-subunit could be also obtained in pure
form from E. coli. The properties of the recombinant
dehydratase (V
max
and K
m
, see below) were identical
to those of the enzyme from C. difficile. ICP-AES
analysis revealed 5.3 iron, 3.2 zinc, 0.2 nickel and
0.08 cobalt mol ⁄ mol enzyme, but no molybdenum
(< 0.01).
(R)-2-Hydroxyisocaproyl-CoA dehydratase activity
2-Hydroxyisocaproyl-CoA dehydratase activity was
measured in the presence of ATP, MgCl
2
, dithionite,
dithiothreitol, serum albumin and activator. Addition
of (R)-2-hydroxyisocaproyl-CoA started this assay and
the formation of isocaprenoyl-CoA was followed at
290 nm (De ¼ 2.2 mm
)1
Æcm
)1
). Due to the high
absorbance of the adenine moiety of CoA, the absorb-
ance maximum at 263 nm was not used. The product
isocaprenoyl-CoA was identified by MALDI-TOF

mass spectrometry (M
r
¼ 865) and by comparison
with the chemically synthesized compound. The
enzyme accepted only (R)-2-hydroxyisocaproyl-CoA
with the S-isomer showing less than 10% activity,
which might be due to a contamination of the R-iso-
mer. As an equal mixture of (R)- and (S)-2-hydroxy-
isocaproyl-CoA gave only one-half of the enzymatic
activity, we assume that that the S-isomer was also
able to bind at the active site of the enzyme but could
not be dehydrated. The apparent K
m
value for
(R)-2-hydroxyisocaproyl-CoA was 50–80 lm and V
max
was determined as 110–150 UÆmg
)1
(160–220 s
)1
) using
different dehydratase preparations. In assays using
(E)-isocaprenoyl-CoA as substrate, no activity could
be observed suggesting that the dehydration is irrevers-
ible under these conditions or the Z-isomer is the cor-
rect product. (E)-Isocaprenoyl-CoA (400 lm) was
shown to decompose slowly (5 nmolÆmin
)1
) under the
assay conditions regardless whether the dehydratase

was present (compare Fig. 9, in which an absorbance
maximum is observed after addition 3). The product
could not be identified by MALDI-TOF spectrometry.
Recombinant HadI activated the dehydratase in the
presence of ATP, MgCl
2
and a one-electron reducing
agent titanium(III) citrate or dithionite, but the initial
experiments revealed a dependence of the activity on
the applied amount of activator (Fig. 8). Hence, it
appeared that each dehydratase molecule required one
activator molecule and ATP is hydrolyzed during every
turnover. A true activator, however, should act catalyt-
ically; it should be able to serve many dehydratase
molecules, each of which catalyses many turnovers
without further hydrolysis of ATP. Subsequent experi-
ments indicated that the low dehydratase ⁄ activator
ratio £ 1 required to get high activity was due to the
instability of the activator in the assay mixture. The
activator HadI could be stabilized with 5 mm dithio-
threitol and 1 lm bovine serum albumin, probably by
removing trace amounts of oxygen and preventing dis-
sociation into subunits. Under these conditions a dehy-
dratase ⁄ activator ratio of 10 gave an even higher
dehydratase activity than a ratio of 0.2 in the absence
Fig. 7. UV-visible spectra of 2-hydroxyisocaproyl-CoA dehydratase.
Solid line, as isolated (1.2 mgÆmL
)1
); activated dehydratase separ-
ated from activator (1.2 mgÆmL

)1
). The insert shows the difference
spectrum of activated dehydratase (dashed line) minus isolated de-
hydratase. The peak at 320 nm stems from dithionite.
2-Hydroxyisocaproyl-CoA dehydratase J. Kim et al.
554 FEBS Journal 272 (2005) 550–561 ª 2004 FEBS
of the stabilisators (Fig. 8). The experiments indicate,
however, that at a dehydratase ⁄ activator ratio of 10 a
preincubation time of at least 40 min is required to
reach full activity. Immediate activation was only
obtained by using dehydratase⁄ activator ratios £ 0.1
(Fig. 8).
In a critical experiment, 4.4 mg dehydratase was
activated for 30 min in the presence of 1.0 mg activa-
tor (dehydratase ⁄ activator ¼ 3), 0.4 mm ATP, 10 mm
MgCl
2
,5mm dithiothreitol and 0.1 mm dithionite in
50 mm Mops pH 7.0 (total volume 2 mL). Dehydra-
tase activity was assayed by diluting a 1.0 lL sample
into 0.5 mL 0.4 mm (R)-2-hydroxyisocaproyl-CoA in
50 mm Tris ⁄ HCl pH 8.0 (139 s
)1
). The active dehy-
dratase was separated from its activator through a
Strep-Tactin column. The tagged activator bound to
the column, while the active dehydratase passed
through. A 2.0 lL sample of the flow-through was
assayed in the same manner as above (69 s
)1

); after
two successive substrate additions the activity was
almost completely lost (Fig. 9). SDS ⁄ PAGE revealed
the double band of the dehydratase around 43 kDa
but no band at 30 kDa indicating that > 95% of
activator was removed. Activation by 0.4 mm ATP,
0.1 mm dithionite and a > 10-fold molar excess of
activator immediately restored the complete activity
(68 s
)1
). Hence the activated dehydratase irreversibly
lost 50% of its activity during passage through the
Strep-Tactin column; the remaining 13.5 pmol active
enzyme dehydrated 103 nmol (R)-2-hydroxyisocap-
royl-CoA (7600 turnovers) until activity ceased. After-
wards by addition of activator and ATP the enzyme
regained the same activity, which was measured after
the passage through the affinity column. This experi-
ment showed that the activated dehydratase retained
its activity (a) in the absence of 0.4 mm ATP, which
was diluted in the assay prior to the affinity chroma-
tography to 0.8 lm; (b) after affinity chromatography
at < 0.8 lm ATP and in the absence of at least
95% of the activator (dehydratase ⁄ activator > 60 and
absence of stabilisators); (c) turnover causes rapid
inactivation; and (d) activator and ATP recovered the
activity, which was lost during turnover. The UV-vis-
ible spectra between 300 and 700 nm of the dehydra-
tase as isolated and after activation and affinity
chromatography revealed the absorbance of a

[4Fe)4S]
2+
cluster around 400 nm (Fig. 7). In the
difference spectrum (insert of Fig. 7) the peak at
320 nm stems from dithionite, whereas the increase in
absorbance around 400 nm may be caused by the
irreversible inactivation of 50% of the dehydratase
during affinity chromatography. In another experiment,
in which the Strep-Tactin column was not treated with
dithionite prior to the affinity chromatography
(see below), the yield of active dehydratase was only
10%, but the absorbance increase around 400 nm was
higher. In contrast to that expected for a reduction of a
[4Fe)4S]
2+
cluster, no decrease in absorbance was
observed.
Fig. 8. Activation of the dehydratase by its activator. Dehydratase
activities were measured in 50 m
M Tris ⁄ HCl pH 8.0, 5 mM MgCl
2
,
0.1 m
M dithionite, and 0.4 mM ATP, and the reactions were started
by adding 200 l
M (R)-2-hydroxyisocproyl-CoA at the indicated pre-
incubation times. The molar ratios of dehydratase ⁄ activator were:
0.2, n; 1.0, h;10,s; 10 in the presence of 5 m
M dithiothreitol and
1 l

M bovine serum albumin, d.
Fig. 9. The activity assay of activated dehydratase separated from
the activator by passage through a Strep-Tactin column. The assay
(total volume 500 lL) contained 27 pmol active dehydratase in
50 m
M Tris ⁄ HCl pH 8.0 in absence of ATP, MgCl
2
, dithionite and di-
thiothreitol. The reaction was started by adding 0.2 lmol of the
substrate (R)-2-hydroxyisocaproyl-CoA (arrow 1). After the substrate
was consumed (DA
290nm
¼ 0.455), further 2 · 0.2 lmol substrate
was added at arrows 1 and 2. The activity was recovered by addi-
tion of an excess amount of activator (> 10-fold), 0.4 m
M ATP,
5m
M MgCl
2
,5mM dithiothreitol and 0.1 mM dithionite (arrow 3).
The decrease in absorbance after 20 min was due to the instability
of the product isocaprenoyl-CoA.
J. Kim et al. 2-Hydroxyisocaproyl-CoA dehydratase
FEBS Journal 272 (2005) 550–561 ª 2004 FEBS 555
It was suggested that 2-hydroxyisocaproyl-CoA
dehydratase could be the most sensitive target of met-
ronidazole [14], which has been used as an antibiotic
for C. difficile infections in the human body [20]. Met-
ronidazole inhibited effectively cell growth (50% inhi-
bition at 10 lm) and the dehydratase activity was

completely abolished at 20 lm, probably by oxidation
of the activated enzyme by the nitro group of the inac-
tivator [21].
Discussion
The experiments described in this work clearly show
that the hadIBC-genes of C. difficile encode a novel
2-hydroxyacyl-CoA dehydratase (HadBC) and its acti-
vator (HadI), probably specific for the dehydration of
(R)-2-hydroxyisocaproyl-CoA to isocaprenoyl-CoA,
but besides the S-isomer no other substrate was tested.
As the known enzymatic eliminations of water
from (R)-2-hydroxyacyl-CoA to (E)-2-enoyl-CoA
(R)-2-hydroxyglutaryl-CoA to (E)-glutaconyl-CoA [22]
(R)-lactyl-CoA to acryloyl-CoA [23] and (R)-phenyllac-
tyl-CoA to (E)-cinnamoyl-CoA [24] all occur in a syn-
fashion, we assume that this will also be the case for
(R)-2-hydroxyisocaproyl-CoA to (E)-2-isocaprenoyl-
CoA, which, however, remains to be determined. The
inability to measure the hydration of the chemically
synthesized (E)-2-isocaprenoyl-CoA could be either
due to the unfavourable equilibrium or due to the
Z-isomer being the correct substrate. It has been
shown that 2-hydroxyglutaryl-CoA dehydratase indeed
catalyzed the reverse reaction. The conditions, how-
ever, were different; this experiment was performed in
the cell-free extract using (E)-glutaconate in the pres-
ence of acetyl-CoA as substrate and the formed (R)-2-
hydroxyglutarate was determined enzymatically [22].
The 2-hydroxyisocaproyl-CoA dehydratase fits well
into the proposed pathway of leucine fermentation by

C. difficile. In addition we showed that ldhA encodes a
fairly specific NAD-dependent (R)-2-hydroxyisocapro-
ate dehydrogenase (GenBank accession number
AY772817) and hadA a highly specific class III [25]
2-hydroxyisocaproate CoA-transferase using (R)-2-hy-
droxyisocaproyl-CoA and (E)-isocaprenoate, probably
as well as isocaproate as substrates (GenBank acces-
sion number AY772818) [26]. The genes acdB, etfB
and etfA, downstream of hadBC, are related to those
of an acyl-CoA dehydrogenease and an electron-trans-
ferring flavoprotein, which most likely are involved in
the reduction of isocaprenoyl-CoA to isocaproyl-CoA.
Finally the CoA-transferase HadA may liberate the
product isocaproate (Figs 1 and 2). An ambiguous step
is the conversion of leucine to 2-oxoisocaproate, which
may proceed via amino transfer to 2-oxoglutarate fol-
lowed by dehydrogenation of the formed glutamate
(Fig. 2) or by a direct one-step oxidative deamination
of leucine. Although the arrangement of the hadAIBC
genes are very similar to those involved in the dehy-
dration of (R)-phenyllactate to (E)-cinnamate [13], a
stable complex of the 2-hydroxyisocaproyl-CoA dehy-
dratase (HadBC) with the CoA-transferase (HadA)
could not be detected, as both enzymes separate during
purification.
The requirement of activator (HadI), dehydratase
(HadBC), ATP, Mg
2+
, dithiothreitol and dithionite
for the activity of 2-hydroxyisocaproyl-CoA dehydra-

tase indicates that this enzyme acts by the same mech-
anism as that proposed for 2-hydroxyglutaryl-CoA
dehydratase [27] (Fig. 10). The reduced activator trans-
fers one electron to the dehydratase concomitant with
hydrolysis of ATP. Although the stoichiometry of 1 or
2 ATP ⁄ electron remains to be determined, the homo-
dimeric structure of the activator with one [4Fe)4S]
cluster and two ATP binding sites strongly suggests 2
ATP ⁄ electron as observed with nitrogenase [28]. The
reduced dehydratase transfers the electron further to
the substrate to generate the ketyl radical anion I,
which expels the adjacent hydroxyl group. The formed
enoxy radical can now be deprotonated at the b-posi-
tion to the product-related ketyl radical anion II,
which is oxidized to isocaprenoyl-CoA by the next
incoming substrate 2-hydroxyisocaproyl-CoA, whereby
the electron is recycled. It has been calculated that the
extremely high pK of the b-protons of 2-hydroxyiso-
caproyl-CoA (% 40), is lowered by 26 units to pK ¼
14 in the enoxy radical [29]. This fairly low pK could
be even further decreased to about 7 by hydrogen
Fig. 10. Proposed mechanism of dehydration from (R)-2-hydroxyiso-
caproyl-CoA to (E)-2-isocaprenoyl-CoA. For protein abbreviations,
see Fig. 2.
2-Hydroxyisocaproyl-CoA dehydratase J. Kim et al.
556 FEBS Journal 272 (2005) 550–561 ª 2004 FEBS
bonds from backbone amides of the enzyme to the
carbonyl oxygen and thus gets into the range of
the pK of carboxylates or imidazolyl residues of the
enzyme [30].

One major support for this mechanism comes from
experiments described in this work. For the first time
it has been shown that catalytic amounts of activator
(HadBC ⁄ HadI ¼ 10 mol ⁄ mol) are sufficient to get
maximum dehydratase activity. This important finding
was due to the development of a direct spectrophoto-
metric assay of the dehydratase and to the improved
stability of HadI through the addition of serum albu-
min and dithiothreitol. In previous work an assay with
six auxiliary enzymes was used and hence gave only
qualitative data [8,31]. Furthermore, the activated
dehydratase could be separated from the activator and
retained its activity for almost 10
4
turnovers. This
experiment clearly demonstrated that ATP and Mg
2+
are only required for activation and ATP is not used
to phosphorylate the hydroxyl group in order to facili-
tate the elimination as suggested in the early work on
lactyl-CoA dehydratase. The authors Anderson and
Wood [32] have already addressed the energetic enigma
if each dehydration would require one ATP, this
means in the case of C. difficile that generation of one
ATP by substrate-level phosphorylation consumes two
ATP (Fig. 2 and Eqn 1). Therefore it was proposed
that one ATP must be sufficient to activate the dehy-
dratase for at least 100 turnovers [1], which has now
been experimentally verified. The activated enzyme
may become inactivated simply by one-electron oxida-

tion with traces of oxygen or by a second electron
transfer to a radical intermediate, which would result
in isocaproyl-CoA rather than isocaprenoyl-CoA as
product, but according to the measured turnover only
one in 10
4
. The inactivation by substrate is reminiscent
of coenzyme B
12
-dependent mutases. The suicide inac-
tivation of b-lysine 5,6-aminomutase is caused by
the substrate-induced one electron transfer from
cob(II)alamin to the 5¢-deoxadenosyl radical resulting
in the inactive pair of cob(III)alamin and
5¢-deoxyadenosine [33].
In our publications on 2-hydroxyglutaryl-CoA dehy-
dratase [1,3,4], the terms component A and component
D were used, A for activator and D for dehydratase,
implicating that only both components together are
able to form an active enzyme. The important result
that even in the absence of activator the activated
2-hydroxyisocaproyl-CoA dehydratase is catalytically
active has consequences for the nomenclature. From
this paper onward we will call component A just
activator or archerase [1] and component D just
dehydratase.
Previous work on 2-hydroxyglutaryl-CoA dehydra-
tase showed that the activator alone had ATPase activ-
ity (4–6 s
)1

) but only in the oxidized state [3]. The
results in this work, which revealed low ATPase activ-
ities of the activator HadI regardless of its oxidation
state, question those data. Therefore the original data
obtained with the activator of 2-hydroxyglutaryl-CoA
dehydratase have been re-calculated and found too
high by a factor of 10. Furthermore, repetition of the
ATPase measurements with the activator from A. fer-
mentans by applying the conditions used in this work
also gave only low activities (M Hetzel & W Buckel,
unpublished results). Addition of dehydratase to the
corresponding reduced activator, however, gave high
ATPase activities; in case of HadI + HadBC up to
50 UÆmg
)1
activator was achieved. These results fit
much better to the proposed mechanism, as the elec-
tron should only be transferred in a complex of both
proteins driven by ATP hydrolysis.
Another important result of this work is the finding
that 2-hydroxyisocaproyl-CoA dehydratase, the 2-hyd-
roxyacyl-CoA dehydratase with highest ever-observed
activity (up to 220 s
)1
), contains no molybdenum and
hardly any flavin. Therefore these two cofactors seem
not to play important roles also in t he other 2 -hydroxy-
acyl-CoA dehydratases. Molybdenum may be an
impurity that could not be separated from 2-hydroxy-
glutaryl-CoA dehydratase and flavin (FMN and ⁄ or

riboflavin) could bind fortuitously. Interestingly, crude
preparations of 2-hydroxyisocaproyl-CoA dehydratase
obtained from C. difficile do contain molybdenum,
which is removed during further purification without
decreasing the activity. The only prosthetic group of
the dehydratase, which after activation could carry the
catalytic electron, is a putative [4Fe)4S] cluster, whose
structure remains to be determined by spectroscopic
and crystallographic methods. The failure to see the
reduction of the cluster in the active dehydratase by a
decrease in absorbance at 400 nm may be due to the
concomitant increase in absorbance of half of the
dehydratase irreversibly inactivated during separation
from its activator. This cluster must have a very negat-
ive redox potential (E
0
¢ ¼ <) 600 mV), as no activity
could be observed after treatment of the inactive dehy-
dratase with excess dithionite or titanium(III) citrate in
the absence of the activator HadI and ATP. On the
other hand this cluster cannot be very unusual, as it is
synthesized by enzymes not only present in C. difficile
but also in E. coli [34] as shown by the functional
heterologous expression of the hadBC genes. The role
of zinc, if any, remains to be established. Hence,
2-hydroxyisocaproyl-CoA dehydratase and its activator
appear as simple iron–sulfur proteins without any
J. Kim et al. 2-Hydroxyisocaproyl-CoA dehydratase
FEBS Journal 272 (2005) 550–561 ª 2004 FEBS 557
special cofactors or rare elements. Owing to this sim-

plicity, one may conclude that 2-hydroxyacyl-CoA de-
hydratases have evolved very early during the
emergence of life [35], probably with an unknown ana-
bolic rather than a catabolic function.
Experimental procedures
Materials
C. difficile (DSMZ 1296
T
) was purchased from the Deut-
sche Sammlung fu
¨
r Mikroorganismen und Zellkulturen
(DMSZ, Braunschweig, Germany) and E. coli, BL21-Co-
donPlus(DE3)-RIL strain for gene expression was obtained
from Stratagene (Heidelberg, Germany). The affinity col-
umn, Strep-Tactin MacroPrep was purchased from IBA
GmbH (Go
¨
ttingen, Germany). The enzymes for molecular
biology were obtained from New England Biolabs (Frank-
furt am Main, Germany), ABgene (Hamburg, Germany)
and Amersham Biosciences (Freiburg, Germany). Primers
were purchased from MWG (Ebersberg, Germany). Protein
molecular mass markers and DNA size markers were
obtained from Amersham Biosciences.
Experiments under anoxic conditions
Purification of the activator and 2-hydroxyisocaproyl-CoA
dehydratase were performed at 15–20 °C in an ‘Anaerobic
Chamber’ (Coy Laboratories, Ann Arbor, MI, USA) under
a nitrogen atmosphere containing 5% H

2
. Oxygen was
removed from buffers for enzyme purification by boiling
and cooling under vacuum. Afterwards the buffers were
flushed with nitrogen, transferred to the anaerobic chamber,
and stirred overnight. In the chamber, 2 mm dithiothreitol
was added to each buffer. Enzyme activity was determined
inside the anaerobic chamber with an Ultrospec 4000 spec-
trophotometer from Amersham Biosciences.
Chemicals and synthesis of CoA-esters
(R)-2-Hydroxyisocaproate and (S)-2-hydroxyisocaproate
were obtained from d- and l-leucine, respectively, by treat-
ment of the corresponding amino acids with sodium nitrite
in dilute sulfuric acid [36]. (E)-2-Isocaprenoate (4-methyl-
trans-2-pentenoic acid) was synthesized from isobutyralde-
hyde and malonic acid in pyridine-piperidine [37]. (R)- and
(S)-2-Hydroxyisocaproyl-CoA and (E)-2-isocaprenoyl-CoA
were prepared from the corresponding acids following the
modified anhydrous 1,1¢-carbonyldiimidazole synthesis [38].
Gene cloning
Routine manipulation of plasmid DNA, PCR, the construc-
tion of recombinant plasmids and isolation of chromosomal
DNA from C. difficile were performed using standard
techniques [39]. The ORF hadI was amplified with follow-
ing primers: FhadI, 5¢-ATGGTAGGTCTCAAATGTACA
CAATGGGATTAGATATAGGTTC-3¢; RhadI, 5¢-ATGG
TAGGTCTCAGCGCTTATATTTTTCACTTCTTTTTGT
GATTCT-3¢.
PCR was performed using proof reading polymerase,
Extensor Hi-Fidelity PCR Enzyme Mix (ABgeneÒ, Ham-

burg, Germany) and the amplified fragment was cloned into
the BsaI restriction site [GGTCTC(N)
1
] of the expression
vector pASK-IBA3 providing a C-terminal Strep-tag II pep-
tide (Trp-Ser-His-Pro-Gln-Phe-Glu-Lys) fused protein. The
plasmid construct, pASK-IBA3::hadI, was named Np3I. The
ORF hadBC was amplified with following primers: FhadBC,
5¢-ATGGTAGGTCTCAAATGTCTGAAAAAAAAGAAG
CTAGAGTAGT-3¢; RhadBC, 5¢-ATGGTAGGTCTCAG
CGCTCGCTAAACTCATCATCTCAGCAAA-3¢.
The amplified fragment using proofreading polymerase
was cloned into the BsaI restriction site of pASK-IBA3 giv-
ing Np3 bc. In order to exclude reading errors of the
polymerase, three different clones from three different PCR
products were sequenced. The sequencing primers labelled
at their 5¢ end with the infrared dye IRD-41 were obtained
from MWG-Biotech (Ebersberg, Germany).
Gene expression and purification of the activator
Plasmid constructs, Np3I or Np3 BC, were transformed
into E. coli BL21-CodonPlus(DE3)-RIL harbouring addi-
tional rare codon tRNA genes (arg, ileY and leuW), in
order to express the relevant genes. An overnight preculture
(100 mL) of a fresh single colony was used to inoculate 2 L
Standard I medium (Merck, Darmstadt, Germany) contain-
ing antibiotics (ampicillin, 100 lgÆmL
)1
, and chlorampheni-
col, 50 lgÆmL
)1

)at30°C (or room temperature for
Np3 BC) under anoxic conditions. When the culture
reached the mid-exponential phase, A
590
¼ 0.5–0.7, gene
expression was induced with anhydrotetracycline
(200 lgÆL
)1
). After another 3 h growth, the culture was
transferred to the anaerobic chamber. Cells were harvested
by centrifugation in airtight bottles, washed and suspended
in 50 mm Mops pH 7.0, 300 mm NaCl, 10 mm MgCl
2
, and
5mm dithiothreitol. Cells in serum bottles tightly closed
with rubber stoppers were transferred through a needle into
the French Press operating at 140 MPa. After the cell deb-
ris had been removed by ultracentrifugation at 100 000 g
for 1 h, the supernatant was loaded on a 5 mL Strep-Tactin
MacroPrep column, which was equilibrated with the buffer
used for suspending the cells. After loading, the column
was washed with at least 10 column volumes of equilibra-
tion buffer and the enzyme was eluted with equilibration
buffer containing 3 mm d-desthiobiotin and 1 mm ADP.
Afterwards d-desthiobiotin was removed by gel filtration on
Sephadex G-25 equilibrated with 50 mm Mops pH 7.0,
1mm ADP, 10 mm MgCl
2
and 5 mm dithiothreitol.
2-Hydroxyisocaproyl-CoA dehydratase J. Kim et al.

558 FEBS Journal 272 (2005) 550–561 ª 2004 FEBS
Purification of 2-hydroxyisocaproyl-CoA
dehydratase from C. difficile
C. difficile cells were cultivated as described before [40] in
2 L tightly closed bottles containing anoxic defined medium
[41] supplemented with l-leucine (1 gÆL
)1
; 7.6 mm). Cells
were harvested, washed and suspended in buffer A contain-
ing 50 mm Mops pH 7.0 and 2 mm dithiothreitol, yield 3 g
wet cell paste. The preparation of the cell free extract was
performed as that described in the activator purification.
The cell free extract was filtered (0.45 lm pore size) and
loaded a DEAE-Sepharose fast-flow column (3 · 10 cm)
equilibrated with buffer A. The column was washed with
70 mL buffer A and the proteins were eluted at a rate of
3mLÆmin
)1
with a linear gradient of 0–1.0 m NaCl in buf-
fer A. The active brown fractions were eluted around 0.4 m
NaCl. An equal volume of 2.0 m (NH
4
)
2
SO
4
in buffer A
was added to the pooled fractions from the first column,
which were then loaded on a phenyl-Sepharose column
(3 · 10 cm) equilibrated with buffer B, 50 mm Mops

pH 7.0, 1.0 m (NH
4
)
2
SO
4
,2mm dithiothreitol. After wash-
ing the column with 70 mL buffer B, the active brown
dehydratase eluted around 0.1 m (NH
4
)
2
SO
4
with a linear
gradient of 1.0–0 m (NH
4
)
2
SO
4
in buffer B at a rate of
3mLÆmin
)1
. The dehydratase fractions were concentrated
on an Amicon PM 30 cell and desalted against buffer A,
then loaded on a Q-Sepharose column (1.8 · 10 cm) equili-
brated with buffer A. After a washing step with 60 mL buf-
fer A, the dehydratase was eluted around 0.5 m NaCl with
a linear gradient of 0–1.0 m NaCl in buffer A at a rate of

3mLÆmin
)1
. The dehydratase was finally concentrated with
an Amicon Ultra-4 PLTK Ultracel-Pl (30 kDa cut-off).
The recombinant 2-hydroxyisocaproyl-CoA dehydratase
from E. coli was purified by the same method, as the
enzyme was not absorbed at the Strep-Tactin MacroPrep
column. After the phenyl-Sepharose column the enzyme
was already pure and therefore the Q-Sepharose column
could be omitted.
Determination of enzyme activity
2-Hydroxyisocaproyl-CoA dehydratase activity was meas-
ured using a continuous direct assay based on the difference
between the extinction coefficients of 2-hydroxyisocaproyl-
CoA and 2-isocaprenoyl-CoA at 290 nm (De ¼ 2.2 mm
)1
Æ
cm
)1
). The dehydratase was incubated for 5 min with an
equal molar amount of recombinant activator in the
presence of 50 mm Tris ⁄ HCl pH 8.0, 5 mm MgCl
2
, 0.4 mm
ATP, 0.1 mm dithionite or titanium(III) citrate, 5 mm
dithiothreitol and 1 lm bovine serum albumin in a total
volume of 0.5 mL. The assay was started by the addition of
200 lm (R)-2-hydroxyisocaproyl-CoA and the absorbance
increase was followed at 290 nm. The ATPase activity of
the activator was measured using a coupled assay with

pyruvate kinase and lactate dehydrogenase [42]. The
cuvette, total volume 0.5 mL, contained 50 mm Tris ⁄ HCl
pH 8.0, 1 mm phosphoenolpyruvate, 10 mm MgCl
2
,1mm
ATP, 0.2 mm NADH, 2 U pyruvate kinase and 2 U lactate
dehydrogenase. After adding the activator, the absorbance
decrease of NADH was followed at 340 nm (e ¼
6.3 mm
)1
Æcm
)1
[43]).
Analysis of CoA-thiol esters by MALDI-TOF
mass spectrometry
The molecular mass of a CoA ester produced in an enzymatic
or chemical reaction was confirmed by MALDI-TOF mass
spectrometry. The reaction was acidified with 1 m HCl to
pH < 4.0 and loaded on Sep-pak Ò C
18
cartridge (Waters,
Eschborn, Germany), which was equilibrated with 0.1% tri-
fluoroacetic acid. The column was washed with five column-
volumes of 0.1% trifluoroacetic acid and the CoA-thiol ester
was eluted with 5 mL 1% trifluoroacetic acid in 50% aceto-
nitrile. After evaporation of the acetonitrile under vacuum, a
drop of the CoA-thiol ester solution was applied on a thin
layer of indole-2-carboxylic acid on a golden plate prepared
from a solution of 300 mm indole-2-carboxylic acid in acet-
one and measured under the described conditions [44].

Separation of activated dehydratase from its
activator
Dehydratase (4.4 mg) was activated by 1.0 mg activator in
the presence of 50 mm Mops pH 7.0, 0.4 mm ATP, 5 mm
MgCl
2
,5mm dithiothreitol, and 0.1 mm dithionite (total
volume 2.0 mL) as described in activity assay but in the
absence of bovine serum albumin. After 30-min incubation
at room temperature, 1.0 lL was assayed for activity with-
out further activation and the reaction mixture was loaded
on a 5 mL Strep-Tactin MacroPrep column, previously
reduced with 50 mm Mops pH 7.0, 5 mm dithiothreitol and
0.1 mm dithionite and equilibrated with 50 mm Mops
pH 7.0, 300 mm NaCl, 10 mm MgCl
2
and 5 mm dithiothre-
itol. The tagged activator was bound to the column while
the dehydratase-containing flow through was collected in
1 mL fractions. An UV-visible spectrum was taken from
the peak fraction (1.2 mg dehydrataseÆmL
)1
), which was
also analyzed for activity. Therefore a 2 lL aliquot
was added to 50 mm Tris ⁄ HCl pH 8.0 and the reaction was
started with 0.2 lmol (R)-2-hydroxyisocaproyl-CoA, total
volume 0.5 mL, d ¼ 1 cm. After the reaction had ceased,
two additional 0.2 lmol (R)-2-hydroxyisocaproyl-CoA
aliquots were added. Finally the enzyme was reactivated by
0.1 mm dithionite, 0.4 mm ATP, 5 mm MgCl

2
and 5 mm
dithiothreitol and 30 lg activator (added last). On an
SDS ⁄ polyacrylamide gel, to which 20 lL of the separated
dehydratase were applied, the double band of the dehydra-
tase (40 kDa) but no trace of the activator (30 kDa) was
visible.
J. Kim et al. 2-Hydroxyisocaproyl-CoA dehydratase
FEBS Journal 272 (2005) 550–561 ª 2004 FEBS 559
Determination of the molecular mass
The apparent molecular masses of the enzymes were deter-
mined by gel filtration on a Superdex 200 column
(1 · 30 cm) in 150 mm NaCl and 50 mm Tris ⁄ HCl, pH 8.0
at a flow rate of 0.5 mLÆmin
)1
. Amylase, aldolase, bovine
serum albumin, catalase and cytochrome c were used for
calibration. The molecular mass standards were obtained
from Roche Molecular Biochemicals (Mannheim, Ger-
many).
Other biochemical methods
Protein concentration was determined with the Bio-Rad
Protein Assay. Bovine serum albumin was used as stand-
ard. SDS ⁄ polyacrylamide gels were stained with Coomassie
brilliant blue. The subunits of the dehydratase were separ-
ated by SDS ⁄ PAGE, blotted and their N-termini were
sequenced by Edman degradation [45]. Non-heme iron [46]
and acid labile sulfur [47] were determined as described.
Absorption spectra were recorded using a HP 8453 UV-vis-
ible spectrophotometer (Hewlett Packard, Bo

¨
blingen,
Germany). The determination of the flavin was performed
by HPLC as described earlier [48]. The molybdenum con-
tent was quantified by atomic absorption spectroscopy [3].
Acknowledgements
This work was supported by grants from the Deutsche
Forschungsgemeinschaft (DFG), European Commis-
sion, Cargill Inc. Minneapolis, MN, USA, and the
Fonds der Chemischen Industrie. We thank Dr Thor-
sten Selmer and Dr Antonio J. Pierik (both at the Phili-
pps-Universita
¨
t Marburg) for helpful discussions. We
are indebted to Dr Dietmar Linder (Universita
¨
t Gies-
sen, Germany) for N-terminal amino acid sequencing,
and to Professor Holger Dobbek (Universita
¨
t Bay-
reuth, Germany) for metal analysis by ICP-AES (Bito
¨
k,
BMBF project 0339476 D). Part of the DNA sequence
data was produced by the Clostridium difficile Sequen-
cing group at the Sanger Centre and can be obtained
from Sev-
eral major improvements of the manuscript are due to
the very helpful advice of two anonymous reviewers.

References
1 Kim J, Hetzel M, Boiangiu CD & Buckel W (2004)
Dehydration of (R)-2-hydroxyacyl-CoA to enoyl-CoA
in the fermentation of alpha-amino acids by anaerobic
bacteria. FEMS Microbiol Rev 28, 455–468.
2 Buckel W (2001) Unusual enzymes involved in five
pathways of glutamate fermentation . Appl Microbiol
Biotechnol 57, 263–273.
3 Hans M, Bill E, Cirpus I, Pierik AJ, Hetzel M, Alber D
& Buckel W (2002) Adenosine triphosphate-induced
electron transfer in 2-hydroxyglutaryl-CoA dehydratase
from Acidaminococcus fermentans. Biochemistry 41,
5873–5882.
4 Buckel W, Hetzel M & Kim J (2004) ATP-driven elec-
tron transfer in enzymatic radical reactions. Curr Opin
Chem Biol 8, 462–467.
5 Locher KP, Hans M, Yeh AP, Schmid B, Buckel W &
Rees DC (2001) Crystal structure of the Acidaminococ-
cus fermentans 2-hydroxyglutaryl-CoA dehydratase com-
ponent A. J Mol Biol 307, 297–308.
6 Tamannaei A (2003) Untersuchungen zum Katalysemech-
nismus der 2-Hydroxyglutaryl-CoA-Dehydratase aus
Fusobacterium nucleatum, Diploma Thesis, Philipps-
Universita
¨
t Marburg, Germany.
7 Hans M, Buckel W & Bill E (2000) The iron–sulfur
clusters in 2-hydroxyglutaryl-CoA dehydratase from
Acidaminococcus fermentans: biochemical and spectro-
scopic investigations. Eur J Biochem 267, 7082–7093.

8 Klees AG, Linder D & Buckel W (1992) 2-Hydroxyglu-
taryl-CoA dehydratase from Fusobacterium nucleatum
(subsp. nucleatum): an iron–sulfur flavoprotein. Arch
Microbiol 158 , 294–301.
9 Hans M, Sievers J, Mu
¨
ller U, Bill E, Vorholt JA, Linder
D & Buckel W (1999) 2-Hydroxyglutaryl-CoA dehydra-
tase from Clostridium symbiosum. Eur J Biochem 265,
404–414.
10 Kapatral V, Anderson I, Ivanova N, Reznik G, Los T,
Lykidis A, Bhattacharyya A, Bartman A, Gardner W,
Grechkin G, Zhu L, Vasieva O, Chu L, Kogan Y,
Chaga O, Goltsman E, Bernal A, Larsen N, D’Souza
M, Walunas T, Pusch G, Haselkorn R, Fonstein M,
Kyrpides N & Overbeek R (2002) Genome sequence
and analysis of the oral bacterium Fusobacterium nuclea-
tum strain ATCC 25586. J Bacteriol 184, 2005–2018.
11 Kuchta RD & Abeles RH (1985) Lactate reduction in
Clostridium propionicum: purification and properties of
lactyl-CoA dehydratase. J Biol Chem 260, 13181–13189.
12 Kuchta RD, Hanson GR, Holmquist B & Abeles RH
(1986) Fe–S centers in lactyl-CoA dehydratase. Bio-
chemistry 25, 7301–7307.
13 Dickert S, Pierik AJ, Linder D & Buckel W (2000) The
involvement of coenzyme A esters in the dehydration of
(R)-phenyllactate to (E)-cinnamate by Clostridium sporo-
genes. Eur J Biochem 267, 3874–3884.
14 Dickert S, Pierik AJ & Buckel W (2002) Molecular char-
acterization of phenyllactate dehydratase and its initiator

from Clostridium sporogenes. Mol Microbiol 44, 49–60.
15 Bader J, Rauschenbach P & Simon H (1982) On a
hitherto unknown fermentation path of several amino
acids by proteolytic clostridia. FEBS Lett 140, 67–72.
16 Elsden SR & Hilton MG (1978) Volatile acid produc-
tion from threonine, valine, leucine and isoleucine by
clostridia. Arch Microbiol 117, 165–172.
2-Hydroxyisocaproyl-CoA dehydratase J. Kim et al.
560 FEBS Journal 272 (2005) 550–561 ª 2004 FEBS
17 Britz ML & Wilkinson RG (1982) Leucine dissimilation
to isovaleric and isocaproic acids by cell suspensions of
amino acid fermenting anaerobes: the Stickland reaction
revisited. Can J Microbiol 28, 291–300.
18 Thauer RK, Jungermann K & Decker K (1977) Energy
conservation in chemotrophic anaerobic bacteria. Bac-
teriol Rev 41, 100–180.
19 Shultzaberger RK, Bucheimer RE, Rudd KE & Schnei-
der TD (2001) Anatomy of Escherichia coli ribosome
binding sites. J Mol Biol 313, 215–228.
20 Stoddart B & Wilcox MH (2002) Clostridium difficile.
Curr Opin Infect Dis 15, 513–518.
21 Mu
¨
ller U & Buckel W (1995) Activation of (R)-2-hydro-
xyglutaryl-CoA dehydratase from Acidaminococcus
fermentans. Eur J Biochem 230, 698–704.
22 Buckel W (1980) The reversible dehydration of (R)-2-
hydroxyglutarate to (E)-glutaconate. Eur J Biochem 106,
439–447.
23 Brunelle SL & Abeles RH (1993) The stereochemistry of

hydration of acrylyl-CoA catalyzed by lactyl-CoA dehy-
dratase. Bioorganic Chem 21, 118–126.
24 Pitsch C & Simon H (1982) The stereochemical course
of the water elimination from (2R)-phenyllactate in the
amino acid fermentation of Clostridium sporogenes,
Hoppe Seyler’s Z. Physiol Chem 363, 1253–1257.
25 Heider J (2001) A new family of CoA-transferases.
FEBS Lett 509, 345–349.
26 Kim J (2004) On the enzymatic mechanism of 2-hydroxy-
isocaproyl-CoA dehydratase from Clostridium difficile,
PhD Thesis, Philipps-Universita
¨
t Marburg, Marburg,
Germany.
27 Buckel W (1996) Unusual dehydrations in anaerobic bac-
teria: considering ketyls (radical anions) as reactive inter-
mediates in enzymatic reactions. FEBS Lett 389, 20–24.
28 Rees DC & Howard JB (2000) Nitrogenase: standing at
the crossroads. Curr Opin Chem Biol 4, 559–566.
29 Smith DM, Buckel W & Zipse H (2003) Deprotonation
of enoxy radicals: theoretical validation of a 50-year-old
mechanistic proposal. Angew Chem Int Ed Engl 42,
1867–1870.
30 Engst S, Vock P, Wang M, Kim JJ & Ghisla S (1999)
Mechanism of activation of acyl-CoA substrates by
medium chain acyl-CoA dehydrogenase: interaction of
the thioester carbonyl with the flavin adenine dinucleo-
tide ribityl side chain. Biochemistry 38, 257–267.
31 Buckel W (1986) Biotin-dependent decarboxylases as
bacterial sodium pumps: purification and reconstitution

of glutaconyl-CoA decarboxylase from Acidaminococcus
fermentans. Methods Enzymol 125, 547–558.
32 Anderson RL & Wood WA (1969) Carbohydrate meta-
bolism in microorganisms. Annu Rev Microbiol 23,
539–578.
33 Tang KH, Chang CH & Frey PA (2001) Electron
transfer in the substrate-dependent suicide inactivation
of lysine 5,6-aminomutase. Biochemistry 40, 5190–
5199.
34 Frazzon J & Dean DR (2003) Formation of iron-sulfur
clusters in bacteria: an emerging field in bioinorganic
chemistry. Curr Opin Chem Biol 7, 166–173.
35 Wa
¨
chtersha
¨
user G (1992) Groundworks for an evolu-
tionary biochemistry: the iron–sulphur world. Prog Bio-
phys Mol Biol 58, 85–201.
36 Degerbeck F, Fransson B, Grehn L & Ragnarsson U
(1993) Synthesis of
15
N-labelled chiral Boc-amino acids
from triflates: enantiomers of leucine and phenylalanine.
J Chem Soc, Perkin Trans 1, 11–14.
37 Gibson TW & Erman WF (1972) The photochemistry
of substituted 1,5-hexadien-3-ones. J Org Chem 37,
1148–1154.
38 Kawaguchi A, Yoshimura T & Okuda S (1981) A new
method for the preparation of acyl-CoA thioesters.

J Biochem (Tokyo) 89, 337–339.
39 Sambrook J & Russell DW (2001) Molecular Cloning: a
Laboratory Manual, 3rd edn. Cold Spring. Harbor
Laboratory Press, Cold Spring Harbor, NY.
40 Selmer T & Andrei PI (2001) p-Hydroxyphenylacetate
decarboxylase from Clostridium difficile: a novel glycyl
radical enzyme catalysing the formation of p-cresol. Eur
J Biochem 268, 1363–1372.
41 Karasawa T, Ikoma S, Yamakawa K & Nakamura S
(1995) A defined growth medium for Clostridium diffi-
cile. Microbiology 141, 371–375.
42 Jaworek D & Welsch J (1985) Adenosine 5¢-diphosphate
and adenosine 5 ¢-monophosphate. Methods of Enzymat-
ic Analysis (Bergmeyer HU, ed.), pp. 365–370. VCH-
Verlagsgesellschaft, Weinheim, Germany.
43 Ziegenhorn J, Senn M & Bu
¨
cher T (1976) Molar
absorptivities of beta-NADH and beta-NADPH. Clin
Chem 22, 151–160.
44 Selmer T & Buckel W (1999) Oxygen exchange between
acetate and the catalytic glutamate residue in glutaco-
nate CoA-transferase from Acidaminococcus fermentans.
Implications for the mechanism of CoA-ester hydrolysis.
J Biol Chem 274, 20772–20778.
45 Michel C, Buckel W & Linder D (1991) Purification of
the coenzyme B
12
-containing 2-methyleneglutarate
mutase from Clostridium barkeri by high-performance

liquid chromatography. J Chromatogr 587, 93–99.
46 Fish WW (1988) Rapid colorimetric micromethod for
the quantitation of complexed iron in biological sam-
ples. Methods Enzymol 158, 357–364.
47 Cline JD (1969) Spectrophotometric determination of
hydrogen sulfide in natural waters. Limnol Oceanogr 14,
454–458.
48 Eikmanns U & Buckel W (1991) Crystalline green
5-hydroxyvaleryl-CoA dehydratase from Clostridium
aminovalericum. Eur J Biochem 197, 661–668.
J. Kim et al. 2-Hydroxyisocaproyl-CoA dehydratase
FEBS Journal 272 (2005) 550–561 ª 2004 FEBS 561