Twobeta-alanyl-CoA:ammonialyasesin Clostridium
propionicum
Gloria Herrmann
1
, Thorsten Selmer
1
, Holly J. Jessen
2
, Ravi R. Gokarn
2
, Olga Selifonova
2
,
Steve J. Gort
2
and Wolfgang Buckel
1
1 Laboratorium fu
ă
r Mikrobiologie, Fachbereich Biologie, Philipps-Universita
ă
t, Marburg, Germany
2 Biotechnology Development Center, Cargill Incorporated, Minneapolis, MN, USA
b-Alanine (3-aminopropionate) is formed as an end
product of uracil fermentation [1] and by decarboxyla-
tion of l-aspartate catalysed by the bacterial enzyme as-
partate 1-decarboxlyase (a-decarboxlyase; EC 4.1.1.11)
as a precursor of (R)-pantothenate and hence coen-
zyme A [2,3]. Furthermore, b-alanine, which can be
considered as the homologue of glycine, is a constitu-
ent of homoglutathione from pea nodules and is
important for nitrogen xation [4]. The hydroxyradical
scavenging peptides l-carnosine and carnicine [5] as
well as the antibiotics efrotomycin [6] and destruxin [7]
also contain b-alanine. The only organism known to
ferment this compound is Clostridiumpropionicum [8].
3 b-Alanine ỵ 2H
2
O ẳ 3NH
ỵ
4
ỵ Acetate
ỵ CO
2
ỵ 2 Propionate
1ị
where, DGÂ ẳ )155 kJặmol acetate
)1
(for DGÂ see [9]
and Results).
C. propionicum has been isolated from black mud of
the San Francisco Bay as an organism fermenting
Keywords
acryloyl-CoA; beta-alanyl-CoA; CoA-
transferase;
L-alanine and beta-alanine
fermentation; pentamer
Correspondence
W. Buckel, Laboratorium fu
ă
r Mikrobiologie,
Fachbereich Biologie, Philipps-Universita
ă
t,
35032 Marburg, Germany
Fax: +49 64212828979
Tel: +49 6421 28 21527
E-mail: buckel@staff.uni-marburg.de
(Received 15 October 2004, revised 24
November 2004, accepted 7 December
2004)
doi:10.1111/j.1742-4658.2004.04518.x
The fermentation of b-alanine by Clostridiumpropionicum proceeds via
activation to the CoA-thiol ester, followed by deamination to acryloyl-
CoA, which is also an intermediate in the fermentation of l-alanine. By
shifting the organism from the carbon and energy source a-alanine to
b-alanine, the enzyme b-alanyl-CoA:ammonia lyase is induced 300-fold
(% 30% of the soluble protein). The low basal lyase activity is encoded by
the acl1 gene, whereas the almost identical acl2 gene (six amino acid substi-
tutions) is responsible for the high activity after growth on b-alanine. The
deduced b-alanyl-CoA:ammonia lyase proteins are related to putative
b-aminobutyryl-CoA ammonia lyases involved in lysine fermentation and
found in the genomes of several anaerobic bacteria. b-Alanyl-CoA:ammo-
nia lyase 2 was puried to homogeneity and characterized as a heteropen-
tamer composed of 16 kDa subunits. The apparent K
m
value for acryloyl-
CoA was measured as 23 4 lm, independent of the concentration of the
second substrate ammonia; k
cat
K
m
was calculated as 10
7
m
)1
ặs
)1
. The
apparent K
m
for ammonia was much higher, 70 5 mm at 150 lm acry-
loyl-CoA with a much lower k
cat
K
m
of 4 ã 10
3
m
)1
ặs
)1
. In the reverse
reaction, a K
m
of 210 30 lM was obtained for b-alanyl-CoA. The elim-
ination of ammonia was inhibited by 70% at 100 mm ammonium chloride.
The content of b-alanyl-CoA:ammonia lyase in b-alanine grown cells is
about 100 times higher than that required to sustain the growth rate of
the organism. It is therefore suggested that the enzyme is needed to bind
acryloyl-CoA, in order to keep the toxic free form at a very low level. A
formula was derived for the calculation of isomerization equilibra between
l-alanine b-alanine or d-lactate 3-hydroxypropionate.
Abbreviations
acl1 and acl2, genes encoding b-alanyl-CoA:ammonia lyase 1 and 2.
FEBS Journal 272 (2005) 813821 ê 2005 FEBS 813
l-alanine to ammonia, acetate, CO
2
and propionate
[10,11], according to the so-called nonrandomising
pathway [12,13], with acryloyl-CoA as the characteristic
intermediate [14,15]. As outlined in Fig. 1, 3 mol
l-alanine are oxidatively deaminated to pyruvate by the
combined action of alanine transaminase (EC 2.6.1.2)
and glutamate dehydrogenase (EC 1.4.1.2). Pyruvate
(1 mol) is further oxidized to acetyl-CoA used for sub-
strate level phosphorylation. In order to balance the
redox equilibrium, 2 mol pyruvate are reduced to pro-
pionyl-CoA via (R)-lactate, (R)-lactyl-CoA and acry-
loyl-CoA. Propionate is liberated by CoA-transfer to
(R)-lactate mediated by propionate (lactate) CoA-
transferase (EC 2.8.3.1), which also uses acrylate and
acetate as substrates [16]. While the NAD-dependent
reduction of acryloyl-CoA to propionyl-CoA is irre-
versible under physiological conditions [17], the other
reactions involved in the transformation of pyruvate to
acryloyl-CoA are reversible and may therefore also
operate in the opposite direction. Indeed, it has been
shown that washed cells of C. propionicum not only
ferment pyruvate and (R)-lactate, but also acrylate to
acetate, CO
2
and propionate [11,18].
Considering the ability of C. propionicum to oxidize
acrylate to acetate, only two additional enzymes are
required in order to allow the organism to grow on
b-alanine: the substrate is activated to its CoA-thiol
ester and subsequently ammonia is eliminated from
b-alanyl-CoA to yield acryloyl-CoA. One of these
enzymes, b-alanyl-CoA:ammonia lyase has been parti-
ally purified and initially characterized with artificial
substrates as an acryloyl-CoA aminase [19], while
nothing is known at present about the activation of
b-alanine. In this communication we report the purifi-
cation of b-alanyl-CoA:ammonia lyase to homogeneity
and its biochemical characterization with natural sub-
strates. Two genes encoding b-alanyl-CoA:ammonia
lyases and the gene of a putative b-alanine CoA-trans-
ferase have been found in C. propionicum.
Results
b-Alanyl-CoA:ammonia lyase was purified from cells
of C. propionicum grown on b-alanine as sole carbon
and energy source. The decrease in the absorbance at
259 nm due to the addition of ammonia to the double
bond of acryloyl-CoA (De ¼ 6.4 mm
)1
Æcm
)1
) was used
to monitor activity. The specific activities in cell-free
extracts of b-alanine grown cells (143 UÆmg
)1
) were
over 300-fold higher than in those obtained
for dl-alanine grown cells (0.44 UÆmg
)1
). As shown in
the SDS ⁄ PAGE of Fig. 2, a strong Coomassie-stained
band below 20 kDa accompanied the high specific
activity. The enzyme was purified by single anion
Fig. 1. Proposed pathway of L-alanine and
b-alanine fermetation. Fd and Fd
–
, oxidized
and reduced ferredoxin, respectively; Pct,
propionate (lactate) CoA-transferase; Act,
b-alanine CoA-transferase; Acl, b-alanyl-
CoA:ammonia lyase.
Beta-alanyl-CoA:ammonia lyases G. Herrmann et al.
814 FEBS Journal 272 (2005) 813–821 ª 2005 FEBS
exchange chromatography on Source 15Q in very high
yields, usually more than 50 mg from 10 g of wet
packed cells (Fig. 2 and Table 1). The subunit mole-
cular mass of 15 999 ± 8 Da was determined by
MALDI-TOF MS and the native molecular mass of
75 kDa was estimated by size exclusion chromatogra-
phy, suggesting a homopentameric structure (a
5
) of the
enzyme, which has been recently confirmed by X-ray
crystallography (K Reuter & T Selmer, unpublished
data).
The N-terminal amino acid sequence of purified
b-alanyl-CoA:ammonia lyase was used to clone its
gene. Surprisingly two almost identical genes were
thereby detected. The gene acl1 encoded a 145 amino
acid protein and acl2 encoded a 144 amino acid pro-
tein. Both genes were heterologously expressed in
Escherichia coli and cell-free extracts of these cells were
assayed for enzymatic activity. The mass spectrometric
detection of b-alanyl-CoA formed in the presence of
acryloyl-CoA and ammonia strongly suggested that
the proteins encoded by acl1 and acl2 are both func-
tional b-alanyl-CoA:ammonia lyases (lyase 1 and 2).
The observed mass of the purified enzyme
(15 999 ± 8 Da) is in accord with the predicted mass
of 16 004 Da for lyase 2, the protein encoded by acl2,
whereas the mass of lyase 1 was calculated as
16 205 Da. Furthermore, the peptide map of the isola-
ted protein also matched the sequence of lyase 2. How-
ever, we cannot exclude the presence of some lyase 1
(< 0.5%) in this purified enzyme. Attempts to charac-
terize the ammonia lyase from dl-alanine grown cells
were only partially successful, certainly due to the low
expression level of acl1. Mass spectrometry revealed
the presence of a 16.2 kDa protein in the impure prep-
arations, indicating that lyase 1 accounts for the basal
activity in the absence of b-alanine.
Both sequences of the b-alanyl-CoA:ammonia lyases
from C. propionicum are very similar to those of pro-
teins with unknown function found in the genomes of
Fusobacterium nucleatum [20], Clostridium tetani [21],
Thermoanaerobacter tengcongensis [22] and Porphyro-
monas gingivalis [23] (41–46% identity). While the gene
for acl1 appears to be devoid of neighbouring genes in
the genomic DNA of C. propionicum, the acl2 gene is
apparently expressed in context with two other genes.
Downstream of the acl2 gene a second ORF (act1)
was identified, encoding a 397 amino acid protein with
a predicted molecular mass of 43 853 Da. The encoded
protein showed significant similarities (35–51% iden-
tity) to several proteins in the genomic sequences
databases of unknown function and to formyl-CoA-
oxalate CoA-transferase from Oxalobacter formigenes
(up to 38% identity [24]). Therefore, the gene probably
encodes b-alanine CoA-transferase, which is required
for the initial activation of b-alanine. Downstream of
the transferase gene, a third ORF was partially se-
quenced, which showed some similarity to merR of
Bacillus subtillis and may be the transcription initi-
ator ⁄ regulator for b-alanine fermentation (bar). The
coding DNA sequences for acl1, acl2 and act1 as well
as the partial sequence of the regulator have been
deposited in the EMBL nucleotide sequence database
under the accession numbers AJ715481 (acl1) and
AJ715482 (acl2, act and bar).
b-Alanyl-CoA:ammonia lyase 2 from C. propionicum
was used to determine the kinetic properties with the
natural substrates acryloyl-CoA, ammonia and b-ala-
nyl-CoA (Table 2). In order to allow a wider variation
of the acryloyl-CoA concentration, the disappearance
of the double bond in acryloyl-CoA was monitored at
Fig. 2. SDS ⁄ PAGE of b-alanyl-CoA:ammonia lyase preparations.
The lanes contained: 1, marker proteins with molecular masses
(kDa); 2, cell free-extract of
L-alanine grown cells (18 lg protein); 3,
cell free-extract of b-alanine grown cells (18 lg protein); 4, eluate
from Source 15Q (2.8 lg of pure b-alanyl-CoA ammonia lyase).
Table 1. Purification of the b-alanyl-CoA:ammonia lyase from
C. propionicum. Cell-free extract prepared from 2 g b-alanine grown
wet cell paste; 1 U total activity ¼ amination of 1 lmol acryloyl-
CoAÆmin
)1
.
Purification
step
Total
Protein (mg)
Total
activity (U)
Specific activity
(UÆmg protein
)1
)
Yield
(%)
Cell-free extract 183 22661 124 100
Source 15Q 10 10208 1033 45
G. Herrmann et al. Beta-alanyl-CoA:ammonia lyases
FEBS Journal 272 (2005) 813–821 ª 2005 FEBS 815
280 nm (De ¼ 3.5 mm
)1
Æcm
)1
) in potassium phosphate
pH 7.5 using ammonium chloride at concentrations up
to 100 mm. Irrespectively of the ammonium chloride
concentration, K
m
¼ 23 ± 4 lm for acryloyl-CoA was
observed, while the specific activities increased with
increasing ammonium chloride concentrations (up to
% 1000 UÆmg
)1
); at 100 mm NH
4
Cl the k
cat
⁄ K
m
¼
10
7
m
)1
Æs
)1
. A much higher K
m
¼ 70 ± 5 mm and a
2.3 · 10
3
-times lower k
cat
⁄ K
m
was measured for NH
4
Cl
(at 150 lm acryloyl-CoA). While the addition of
ammonia to the double bond of acryloyl-CoA was
readily measurable, the elimination of ammonia from
b-alanyl-CoA was not easily detected. Starting with
b-alanyl-CoA, only trace amounts of acryloyl-CoA
(< 1%) have been observed by reverse phase-HPLC.
In contrast, a rapid incorporation of deuterium into
b-alanyl-CoA has been found by mass spectrometry
when the reaction was carried out in deuterium-
enriched water (data not shown). These observations
suggest that the equilibrium of the reaction strongly
favours b-alanyl-CoA over acryloyl-CoA even in the
initial absence of ammonia.
The elimination of ammonia from b-alanyl-CoA,
however, was readily measured when the acryloyl-CoA
reductase complex was used to trap acryloyl-CoA [17].
In order to avoid NADH-oxidase activity of the reduc-
tase complex under air, the assay was performed under
strict anoxic conditions using reduced methylviologen as
the electron donor. Because the reduction of acryloyl-
CoA to propionyl-CoA is irreversible under physiologi-
cal conditions, the acryloyl-CoA reductase allowed the
ammonia lyase-dependent conversion of b-alanyl-CoA
to propionyl-CoA, which was coupled to the oxidation
of the dye (e
604nm
¼ 2 · 13.6 mm
)1
Æcm
)1
). Using this
assay, K
m
¼ 210 ± 30 lm was obtained for b-alanyl-
CoA. The elimination of ammonia appeared to be inhib-
ited by ammonium chloride; 70% inhibiton was
observed at 100 mm, whereas the K
m
for b-alanyl-CoA
remained constant. The addition of a nucleophile to the
double bond of acryloyl-CoA seems to be restricted to
ammonia and methylamine. The products were exam-
ined by MALDI-TOF mass spectrometry, which gave
the expected masses for b-alanyl-CoA (840 Da) and
b-(methylamino)propionyl-CoA (854 Da ¼ 840 + 14).
Hydroxylamine, glycine, water and hydrogensulfide
were not suitable substrates for lyase 2. In the incuba-
tions with water, glycine and hydrogensulfide, the
unchanged acryloyl-CoA could be recovered, whereas
with hydroxylamine cleavage of the thiol ester was
observed. As judged by enzymatic assays, the enzyme
exhibited activity with both crotonyl-CoA and meth-
acryloyl-CoA. The formation of the predicted products
3-aminobutyryl-CoA and 3-aminoisobutyryl-CoA was
confirmed by mass spectrometry.
Calculation of the Gibbs free energy of b-alanine
formation (DGƒ°)
The two alanine isomers differ mainly by the pK of
the carboxyl group (pK
A
¼ 2.35 for l-a-alanine and
pK
B
¼ 3.60 for b-alanine [25]) and by the stereochem-
istry. Hence the equilibrium between the protonated
forms dl-a-alanine and b-alanine should be close to
1.0, whereas that between chiral l-a-alanine (AH
+
)
and achiral b-alanine (BH
+
) should be 2.0, whereby
the factor 2 takes into account that the formations of
l- and d-a-alanine from b-alanine are equal [26]:
AH
þ
! BH
þ
; K ¼½BH
þ
=½AH
þ
¼2:
Equilibrium between the zwitterionic forms of l-a-
alanine (A) and b-alanine (B):
A ! B; K
eq
¼½B=½A;
with K
B
¼ [B] · [H
+
] ⁄ [BH
+
], K
A
¼ [A] · [H
+
] ⁄ [AH
+
]
and [BH
+
] ⁄ [AH
+
] ¼ 2:
K
eq
¼ 2 K
B
=K
A
pK
eq
¼Àlog2þpK
B
ÀpK
A
¼À0:30þ3:60 À2:35 ¼þ0:95
K
eq
¼ 0.11, favouring l-a-alanine.
With DG° ¼ –RT ln K
eq
¼ )5.7 log K
eq
(kJÆmol
)1
)the
DGƒ° of b-alanine ()366.1 kJÆmol
)1
) is more positive
Table 2. Kinetic constants of b -alanyl-CoA:ammonia lyase 2 of C. propionicum. Apparent molecular mass of 16 kDa was used to calculate
k
cat
values.
Substrate Cosubstrate ⁄ Inhbitor Apparent K
m
(mM) V
max
(UÆmg
)1
) k
cat
(s
)1
) k
cat
⁄ K
m
(lM
)1
s
)1
)
Acryloyl-CoA 50 m
M NH
4
Cl 0.023 ± 0.004 730 195 8.5
Acryloyl-CoA 100 m
M NH
4
Cl 0.026 ± 0.004 956 255 9.8
NH
4
Cl 150 lM Acryloyl-CoA 70 ± 5 1127 301 0.0043
b-Alanyl-CoA 0.21 ± 0.03 95 25 0.12
b-Alanyl-CoA 10 m
M NH
4
Cl 0.23 ± 0.03 81
b-Alanyl-CoA 100 m
M NH
4
Cl 0.16 ± 0.07 27
Beta-alanyl-CoA:ammonia lyases G. Herrmann et al.
816 FEBS Journal 272 (2005) 813–821 ª 2005 FEBS
by about +5.4 kJÆmol
)1
than that of l-a-alanine
()371.5 kJÆmol
)1
[9]).
Discussion
The results presented in this work show clearly that
the introduction of b-alanine into the acryloyl-CoA
fermentation pathway of C. propionicum occurs at the
CoA-thiol ester level. This is demonstrated by the
300-fold induction of b-alanyl-CoA:ammonia lyase 2
(Acl2) after growth of C. propionicum on b-alanine.
The high lyase activity is caused by the expression of
the acl2-gene, whereas the low activity found in
dl-alanine grown cells stems from the acl1-gene, which
is expressed only to a low level. The ammonia lyase 2
appears to be coinduced with a putative b-alanine
CoA-transferase required for activation of the b-amino
acid (Act). The enzyme is a class III CoA-transferase
[27] different from the already known class I propion-
ate CoA-transferase from C. propionicum [16], which
does not accept b-alanine as substrate. Preliminary
results indeed showed the expression of the act-gene
in C. propionicum, but only in b-alanine grown cells
(T Selmer, unpublished observations). Thus the physio-
logical function of the b-alanyl-CoA ammonia lyase 1
remains to be established. Possibly the lyase 1 could
provide b-alanyl-CoA from acryloyl-CoA and ammo-
nia for the synthesis of unknown peptides. The synthe-
sis of pantothenate requires free b-alanine [2,3], but
how it is formed from the thiol ester in the absence of
b-alanine CoA-transferase is not known.
Among the most striking features of b-alanine fer-
mentation in C. propionicum is the extraordinarily high
expression level of the acl2-gene in media supplemented
with b-alanine. In fact, it can be estimated that the lyase
2 accounts for 30% of the total cell protein under these
conditions. It still remains an enigma why the organism
produces such a high level of a functional enzyme. The
cell-free extract catalyses the elimination of ammonia
with the high specific activity of 144 UÆmg
)1
; therefore
100-times lower concentrations of the enzyme could eas-
ily account for the growth rates observed on b-alanine.
It is known that often the activity of enzymes exceeds
that necessary for the observed substrate fluxes, but the
occurrence of a highly active enzyme, which comprises
30% of the total soluble protein, is unusual and requires
a special explanation. The product of ammonia elimin-
ation, acryloyl-CoA, is a very strong electrophile, much
stronger than crotonyl-CoA, and reacts readily with
nucleophilic groups in essential biomolecules (thiols,
enzymes and nucleic acids). The large quantities of
b-alanyl CoA:ammonia lyase may be necessary to bind
acryloyl-CoA, in order to keep the free form of this
reactive thiol ester in the cell at a very low concentra-
tion. In line with this proposal it has been shown that
the K
m
values of acryloyl-CoA reductase (2 ± 1 lm)
[17], and ammonia lyase 2 (23 ± 4 lm) (this paper), are
very low, one to two orders of magnitude lower than the
K
m
value of b-alanyl-CoA (210 ± 30 lm). On the other
hand the relatively high amount of acryloyl-CoA, which
may be stored by the ammonia lyase, ensures an effi-
cient hydration to (R)-lactyl-CoA, necessary for the
oxidative branch of the fermentation pathway (Fig. 1).
The observed promiscuity of the ammonia lyase 2
with respect to the enoyl-CoA derivatives used as
ammonia (and methylamine) acceptors may offer inter-
esting technical applications for the enzyme. Enantio-
merically pure, bifunctional b-amino C
4
-compounds
like 3-aminobutyrate or 3-aminoisobutyrate are inter-
esting building blocks in preparative organic chemistry.
The enzyme-mediated addition of ammonia to meth-
acryloyl-CoA will very likely result in only one of the
two possible enantiomers of 3-aminoisobutyryl-CoA
and should therefore offer a synthetic route for de novo
synthesis of chiral centres starting from commercially
available bulk chemicals. However, the purification
and characterization of suitable CoA-transferases will
be required, in order to allow the synthesis to operate
on a commercially meaningful scale. The analysis of
b-alanine CoA-transferase from C. propionicum may
be helpful in this respect and its purification and char-
acterization is currently in process in our laboratory.
F. nucleatum, whose genome has been sequenced [20],
is known to ferment l-lysine via 3,5-diaminohexanoate
to ammonia, acetate and butyrate [28,29]. This pathway
involves the reversible deamination of 3-aminobutyryl-
CoA to crotonyl-CoA [30], a reaction also catalysed by
b-alanine ammonia lyase. It is therefore most likely that
the deduced fusobacterial protein, which is homologous
to the lyases 1 and 2, functions as b-aminobutyryl-CoA
ammonia lyase. Probably the deduced homologous
amino lyasesin C. tetani [21], T. tengcongensis [22] and
P. gingivalis [23] also use b-aminobutyryl-CoA as pre-
ferred substrate, as these organisms contain additional
genes coding for enzymes of the lysine fermentation
pathway.
The Gibbs free energy DG°¢ of fermentations like
that of b -alanine to ammonia, CO
2
, acetate and
propionate (Eqn 1) can be calculated readily from the
DGƒ° values (Gibbs free energies of formation) of the
reaction partners. Unfortunately, only the DGƒ° value
of l-a-alanine but not that of b-alanine is available [9].
Therefore, we assume that DGƒ° of the protonated
forms of dl-a-alanine and b-alanine are equal, i.e. that
the interactions of the ammonium groups with the
carboxylic acid groups mainly influence the pK values
G. Herrmann et al. Beta-alanyl-CoA:ammonia lyases
FEBS Journal 272 (2005) 813–821 ª 2005 FEBS 817
of the carboxylates (pK
a-alanine
¼ 2.35; pK
b-alanine
¼
3.60). In addition the difference in stereochemistry has
to be taken into account [26]. These considerations
lead towards the equilibrium between chiral l-a- and
achiral b-alanine to K
eq
¼ 2 · K
b-alanine
⁄ K
a-alanine
¼
0.11, DGƒ° ¼ )366.1 kJÆmol
)1
for b-alanine and
DG°¢ ¼ )155 kJÆmol
)1
acetate (Eqn 1). A similar calcu-
lation can be made for the equilibrium between chiral
d-lactate (pK ¼ 3.8) and achiral 3-hydroxypropionate
(pK ¼ 4.5), yielding K
eq
¼ 0.4. This value might be
useful for the design of a 3-hydroxypropionate produc-
tion pathway from carbohydrates [31].
The maximum possible amount of conserved ATP in
fermentations can be calculated by dividing the DG°¢
value by )75 kJÆmol
)1
[9]. Hence, in the case of
b-alanine fermentation (Eqn 1) up to )155 ⁄ )75 %
2 mol ATPÆmol acetate
)1
can be conserved. As shown
in Fig. 1, however, only one ATP is conserved via sub-
strate level phosphorylation from acetyl-CoA obtained
by pyruvate oxidation. Therefore, either the other
possible ATP is wasted or there is an additional elec-
tron transport phosphorylation. A possible function of
acryloyl-CoA as terminal acceptor in an electron trans-
port phosphorylation similar to fumarate reductase
[32] appears unlikely as it has been shown that acry-
loyl-CoA reductase is a soluble enzyme, which uses
NADH rather than menaquinol as electron donor.
Growth experiments with two lactate fermenting
organisms confirm the conservation of only one
ATP ⁄ acetate via the acryloyl-CoA pathway [33]. Clos-
tridium homopropionicum ferments lactate via acryloyl-
CoA (Fig. 1) and its growth yield (2.5 g dried cellsÆmol
lactate
)1
) is less than half as high that of Propionibac-
terium freudenreichii (6.0 gÆmol lactate
)1
), which uses
another pathway via the coenzyme B
12
-dependent
methylmalonyl-CoA mutase. The latter pathway com-
prises additional electron transport phosphorylation
via the oxidation of reduced ferredoxin and NADH by
fumarate via menaquinol [32]. Hence, the low ATP
yield in the b-alanine fermentation by C. propionicum
should increase the catabolic substrate flux, which in
the presence of high b-alanine concentrations enhances
the growth rate [33] and probably keeps the concentra-
tion of the toxic acryloyl-CoA at a low level.
Experimental procedures
Materials
Clostridium propionicum (DSM 1682) was obtained from the
Deutsche Sammlung fu
¨
r Mikroorganismen und Zellkulturen
(DSMZ, Braunschweig, Germany). Sequencing grade pro-
teases and citrate synthase were from Roche (Mannheim,
Germany). Coenzyme A (tri-lithium salt) was from ICN
Biomedicals (Eschwege, Germany) and crotonyl-CoA from
Sigma-Aldrich Chemie GmbH (Mu
¨
nchen, Germany). All
other chemicals were of the highest available grade.
Synthesis of acyl-CoA substrates
Acetyl-CoA, butyryl-CoA and propionyl-CoA were syn-
thesized from CoASH with a small excess of their corres-
ponding anhydrides according to the method of Simon &
Shemin [34]. Acryloyl-CoA and methacryloyl-CoA were
prepared enzymatically from 1 mm acetyl-CoA and freshly
neutralized acid (100 mm) using propionate CoA-trans-
ferase [14] (10 mUÆmL
)1
)in50mm potassium phosphate,
pH 7.0. b-Alanyl-CoA and b-aminoisobutyryl-CoA
(3-amino-2-methylpropionyl-CoA) were prepared from
1mm acetyl-CoA, 100 mm acrylate or methacrylate and
ammonium chloride (100 mm) using propionate CoA-
transferase (10 mUÆmL
)1
) and b-alanyl-CoA:ammonia lyase
(1 UÆmL
)1
)in50mm potassium phosphate, pH 7.0. All
CoA derivatives were purified using a C-18 cartridge as des-
cribed earlier [35], freeze-dried and stored at )80 °C.
Cultivation and storage of C. propionicum
The microorganism was cultivated in a complex medium
containing either dl-alanine or b-alanine as major energy
source as described previously [14]. Freshly prepared
anoxic media were inoculated with 5–20% stationary or
late exponential precultures and grown for 24–36 h at
37 °C. The cells were harvested by centrifugation and
stored at )80 °C.
Preparation of cell free extracts
Frozen cells (2 g) were suspended in 25 mm Tris ⁄ HCl,
pH 7.5, 1 mm dithiothreitol, 1 mm MgCl
2
,1mm EDTA
(buffer A). Cells were broken by sonication on ice and cell
debris was removed by centrifugation for 1 h at 100 000 g
at 4 °C. The clear supernatant was used immediately.
Purification of b-alanyl-CoA:ammonia lyase
All purification steps were carried out at 4 °C under air.
Cell free extracts of b-alanine grown cells were applied on
a Source 15Q (1.6 ⁄ 20) column (Amersham-Pharmacia,
Freiburg, Germany) equilibrated with buffer A. The col-
umn was developed using a linear gradient of 0–500 mm
NaCl in buffer A. Fractions containing activity were
pooled, concentrated and desalted by repetitive membrane
centrifugation using Vivaspin 10 kDa cut-off concentrators
(Vivascience, Hannover, Germany). The final preparation
was stored at )20 °C. b-Alanyl-CoA:ammonia lyase activ-
ity was measured following the decrease in absorbance of
Beta-alanyl-CoA:ammonia lyases G. Herrmann et al.
818 FEBS Journal 272 (2005) 813–821 ª 2005 FEBS
acryloyl-CoA at 259 nm. The assay mixture contained
100 mm triethanolamine ⁄ HCl, pH 8.0, 100 mm ammonium
chloride and 20 lm acryloyl-CoA. The catalytic activity
was calculated using a De
259nm
¼ 6.4 m m
)1
Æcm
)1
between
acryloyl-CoA (22.3 mm
)1
Æcm
)1
) and b-alanyl-CoA
(15.9 mm
)1
Æcm
)1
).
N-terminal sequencing
Purified protein (1 nmol) was loaded on a Supelcosil-
LC3DP (4.6 · 250 mm) column (Sigma-Aldrich) equili-
brated with 0.1% (v ⁄ v) trifluoroactetic acid. The column
was developed using a linear gradient from 0 to 85%
(v ⁄ v) acetonitrile. Elution of the protein was monitored
at 280 nm. The protein was collected manually and aliqu-
ots (100 pmol) were N-terminally sequenced by Edman
degradation and analysed by MALDI-TOF MS.
Peptide mapping
HPLC-purified protein was subjected to reduction with di-
thiothreitol and S-carboxymethylation with sodium iodo-
acetate [36]. The proteins were desalted by size exclusion
chromatography on Sephadex G25 equilibrated with 10%
(v ⁄ v) acetonitrile in 50 mm ammonium acetate, pH 8.0.
Aliquots (100 pmol) were digested with 1% (w ⁄ w) sequen-
cing grade endoproteases (trypsin, AspN, GluC, LysC,
ArgC) at 37 °C for 4 h. The samples were analysed by
MALDI-TOF MS [16].
Determination of the gene sequences of
b-alanyl-CoA:ammonia lyases 1 and 2
The 35 amino acid N-terminal sequence of the b-alanyl-
CoA:ammonia lyase 2 was used to design degenerate prim-
ers with which the corresponding DNA from genome of
C. propionicium was amplified. Genomic DNA from C. pro-
pionicum was isolated using the Gentra Gram-positive
genomic DNA isolation procedure (Gentra Systems, Min-
neapolis, MN, USA). The primers: 5¢-ATGGTWGGY
AARAARGTWGT-3¢ and 5¢- TCRCCCCAYTGRTTWA
CRAT-3¢ were used in a touchdown PCR program with
annealing temperatures of 58 °Cto52°C. A 100 bp PCR
product was purified from an agarose gel (Qiagen, Valen-
cia, CA, USA), ligated into pCRII-TOPO vector, trans-
formed into TOP10 E. coli cells using a TOPO cloning
procedure (Invitrogen, Carlsbad, CA, USA), and sequenced
using vector primers.
Primers for genome walking both upstream and down-
stream of the gene fragment were designed using the por-
tion of the nucleic acid sequence internal to the degenerate
primers. The primer sequences GSP1F: 5 ¢-GTACATCATT
TAATGATGAGCGCAAAAGATG-3¢; GSP2F: 5¢-GAT
GCTCACTATACTGGAAACTTAGTAAAC-3¢; GSP1R:
5¢-ATTCTAGCGCCGTTTACTAAGTTTCCAG-3¢; and
GSP2R: 5¢-CCAGTATAGTGAGCATCTTTTGCGCTCA
TC-3¢ were used, where GSP1F and GSP2F are primers
facing downstream, GSP1R and GSP2R are primers facing
upstream, and GSP2F and GSP2R are primers nested
inside GSP1F and GSP1R, respectively. Genome walking
libraries were constructed according to the Clontech Uni-
versal GenomeWalker
TM
Kit User Manual (Clontech
Laboratories, Palo Alto, CA, USA), with the exception that
the restriction enzymes SspI and HincII were used in addi-
tion to DraI, EcoRV, and PvuII. PCR was conducted in a
PerkinElmer 9700 Thermocycler using the following reac-
tion mix: 1· XL Buffer II, 0.2 mm each dNTP, 1.25 mm
Mg(OAc)
2
, 0.2 lm each primer, 2 units of rTth DNA
polymerase XL (Applied Biosystems, Foster City, CA,
USA), and 1 lL of library per 50 lL reaction. Second
round PCR product was separated on an agarose gel, puri-
fied, and cloned as described previously: a 1.4 kb DraI
band and 1.5 kb HincII band were cloned for the forward
reactions, and a 0.8 kb EcoRV band and 2.0 Kb HincII
band were cloned for the reverse reaction. Sequencing of
the clones showed that two very homologous b-alanyl-
CoA:ammonia lyase genes had been amplified.
A second round of genome walking and cloning was con-
ducted, as described above, to obtain the full sequence of the
second b-alanyl-CoA:ammonia lyase homologue. The prim-
ers GW2GSP1R, 5¢-TTATTGAGGGTGCTTTGCATCCT
TGAAG-3¢, and GW2GSP2R, 5¢-AAGGCTGCCTGTTG
CAGTACCACAAAG-3¢ were used. Bands (1.7 kb) were
cloned from second round PCR product from the DraI and
SspI libraries. The sequence of the second homolog corres-
ponded to the purified b-alanyl-CoA:ammonia lyase protein.
Both acl genes were amplified by PCR with the follow-
ing primers: aclNdeF: 5¢-GGGAATTCCATATGGTAGG
TAAAAAGGTTGTACATC-3¢, acl1BamR: 5¢-CGACG
GA-TCCATTCGTCCGCTTGAATAACTAAAG-3¢, acl2-
BamR: 5¢-CGACGGATCCCGAAAA-TGTCACCAAAA
ATTATTGAG-3¢. The aclNdeF forward primer was the
same for both genes as the genes were identical at the
beginning. PCR was conducted in a PerkinElmer 2400
Thermocycler using Pfu Turbo polymerase (Stratagene,
La Jolla, CA, USA) according to the manufacturer’s
instructions. PCR was performed under the following
conditions: initial denaturation step 94 °C for 2 min; Fol-
lowed by 25 cycles of 94 °C for 30 s, 55 °C for 30 s,
72 °C for 2 min; final extension at 72 ° C for 7 min.
Obtained PCR products were digested with NdeI and
BamHI restriction enzymes and gel purified using a
Qiagen Gel Extraction Kit (Qiagen). The genes were cloned
in pET11a vector digested with NdeI and BamHI as well.
Resulting plasmids pACL-1 and pACL-2 were transformed
into BL21(DE3) cells to study gene expression.
DNA sequence obtained downstream of the second
b-alanyl-CoA:ammonia lyase homologue indicated the
G. Herrmann et al. Beta-alanyl-CoA:ammonia lyases
FEBS Journal 272 (2005) 813–821 ª 2005 FEBS 819
possible presence of a partial coding sequence for a CoA-
transferase. A third round of genome walking was conduc-
ted to obtain the full gene sequence of this downstream
open reading frame. Genome walking was conducted using
primers GW2GSP1F, 5¢-CTATGTAAAGCAATGGGCAG
AGAGGATTTG-3¢, and GW2GSP2F, 5¢-TCCTCGTTTC
AATACAAACCTGAATCGTTG-3¢. A 0.9 Kb SspI band
was cloned and sequenced to obtain the full sequence of
the open reading frame.
Gene expression
BL21(DE3) carrying pET11a (control), pACL-1 and pACL-
2 were grown in 10 mL LB medium supplemented with car-
benicillin (50 lgÆmL
)1
) to an attenuance at 600 nm (D
600
)of
% 0.5 and induced with 100 lm isopropyl thio-b-d-galacto-
side for 4 h. The induced cells were collected by centrifuga-
tion at % 2000 g in an Avanti J20 centrifuge (Beckman,
Fullerton, CA, USA) and treated with Bug Buster (Nov-
agen, Madison, WI, USA) according to the manufacturer’s
instructions. Obtained cell extract was used for the enzyme
assay that followed the conversion of acryloyl-CoA to
b-alanine-CoA by mass spectrometry. The assay mixture
was the same as described above for purified protein.
Enzyme activities
The K
m
values for acryloyl-CoA and ammonia were deter-
mined following the decrease in absorbance of acryloyl-
CoA at 280 nm (De ¼ 3.5 mm
)1
Æcm
)1
)in50mm potassium
phosphate, pH 7.5. The concentrations of acryloyl-CoA
were varied between 5 lm and 50 lm at 50 mm ammonium
chloride and the ammonium chloride concentrations were
varied between 10 mm and 70 mm at 150 lm acryloyl-CoA.
The elimination of ammonia from b-alanyl-CoA was cou-
pled to the acryloyl-CoA reductase-mediated oxidation of
reduced methylviologen in 50 mm potassium phosphate at
pH 7.0 under strict anoxic conditions. The assay contained
1mm methylviologen, which was titrated with 10 mm
Ti(III)citrate to an absorbance at 604 nm (A
604
)of% 0.8.
Considering the need of two molecules of methylviologen
being oxidized per mole of acryloyl-CoA reduced, e
604
¼
27.2 mm
)1
Æcm
)1
was used to calculate activities [37]. In
order to determine the specificity of the nucleophile,
0.1 mm acryloyl-CoA was incubated with ammonium chlor-
ide, hydroxylammonium chloride, glycine, or methylamo-
nium chloride, 100 mm each, or 5 mm sodium sulfide in
50 mm potassium phosphate pH 7.5 with b-alanyl-CoA
ammonia lyase (10 UÆmL
)1
) for 30 min at ambient
temperature. The resulting CoA-thiol esters were isolated as
described above and subjected MALDI-TOF mass spectro-
metry. In a similar way methylacryloyl-CoA or crotonyl-
CoA, 0.1 mm each, were incubated with 100 mm
ammonium chloride and the products were analysed by
mass spectrometry.
Acknowledgements
This work was supported by grants from the Deutsche
Forschungsgemeinschaft and the Fonds der chemis-
chen Industrie. Dr Bernhard Schmidt and Klaus Neifer
(Universita
¨
tGo
¨
ttingen, Germany) sequenced the
N-terminus of b-alanine ammonia lyase. We thank
Dr Antonio J. Pierik (Philipps-Universita
¨
t Marburg)
for helpful advice.
References
1 Barker HA (1961) Fermentations of nitrogenous organic
compounds. In The Bacteria (Gunsalus IC, ed.), pp.
151–207. Academic Press Inc, New York.
2 Cronan JE Jr (1980) Beta-alanine synthesis in Escheri-
chia coli. J Bacteriol 141, 1291–1297.
3 Cronan JE Jr, Littel KJ & Jackowski S (1982) Genetic
and biochemical analyses of pantothenate biosynthesis
in Escherichia coli and Salmonella typhimurium. J Bac-
teriol 149, 916–922.
4 Iturbe-Ormaetxe I, Heras B, Matamoros MA, Ramos J,
Moran JF & Becana M (2002) Cloning and functional
characterization of a homoglutathione synthetase from
pea nodules. Physiol Plant 115, 69–73.
5 Babizhayev MA, Seguin MC, Gueyne J, Evstigneeva
RP, Ageyeva EA & Zheltukhina GA (1994) l-carnosine
(beta-alanyl-l-histidine) and carcinine (beta-alanylhista-
mine) act as natural antioxidants with hydroxyl-radical-
scavenging and lipid-peroxidase activities. Biochem J
304, 509–516.
6 Darland G, Arison B & Kaplan L (1991) The biosyn-
thetic origin of the pyridone ring of efrotomycin. J Ind
Microbiol 8, 265–271.
7 Liu BL, Chen JW & Tzeng YM (2000) Production of
cyclodepsipeptides destruxin A and B from Metarhizium
anisopliae. Biotechnol Prog 16, 993–999.
8 Goldfine H & Stadtman ER (1960) Propionic acid meta-
bolism. 5. The conversion of beta-alanine to propionic
acid by cellfree extracts of Clostridium propioncum.
J Biol Chem 235, 2238–2245.
9 Thauer RK, Jungermann K & Decker K (1977) Energy
conservation in chemotrophic anaerobic bacteria.
Bacteriol Rev 41, 100–180.
10 Cardon BP & Barker HA (1946) Two new amino-acid-
fermenting bacteria, Clostridiumpropionicum and Diplo-
coccus glycinophilus. J Bacteriol 52, 629–634.
11 Cardon BP & Barker HA (1947) Amino acid fermenta-
tions by Clostridiumpropionicum and Diplococcus glyci-
nophilus. Arch Biochem Biophys 12, 165–171.
12 Johns AT (1952) The mechanism of propionic acid for-
mation by Clostridium propionicum. J General Microbiol
6, 123–127.
13 Leaver FW, Wood HG & Stjernholm R (1955) The
fermentation of three carbon substrates by Clostridium
Beta-alanyl-CoA:ammonia lyases G. Herrmann et al.
820 FEBS Journal 272 (2005) 813–821 ª 2005 FEBS
propionicum and Propionibacterium. J Bacteriol 70,
521–530.
14 Schweiger G & Buckel W (1984) On the dehydration of
(R)-lactate in the fermentation of alanine to propionate
by Clostridium propionicum. FEBS Lett 171, 79–84.
15 Schweiger G & Buckel W (1985) Identification of acry-
late, the product of the dehydration of (R)-lactate cata-
lysed by cell-free extracts from Clostridium propionicum.
FEBS Lett 185, 253–256.
16 Selmer T, Willanzheimer A & Hetzel M (2002) Propio-
nate CoA-transferase from Clostridium propionicum.
Cloning of gene and identification of glutamate 324 at
the active site. Eur J Biochem 269, 372–380.
17 Hetzel M, Brock M, Selmer T, Pierik AJ, Golding BT
& Buckel W (2003) Acryloyl-CoA reductase from
Clostridium propionicum. An enzyme complex of propio-
nyl-CoA dehydrogenase and electron-transferring flavo-
protein. Eur J Biochem 270, 902–910.
18 Sinskey A, Akedo M & Cooney C (1981) Acrylate Fer-
mentations. Plenum Press, New York and London.
19 Vagelos PR, Earl JM & Stadtman ER (1959) Propionic
acid metabolism. I. The purification and properties of
acrylyl coenzyme A aminase. J Biol Chem 234, 490–497.
20 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.
21 Bru
¨
ggemann H, Ba
¨
umer S, Fricke WF, Wiezer A, Liese-
gang H, Decker I, Herzberg C, Martinez-Arias R,
Merkl R, Henne A & Gottschalk G (2003) The genome
sequence of Clostridium tetani, the causative agent of
tetanus disease. Proc Natl Acad Sci USA 100, 1316–
1321.
22 Bao Q, Tian Y, Li W, Xu Z, Xuan Z, Hu S, Dong W,
Yang J, Chen Y, Xue Y, Xu Y, Lai X, Huang L, Dong
X, Ma Y, Ling L, Tan H, Chen R & Wang J., Yu, J &
Yang H (2002) A complete sequence of the T. tengcon-
gensis genome. Genome Res 12, 689–700.
23 Nelson KE, Fleischmann RD, DeBoy RT, Paulsen IT,
Fouts DE, Eisen JA, Daugherty SC, Dodson RJ, Dur-
kin AS, Gwinn M, Haft DH, Kolonay JF, Nelson WC,
Mason T, Tallon L, Gray J, Granger D, Tettelin H,
Dong H, Galvin JL, Duncan MJ, Dewhirst FE &
Fraser CM (2003) Complete genome sequence of the
oral pathogenic bacterium Porphyromonas gingivalis
strain W83. J Bacteriol 185, 5591–5601.
24 Sidhu H, Allison M & Peck AB (1997) Identification
and classification of Oxalobacter formigenes strains by
using oligonucleotide probes and primers. J Clin Micro-
biol 35, 350–353.
25 Merck (1989) The Merck Index, an Encyclopedia of
Chemicals, Drugs and Biologicals, 11th edn. Merck, Inc,
Rahway, NJ, USA.
26 Buckel W & Miller SL (1987) Equilibrium constants of
several reactions involved in the fermentation of gluta-
mate. Eur J Biochem 164, 565–569.
27 Heider J (2001) A new family of CoA-transferases.
FEBS Lett 509, 345–349.
28 Jackins HC & Barker HA (1951) Fermentative processes
of the fusiform bacteria. J Bacteriol 61 , 101–114.
29 Barker HA, Kahn JM & Hedrick L (1982) Pathway of
lysine degradation in Fusobacterium nucleatum. J Bacter-
iol 152, 201–207.
30 Jeng IM & Barker HA (1974) Purification and proper-
ties of l -3-aminobutyryl coenzyme A deaminase from a
lysine-fermenting Clostridium. J Biol Chem 249, 6578–
6584.
31 Gokarn RR, Selifonova OV, Jessen HJ, Gort SJ,
Selmer T & Buckel W (2004) 3-Hydroxypropionic acid
and other organic compounds. US Patent Application
20040076982, Cargill Incorporated, Minneapolis, MN,
USA.
32 Kro
¨
ger A, Biel S, Simon J, Gross R, Unden G & Lan-
caster CR (2002) Fumarate respiration of Wolinella
succinogenes: enzymology, energetics and coupling
mechanism. Biochim Biophys Acta 1553, 23–38.
33 Seeliger S, Janssen PH & Schink B (2002) Energetics
and kinetics of lactate fermentation to acetate and pro-
pionate via methylmalonyl-CoA or acrylyl-CoA. FEMS
Microbiol Lett 211, 65–70.
34 Simon E & Shemin D (1953) The preparation of S-suc-
cinyl-coenzyme A. J Am Chem Soc 75, 2520.
35 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.
36 Schmidt B, Selmer T, Ingendoh A & von Figura K
(1995) A novel amino acid modification in sulfatases
that is defective in multiple sulfatase deficiency. Cell 82,
271–278.
37 Mayhew SG (1978) The redox potential of dithionite
and SO-2 from equilibrium reactions with flavodoxins,
methyl viologen and hydrogen plus hydrogenase. Eur J
Biochem 85, 535–547.
G. Herrmann et al. Beta-alanyl-CoA:ammonia lyases
FEBS Journal 272 (2005) 813–821 ª 2005 FEBS 821
. characteristic
intermediate [14,15]. As outlined in Fig. 1, 3 mol
l-alanine are oxidatively deaminated to pyruvate by the
combined action of alanine transaminase. hydroxyradical
scavenging peptides l-carnosine and carnicine [5] as
well as the antibiotics efrotomycin [6] and destruxin [7]
also contain b-alanine. The only