Báo cáo khoa học: Fragments of pro-peptide activate mature penicillin amidase of Alcaligenes faecalis pdf

Fragments of pro-peptide activate mature penicillin amidase
of
Alcaligenes faecalis
Volker Kasche, Boris Galunsky and Zoya Ignatova
Institute of Biotechnology II, Technical University Hamburg-Harburg, Hamburg, Germany
Penicillin amidase from Alcaligenes faecalis is a recently
identified N-terminal nucleophile hydrolase, which possesses
the highest specificity constant (k
cat
/K
m
)forthehydrolysis
of benzylpenicillin compared with penicillin amidases from
other sources. Similar to the Escherichia coli penicillin ami-
dase, the A. faecalis penicillin amidase is maturated in vivo
from an inactive precursor into the catalytically active
enzyme, containing one tightly bound Ca
2+
ion, via a
complex post-translational autocatalytic processing with a
multi-step excision of a small internal pro-peptide. The
function of the pro-region is so far unknown. In vitro addi-
tion of chemically synthesized fragments of the pro-peptide
to purified mature A. faecalis penicillin amidase increased its
specific activity up to 2.3-fold. Mutations were used to block
various steps in the proteolytic processing of the pro-peptide
to obtain stable mutants with covalently attached fragments
of the pro-region to their A-chains. These extensions of the
A-chainraisedtheactivityupto2.3-foldandincreasedthe
specificity constants for benzylpenicillin hydrolysis mainly
by an increase of the turnover number (k

cat
).
Keywords: Alcaligenes faecalis; pro-peptide; enzyme activa-
tion; penicillin amidase; site-directed mutagenesis.
Penicillin amidases (PA, EC 3.5.1.11) are biotechnologically
important enzymes used in the production of semisynthetic
b-lactam antibiotics. Penicillin amidases are present in a
variety of organisms including bacteria, yeast and fungi, and
they all diverge from a common evolutionary ancestor [1].
The physiological function of penicillin amidases in vivo is
not yet known. It has been speculated that they are involved
in the metabolism of aromatic compounds as carbon
sources [2], as the pac gene is localized in the proximity of
genes coding for enzymes involved in degradation of
4-hydroxyphenylacetic acid [3].
PA belongs to the structural superfamily, the Ntn
(N-terminal nucleophile) hydrolases, in which all members
are related in that the first event in the autocatalytic
processing of the inactive precursor reveals a catalytic serine,
threonine or cysteine at the N-terminal position [4]. The
processing of the inactive PA precursor to mature periplas-
mic enzyme has been studied in detail for the Escherichia
coli enzyme. The nascent pac gene encodes a prepro-PA
(97 kDa) containing an N-terminal signal peptide (pre-
sequence, 26 amino acids) that is cleaved upon crossing the
cytoplasmic membrane via the Tat pathway [5]. The crystal
structures of E. coli PA [6], of Providencia rettgeri PA [7],
as well as the mutant slow processing E. coli pro-PA [8]
provide insight into the catalytic mechanism and clarify the
role of the N-terminal serine of the B-chain as a single

catalytic residue. The inactive pro-PA (92 kDa) is activated
by multiple proteolytic cleavages starting with an intra-
molecular autocatalytic step between Thr263 and Ser264,
which generates the B-chain (62 kDa) [4,6,8,9]. The pro-
peptide (known also as linker or spacer peptide, 54 amino
acids) is further sequentially removed from the C-terminus
of the A-chain in intra- and intermolecular processing
events, resulting in a release of the A-chain (23 kDa) [8],
found as a dominating form in the commercial PA
preparations. While the presequence mediates translocation
through the membrane, the function of the pro-region is still
unknown. Even though such an exclusion mechanism of
short peptides from inactive precursors in the maturation
process is a widely spread in living systems, the exact role of
the pro-domain is not completely understood. For some
proteases such as subtilisin [10], nerve growth factor [11] and
a-lytic protease [12], the pro-region is required for correct
folding in vivo or refolding in vitro. Furthermore, the pro-
domain accelerates the structure formation by facilitating
formation of correct disulfide bonds [11]. Partial or whole
deletions in the pro-sequence affect maturation and correct
processing of nerve growth factor [13].
While Alcaligenes faecalis PA shares the lowest
sequence homology to E. coli PA in the penicillin amidase
family from the Gram-negative bacteria, the precursor
organization resembles that of the E. coli PA, starting
with an N-terminal presequence (26 amino acids), fol-
lowed by the A-chain (202 amino acids), pro-region (37
amino acids), and B-chain (551 amino acids) [14]. As both
enzymes possess the same substrate specificity and share

extensive similarities in functionally important amino acid
residues, it is expected that their molecular mechanisms of
processing are similar, e.g. the pro-peptide is step-wise
proteolytically removed in the maturation process yielding
Correspondence to V. Kasche, Institute of Biotechnology II, Technical
University Hamburg-Harburg, Denickestr. 15, 21073 Hamburg,
Germany. Fax: + 49 40 42787 2127, Tel.: + 49 40 42878 3018,
E-mail:
Abbreviations: IEF, isoelectric focusing; NIPAB, 6-nitro-3-phenyl-
acetamido benzoic acid; Ntn, N-terminal nucleophile;
PA, penicillin amidase.
Enzyme: penicillin amidase (EC 3.5.1.11).
(Received 9 July 2003, revised 4 September 2003,
accepted 6 October 2003)
Eur. J. Biochem. 270, 4721–4728 (2003) Ó FEBS 2003 doi:10.1046/j.1432-1033.2003.03871.x
the mature two-chain enzyme [14,15]. Recently, we gave
experimental evidence for the step-wise shortening of the
pro-peptide of A. faecalis PA by the isolation of the last
two active forms with different length of the A-chains [15].
Comparative studies of the E. coli and A. faecalis PA
showed that the specific activity of A. faecalis enzyme in
the cell homogenate is about fivefold higher. After
purification to homogeneity only twofold higher specific
activity of A. faecalis PA compared to E. coli PA was
measured [15]. The difference in the specific activity of the
A. faecalis PA in the homogenate and as a purified
protein indicates that an activating compound is lost
during the purification of this enzyme. This is verified in
this study on wild-type A. faecalis PA, where we demon-
strate that fragments from the pro-peptide act as activa-

tors in vitro. Furthermore, our results show that inhibiting
the later steps of the pro-peptide removal in vivo by
introduction of specific point mutations in the pro-domain
increased the specific activity of the mutant enzymes with
extended A-chains. The observed higher specificity con-
stants of the mutants for benzylpenicillin hydrolysis are
mainly due to an increase in the turnover number (k
cat
).
Experimental procedures
Bacterial strains, plasmid construction
and growth conditions
Plasmid pPAAF for the in vivo synthesis of A. faecalis
prepro-PA was constructed as follows. A 2360 bp PCR
fragment covering the region from 13 nucleotides upstream
from the start codon of A. faecalis pac with the altered RBS
and Shine–Dalgarno sequence was amplified using the
following primers
5¢-CGAATTCTGAGGAGGTAGTAATGCAGAAAGG
GCT-3¢ and
5¢-CCTCCAAGCTTAAGGCAGAGGCTG-3¢ (ARK-
Scientific GmbH, Germany)
with chromosomal DNA from A. faecalis ATCC 1908 as a
template. The product was double digested with EcoRI and
HindIII and cloned into the multiple cloning site of
pMMB207 [16] yielding a pPAAF plasmid. The last was
used as a template for the introduction of site-specific
mutations (Table 1) into the pro-peptide coding sequence
using QuickChange Mutagenesis Kit (Stratagene, the
Netherlands). All mutations were verified by DNA sequen-

cing (SeqLab, Germany).
The A. faecalis pac gene was expressed under the tac-
promoter and therefore induced by 0.5 m
M
isopropyl
thio-b-
D
-galactoside. During all genetic manipulations the
host cells E. coli DH5a were grown aerobically in Luria–
Bertani medium supplemented with 25 lgÆmL
)1
chloram-
phenicol as a selection marker [17]. Transformed E. coli
DH5a cells were plated on LB agar medium with a nitro-
cellulose filter. Positive clones harboring the A. faecalis pac
gene were screened phenotypically for PA-activity with the
chromogenic substrate 6-nitro-3-phenylacetamido benzoic
acid (NIPAB) [18].
Purification of wild-type
A. faecalis
penicillin amidase
and its pro-peptide mutants
For expression E. coli BL21(DE3) cells were transformed
with either pPAAF or plasmids carrying mutations in the
pro-peptide and were cultivated at 28 °C in minimal M9
medium, containing 2.5 gÆL
)1
glucose. Six hours after
induction with isopropyl thio-b-
D

-galactoside (0.5 m
M
)the
cells were harvested by centrifugation at 1700 g for 15 min.
Furthermore, they were fractionated into periplasmic and
cytoplasmic fractions by cold mild osmotic shock procedure
as described previously [19].
The wild-type A. faecalis PA and the pro-peptide
mutants were purified from the concentrated supernatant
by anion-exchange chromatography using the same proce-
dure as described in [15,20]. All eluted protein fractions were
desalted into 30 m
M
Tris buffer, pH 7.5, and concentrated
using Amicon centrifugal filters (cut-off 10 kDa). The
homogeneity of the enzyme forms was analyzed by isoelec-
tric focusing (IEF) and SDS/PAGE [21]. In the IEF
experiments ready for use ServalytÒPrecotesÒ 3–10 gels
with supplied buffer systems (Serva, Germany) were run
according to the instructions of the manufacturer on a
Multiphor II (LKB Bromma, Sweden) apparatus.
Assay for penicillin amidase activity and active site
titration
The PA activity was measured by a spectrophotometric
assay with the chromogenic substrate NIPAB [20]. Under
standard conditions (pH 7.5, 25 °C, 125 l
M
NIPAB), the
specific activity is defined as a change in the absorbance
at 380 nmÆmin

)1
, per protein content expressed as an
absorbance at 280 nm (DA
380
min
)1
ÆA
280
)1
). Pure E. coli
PA with a concentration 1 mgÆmL
)1
possesses an A
280
value
of 2.0 [20]. The formula for recalculation of the acti-
vity measured with the same substrate at 405 nm is:
DA
405
min
)1
¼ 0.94 · DA
380
min
)1
.
The molar concentrations of the enzymes were determined
by active site titration [22]. Equal amounts of wild-type
Table 1. Amino acid substitutions in the pro-peptide generated by site-directed mutagenesis. The amino acids introduced by mutagenesis are shown in
bold. The sequences of the mutated pro-peptides start from the N-terminus.

Short assignment of the
mutated pro-peptides Amino acid sequences of the pro-peptides
Wild-type pro-sequence QAGTQDLAHVSSPVLATELERQDKHWGGRGPDFAPKA
T206P QAGPQDLAHVSSPVLATELERQDKHWGGRGPDFAPKA
T206G QAGGQDLAHVSSPVLATELERQDKHWGGRGPDFAPKA
T206GS213G QAGGQDLAHVGSPVLATELERQDKHWGGRGPDFAPKA
T206GS213GT219G QAGGQDLAHVGSPVLAGELERQDKHWGGRGPDFAPKA
4722 V. Kasche et al. (Eur. J. Biochem. 270) Ó FEBS 2003
A. faecalis PA or enzymes with point mutations in the pro-
peptide were incubated with different amounts of phenyl-
methanesulfonyl fluoride in phosphate buffer pH 7.5,
I ¼ 0.2
M
for 30 min. The residual activity was measured
spectrophotometrically using NIPAB as a substrate.
Determination of the kinetic parameters
The PA-catalyzed hydrolysis of benzylpenicillin was per-
formed at 25 °C and pH 7.5 (phosphate buffer I ¼ 0.2
M
).
The used substrate concentrations were 5, 10, 20, 40, 60 and
80 l
M
. Enzyme concentration in the reaction mixture were
between 3.2 · 10
)11
M
and 10 · 10
)11
M

. Periodically aliqu-
ots were withdrawn and immediately analyzed by HPLC as
described previously [23]. The initial rates (about 10%
substrate exhausting) were determined on the basis of the
increase of phenylacetic acid concentration as a function of
time. Five to six points were measured. The initial rates were
calculated by linear regression analysis using
PLOTIT
software, version 3.14 (Scientific Programming Interfaces,
1994). The initial rates at each substrate concentration were
average values of three independent experiments. The values
of the steady-state kinetic parameters K
m
and k
cat
for
A. faecalis PA and pro-peptide mutants were calculated
using reversed Eadie–Hofstee plots.
Determination of the bound calcium ion
The calcium ion content in the purified A. faecalis forms
(protein concentration 1 mgÆmL
)1
) was measured by
Induced Coupled Plasma-Atom Emission Spectroscopy
(ICP-AE-spectrophotometer, Perkin-Elmer). In order to
rule out any unspecific bound calcium ions, the purification
was performed with calcium-free buffers and additionally
before the measurement the purified proteins were trans-
ferred into double distilled water with Bio-Rad HR 10/10
desalting column. Calcium ion content of the blank (double

distilled water treated on the same way as the sample) was
zero.
In vitro
influence of the pro-peptide and fragments
of it on the activity of purified
A. faecalis
penicillin amidase
The activation of A. faecalis PA in vitro was tested with
chemically synthesized fragments of the pro-peptide (11-
mer, 20-mer, 29-mer and the whole pro-peptide 37-mer;
ARK-Scientific GmbH, Germany). The sequences of all
oligopeptides were derived from the pro-peptide as
presented in Table 2. Purified A. faecalis PA with an
isoelectric point (pI) of 5.3 (15 n
M
) was incubated for
15 min at 25 °C in phosphate buffer pH 7.5 I ¼ 0.2
M
with
the above oligopeptides in the concentration range 0–75 n
M
.
Then the mixture was subjected to activity measurements
using NIPAB as a substrate.
Results and discussion
Sequence alignment and comparison with
E. coli
penicillin amidase
The A. faecalis PA shows 40% protein sequence identity
with the E. coli PA (Fig. 1). Taking conservative substitu-

tions into account, the homology rises above 48%. The key
catalytic and oxyanion hole forming residues [24] (Ser264,
Gln286, Ala332, Asn504, Asn505, Arg526; numbering is
according to the amino acid sequence of E. coli pro-PA [25])
are strictly conserved in the A. faecalis PA (Fig. 1) and in
the other members of the PA family [14,26]. Another
interesting aspect of this comparison is that the most of the
conserved clusters, e.g. residues 133–148, 284–316, 440–446,
490–507, and 739–751, are in the vicinity of the active site.
While the enzymes of the PA family do not require a
calcium ion as a cofactor, the crystal structures of E. coli
PA (PDB access number 1PNK), of the slow processing
Gly263Thr mutant E. coli pro-PA (PDB access number
1E3A), and of the P. rettgeri mutant Bro1 PA [7] reveal a
tight bound calcium ion in the structure. ICP-AES analysis
confirmed the presence of one calcium ion in the A. faecalis
PA molecule. Five of the six calcium co-ordinating residues
identifiedintheE. coli PA (Glu152, Asp336, Val338,
Asp339, and Asp515) are fully conserved in the A. faecalis
PA (Fig. 1). These residues are also conserved among the
other PA members of the Enterobacteriaceae Kluyvera
cytrophila and P. rettgeri (see the alignment published by
Verhaert et al. [14]).
The largest divergence exists in the pro-peptide removed
during maturation. The crystal structure of mature E. coli
PA reveals that both chains form a pyramid with the active
site serine located at the base of a deep cone [6]. In the E. coli
pro-PA the active site cleft is covered by the pro-peptide [8],
localized on the surface of the pro-enzyme molecule and
flanking the superficial C-terminal part of the A-chain and

the deep concealed N-terminus of the B-chain. The pro-
region of A. faecalis pro-PA is 17 amino acids shorter than
the E. coli pro-PA. This deletion is localized in the first
superficial part of the pro-peptide, in the loop before the
a-helix structure. Loops as flexible structural elements easily
tolerate deletions or insertion of extra residues without
perturbation of the entire structure [27].
Until now, no direct evidence exists about all amino acids
participating in the autocatalytic maturation process.
Table 2. Amino acid sequences of the synthetic oligopeptides. The sequences of the synthetic oligopeptides correspond to the (fragment) sequence in
the wild-type A. faecalis pro-peptide starting from the N-terminus.
Length of the oligopeptide Amino acid sequence in a single letter code
11-mer QAGTQDLAHVS
20-mer QAGTQDLAHVSSPVLATELE
29-mer QAGTQDLAHVSSPVLATELERQDKHWGGR
37-mer QAGTQDLAHVSSPVLATELERQDKHWGGRGPDFAPKA
Ó FEBS 2003 Penicillin amidase activation by pro-peptide fragments (Eur. J. Biochem. 270) 4723
Possible candidates, such as the N-terminal SN sequence
(Ser264, Asn265) of the B-chain, and Gly284 [28], are fully
conserved in both penicillin amidases (Fig. 1). Lys273,
described as a residue responsible for a pH-dependent
processing [29], is conservatively substituted in the A. fae-
calis sequence with an arginine which provides the necessary
sidechainwithabasicpK
a
. The catalytically active serine
at the N-terminus of the B-chain being totally conserved
(Fig. 1) reveals the necessary requirement for an efficient
self-processing prerequisite for the PA activity [4]. These
sequence considerations support the assumption for similar

processing mechanism of both A. faecalis and E. coli PA.
Moreover, our previous study with A. faecalis PA [15]
supports with experimental evidence the assumption for a
sequential removal of the pro-peptide from its C-terminus,
similar to the maturation of E. coli PA [9].
In vitro
influence of the pro-peptide and its fragments
on the activity of
A. faecalis
PA
The stable processed form of A. faecalis PA, expressed in
E. coli, was produced and purified as already described [15].
Typically, the purified final mature form of the enzyme with
a completely removed pro-peptide appeared homogeneous
with respect to IEF and SDS/PAGE analysis with an
isoelectric point (pI) of 5.3. The total activity, used to
evaluate the purification yield showed a 57% loss after the
first purification step, the concentration of the periplasmic
fraction by ultrafiltration (molecular size cut-off 10 kDa)
[15]. An addition of this filtrate to purified A. faecalis PA led
to more than twofold increase of specific activity and 86%
of the total activity was restored (data not shown). The pro-
peptide (37 amino acids) is sequentially shortened during the
maturation process and the resulting fragments, acting
obviously as activators, are probably removed from the
active enzyme in this step, remaining in the ultrafiltrate. This
prompted us to investigate the possible influence of the
whole pro-region or fragments of it with random lengths on
the activity of the A. faecalis PA. The incubation of the
chemically synthesized oligopeptides with purified A. fae-

calis PA (pI 5.3) at different molecular ratios led to an
activation of PA and the specific enzyme activity increased
up to 2.3-fold (Fig. 2). The highest activation was measured
for the shortest oligopeptide (11mer) with an activation
effect being concentration dependent. Increasing the
amount of the 29-mer over the stoichiometric ratio had
hardly any significant effect. In the case of 11-mer oligopep-
tide the activity raised up to a ratio 1 : 2 (PA/11-mer)
Fig. 1. Amino acid sequence alignment of E. coli PA and A . faecalis PA. Identical residues are shadowed, similar substitutions are framed.
Numbering is according to the amino acid sequence of the E. coli pro-PA [25] starting with the first amino acid of the A-chain. The signal peptide
cleaved off after translocation is numbered in the opposite direction. d, catalytic residues and residues from the oxyanion hole; h,calciumion
coordinating residues. The residues of the pro-peptides in both sequences are underlined.
4724 V. Kasche et al. (Eur. J. Biochem. 270) Ó FEBS 2003
(Fig. 2), although over the physiological ratio 1 : 1 the
activity increased only with additional 20%. The non-
covalent interactions between the shortest oligopeptide and
PA seem to be dynamic and reverse, therefore concentra-
tions over the stoichiometric ratio increase the number of
oligopeptide bound to the enzyme resulting in a higher
enzymatic activity.
In the experiments with the oligopeptide with a length of
the whole pro-peptide (37-mer), the PA activity measure-
ments were problematic. During the first 10 s after mixing
with the substrate NIPAB an increase of the absorbance at
380 nm was detected, followed by a phase where practically
no absorbance change was observed, even when the substrate
was not exhausted (data not shown). Most probably, during
the preincubation of purified A. faecalis PA(pIof5.3)with
the 37-mer oligopeptide (representing the whole pro-pep-
tide), it fits once again into the entrance of the cone and

covers the active site, which results in restricted diffusion of
the substrate molecules to the catalytic serine.
Effects of the inhibition of the complete proteolytic
processing of the pro-peptide on the penicillin
amidase activity
In a previous study we succeeded in isolating the last two
active forms of A. faecalis PA. ICP-AES analysis confirmed
that both forms contained one tightly bound calcium ion
leading to the conclusion that calcium ion binding precedes
the processing of pro-PA. By mass-spectrometry analysis we
showed that the observed higher molecular mass of the
A-chain of the form with pI 5.5 compared to the A-chain of
the last maturation form with pI 5.3, is due to the four
amino acids from the pro-peptide still remaining covalently
attached to the A-chain [15]. Therefore, the first position
mutated was Thr206 and we exchanged it with Pro and Gly
(Table 1). The resulting mutant PA-precursors were con-
cisely assigned by the single letter code of the substituted
amino acid and its position, followed by the code of the
replacing amino acid. The numbering is according to the
published primary structure of A. faecalis pro-PA [14],
starting with the N-terminal amino acid of the A-chain. The
measured specific activity of the T206P mutant was lower,
being about 85% of the specific activity of the wild-type
completely processed A. faecalis PA (pI 5.3) (Table 3). The
T206P mutant appeared to undergo further normal pro-
teolytic processing leading to a completely processed PA
form with pI 5.3 (Fig. 3A, lane 3).
Table 3. Specific activity of the wild-type A. faecalis PA (pI 5.3) and the
site-directed mutants. Activity was measured with purified proteins.

Each specific activity value is an average of three measurements. The
k
cat
values for NIPAB hydrolysis were estimated from the specific
activity and the active site titration data and were calculated to be:
wild-type A. faecalis PA 82 s
)1
(see also [15]), T206G mutant 131 s
)1
,
T206GS213G mutant 152 s
)1
, T206GS213GT219G 185 s
)1
.
A. faecalis PA forms
Specific activity
DA
380
min
)1
ÆA
280
)1
Wild-type (pI 5.3) 2.0 ± 0.1
T206G mutant 3.2 ± 0.2
T206P mutant 1.7 ± 0.1
T206GS213G mutant 3.7 ± 0.2
T206GS213GT219G mutant 4.5 ± 0.3
Fig. 3. Processing patterns of purified mutant A. faecalis PA precursors

with alterations at positions 206, 213 and 219. (A) IEF stained with
Coomassie blue, Lanes: M, isoelectic point marker; 1, purified last two
maturation forms of the wild-type A. faecalis PA with pI 5.3 and 5.5; 2,
T206G mutant; 3, T206P mutant. (B) SDS/PAGE stained with Coo-
massie blue. Lanes: 1, purified last maturation form of the wild-type
A. faecalis PA (pI 5.3); 2, T206GS213GT219G mutant; 3,
T206GS213G mutant; 4, A. faecalis PA (pI 5.5); 5, T206G mutant.
Fig. 2. In vitro influence of fragments of the pro-peptide on the A. fae-
calis PA (pI 5.3) activity. The activity measurements were performed as
described in Materials and methods with 15 n
M
enzyme and oligo-
peptides in the concentration range 0–75 n
M
. The starting point is the
activity of A. faecalis PA (pI 5.3) without oligopeptides, which was
taken as 1. d, 11-mer; m,20-mer;s, 29-mer.
Ó FEBS 2003 Penicillin amidase activation by pro-peptide fragments (Eur. J. Biochem. 270) 4725
Our previous mutational experiments showed that the
replacement of the original Thr in the pro-sequence of
E. coli pro-PA by Gly retards the rate of its processing
which allowed isolation of the precursor [8,9]. Furthermore,
the Thr206 was also mutated to Gly (T206G), which led to
the predominating active form of PA with pI 5.5 (Fig. 3A,
lane 2). The specific activity of T206G mutant was 60%
higher compared to the wild-type A. faecalis PA with pI of
5.3 (Table 3). SDS/PAGE analysis under denaturing con-
ditions gave a double band for the A-chain (Fig. 3B, lane 5).
The lower band corresponds to the size of the completely
processed A-chain of A. faecalis PA with pI 5.3 and the

upper one (marked as A + P) is of the approximate size of
the A-chain plus fragment of the pro-peptide. The further
cleavage of the remaining four amino acids from the pro-
peptide at 25 °C and pH 7.5 was a relatively slow process
and even after 312 h incubation approximately 30% was
not converted into the form with pI of 5.3 (Fig. 4A).
The question arose, whether the extended length of the
A-chain by four amino acids affects the catalytic or the
binding properties of the enzyme. The steady-state kinetic
parameters K
m
and k
cat
for benzylpenicillin hydrolysis are
summarized in Fig. 5. Whereas the K
m
values for benzyl-
penicillin hydrolysis by A. faecalis PA (pI 5.3) and T206G
mutant were equal, the k
cat
value for T206G mutant was
about 1.5-fold higher (Fig. 5). The similarity in the K
m
values was not surprising, as the remaining four amino acids
from the pro-peptide cannot cover the entrance to the active
site and therefore do not influence the substrate binding
properties of the enzyme. Fragments of the pro-peptide still
remaining covalently attached to the mature A. faecalis PA
can probably influence the stabilization of the transition
state of the rate limiting step (formation of the acyl-enzyme

intermediate) thus leading to higher k
cat
values. A similar
effect was observed for cephalosporin acylase from Pseu-
domonas sp. 130 [30].
Although the replacement of T206 by Gly led to a
retarded processing of the mutant precursor, the further
removal of the pro-peptide could not be blocked completely.
All purified samples of T206G contained traces of the
completely processed PA with pI 5.3 (Fig. 3), therefore
additional site-specific amino acid substitutions were intro-
duced into the pro-peptide coding region of A. faecalis
PA (Table 1). In the in vitro experiments with chemically
synthesized oligopeptides the highest activation was
observed with the 11-mer peptide (Fig. 2), thus the position
of Ser213 was chosen for the next replacement.
The processing of E. coli pro-PA starts with an intra-
molecular autoproteolytic cleavage between Thr263 and
Ser264 yielding the free N-terminal serine of the B-chain [6].
Detailed mapping of some of the further shortening of the
pro-region revealed Asn241-Arg242 and Asp223-Arg224 to
be the next cleavages in the maturation process [9]. The
Asn241-Arg242 bond is within the a-helical region (resi-
dues 240–251 [8]). The a-helix propensity analysis of the
Fig. 4. Stability of purified mutant A. faecalis PA precursors monitored
by IEF. (A) Purified T206G mutant dissolved in 1 m
M
Tris/HCl
pH 7.5 was incubated at 25 °C for 24 h (lane 2), 48 h (lane 3) and
312 h (lane 4), Lanes: M, isoelectric point marker; 1, purified last two

maturation forms of the wild-type A. faecalis PA(pI5.3andpI5.5).
(B) Purified T206GS213G and T206GS213GT219G mutants were
incubatedin1m
M
Tris/HCl pH 7.5 at 25 °Cfor0h(lanes1and4)
and 192 h (lanes 2 and 5). Purified last two maturation forms of the
wild-type A. faecalis PA (pI 5.3 and pI 5.5) served as references
(lane 3).
Fig. 5. Reversed Eadie–Hofstee plots for the hydrolysis of benzylpeni-
cillin catalyzed by A. faecalis PA (pI 5.3) and A. faecalis PA mutants.
Phosphate buffer pH 7.5, I ¼ 0.2
M
,25°C; substrate concentrations in
the range 5 · 10
-6
to 80 · 10
)6
M
; enzyme concentrations in the range
3.2 · 10
)11
to 10 · 10
)11
M
. The initial rates used to determine the
steady-state kinetic parameters were average values of three inde-
pendent experiments at each concentration. The standard deviations
are given by error bars.
4726 V. Kasche et al. (Eur. J. Biochem. 270) Ó FEBS 2003
pro-sequence of A. faecalis PA revealed that residues

Val216 to Lys226 are likely to adopt an a-helical confor-
mation. Assuming a similar processing pathway as for
E. coli PA (based on sequence homology, Fig. 1), the third
residue for mutation, Thr219, was chosen to be a residue
within the a-helix proportionally at the same position of the
Asn241 in the a-helix of pro-peptide of E. coli PA.
The processing phenotypes of all altered A. faecalis PA
pro-peptide mutant precursors were analyzed by SDS/
PAGE (Fig. 3B). Introduction of an additional mutation
at position 213 (T206GS213G) stabilized the precursor
and in the processing patterns only PA-forms with longer
A-chain (A + P*) corresponding to the 13 amino acids
extension were detected (Fig. 3B, lane 3). Thus, the
purified mutant appeared as a single stable band on the
IEF-gels with a pI of 5.6 and was not further converted
even after incubation at room temperature for 192 h
(Fig. 4B). A third mutation in the pro-peptide at position
219 (T206GS213GT219G) showed quite diverse effects.
The SDS-processing pattern of this mutant revealed an
appearance of an unstable intermediate with a larger
A-chain (A + P* form, Fig. 3B, lane2), which after 72 h
is further converted to the A + P form (data not shown).
This suggests that the introduced mutation at position 219
causes only retardation, and not complete blockage of this
cleavage. Meanwhile, mutagenized Thr219 also seems to
destabilize the peptide chain at the other exchanged
(T206G and S213G) positions and a band corresponding
to the completely processed A. faecalis PA (pI 5.3) was
detected on the IEF gels, even immediately after purifi-
cation (Fig. 4B, lanes 4, 5). Nevertheless, both pro-peptide

mutants (T206GS213G and T206GS213GT219G) exhi-
bited increased specific activities (1.9- and 2.3-fold,
respectively) compared with the completely processed
A. faecalis PA with pI 5.3 (Table 3). These results are in
good agreement with the observed in vitro activation of
A. faecalis PA (pI 5.3) by fragments of the pro-peptide
with a corresponding length (11-mer and 20-mer) (Fig. 2).
The k
cat
value for benzylpenicillin hydrolysis catalyzed
by the T206GS213G mutant was higher than the value
for the T206G mutant (Fig. 5). The introduced third
mutation in the pro-peptide of A. faecalis PA in the
T206GS213GT219G mutant resulted in a 2.9-fold increase
of the specificity constant compared with A. faecalis PA,
mainly due to the higher turnover number (Fig. 5).
Pro-domains of many zymogenes have been shown to
accelerate 3D-structure formation [31] or to influence the
folding as an intramolecular chaperone [32,33]. The mech-
anism by which fragments of a pro-peptide function as
activating factors is presently unknown. Based on the results
presented in this study, we assume that fragments of the
pro-peptide of A. faecalis PA activate the enzyme by
stabilizing the transition state of acyl-enzyme formation
resulting in enhanced catalytic constants for all of the
mutants with extended A-chains. Even though the observed
activation of A. faecalis PA in cell homogenate has been
explained by the results so far obtained, many questions
remain to be answered: What is the biological significance in
generating enzymes for which activity decreases in the

maturation process? What is the molecular mechanism by
which fragments of the pro-peptide exactly influence the
catalytic constant of the enzyme?
Acknowledgements
We thank Dr Frank Meyberg, Institut fu
¨
r Anorganische und
Angewandte Chemie, Universita
¨
t Hamburg, for performing the ICP-
AES analyses.
References
1. Alkema, W.B.L., De Vries, E.J., Hensgens, C.M.H., Polderman-
Tijmes, J.J., Dijkstra, B.W. & Jansen, D.B. (2002) Engineering
enzymes for the synthesis of semi-synthetic antibiotics. In Synth-
esis of B-Lactam Antibiotics (Bruggink, A., ed.), pp. 250–279.
Kluwer Academic Publishers, Dordrecht, the Netherlands.
2. Valle, F., Balbas, P., Merino, E. & Bolivar, F. (1991) The role of
the penicillin amidases in nature and industry. Trends Biochem.
Sci. 16, 36–40.
3. Prieto, M.A., Diaz, E. & Garcia, J.L. (1996) Molecular char-
acterization of the 4-hydroxyphenylacetate catabolic pathway of
Escherichia coli W: engineering a mobile aromatic degradative
cluster. J. Bacteriol. 178, 111–120.
4. Brannigan, J.A., Dodson, G., Duggelby, H.J., Moody, P.C.,
Smith, J.L., Tomchick, D.R. & Murzin, A.G. (1995) A protein
catalytic framework with an N-terminal nucleophile is capable of
self-activation. Nature 378, 416–419.
5. Ignatova, Z., Ho
¨

rnle, C., Nurk, A. & Kasche, V. (2002) Unusual
signal peptide directs penicillin amidase from E. coli to the Tat
translocation machinery. Biochem. Biophys. Res. Commun. 291,
146–149.
6. Duggleby, H.J., Tolley, S.P., Hill, C.P., Dodson, E.J., Dodson, G.
& Moody, P.C. (1995) Penicillin acylase has a single-amino-acid
catalytic centre. Nature 373, 264–268.
7. McDonough, M.A., Klei, H.E. & Kelly, J.A. (1999) Crystal
structure of penicillin G acylase from the Brol mutant strain of
Providencia rettgeri. Prot. Sci. 8, 1971–1981.
8. Hewitt, L., Kasche, V., Lummer, K., Lewis, R.J., Murshudov,
G.N., Verma, C.S., Dodson, G.G. & Wilson, K.S. (2000) Struc-
ture of a slow processing precursor penicillin acylase from
Escherichia coli reveals the linker peptide blocking the active-cleft.
J. Mol. Biol. 302, 887–898.
9. Kasche, V., Lummer, K., Nurk, A., Piotraschke, E., Rieks, A.,
Stoeva, S. & Voelter, W. (1999) Intramolecular autoproteolysis
initiates the maturation of penicillin amidase from Escherichia coli.
Biochim. Biophys. Acta 1433, 76–86.
10. Li, Y., Hu, Z., Jordan, F. & Inouye, M. (1995) Functional analysis
of the propeptide of subtilisin E as an intramolecular chaperone
for protein folding: refolding and inhibitory abilities of propeptide
mutants. J. Biol. Chem. 270, 25127–25132.
11. Rattenholl, A., Lilie, H., Grossmann, A., Stern. A., Schwarz, E. &
Rudolph, R. (2001) The pro-sequence facilitates folding of human
nerve growth factor from Escherichia coli inclusion bodies. Eur. J.
Biochem. 268, 3296–3303.
12. Silen, J.L., Frank, D., Fujishige, A., Bone, R. & Agard, D.A.
(1989) Analysis of the prepro-alpha-lytic protease expression in
Escherichia coli reveals that the pro region is required for activity.

J. Bacteriol. 171, 1320–1325.
13. Suter,U.,Heymach,J.V.Jr&Shooter,E.M.(1991)Twocon-
served domains in the NGF propeptide are necessary and suffi-
cient for the biosynthesis of correctly processed and biologically
active NGF. EMBO J. 10, 2395–2400.
14. Verhaert, R.M., Riemens, A.M., van der Laan, J M., van Duin, J.
& Quax, W.J. (1997) Molecular cloning and analysis of the gene
encoding the thermostable penicillin G acylase from Alcaligenes
faecalis. Appl. Environ. Microbiol. 63, 3412–3418.
15. Ignatova, Z., Stoeva, S., Galunsky, B., Ho
¨
rnle, C., Nurk, A.,
Piotraschke,E.,Voelter,W.&Kasche,V.(1998)Proteolytic
processing of penicillin amidase from Alcaligenes faecalis cloned
Ó FEBS 2003 Penicillin amidase activation by pro-peptide fragments (Eur. J. Biochem. 270) 4727
in E. coli yields several active forms. Biotechnol. Lett. 20, 977–
982.
16. Morales, V.M., Ba
¨
ckmann, A. & Bagdasarian, M. (1991) A series
of wide-host-range low-copy-number-vector that allow direct
screening for recombinants. Gene 97, 39–47.
17. Maniatis, T., Fritsch, E.F. & Sambrook, J. (1992) Molecular
cloning: A Laboratory Manual,(Nolan,C.,ed.).ColdSpring
Harbor Laboratory Press, New York, USA.
18. Zhang, Q.J., Zhang, L.F., Hau, H.X. & Zhang, Y.Y. (1986) A
method for screening penicillin G acylase-producing bacteria by
means of 2-nitro-5-phenylacetamidobenzoic acid test paper. Anal.
Biochem. 156, 413–416.
19. Ignatova, Z., Enfors, S., Hobbie, M., Taruttis, S., Vogt, C. &

Kasche, V. (2000) The relative importance of intracellular pro-
teolysis and transport on the yield of the periplasmic enzyme
penicillin amidase in Escherichia coli. Enzyme Microb. Technol. 26,
165–170.
20. Kasche,V.,Haufler,U.,Markowsky,D.,Melnyk,S.,Zeich,A.&
Galunsky, B. (1987) Penicillin amidase from E. coli:enzymehet-
erogeneity and stability. Ann. N.Y. Acad. Sci. 501, 97–102.
21. Laemmli, U.K. (1970) Cleavage of structural proteins during the
assembly of the head of bacteriophage T4. Nature 227, 680–685.
22. Svedas, V.K., Margolin, A.L., Sherestyuk, C.F., Klyosov, A.A. &
Berezin, I.V. (1977) Phenylmethansulfonylfloride inactivation
of soluble and immobilized pernicillin amidase from E. coli.
Biotechnol. Lett. 7, 877–882.
23. Galunsky,B.,Lummer,K.&Kasche,V.(2000)Comparative
study of the substrate- and stereospecificity of penicillin G ami-
dases from different sources and hybrid isoenzymes. Chem.
Monthly 131, 623–632.
25. Schumacher, G., Sizmann, D., Hang, H., Buckel, P. & Bo
¨
ck, A.
(1986) Penicillin acylase from E. coli: unique gene-protein relation.
Nucleic Acids Res. 14, 5713–5727.
24. McVey, C.E., Walsh, M.A., Dodson, G.G., Wilson, K.S. &
Brannigan, J.A. (2001) Crystal strucures of penicillin acylase
enzyme–substrate complexes: structural insights into the catalytic
mechanism. J. Mol. Biol. 313, 139–150.
26.Piotraschke,E.,Nurk,A.,Galunsky,B.&Kasche,V.(1994)
Genetic construction of catalytically active cross-species hetero-
dimer penicillin G amidase. Biotechnol. Lett. 16, 119–124.
27. Sibanda, B.L. & Thornton, J.M. (1993) Accomodating sequence

changesinbeta-hairpininproteins.J. Mol. Biol. 229, 428–447.
28. Prieto,I.,Rodriguez,M.C.,Marquez,G.,Perez-Aranda,A.&
Barbero, J.L. (1992) Changing glycine 21 for glutamic acid in the
beta-subunit of penicillin G acylase from Kluyvera cytrophila
prevents protein maturation. Appl. Microbiol. Biotechnol. 36,
659–662.
29. Lee, H., Park, O.K. & Kang, H.S. (2000) Identification of a new
active site for autocatalytic processing of penicillin acylase pre-
cursor in Escherichia coli ATCC11105. Biochem. Biophys. Res.
Commun. 272, 199–204.
30. Li,Y.,Chen,J.,Jiang,W.,Mao,X.,Zhao,G.&Wang,E.(1999)
In vivo post-translational processing and subunit reconstitution of
cephalosporin acylase from Pseudomonas sp. 130. Eur. J. Biochem.
262, 713–719.
31. Rattenholl, A., Ruopollo, M., Flagellio, A., Monti, M., Vinci, F.,
Marino, G., Lilie, H., Schwarz, E. & Rudolph, R. (2001) Prose-
quence assisted folding and disulfide bond formation of human
nerve growth factor. J. Mol. Biol. 305, 523–533.
32. Beer, H D., Wohlfahrt, G., Schmid, R.D. & McCarthy, J.E.
(1996) The folding and activity of the extracellular lipase of Rhy-
zopus oryzae are modulated by a prosequence. Biochem. J. 319,
351–359.
33. Baker, D., Shiau, A.K. & Agard, D.A. (1993) The role of the pro
regions in the protein folding. Curr. Opin. Cell Biol. 5, 966–970.
4728 V. Kasche et al. (Eur. J. Biochem. 270) Ó FEBS 2003