Oxidation of propionate to pyruvate in
Escherichia coli
Involvement of methylcitrate dehydratase and aconitase
Matthias Brock
1,
*, Claudia Maerker
1,
*, Alexandra Schu¨tz
1
,UweVo¨ lker
1,2,†
and Wolfgang Buckel
1
1
Laboratorium fu
¨
r Mikrobiologie, Fachbereich Biologie, Philipps-Universita
¨
t, Marburg, Germany;
2
Abteilung Biochemie,
Max-Planck-Institut fu
¨
r terrestrische Mikrobiologie, Marburg, Germany
The pathway of the oxidation of propionate to pyruvate in
Escherichia coli involves five enzymes, only two of which,
methylcitrate synthase and 2-methylisocitrate lyase, have
been thoroughly characterized. Here we report that the
isomerization of (2S,3S)-methylcitrate to (2R,3S)-2-methyl-
isocitrate requires a novel enzyme, methylcitrate dehydratase
(PrpD), and the well-known enzyme, aconitase (AcnB), of

showed the production of all proteins encoded by the prp
operon during growth on propionate as sole carbon and
energy source, except PrpE, which seems to be replaced by
acetyl-CoA synthetase. This is in good agreement with
investigations on Salmonella enterica LT2, in which disrup-
tion of the prpE gene showed no visible phenotype.
Keywords: 2-methylisocitrate; aconitase; methylcitrate dehy-
dratase; propionate metabolism; prp operon.
Several bacteria and fungi are able to oxidize propionate via
methylcitrate to pyruvate. Initially propionyl-CoA conden-
ses with oxaloacetate to (2S,3S)-methylcitrate, which iso-
merizes to (2R,3S)-2-methylisocitrate. Cleavage leads to
pyruvate and succinate. The consecutive oxidative regener-
ation of oxaloacetate from succinate completes the methyl-
citrate cycle. Initially this cycle was discovered by growing a
mutant strain of the yeast Candida lipolytica on odd-chain
fatty acids. The accumulation of a tricarboxylic acid was
observed during growth and identified as methylcitrate [1].
Further investigations revealed other enzymes necessary for
a functional methylcitrate cycle. The enzymes, however,
were only partially characterized and no genomic sequences
were identified [2–6]. More recently it was discovered that
propionate oxidation in aerobically growing Gram-negative
bacteria, especially Escherichia coli [7] and Salmonella
enterica serovar Thyphimurium LT2 [8], also proceeds via
methylcitrate. The purification of one of the key enzymes of
the methylcitrate cycle, methylcitrate synthase, led to the
identification of an operon necessary for propionate degra-
dation. In E. coli and S. enterica this prp operon is
composed of the genes prpB, prpC, prpD and prpE.PrpB

*Present address: Institut fu
¨
r Mikrobiologie der Universita
¨
t,
Herrenha
¨
user Str. 2, D-30167 Hannover, Germany. These two authors
contributed equally to this work.
Present address: Funktionelle Genomforschung, Medizinische
Fakulta
¨
t, Ernst-Moritz-Arndt-Universita
¨
t, Walther-Rathenau-Str.
49A, D-17489 Greifswald, Germany.
(Received 28 July 2002, revised 24 October 2002,
accepted 28 October 2002)
Eur. J. Biochem. 269, 6184–6194 (2002) Ó FEBS 2002 doi:10.1046/j.1432-1033.2002.03336.x
catalysed by the PrpD protein [11]. However, the product of
this reaction was not further analysed. It was suggested that
2-methyl-cis-aconitate was formed. Interestingly, this reac-
tion would involve the unusual syn elimination of water,
whereas in all other analysed derivatives of malate this
b-elimination occurs in an anti manner; for a review see [12].
Aconitase from bovine heart follows this rule by dehydra-
ting both substrates, citrate and (2R,3S)-isocitrate, in an anti
manner. Furthermore the enzyme is able to hydrate
2-methyl-cis-aconitate to threo-2-methylisocitrate in an anti
manner, but cannot use methylcitrate as substrate [13].

k
sensitive) was used [15]. For overexpression of the genes prpD
and prpB, E. coli TOP10 cells (Invitrogen) were used,
containing plasmids with the corresponding genes and an
N-terminal cloned histidine tag. For purification of wild-type
enzymes and expression studies, cells were grown aerobically
at 37 °C in minimal medium containing 60 m
M
K
2
HPO
4
,
33 m
M
KH
2
PO
4
,76m
M
(NH
4
)
2
SO
4
,2m
M
trisodium

in an anaerobic chamber (95% N
2
,5%H
2
). Cells were
thawed on ice and suspended in 20 mL anaerobic buffer I
(20 m
M
potassium phosphate, pH 7.5, 1 m
M
trisodium
citrate and 1 m
M
dithiothreitol). Cells were broken by
sonication (Branson sonifier; 3 · 5 min at 60% pulse and
80% of full power). Cell debris was removed by ultracen-
trifugation at 96 000 g for 45 min. This crude extract was
filtered (0.45 lm pore size; Sarsted, Nu
¨
mbrecht, Germany)
and loaded on to a hydroxyapatite column (20 mL bed
volume) equilibrated with buffer I. Unless otherwise indi-
cated, the FPLC system and columns from Amersham
Biosciences were used. The hydroxyapatite column was
washed with buffer I. The flow through was concentrated
with an Amicon chamber over a PM 30 size-exclusion filter
(Millipore) and diluted in buffer II (20 m
M
Tris/HCl,
pH 7.5, with 1 m

volume 30 mL), previously equilibrated with buffer IV
(20 m
M
Tris/citrate, pH 8.0, with 1 m
M
dithiothreitol and
1
M
(NH
4
)
2
SO
4
). The enzyme was eluted with a linear
(NH
4
)
2
SO
4
gradient of 1.0–0
M
in buffer V (20 m
M
Tris/
citrate, pH 8.0, with 1 m
M
dithiothreitol) between 0.2 and
0

¼ 0.8
and induction with 1 m
M
of isopropyl thio-b-
D
-galactoside
followed by incubation overnight. Overproduction of
methylcitrate dehydratase in four different clones was
confirmed by SDS/PAGE. All clones exhibted an induced
protein at 54 kDa (data not shown).
Cells from a 1.2-L culture (D
578
 3) were induced for
10 h and harvested by centrifugation. Cells were washed
with 50 m
M
potassium phosphate, pH 7.0, centrifuged, and
suspended in the same buffer. Cells were broken by
sonication and centrifuged at 96 000 g. The resulting cell-
free extract was loaded on to a gravity flow Ni/nitrilotri-
acetic acid/agarose column with a bed volume of 5 mL.
The column was washed with 20 mL 50 m
M
potassium
Ó FEBS 2002 Isomerization of methylcitrate to 2-methylisocitrate (Eur. J. Biochem. 269) 6185
phosphate, pH 7.0, containing 20 m
M
histidine to remove
unspecifically bound proteins. PrpD was eluted with 50 m
M

CoASH, 500 U phosphotransacetylase and 50 U methyl-
citrate synthase. The reaction was buffered at pH 7.5 in
20 m
M
potassium phosphate. After incubation, the enzymes
were denatured by heat treatment for 20 min at 80 °Cand
centrifuged at 10 000 g for 10 min. The supernatant was
concentrated to a final volume of 10 mL in a rotary
evaporator. Precipitated salts were removed by centrifuga-
tion as described above, and the supernatant was loaded on
to a Dowex 1x8 column (Cl
–
form, bed volume 10 mL).
Methylcitrate was eluted with 1
M
HCl. The methylcitrate-
containing fractions, as tested enzymatically with the PrpD
protein, were concentrated by evaporation. The residual
brownish oil was checked for purity by
1
H-NMR
(500 MHz, CDCl
3
): d ¼ 1.19 (3H, d,
3
J ¼ 6.9 Hz CH
3
),
2.90 (1H, q,
3

M
potassium phosphate, pH 7.5, and 1.3 m
M
methylcitrate in a final volume of 1 mL.
The racemic mixture of chemically synthesized threo-2-
methylisocitrate [9] was used to follow the dehydration and
the formation of the double bond in 2-methyl-cis-aconitate
at 240 nm; e
240
¼ 4.5 m
M
)1
Æcm
)1
[4]. The composition of
the assay was 50 m
M
potassium phosphate, pH 7.5, and
0.3 m
M
threo-2-methylisocitrate in a final volume of 1 mL.
To measure 2-methylisocitrate dehydratase (AcnB), a
coupled assay was performed in the reverse direction. The
reaction was followed at 340 nm under anaerobic condi-
tions with e
340
¼ 6.2 m
M
)1
Æcm

then N-terminally sequenced by Edman degradation (kindly
performed by D. Linder, Universita
¨
t Gießen, Germany).
Re-activation and inactivation of AcnB
AcnB was inactivated by exposure to air and by addition
of either EDTA or o-phenanthroline (both 2.5 m
M
final concentration). For reactivation, 98.3 mg FeSO
4
·
(NH
4
)
2
SO
4
· 6H
2
O (final concentration 5 m
M
)and136mg
cysteine hydrochloride (monohydrate) (15 m
M
)weredis-
solved under anaerobic conditions in 45 mL water, and the
pH was adjusted to 7.5 by dropwise addition of 1
M
NaOH.
Water was added to a final volume of 50 mL. One part of

Synthesis of digoxygenin-labelled RNA probes
For the detection of mRNAs of the genes acs, acnB, prpD
and prpE, specific RNA probes labelled with digoxygenin
6186 M. Brock et al.(Eur. J. Biochem. 269) Ó FEBS 2002
were produced using the T7 polymerase. Oligonucleotides
were designed that contained the sequence of the T7
promoter in the reverse primer (the full sequences of all
primers are shown in Table 1). A PCR was performed with
Taq polymerase, and genomic DNA of E. coli W3350 was
used as a template. PCR products were separated by
electrophoresis in a 1% agarose gel and purified by the
Geneclean Kit II (BIO 101) as described in the manufac-
turer’s protocol. For in vitro transcription, 0.5–1.0 lgPCR
product was mixed with 2 lL NTP labelling mixture
containing UTP (Dig RNA Labelling Kit T7; Roche),
2 lL reaction buffer (Ambion), 1 lL RNase inhibitor (Dig
RNA Labelling Kit T7; Roche), 2 lL T7 polymerase
(20 UÆlL
)1
; Ambion) and diethyl pyrocarbonate-treated
water to a final volume of 20 lL. Transcription was carried
out at 37 °C for 1 h. RNase-free DNase I was added, and
the mixture was incubated at 37 °C for a further 15 min.
RNA was precipitated by the addition of 2.5 lL1
M
LiCl
and 90 lL 100% ethanol and incubated for 1 h at )80 °C.
After centrifugation (12 000 g,4°C), RNA pellets were
dried and dissolved in 100 lL nuclease-free water. The
intensity of the digoxygenin label of the probes was checked

for 2000 Vh, a linear increase from 500 to 3500 V for 10 000
Vh and a final phase of 3500 V for 35 000 Vh (pH 4–7) or
for 21 000 Vh (pH 3–10). IPG-strips were consecutively
incubated for 15 min each in equilibration solution A and
B. Solution A contained 50 m
M
Tris/HCl, pH 6.8, 6
M
urea,
30% glycerol, 4% SDS and dithiothreitol (3.5 mgÆmL
)1
).
Solution B contained iodoacetamide (45 mgÆmL
)1
)instead
of dithiothreitol. In the second dimension, proteins were
separated on SDS/12.5% polyacrylamide gels with the
Investigator
TM
System (Perkin–Elmer Life Sciences,
Cambridge, UK) at 2 W per gel. Gels were stained with
PhastGel BlueR according to the manufacturer’s (Amer-
sham Biosciences) instructions. After scanning, the 2D
PAGE images were analysed with the Melanie3Ò software
package (Bio-Rad Laboratories GmbH). Three separate
gels of each condition and two independent cultivations
were analysed, and only spots displaying the same pattern in
all parallels were selected for further characterization.
Protein identification by peptide mass fingerprinting
Protein spots were excised from PhastGel BlueR-stained 2D

M
MgCl
2
,20m
M
NaN
3
; diethyl pyrocarbonate treated)
and centrifuged for 10 min at 4000 g.Cellpelletswere
suspended in 200 lL killing buffer, and frozen in
liquid nitrogen. Cells were broken in a frozen state in a
Table 1. Oligonucleotides used for the generation of RNA probes. The reverse primer contains the promoter region for the T7 polymerase at the 5¢
end. An asterisk denotes the end of the promoter region.
Probe Reverse primer Forward primer
Acs 5¢-
TAATACGACTCACTATAGGGA*5¢-AACACACCATTCCTGCCAAC-3¢
CCACCACAGGTCGCGCC-3¢
AcnB 5¢-
TAATACGACTCACTATAGGGA*5¢-CTCACACGCTGCTGATGTTC-3¢
CGTGGTTACGCACTTCACC-3¢
PrpD 5¢-
TAATACGACTCACTATAGGGA*5¢-AACATCGGCGCGATGATCC-3¢
TCGCTGCTTCAACTGCCG-3
PrpE 5¢-
TAATACGACTCACTATAGGGA*5¢-ACCGGAGCAGTTCTGGGC-3¢
GATTCCAGCCACGCCACC-3¢
Ó FEBS 2002 Isomerization of methylcitrate to 2-methylisocitrate (Eur. J. Biochem. 269) 6187
Micro-Dismembrator (Braun Biotech International) at
2600 r.p.m. for 2 min. Cell extracts were mixed with 4 mL
lysis solution containing 4

Mops, 50 m
M
sodium acetate,
10 m
M
EDTA, dissolved in diethyl pyrocarbonate-treated
water and adjusted to pH 7.0)] and loaded on to a 1.4%
(w/v) agarose gel containing 1.8
M
formaldehyde. RNA was
separatedat70 Vfor3handtransferredtoanylontransfer
membrane (Schleicher and Schuell) [22]. The RNA blot was
saturated with blocking reagent and hybridized with the
digoxygenin-labelled antisense RNA probes overnight.
After a wash, specific hybridization signals were detected
by incubation with alkaline phosphatase-conjugated anti-
digoxygenin Ig (Roche) and monitoring the conversion of
the ECF-vistra substrate with a STORM 860 fluorimager
(Amersham Biosciences).
Determination of 2-methyl-
cis
-aconitate by anhydride
formation
Enzymatically synthesized methylcitrate (0.8 m
M
)wasdis-
solved in a final volume of 5 mL 20 m
M
Hepes, pH 7.5, and
incubated with 0.5 U PrpD. A 1-mL aliquot was taken, and

(2S,3S)-methylcitrate as substrate and with PrpD, PrpB and
lactate dehydrogenase as auxiliary enzymes (Fig. 1). Start-
ing from 18 g wet cells, the protein was purified from a
specific activity in crude extracts of 0.16 UÆmg
)1
to
1.9 UÆmg
)1
with a yield of 3.6%. Purification was per-
formed by chromatography on hydroxyapatite, Q-Seph-
arose, phenyl-Sepharose and UnoQ (Table 2). A major
band was revealed in the resulting protein fractions by
SDS/PAGE (Fig. 2, lanes 6 and 7) with an apparent
molecular mass of 94 kDa and a turnover number of 3 s
)1
.
Comparison of the forward reaction (hydration of
Fig. 1. Pathway of propionate oxidation to pyruvate. The enzymes are
indicatedinitalics.
Table 2. Purification protocol for AcnB. A unit is defined as the oxidation of 1 lmol NADHÆmin
)1
in the coupled assay.
Purification step
Activity
(U)
Protein
(mg)
Specific activity
(UÆmg
)1

protein by Edman degradation revealed the peptide
sequence, MLEEYXKXVAEXAAE, where X denotes
unclear amino acids. Comparison of this sequence with
the databases showed 100% identity with the N-terminal
sequence of the E. coli citrate cycle aconitase, AcnB
(
SWISSPROT
P36683), with the sequence, MLEEYRKH
VAERAAE. The calculated molecular mass of AcnB from
its genomic sequence is 93 498 Da, which is in good
agreement with the apparent molecular mass of 94 kDa
derived from the SDS/PAGE analysis.
The enzyme was rapidly inactivated by exposure to air,
which is already known for aconitases [24], as well as during
the purification procedure, especially during chromatogra-
phy on the phenyl-Sepharose column. Activity was partially
restored by incubation in re-activation mixture under
anaerobic conditions as described in Experimental proce-
dures. Addition of EDTA or o-phenanthroline totally
inactivated enzymatic activity. This is in good agreement
with the requirement for a functional [4Fe)4S] cluster for
aconitase activity.
Cloning and characterization of PrpD
The prpD gene was cloned and overexpressed as described
in Experimental Procedures. The overproduced protein was
purified to a specific activity of 11.4 UÆ(mg protein)
)1
by
chromatography on a Ni/nitrilotriacetate/agarose column
(Fig. 3). PrpD showed maximum activity with enzymati-

observed in 2,3-dimethylmaleate, which is formed by a
d-isomerase reaction from (R)-3-methylitaconate (2-methy-
lene-3-methylsuccinate) during the nicotinate fermentation
Fig. 3. Analysis of purified PrpD by SDS/PAGE. The protein was
overproduced with an N-terminal His tag and purified by chroma-
tography on a Ni/nitrilotriacetate/agarose column. Lane 1, sample of
purified PrpD; lane M, molecular mass standard.
Fig. 2. Analysis of the purification of AcnB by SDS/PAGE. Lane 1,
crude extract (21 lg); lane 2, hydroxyapatite (30 lg); lane 3, Q-Seph-
arose (14 lg); lane 4, phenyl-Sepharose (5 lg); lane M, molecular mass
standard; lane 5, UnoQ column fraction 40 (2 lg); lane 6, UnoQ
column fraction 48 (2 lg); lane 7, UnoQ column fraction 49 (1 lg).
Lanes 6 and 7 show the purified AcnB protein at 94 kDa as determined
by Edman degradation.
Table 3. Substrate specificity of PrpD. No activity (< 0.01 UÆmg
)1
)
was found with threo-2-methylisocitrate and erythro-2-methylisoci-
trate, trans-aconitate,
D
-malate and
L
-malate, fumarate, maleate,
D
-tartrate and meso-tartrate,
D
-citramalate and
L
-citramalate,
mesaconate, citraconate, itaconate, and (R,S)-3-methylitaconate.

to (2R,3S)-2-methylisocitrate. This product was cleaved by
PrpB into succinate and pyruvate (Fig. 1). To monitor the
reaction and to pull the equilibrium to the side of pyruvate
formation, lactate dehydrogenase and NADH as cosub-
strate were used. This coupled assay was also used to
monitor the purification of the aconitase AcnB as described
above. Absence of the aconitase or any other enzyme
resulted in a loss of pyruvate formation. This result clearly
demonstrates that both proteins, PrpD and AcnB, are
essential for the conversion of methylcitrate into 2-methyl-
isocitrate.
Northern-blot analysis of
prpD
,
prpE
,
acnB
and
acs
transcripts
The four genes were selected for the following reasons.
Transcription levels of acs, the gene coding for acetyl-CoA
synthetase (Acs), were used for comparison of the specificity
of transcription during growth on acetate and propionate,
respectively. Furthermore, this gene was of interest because
of the ability of the Acs to activate propionate to the
corresponding CoA ester. In S. enterica it was shown earlier
thatastraincarryingadeletionoftheacs gene was still able
to grow on propionate but not on acetate. A propionyl-
CoA synthetase mutant was able to grow on propionate as

(kb) are: 2.0 and 5.7* for acs;4.6and5.9*forprpE and prpD.Further
explanations are given in the Results section. The small box on the
right shows the region of the rRNA, to show that the same amount of
RNA was applied to each lane.
Fig. 6. Scheme of the structure of the E. coli and S. enterica prp oper-
ons. All genes are located in the same orientation, and the encoded
proteins show sequence identities of 76–96%. The E. coli operon
contains an additional repetitive extragenic palindromic element
(REP-element) between the prpB and prpC coding sequence. The
DNA sequence of the intergenic region containing the REP-element is
shown in the upper part of the figure. Bold and italic letters highlight
thesinglerepetitiveelements,respectively.
6190 M. Brock et al.(Eur. J. Biochem. 269) Ó FEBS 2002
and therefore comprises a twofold function during growth
on propionate.
For the transcription experiments, RNA was purified
from E. coli W3350 cells grown on glucose, acetate or
propionate as sole carbon and energy source. Cells were
harvested in the early exponential growth phase
(D
578
¼ 0.7–1.2) and broken as described in Experimental
Procedures. Quality and quantity of the RNA used in each
experiment was confirmed by the use of the Agilent 2100
Bioanalyzer. For each probe, the same quantity of RNA
from cells grown on glucose, acetate or propionate was used
(Fig. 5). Arrows denote main transcripts. Those with an
additional asterisk denote larger transcripts, which may be
formed by a read-through and can be observed from high
gene expression or because of an alternative starting point of

observation of acetate excretion and consumption during
growth on glucose medium [27].
2D gel electrophoresis
2D gel electrophoresis was carried out to monitor differ-
ences in the protein pattern of cells grown on acetate or
propionate. E. coli W3350 cells were grown on propionate
or acetate minimal medium and crude extracts were
prepared from exponentially growing cells as described in
Experimental procedures. Figure 7 exemplarily displays the
protein profile of E. coli W3350 grown with either acetate or
propionate as carbon source. Protein spots, which displayed
significantly different intensities under the two growth
conditions, were isolated from the gels and identified by
peptide mass fingerprinting (Table 4). Proteins induced in
Fig. 7. Comparison of the protein profile of E. coli grown in minimal
medium with acetate (A) or propionate (B) as carbon sources. Crude
protein extracts were prepared and separated by 2D gel electrophor-
esis. After staining with PhastGel BlueR, the gels were scanned with an
imaging system and analysed with the Melanie 3.0 software package.
Protein spots induced or repressed by propionate are marked with
arrowheads or boxes, respectively. Proteins identified by peptide mass
fingerprinting are labelled with their gene names. The acnB gene
product was identified by MS analysis of coseparated purified AcnB
and a comparison with previous 2DE data [33]. (C) Alkaline sections
of gels covering the pH range 3–10 and containing PrpC are displayed.
Ó FEBS 2002 Isomerization of methylcitrate to 2-methylisocitrate (Eur. J. Biochem. 269) 6191
the presence of propionate at a higher or lower level than in
the presence of acetate are labelled with arrowheads and
boxes, respectively (Fig. 7). PrpB, PrpC and PrpD encoded
by the prp operon were exclusively produced during growth

propionate via methylcitrate yields pyruvate, which is
converted into acetyl-CoA and funnelled into the citrate
cycle [7]. The observation that AcnB was purified instead of
AcnA is in agreement with the different expression of the
two genes. AcnB was identified as the major citrate cycle
enzyme, whereas AcnA is an anaerobic stationary-phase
enzyme which is specifically induced by iron and redox
stress [29].
Interestingly, two enzymes are involved in the conversion
of methylcitrate into 2-methylisocitrate. PrpD is involved in
the dehydration of (2S,3S)-methylcitrate to 2-methyl-cis-
aconitate. The elimination of water from (2S,3S)-methyl-
citrate to 2-methyl-cis-aconitate is an unusual reaction,
because it displays a syn elimination, which has not
previously been found in any other dehydration of a
derivative of malate. This may explain why PrpD shows no
significant identities with other proteins with known func-
tion except deduced proteins from prp operons of many
proteobacteria, e.g. S. enterica (Fig. 6). In addition, PrpD
shows sequence identities with deduced proteins from the
Gram-positive Bacillus subtilis (61%, Mmge, accession no.
P45859), the eukaroytes Saccharomyces cerevisiae (57%,
Pdh1p, accession no. NP-015326) and Mus musculus (14%,
immune responsive protein 1, accession no. XP-127883), as
well as the archaeon Sulfolobus tokodaii (23%, long
hypothetical Mmge protein, accession no. BAB66901).
The PrpD protein from E. coli possesses high substrate
specificity. The best substrate was stereochemically pure
(2S,3S)-methylcitrate produced by methylcitrate synthases
from E. coli or A. nidulans. Partial activity was also

AceB 5.39 60.3 Malate synthase P08997 49
AcnB 5.24 93.5 Aconitase B P36683 10
Acs 5.50 72.1 Acetyl-CoA synthetase P27550 27
PrpB 5.44 32.1 2-Methylisocitrate lyase
(carboxyphosphoenolpyruvate phosphonomutase)
P77541 52
PrpC 6.66 43.1 Methylcitrate synthase P31660 22
PrpD 5.68 54.0 Methylcitrate dehydratase P77243 49
MglB 5.68 35.7 Galactose-binding protein P02927 38
MalE 5.22 40.7 Maltose-binding protein P02928 71
2. Proteins induced at a lower level as compared with growth on acetate:
AdhP 5.94 35.4 Propanol-preferring alcohol dehydrogenase P39451 54
GpmA 5.86 28.4 Phosphoglycerate mutase 1 P31217 39
PykF 5.77 50.7 Pyruvate kinase P14178 50
6192 M. Brock et al.(Eur. J. Biochem. 269) Ó FEBS 2002
(1.65 UÆmg
)1
protein) was significantly underestimated. The
substrate had been produced with the commercially avail-
able citrate synthase from pig heart, which yielded all four
possible stereoisomers rather than enantiomeric pure
(2S,3S)-methylcitrate as obtained with methylcitrate syn-
thases. Furthermore, the only active stereoisomer is pro-
duced in the lowest amount [13,30]. Our own observations
on the maximum activity of the PrpD protein with a
racemic mixture of all four stereoisomers of chemically
synthesized methylcitrate revealed a 10-fold decrease in
activity. This may also explain the higher relative activities
obtained in the former study with substrates other than
methylcitrate.

to be performed exclusively by the Acs, which was identified
in the 2D gels and Northern-blot experiments of cells grown
on acetate as well as on propionate. Probably prpE
transcripts are translated when the Acs is mutated, as
indirectly shown for S. enterica. In this study an acs mutant
strain was still able to grow on propionate [10].
In conclusion, the prp operon does not harbour all genes
necessary for a functional methylcitrate cycle. However,
propionate catabolism via methylcitrate (Fig. 1) connects
the enzymes of three different pathways to a new functional
unit: AcnB, succinate dehydrogenase, fumarase and malate
dehydrogenase from the citrate cycle, Acs from the glyoxy-
late cycle and three special enzymes, which are capable of
acting on C
7
organic acids (PrpC, PrpD and PrpB).
ACKNOWLEDGEMENTS
The authors thank Professor A. Mosandl, Universita
¨
t Frankfurt/Main,
Germany for performing the enantioselective multidimensional capillar
gas chromatography with our methylcitrate samples, and Dr D. Linder,
Universita
¨
t Gießen, Germany, for the determination of the N-terminus
of aconitase B. The work was supported by grants from Deutsche
Forschungsgemeinschaft and the Fonds der Chemischen Industrie.
REFERENCES
1. Tabuchi, T., Serizawa, N. & Uchiyama, H. (1974) A novel path-
way for the partial oxidation of propionyl-CoA to pyruvate via

8. Horswill, A.R. & Escalante-Semerena, J.C. (1999) Salmonella
typhimurium LT2 catabolizes propionate via the 2- methylcitric
acid cycle. J. Bacteriol. 181, 5615–5623.
9. Brock, M., Darley, D., Textor, S. & Buckel, W. (2001)
2-Methylisocitrate lyases from the bacterium Escherichia coli
and the filamentous fungus Aspergillus nidulans:characterization
and comparison of both enzymes. Eur. J. Biochem. 268, 3577–
3586.
10. Horswill, A.R. & Escalante-Semerena, J.C. (1999) The prpE gene
of Salmonella typhimurium LT2 encodes propionyl-CoA synthe-
tase. Microbiology 145, 1381–1388.
11. Horswill, A.R. & Escalante-Semerena, J.C. (2001) In vitro con-
version of propionate to pyruvate by Salmonella enterica enzymes:
2-methylcitrate dehydratase (PrpD) and aconitase enzymes cata-
lyze the conversion of 2-methylcitrate to 2-methylisocitrate.
Biochemistry 40, 4703–4713.
12. Creighton, D.J. & Murthy, N.R.S.K. (1990) Stereochemistry of
enzyme-catalysed reactions at carbon. In The Enzymes,Vol.19,
3rd edn (Sigman, D.S. & Boyer, P.D., eds), pp. 324–421. Academic
Press Inc., San Diego, USA.
13. Beach, R.L., Aogaichi, T. & Plaut, G.W. (1977) Identification of
D
-threo-alpha-methylisocitrate as stereochemically specific sub-
strate for bovine heart aconitase and inhibitor of TPN-linked
isocitrate dehydrogenase. J. Biol. Chem. 252, 2702–2709.
14. Blank,L.,Green,J.&Guest,J.R.(2002)AcnCofEscherichia coli
is a 2-methylcitrate dehydratase (PrpD) that can use citrate and
isocitrate as substrates. Microbiology 148, 133–146.
15. Campbell, A. (1961) Sensitive mutants of bacteriophage lambda.
Virology 14, 22–32.

24. Kennedy, M.C., Emptage, M.H., Dreyer, J.L. & Beinert, H.
(1983) The role of iron in the activation-inactivation of aconitase.
J. Biol. Chem. 258, 11098–11105.
25. Michel, C., Hartrampf, G. & Buckel, W. (1989) Assay and puri-
fication of the adenosylcobalamin-dependent 2-methylenegluta-
rate mutase from Clostridium barkeri. Eur. J. Biochem. 184,103–
107.
26. Gilson, E., Saurin, W., Perrin, D., Bachellier, S. & Hofnung, M.
(1991) The BIME family of bacterial highly repetitive sequences.
Res. Microbiol. 142, 217–222.
27. Farmer, W.R. & Liao, J.C. (1997) Reduction of aerobic acetate
production by Escherichia coli. Appl. Environ. Microbiol. 63, 3205–
3210.
28. Gawron, O., Glaid, A.J. & Fondy, T.P. (1961) Stereochemistry of
Krebs’ cycle hydrations and related reactions. J. Am. Chem. Soc.
83, 3634–3640.
29. Cunningham, L., Gruer, M.J. & Guest, J.R. (1997) Transcrip-
tional regulation of the aconitase genes (acnA and acnB)of
Escherichia coli. Microbiology 143, 3795–3805.
30. van Rooyen, J.P., Mienie, L.J., Erasmus, E., De Wet, W.J.,
Ketting, D., Duran, M. & Wadman, S.K. (1994) Identification of
the stereoisomeric configurations of methylcitric acid produced by
Si-citrate synthase and methylcitrate synthase using capillary gas
chromatography-mass spectrometry. J. Inherit. Metab. Dis. 17,
738–747.
31. Lauble, H. & Stout, C.D. (1995) Steric and conformational fea-
tures of the aconitase mechanism. Proteins 22, 1–11.
32. Bra
¨
mer, C.O. & Steinbu