ShewasinA,anactivepepsinhomologfromthe bacterium
Shewanella amazonensis
Isaura Simo
˜
es
1,2
, Rosa
´
rio Faro
1
, Daniel Bur
3
, John Kay
4
and Carlos Faro
1,2
1 CNC-Center for Neuroscience and Cell Biology, University of Coimbra, Portugal
2 Biocant, Biotechnology Innovation Center, Cantanhede, Portugal
3 Actelion Pharmaceuticals Ltd, Allschwil, Switzerland
4 School of Biosciences, Cardiff University, UK
Keywords
aspartic proteinase; bacteria; pepsin-like
Correspondence
I. Simo˜ es, Biocant, Parque Tecnolo
´
gico de
Cantanhede, Nu
´
cleo 4, Lote 3, 3060-197
Cantanhede, Portugal
Fax: +351 231 419049
Tel: +351 231 419040
E-mail: isimoes@biocant.pt
(Received 19 April 2011, revised 4 July
2011, accepted 8 July 2011)
doi:10.1111/j.1742-4658.2011.08243.x
The view has been widely held that pepsin-like aspartic proteinases are
found only in eukaryotes, and not in bacteria. However, a recent bioinfor-
matics search [Rawlings ND & Bateman A (2009) BMC Genomics 10, 437]
revealed that, in seven of 1000 completely sequenced bacterial genomes,
genes were present encoding polypeptides that displayed the requisite hall-
mark sequence motifs of pepsin-like aspartic proteinases. The implications
of this theoretical observation prompted us to generate biochemical data to
validate this finding experimentally. The aspartic proteinase gene from one
of the seven identified bacterial species, Shewanella amazonensis, was
expressed in Escherichia coli. The recombinant protein, termed shewasin A,
was produced in soluble form, purified to homogeneity, and shown to dis-
play properties remarkably similar to those of pepsin-like aspartic protein-
ases. Shewasin A was maximally active at acidic pH values, cleaving a
substrate that has been widely used for assessment of the proteolytic activ-
ity of other aspartic proteinases, and displayed a clear preference for cleav-
ing peptide bonds between hydrophobic residues in the P1*P1¢ positions of
the substrate. It was completely inhibited by the general inhibitor of aspar-
tic proteinases, pepstatin, and mutation of one of the catalytic Asp residues
(in the Asp-Thr-Gly motif of the N-terminal domain) resulted in complete
loss of enzymatic activity. It can thus be concluded unequivocally that this
Shewanella gene encodes anactive pepsin-like aspartic proteinase. It is now
beyond doubt that pepsin-like aspartic proteinases are not confined to
eukaryotes, but are encoded within some species of bacteria. The distinc-
tions between the bacterial and eukaryotic polypeptides are discussed and
their evolutionary relationships are outlined.
Structured digital abstract
l
Shewasin A cleaves Oxidized Insulin B chain by protease assay (View Interaction 1, 2)
Introduction
Aspartic proteinases (APs) are widely distributed in
nature, including in a variety of infectious organisms,
such as Plasmodium falciparum, HIV, and a large num-
ber of fungi [1]. However, relatively few have been
Abbreviations
AP, aspartic proteinase; DABCYL, 4-(dimethylaminoazo)benzene-4-carboxylic acid; DNP, 2,4-dinitrophenyl;
E-64, l-trans-epoxysuccinylleucylamide-(4-guanidino)butane; EDANS, 5-[(2-aminoethyl)amino]naphthalene-1-sulfonic acid;
MCA, (7-methoxycoumarin-4-yl) acetic acid; Nbs
2
, 5,5¢-dithio-bis(2-nitrobenzoic acid).
FEBS Journal 278 (2011) 3177–3186 ª 2011 The Authors Journal compilation ª 2011 FEBS 3177
described in bacteria. These enzymes are currently sub-
divided into different families and clans as described in
the MEROPS database [2]. Seven of these families
(A8, A22, A24, A25, A31, A26, and A5) are organized
into distinct clans in which, although Asp residues are
known to be critical for enzymatic activity, they
appear in diverse sequence motifs [2]. Indeed, this set
of families contains the only APs of bacterial origin
that have so far been characterized: these include sig-
nal peptidase II from Escherichia coli, which is the
type peptidase of family A8 [3], the prepilin peptidases
of family A24 [4], GPR endopeptidase from Bacil-
lus megaterium in family A25 [5], omptin from E. coli,
which is the type peptidase of family A26 [6], and
HybD peptidases of family A31 [7].
The remaining families of APs in the MEROPS
database (A1, A2, A3, A9, A11, and A33) all belong
to only one clan (AA), with their members being read-
ily identified by the presence of characteristic hallmark
sequence motifs. These ‘archetypal’ APs include
eukaryotic enzymes such as pepsin and viral retropep-
sins, including HIV-1 retropepsin. Pepsin-like APs
characteristically consist of two internally homologous
domains, each of which provides a catalytic Asp to the
active site. Each Asp is present in the hallmark motif
Asp-Thr ⁄Ser-Gly, followed further downstream by a
hydrophobic-hydrophobic-Gly sequence. Together,
these motifs form a structural feature known as a psi
loop, which serves to locate the two Asp residues nec-
essary for operation of the catalytic machinery [1]. In
contrast, retroviral-type APs are obligate homodimers,
in which each monomer contributes one catalytic motif
to one psi loop. Enzymes with a pepsin-like ‘arche-
typal’ organization are by far the most numerous and
well-characterized APs, and have been thought to be
confined to eukaryotes. This has been supported by
structural evidence suggesting that pepsin-like enzymes
evolved through a gene duplication and fusion event
from a retropepsin-type of ancestral gene [8]. However,
the absolute requirement for the psi loop structural
feature described above provides four landmark motifs
(two Asp-Thr ⁄Ser-Gly and two hydrophobic-hydro-
phobic-Gly) that are required to be present in con-
served locations, and so can be searched for during
data mining operations to identify putative pepsin-like
APs in any newly sequenced genome. In such an
endeavor, contrary to long-held beliefs, pepsin-like
APs were detected within the genomes of a few bacte-
ria [9]. All of the currently sequenced bacterial
genomes ( 1000) were examined, and putative AP-
encoding genes were identified in seven species. Of
these, two pairs of Asp-Thr ⁄ Ser-Gly + hydrophobic-
hydrophobic-Gly motifs were present in the predicted
polypeptides from five species, all marine psychro-
philes, including Shewanellaamazonensis [9]. Prior to
this recent report, other publications suggesting the
presence of archetypal types of AP in bacteria were
somewhat unconvincing [10–12].
Given the potential significance of this detection of
AP-encoding genes in a few species of bacteria, it was
thus considered of particular importance to produce
reliable biochemical data to characterize such putative
bacterial gene products and thus establish unequivo-
cally whether these encoded protein products were
functional enzymes. In this article, we describe the pro-
duction of recombinant shewasinA,the pepsin-like
homolog fromthebacterium S. amazonensis, and dem-
onstrate that it displays all of the enzymatic properties
characteristic of a eukaryotic pepsin-like AP. We dis-
cuss the differences between bacterial and eukaryotic
polypeptides, and consider the evolutionary signifi-
cance of these observations.
Results
Expression and purification of recombinant
shewasin A
In order to characterize one of the bacterial pepsin-like
homologs identified by Rawlings & Bateman [9], the
gene from S. amazonensis (GenBank: ABL98994.1)
was selected. DNA was synthesized (sequence detailed
in Fig. S1) to encode the full-length polypeptide, the
sequence of which is shown in Fig. 1. The synthetic
gene was expressed in E. coli BL21(DE3) as described
Fig. 1. Deduced amino acid sequence of S. amazonensis shewa-
sin A. The hallmark motifs of pepsin-like APs are highlighted in the
sequence, and include: (a) theactive site motifs (DT ⁄ SG) (shown in
bold in gray boxes); (b) the hydrophobic-hydrophobic-Gly motifs of
the psi loops (shown in bold); and (c) the conserved Tyr residue in
the ‘flap’ region (double underlined). The eight Cys residues are
underlined in the sequence. No signal peptide or propart segment
is present in theshewasin A amino acid sequence. The Asp of the
active site motif fromthe N-terminal domain (marked with an aster-
isk) was mutated to an Ala to generate theactive site mutant
shewasin A_(D37A). Although it displays the typical hallmark motifs
of pepsin-like APs, shewasin A shows a low overall percentage of
sequence identity with eukaryotic pepsin-like enzymes, e.g. pep-
sin A (18%), BACE1 (10%), and renin (9%).
Characterization of recombinant shewasin A I. Simo˜es et al.
3178 FEBS Journal 278 (2011) 3177–3186 ª 2011 The Authors Journal compilation ª 2011 FEBS
in Experimental procedures, and initial conditions were
optimized to enhance the accumulation of recombinant
shewasin A, with an N-terminal His-tag, in the soluble
fraction of the cell lysates. Metal ion affinity chroma-
tography was applied (Fig. 2A), and fractions enriched
in shewasin A were pooled and further purified by
size-exclusion chromatography on a HiLoad 26 ⁄60
Superdex 200 column (Fig. 2B). SDS ⁄ PAGE analysis
of the purified fractions under reducing conditions
confirmed the presence of a protein with the predicted
molecular mass of 50 kDa (Fig. 2D, lanes 1–4). The
identity of this band was further confirmed by western
blot analysis with an antibody against His tag (not
shown).
The purified recombinant shewasin A was subjected
to analytical size-exclusion chromatography (Fig. 2C)
under nondenaturing conditions and in the absence of
a reducing agent, and its molecular mass was deter-
mined to be 50 kDa, consistent with the value calcu-
lated for the polypeptide (Fig. 1) encoded by the
bacterial gene. Thus, recombinant shewasin A exists as
a monomeric polypeptide, as observed for the majority
of eukaryotic pepsin-like APs studied previously.
Shewasin A contains eight Cys residues at noncon-
served positions. To evaluate the number of these resi-
dues present in a reduced form, a 5,5¢-dithiobis
(2-nitrobenzoic acid) (Nbs
2
) assay was carried out. The
number of free thiol groups in recombinant shewa-
sin A estimated from Nbs
2
titration was 8.27 ± 0.59.
These results clearly suggest that all sulfhydryl groups
of shewasin A exist as free thiols, in sharp contrast to
its eukaryotic counterparts.
Activity and specificity of recombinant
shewasin A
Recombinant shewasin A was next examined for its
ability to cleave a number of polypeptides typically
used as AP substrates. Fluorogenic substrates
cleaved by renin [(5-[(2-aminoethyl)amino]naphthalene-
Fig. 2. Purification and analysis of recombinant shewasin A. Wild-type shewasin A was produced in E. coli in soluble form, fused to an
N-terminal His-tag. (A) HisTrapHP chromatogram. Recombinant shewasin A was purified by metal ion affinity chromatography with a HisT-
rapHP column. Elution was accomplished by using stepwise increases in concentration of imidazole (50, 100 and 500 m
M). The recombinant
protein was eluted with 100 m
M imidazole, corresponding to fractions highlighted by dotted lines (1 and 2; numbers above the peaks). (B)
S200 chromatogram. HisTrap eluate (fractions 1 and 2) was pooled and further purified by size-exclusion chromatography as described in
Experimental procedures. Purified recombinant shewasin A (dotted lines, sample 3) was used in subsequent characterization assays. (C)
Analytical size-exclusion chromatography of purified recombinant shewasin A. The Superose 12 was equilibrated with 20 m
M Hepes buffer
(pH 7.5) and 100 m
M NaCl. The dots indicate the elution volumes of molecular mass markers used for calibration (from left to right: aldolase,
158 kDa; conalbumin, 75 kDa; ovalbumin, 34 kDa; carbonic anhydrase, 29 kDa; ribonuclease A, 13.7 kDa). The collected fraction is high-
lighted by dotted lines, and pooled as fraction 4. (D) SDS ⁄ PAGE analysis of protein fractions collected fromthe different steps of purification.
Lanes 1 and 2: fractions 1 and 2 in (A). Lane 3: fraction 3 in (B). Lane 4: Superose 12 eluate marked with number 4 in (C). The gel was
stained with Coomassie Brilliant Blue.
I. Simo˜es et al. Characterization of recombinant shewasin A
FEBS Journal 278 (2011) 3177–3186 ª 2011 The Authors Journal compilation ª 2011 FEBS 3179
1-sulfonic acid [EDANS])-Ile-His-Pro-Phe-His-Leu-Val-
Ile-His-Thr-Lys(DABCYL)-Arg], HIV-1 retropepsin
[Arg-Glu(EDANS)-Ser-Gln-Asn-Tyr-Pro-Ile-Val-Gln-Lys-
4-(dimethylaminoazo)benzene-4-carboxylic acid (DAB-
CYL)-Arg] and BACE1 [(7-methoxycoumarin-4-yl)
acetic acid (MCA)Lys-Ser-Glu-Val-Asn-Leu-Asp-Ala-
Glu-Phe-Lys-2,4-dinitrophenyl (DNP)] were not signifi-
cantly processed by shewasin A under the conditions
tested. The failure to cleave the BACE1 substrate may
be noteworthy, in that a phylogenetic analysis of fam-
ily A1 members resolved shewasin A into a cluster
with BACE1 and its human paralog BACE2, suggest-
ing the closest relationship with these eukaryotic
enzymes [9]. I n contrast, the fluorogenic peptide (MCA)
Lys-Leu-His-Pro-Glu-Val-Leu-Phe-Val-Leu-Glu-Lys(DNP),
which was originally designed as a substrate for
CDR1, an atypical AP from Arabidopsis thaliana [13],
was readily hydrolyzed by shewasin A at pH 4. Analy-
sis by MS revealed that the primary cleavage site was
at Leu*Phe (* indicates the cleavage site), with a fur-
ther minor cleavage occurring at the adjacent Phe*Val
(Table 1).
Incubation of shewasin A with the B chain of oxi-
dized insulin at pH 4 was followed by RP-HPLC sepa-
ration of the products (not shown) and analysis by
MS. For this peptide, two major cleavage sites were
identified, Leu15*Tyr16 and Tyr16*Leu17 (Table 1).
Thus, shewasin A reveals a specificity that is intrinsic
to most eukaryotic pepsin-like APs in cleaving prefer-
entially between hydrophobic residues occupying the
substrate P1 and P1¢ positions.
The final substrate tested was a fluorogenic deriva-
tive of the chromogenic peptide Lys-Pro-Ala-Glu-
Phe*Nph-Ala-Leu (where Nph is
L-norleucine) [14].
The quenched fluorescent version of this peptide, (MCA)
Lys-Lys-Pro-Ala-Glu-Phe-Phe-Ala-Leu-Lys(DNP), was
readily cleaved by shewasinA, displaying typical
Michaelis–Menten kinetic behavior. The kinetic
parameters determined for cleavage at pH 4.0 were
K
m
= 5.4 lM, k
cat
= 0.03 s
)1
, and k
cat
⁄ K
m
= 5.6 ·
10
3
M
)1
Æs
)1
, respectively. MS analysis revealed that this
peptide was preferentially cleaved at Phe*Phe, with a
second minor cleavage occurring at Phe*Ala (Table 1).
Maximum activity was observed at temperatures
between 42 and 50 °C, decreasing so drastically above
50 °C that complete loss of activity was detected at
60 °C (Fig. 3A).
The pH dependence of the cleavage of (MCA)Lys-
Lys-Pro-Ala-Glu-Phe-Phe-Ala-Leu-Lys(DNP) by recom-
binant shewasin A is shown in Fig. 3B. The highest
activity was detected at acidic pH values between
pH 3.75 and pH 4.5, and only 50% activity was
retained at pH 5. At pH 6.0, the enzyme showed no
activity towards this substrate (Fig. 3B). This behavior
is typical of many eukaryotic APs mainly acting in
Table 1. Primary specificity of recombinant shewasin A. Three dif-
ferent substrates were incubated with recombinant shewasin A as
described in Experimental procedures. The resulting cleavage prod-
ucts were identified directly by MS analysis or, in the case of oxi-
dized insulin B chain, separated by RP-HPLC prior to identification
by MS. Preferential cleavage sites are indicated by (››) and minor
cleavage sites by (›).
Substrate Sequence
CDR1 peptide (MCA)KLHPEVL››F›VLEK(DNP)
Oxidized insulin
B chain
FVNQHLCGSHLVEAL››Y››LVCGERGFFYTPKA
Typical peptide (MCA)KKPAEF››F›ALK(DNP)
Fig. 3. Effect of temperature and pH on the activity of recombinant
shewasin A. Shewasin A was tested for activity with the synthetic
fluorogenic peptide (MCA)Lys-Lys-Pro-Ala-Glu-Phe-Phe-Ala-Leu-
Lys(DNP) as substrate; the chromogenic version of this has been
used as a model substrate to characterize pepsin-like enzymes
from various sources. (A) Activity studies at different temperatures
were performed by incubating shewasin A in 0.05
M sodium ace-
tate buffer (pH 4) and 0.1
M NaCl at temperatures between 15 and
60 °C. (B) Activities at different pH values were measured by incu-
bating shewasin A at 37 °C with buffers between pH 2.5 and pH 7
containing 0.1
M NaCl (0.05 M sodium citrate, pH 2.5–3.5; 0.05 M
sodium acetate, pH 4–5.5, 0.05 M sodium phosphate, pH 6–7).
Characterization of recombinant shewasin A I. Simo˜es et al.
3180 FEBS Journal 278 (2011) 3177–3186 ª 2011 The Authors Journal compilation ª 2011 FEBS
acidic environments, including vertebrate pepsins,
cathepsin D, and a variety of enzymes of fungal origin
[1,15,16], but contrasts with that observed for more
specialized APs, which are active at elevated pH values
closer to neutrality, e.g. renin and HIV-1 retropepsin.
The fact that maximum activity for shewasin A is
observed between pH 3.75 and pH 4.5 makes the
behavior of the recombinant bacterial AP even more
like that of an archetypal pepsin-like enzyme rather
than like some of the more ‘specialized’ APs, such as
renin and retroviral proteinases.
Inhibition and dependence on conserved catalytic
residues for shewasin A activity
The most frequently applied test employed to classify a
newly identified protease is susceptibility to prototypi-
cal inhibitors such as pepstatin [1]. Consequently, the
effect of pepstatin on the activity of shewasin A was
examined; whereas pepstatin completely blocked its
proteolytic activity at pH 4, all other inhibitors tested
were devoid of inhibitory effect (Table 2). In order to
substantiate this finding further, anactive site mutant
of shewasin A was generated in which the (putative)
catalytic Asp of the Asp-Thr-Gly motif of the N-termi-
nal domain (Fig. 1) was mutated to an Ala (D37A).
This mutant was expressed in E. coli and purified
under similar conditions to those used for wild-type
shewasin A. Purified shewasin A_(D37A) was analyzed
in a size-exclusion chromatography column, and
displayed a molecular mass of 50 kDa (Fig. 4A),
consistent with that described above for the wild type.
Analysis by SDS ⁄ PAGE and western blot with a
His-tag antibody (Fig. 4B) revealed that the mutant
protein migrated identically to the wild-type shewasin A.
In sharp contrast, however, purified shewasin A_(D37A)
was completely inactive towards the fluorogenic
substrate (MCA)Lys-Lys-Pro- Ala-Glu-Phe-Phe-Ala-Leu-
Lys(DNP) at pH 4.0 (Fig. 4C).
Discussion
Shewasin A exists as a monomer, exhibits activity at
acidic pH against a well-documented AP substrate,
Table 2. Effect of prototypical proteinase inhibitors on the activity
of recombinant shewasin A. Recombinant shewasin A was tested
for activity with the synthetic fluorogenic peptide (MCA)Lys-Lys-
Pro-Ala-Glu-Phe-Phe-Ala-Leu-Lys(DNP) as substrate in 0.05
M
sodium acetate (pH 4) and 0.1 M NaCl at 37 °C. The enzyme was
preincubated in the presence of each prototypical inhibitor for
10 min at 37 °C before substrate addition.
Inhibitor Concentration (m
M) Activity (%)
Pepstatin 0.001 0
Pefabloc 1 87.8
EDTA 5 82.8
E-64 0.01 89.5
Amastatin 0.01 97.2
Bestatin 0.01 92.7
Leupeptin 0.01 91.9
Dithiothreitol 2 91.2
Iodoacetamide 0.05 82.1
Fig. 4. Purification and analysis of recombinant shewasin A active
site mutant. Theactive site mutant shewasin A_(D37A) was
expressed in E. coli, purified according to the protocol optimized for
shewasin A described in Experimental procedures, and subse-
quently analyzed by analytical size-exclusion chromatography in a
Superose 12 column (A). (B) Purified shewasin A_(D37A) [fraction
delimited by dotted lines in (A)] was analyzed by SDS ⁄ PAGE and
western blot (WB) with a His-tag antibody. Wild-type shewasin A
(WT) was included for comparison. The gel was stained with Coo-
massie Brilliant Blue. (C) Purified recombinant shewasin A_(D37A)
was tested for activity with the synthetic fluorogenic peptide
(MCA)Lys-Lys-Pro-Ala-Glu-Phe-Phe-Ala-Leu-Lys(DNP) as substrate
in 0.05
M sodium acetate buffer (pH 4) and 0.1 M NaCl at 37 °C.
I. Simo˜es et al. Characterization of recombinant shewasin A
FEBS Journal 278 (2011) 3177–3186 ª 2011 The Authors Journal compilation ª 2011 FEBS 3181
cleaves its substrates preferentially between hydrophobic
amino acids, and is susceptible to inhibition by pepsta-
tin. Furthermore, the presence of four typical motifs
(two Asp-Thr ⁄ Ser-Gly and two hydrophobic-hydro-
phobic-Gly) in its sequence in combination with a total
loss of activity as a result of mutation of one of these
putative catalytic Asp residues provides further strong
evidence that this enzyme is anactive pepsin-like AP.
To the best of our knowledge, this is the first docu-
mentation of such an activity, and establishes beyond
doubt that pepsin-like APs belonging to family A1 are
not confined to eukaryotes but are encoded in certain
species of bacteria.
Whereas shewasin A’s enzymatic properties are in
good agreement with those of its eukaryotic homo-
logs, one obvious molecular feature serves to distin-
guish between the bacterial AP and its eukaryotic
counterparts. Eukaryotic pepsin-like APs from fam-
ily A1 are typically encoded and produced as prep-
roenzymes (e.g. pig pepsinogen), consisting of an
initial signal peptide, a propart segment, and the
mature enzyme region. In sharp contrast, the shewa-
sin A polypeptide encoded within the bacterial gen-
ome is devoid of both a signal peptide and propart
segment (Fig. 1). Eukaryotic AP polypeptides lacking
either a signal peptide or a propart segment have
been described previously in other species (fungi [17]
and oomycetes [18]), but the finding that bacterial
APs such as shewasin A lack both of the compo-
nents is totally unprecedented. In our studies, recom-
binant shewasin A was isolated in anactive form
directly fromthe soluble fraction of E. coli cell ly-
sates, so the absence of a propart segment would
not appear to be detrimental to the folding of this
bacterial AP in the heterologous expression system
chosen. Further experiments will be necessary to
establish the subcellular location and activity of
shewasin A within S. amazonensis cells.
In eukaryotic zymogens, the propart segment is
known to make essential contributions, such as ensur-
ing proper folding and intracellular sorting of the
zymogen polypeptide, and facilitation of its activation
to release the mature enzyme when the appropriate
conditions are encountered [1,19]. Interestingly, the
presence of the propart segment in proBACE1 was
found to have little effect on the intrinsic proteolytic
activity as in a typical AP zymogen, but its inclusion
at the protein’s N-terminus ensured much more rapid
folding of the polypeptide than was observed when
only the mature form was produced [20,21]. As
BACE1 and BACE2 were predicted to be shewasin A’s
closest eukaryotic homologs [9], it is possible that the
propart segment in proBACE1 and proBACE2 may
represent an ancient version of this domain that might
have been acquired throughout proteinase evolution,
developing according to evolutionary pressures to
extend the lifetime of these eukaryotic APs [22].
Indeed, given the strict requirement of the propart
segment for proper folding of the precursors of the
large majority of eukaryotic pepsin-like proteinases,
active bacterial pepsin homologs lacking the propart
segment, such as shewasinA, might represent ‘fossil’
versions of pepsin-like proteinases rather than a
derived state resulting from a horizontal gene transfer
mechanism.
Another interesting feature of the bacterial pepsin
homologs [9] is the difference in their Cys content and
position within the sequence. Shewasin A contains
eight Cys residues (Fig. 1), whereas three, four, seven
or eight are present, at nonconserved locations, in the
six predicted polypeptides fromthe other species of
bacteria [9]. This contrasts sharply with eukaryotic
pepsin-like APs, which commonly contain two, four or
six Cys residues located at totally conserved positions,
and form one, two or three disulfide bonds, respec-
tively. None of the Cys residues in theshewasin A
sequence are present at these conserved positions, and
their localization along the protein sequence suggests
that the Cys residues in this bacterial AP may not
form disulfide bonds, as determined from a model of
shewasin A built on pig pepsin (data not shown).
Determination of shewasin A free sulfhydryl groups by
Nbs
2
titration further confirmed this in silico analysis,
as all of its eight Cys residues were estimated to exist
as free thiols. It is very likely that the Cys residues
remain in their reduced form with free SH side chains,
which would be consistent with the reducing environ-
ment that exists inside bacterial cells. In further
support of this interpretation, shewasin A accumulated
in a soluble monomeric form in E. coli , and addition
of dithiothreitol (at 2 m
M) had no effect on either the
molecular mass of theactive entity or on the activity
observed for the purified recombinant wild-type
shewasin A (Table 2). A similar result was observed
when shewasin A activity was assayed in the presence
of iodoacetamide (at 0.05 m
M) (Table 2).
The absence of a signal peptide at the N-terminus
of bacterial APs such as shewasin A also suggests
that these might be cytosolic proteins [9]. Accord-
ingly, it was expected that shewasin A would be
active at pH values reflecting that of the bacterial
cytoplasm, i.e. close to neutrality. The presence in
shewasin A of an Ala just downstream fromthe Asp-
Thr ⁄ Ser-Gly motif of the C-terminal domain (in the
sequence Asp-Ser-Gly-Ala; Fig. 1) was also suggestive
of such an effect, because the maximum activity of
Characterization of recombinant shewasin A I. Simo˜es et al.
3182 FEBS Journal 278 (2011) 3177–3186 ª 2011 The Authors Journal compilation ª 2011 FEBS
more specialized APs such as renin and HIV-1 retro-
pepsin at pH values closer to neutrality has been
attributed, at least in part, to the presence of this Ala
[23]. This contrasts with the situation in many pepsin-
like enzymes, in which the equivalent residue is Thr.
However, our experimental observations do not sup-
port this interpretation, as shewasin A was shown to
be maximally active around pH 4, and no activity
whatsoever was detected at pH 6. Thus, this differ-
ence in shewasin A catalytic activity from those of
other APs with a similar active site sequence motif
may be the result of subtle variations in subsite bind-
ing pockets.
APs of the pepsin-like (A1) family were believed,
until recently, to be confined to eukaryotic organisms;
our results provide unequivocal experimental substanti-
ation that this type of AP is also encoded, but in
the form of the mature enzyme, in bacteria. The
S. amazonensispepsinhomolog described here is
strongly reminiscent of eukaryotic pepsin-like APs.
Our findings pose challenges for understanding the
evolutionary relationships between bacterial APs and
their eukaryotic counterparts, particularly as shewa-
sin A was shown to be distantly positioned near the
root of the phylogenetic tree derived from family A1
members [9]. The most widely held view has been that
retroviral APs represent the ancestral state, and that
bilobed pepsin-like proteinases are the result of gene
duplication and fusion events [1,8]. As it is now clear
that the S. amazonensis genome does encode an active
pepsin-like proteinase, it would appear that the gene
duplication and fusion may well be very ancient
events, preceding the divergence between bacteria
and eukaryotes. The recent identification of a novel
retroviral-type AP (SpoIIGA) in Bacillus subtilis,a
Gram-positive bacterium [24], further contributes to
the discussion on the evolutionary relationships
between retroviral and pepsin-like APs. This sequence
contained one Asp-Thr-Gly motif, consistent with that
expected of family A2 members, so that dimerization
would be required for activity. Mutational analysis
demonstrated the critical role of the Asp for substrate
processing; however, attempts at inhibition of the
observed activity were rather inconclusive. Further
investigations will thus be necessary to demonstrate
unequivocally that an ancestral gene encoding a single-
lobed AP sequence is present in prokaryotes. However,
the evidence currently available does provide an initial
indication that the hypothetical gene duplica-
tion ⁄ fusion events that may have given rise to the
bi-lobed pepsin-like APs, such as shewasinA, might
have preceded the most recent common ancestor of
prokaryotes and eukaryotes.
Experimental procedures
Cloning of S. amazonensis gene encoding
shewasin A
DNA encoding the S. amazonensispepsinhomolog gene
(gene locus Sama_0787; genomic sequence available at the
EBI Data Bank under the accession number
ABL98994)
was chemically synthesized (Genscript, Piscataway, NJ,
USA), and optimized for codon usage in E. coli to enhance
protein expression. The synthetic gene sequence (detailed in
Fig. S1) included restriction sites for NdeI and XhoI at the
5¢-end and 3¢-end, respectively, to facilitate subsequent
subcloning in pET28a (Novagen, Gibbstown, NJ, USA)
in-frame with an N-terminal His-tag. The positive clones
selected by restriction analysis were confirmed by DNA
sequencing. The QuickChange site-directed mutagenesis kit
(Stratagene, La Jolla, CA, USA) was used to generate the
active site mutant shewasin A_(D37A) in the vector pET28a,
using the primers 5¢-AGCGTGAACCTGATTATT
GCGA
CCGGCAGCAGCACCCTG-3¢ (forward primer) and
5¢-CAGGGTGCTGCTGCCGGT
CGCAATAATCAGGTT
CACGCT-3¢ (reverse primer) (mutation sites underlined).
The positive mutant clones were confirmed by DNA
sequencing.
Expression and purification of recombinant
shewasin A and theactive site mutant in E. coli
Wild-type shewasin A and shewasin A_(D37A) were trans-
formed into E. coli BL21(DE3). The method of recombi-
nant protein expression was optimized to maximize the
yield of protein in soluble form, and the resulting condi-
tions were used in all subsequent experiments as follows.
After growth of the cells at 30 °CtoD
600 nm
of 0.6, gene
expression was induced by the addition of isopropyl thio-b-
D-galactoside (0.05 mM final concentration). After 4 h at
30 °C, cells were harvested by centrifugation at 8983 g at
4 °C for 20 min, resuspended in 20 m
M sodium phosphate
buffer (pH 7.4) containing 10 m
M imidazole and 0.5 M
NaCl (binding buffer for immobilized metal ion affinity
chromatography), and lysed with lysozyme (100 lgÆmL
)1
).
After freezing and thawing, DNase (100 lgÆmL
)1
) and
MgCl
2
(100 mM) were added, and the reaction mixture was
incubated for 2 h at 4 °C. The cell lysate was centrifuged at
12 000 g and 4 °C for 12 min. The soluble fraction was fil-
tered through 0.2-lm filters, and immediately loaded onto a
HisTrapHP 5-mL column (GE Healthcare Life Sciences,
Uppsala, Sweden) previously equilibrated in binding buffer.
After sample loading, the column was connected to an
FPLC system (DuoFlow-BioRad, Hercules, CA, USA), and
extensively washed with binding buffer until A
280 nm
reached a steady baseline. Protein elution was carried out
by increasing the concentration of imidazole stepwise
I. Simo˜es et al. Characterization of recombinant shewasin A
FEBS Journal 278 (2011) 3177–3186 ª 2011 The Authors Journal compilation ª 2011 FEBS 3183
(50, 100 and 500 mM) in the same buffer. Both shewasin A
and shewasin A_(D37A) were eluted with the buffer con-
taining 100 m
M imidazole. Pooled fractions were applied to
a HiLoad 26 ⁄ 60 Superdex 200 gel filtration column (GE
Healthcare Life Sciences) connected to an FPLC system
(DuoFlow-BioRad) equilibrated in 20 m
M Hepes buffer
(pH 7.5) containing 100 m
M NaCl for further purification
and imidazole removal.
Size-exclusion chromatography
The molecular masses of purified recombinant shewasin A
and shewasin A_(D37A) were estimated under nondenatur-
ing conditions by size-exclusion chromatography on a
Superose 12 (GE Healthcare Life Sciences) column con-
nected to an FPLC system (DuoFlow-BioRad). The column
was equilibrated in 20 m
M Hepes buffer (pH 7.5) contain-
ing 100 m
M NaCl, and calibrated with Gel Filtration LMW
and HMW calibration kits (GE Healthcare Life Sciences),
according to the manufacturer’s instructions. The molecular
mass markers used for calibration were aldolase (158 kDa),
conalbumin (75 kDa), ovalbumin (43 kDa), carbonic anhy-
drase (29 kDa), ribonuclease A (13.7 kDa), and aprotinin
(6.5 kDa).
Nbs
2
assay
The sulfhydryl contents of recombinant shewasin A were
determined spectrophotometrically at 412 nm with Nbs
2
[25]. Purified recombinant shewasin A (0.37 lM) was incu-
bated at 30 °C with 3.3 m
M Nbs
2
in 0.1 M sodium phos-
phate buffer (pH 8.0) containing 1 m
M EDTA. A control
without enzyme was performed to measure the spontaneous
breakdown of the reagent, and this value was used to cor-
rect the titration value obtained for recombinant shewa-
sin A. The number of sulfhydryl groups was calculated by
using the molar extinction coefficient of 2-nitro-5-thioben-
zoic acid (14 150 M
)1
Æcm
)1
).
Enzyme assays
The proteolytic activities of purified recombinant shewa-
sin A and shewasin A_(D37A) were tested against several
fluorogenic peptides, initially at concentrations between 1
and 2 l
M in buffers at different pH values containing 0.1 M
NaCl and 8% (v ⁄ v) dimethylsulfoxide. These included the
renin substrate Arg-Glu(EDANS)-Ile-His-Pro-Phe-His-Leu-
Val-Ile-His-Thr-Lys(DABCYL)-Arg and the HIV-1 retropepsin
protease substrate Arg-Glu(EDANS)-Ser-Gln-Asn-Tyr-
Pro-Ile-Val-Gln-Lys(DABCYL)-Arg, both from Sigma (St
Louis, MO, USA), and the BACE1 substrate (MCA)Lys-
Ser-Glu-Val-Asn-Leu-Asp-Ala-Glu-Phe-Lys(DNP), as well
as (MCA)Lys-Leu-His-Pro-Glu-Val-Leu-Phe-Val-Leu-Glu-Lys-
(DNP) and (MCA)Lys-Lys-Pro-Ala-Glu-Phe-Phe-Ala-Leu-
Lys(DNP), all synthesized by Genosphere Biotechnologies
(Paris, France). The last of these peptides was found to be
cleaved readily, and the rate of hydrolysis was monitored at
an excitation wavelength of 328 nm and an emission wave-
length of 393 nm. The relationship between fluorescence
change and peptide concentration was calculated by mea-
suring the total fluorescence change that occurred upon
complete hydrolysis of the peptide. Kinetic parameters for
the cleavage reaction were calculated fromthe Lineweaver–
Burk plot with appropriate software. For activity studies at
different pH values, the following buffers, all containing
0.1
M NaCl, were used between pH 2.5 and pH 7 at 37 °C:
0.05
M sodium citrate (pH 2.5–3.5); 0.05 M sodium acetate
(pH 4–5.5); and 0.05
M sodium phosphate (pH 6–7). For
activity studies at different temperatures, recombinant
shewasin A was incubated in 0.05
M sodium acetate and
0.1
M NaCl (pH 4) at temperatures between 15 and 60 °C.
The effects of various inhibitors on the proteolytic activity
of shewasin A were assayed by preincubating the enzyme
with each compound for 10 min at 37 °C in 0.05
M sodium
acetate buffer (pH 4.0) containing 0.1
M NaCl before deter-
mination of the residual proteolytic activity. Shewa-
sin A_(D37A) was examined for activity under the same
assay conditions.
Digestion of oxidized insulin B chain
Digestion of oxidized insulin B chain by purified recombi-
nant shewasin A was carried out for 4 h at 37 ° C in 0.1
M
sodium acetate buffer (pH 4). The reaction was stopped
with 0.6% (v ⁄ v) trifluoroacetic acid (final concentration)
and, after centrifugation (12 000 g, 5 min), digestion frag-
ments were separated by RP-HPLC on a C18 column,
using a Prominence system (Shimadzu Corporation, Tokyo,
Japan) with a KROMASIL 100 C18 250 · 4.6 mm column.
Elution was carried out with a linear gradient of acetoni-
trile (0–80%) in 0.1% (v ⁄ v) trifluoroacetic acid for 30 min
at a flow rate of 1 mLÆ min
)1
. Absorbance was monitored at
220 nm, and the isolated peptides were collected, freeze-
dried, and submitted to identification with a 4000 QTRAP
system (Proteomics Unit of the Center for Neuroscience
and Cell Biology, University of Coimbra, Portugal).
PAGE and immunoblotting
Protein samples were separated by SDS ⁄ PAGE, using 12%
gels in a Bio-Rad Mini Protean III electrophoresis appara-
tus (Bio-Rad, Hercules, CA, USA). Gels were stained with
Coomassie Brilliant Blue R-250 (Sigma). For immunoblotting
analysis, protein samples were separated by SDS ⁄ PAGE
(12% gels) and transferred to a poly(vinylidene difluoride)
membrane for immunoblotting (40 V, overnight, at 10 °C).
The membranes were blocked for 60 min with 5% (w ⁄ v)
nonfat dry milk plus 0.1% (v ⁄ v) Tween-20 in NaCl ⁄ Tris,
and then incubated at room temperature for 60 min with
the primary antibody, mouse His-tag antibody (GenScript;
Characterization of recombinant shewasin A I. Simo˜es et al.
3184 FEBS Journal 278 (2011) 3177–3186 ª 2011 The Authors Journal compilation ª 2011 FEBS
1 : 5000 dilution). After several washes with 0.5% (w ⁄ v)
nonfat dry milk plus 0.1% (v ⁄ v) Tween-20 in NaCl ⁄ Tris,
the membranes were incubated at room temperature for
60 min with secondary antibody [alkaline phosphatase-con-
jugated goat anti-(mouse IgG+ IgM)] (GE Healthcare;
1 : 10 000 dilution). The membranes were again washed,
and alkaline phosphatase activity was visualized by the
enhanced chemifluorescence method using ECF substrate
(GE Healthcare) on a Molecular Imager FX System
(Bio-Rad).
Acknowledgements
MS analysis was performed in the Proteomic Facility of
the Center for Neuroscience and Cell biology (CNC).
References
1 Dunn BM (2002) Structure and mechanism of the pep-
sin-like family of aspartic peptidases. Chem Rev 102,
4431–4458.
2 Rawlings ND, Barrett AJ & Bateman A (2010) MER-
OPS: the peptidase database. Nucleic Acids Res 38,
D227–D233.
3 Sankaran K (2004) Signal peptidase II. In Handbook of
Proteolytic Enzymes (Barrett AJ, Rawlings ND &
Woessner JF, eds), pp. 201–204. Elsevier Academic
Press, London.
4 Dupuy B, Deghmane AE & Taha MK (2004) Type IV
prepilin peptidase. In Handbook of Proteolytic Enzymes
(Barret AJ, Rawlings ND & Woessner JF, eds), pp.
204–208. Elsevier Academic Press, London.
5 Carroll TM & Setlow P (2005) Site-directed mutagenesis
and structural studies suggest that the germination
protease, GPR, in spores of Bacillus species is an
atypical aspartic acid protease. J Bacteriol 187,
7119–7125.
6 Vandeputte-Rutten L, Kramer RA, Kroon J, Dekker
N, Egmond MR & Gros P (2001) Crystal structure
of the outer membrane protease OmpT from
Escherichia coli suggests a novel catalytic site. EMBO J
20, 5033–5039.
7 Theodoratou E, Huber R & Bock A (2005) [NiFe]-
Hydrogenase maturation endopeptidase: structure and
function. Biochem Soc Trans 33, 108–111.
8 Tang J, James MN, Hsu IN, Jenkins JA & Blundell TL
(1978) Structural evidence for gene duplication in the
evolution of the acid proteases. Nature 271, 618–621.
9 Rawlings ND & Bateman A (2009) Pepsin homologues
in bacteria. BMC Genomics 10, 437–447.
10 Hill J & Phylip LH (1997) Bacterial aspartic proteinas-
es. FEBS Lett 409, 357–360.
11 Kumar AG, Nagesh N, Prabhakar TG & Sekaran G
(2008) Purification of extracellular acid protease and
analysis of fermentation metabolites by Synergistes sp.
utilizing proteinaceous solid waste from tanneries.
Bioresour Technol 99, 2364–2372.
12 Toogood HS, Prescott M & Daniel RM (1995) A pepst-
atin-insensitive aspartic proteinase from a thermophilic
Bacillus sp. Biochem J 307 (Pt 3), 783–789.
13 Simoes I, Faro R, Bur D & Faro C (2007) Character-
ization of recombinant CDR1, an Arabidopsis aspartic
proteinase involved in disease resistance. J Biol Chem
282, 31358–31365.
14 Dunn BM, Jimenez M, Parten BF, Valler MJ, Rolph
CE & Kay J (1986) A systematic series of synthetic
chromophoric substrates for aspartic proteinases.
Biochem J 237, 899–906.
15 Dunn BM, Valler MJ, Rolph CE, Foundling SI,
Jimenez M & Kay J (1987) The pH dependence of the
hydrolysis of chromogenic substrates of the type,
Lys-Pro-Xaa-Yaa-Phe-(NO2)Phe-Arg-Leu, by selected
aspartic proteinases: evidence for specific interactions in
subsites S3 and S2. Biochim Biophys Acta 913, 122–130.
16 Kay J, Tyas L, Humphreys MJ, Hill J, Dunn BM &
Berry C (1996) Aspartic proteinases from parasites.
Adv Exp Med Biol 389, 247–250.
17 ten Have A, Espino JJ, Dekkers E, Van Sluyter SC,
Brito N, Kay J, Gonzalez C & van Kan JA (2010) The
Botrytis cinerea aspartic proteinase family. Fungal Genet
Biol 47, 53–65.
18 Kay J, Meijer HJ, ten Have A & van Kan JA (2011)
The aspartic proteinase family of three Phytophthora
species. BMC Genomics 12, 254–269.
19 Horimoto Y, Dee DR & Yada RY (2009) Multifunc-
tional aspartic peptidase prosegments. N Biotechnol 25,
318–324.
20 Shi XP, Chen E, Yin KC, Na S, Garsky VM, Lai MT,
Li YM, Platchek M, Register RB, Sardana MK et al.
(2001) The pro domain of beta-secretase does not
confer strict zymogen-like properties but does assist
proper folding of the protease domain. J Biol Chem
276, 10366–10373.
21 Emmons TL, Mildner AM, Lull JM, Leone JW, Fischer
HD, Heinrikson RL & Tomasselli AG (2009) Large
scale refolding and purification of the catalytic domain
of human BACE-2 produced in E. coli. Protein Pept
Lett 16, 121–131.
22 Shinde U & Inouye M (2000) Intramolecular chaper-
ones: polypeptide extensions that modulate protein
folding. Semin Cell Dev Biol 11, 35–44.
23 Wlodawer A & Gustchina A (2000) Structural and
biochemical studies of retroviral proteases. Biochim
Biophys Acta 1477, 16–34.
24 Imamura D, Zhou R, Feig M & Kroos L (2008)
Evidence that the Bacillus subtilis SpoIIGA protein is a
novel type of signal-transducing aspartic protease.
J Biol Chem 283, 15287–15299.
I. Simo˜es et al. Characterization of recombinant shewasin A
FEBS Journal 278 (2011) 3177–3186 ª 2011 The Authors Journal compilation ª 2011 FEBS 3185
25 Riddles PW, Blakeley RL & Zerner B (1983)
Reassessment of Ellman’s reagent. Methods Enzymol 91,
49–60.
Supporting information
The following supplementary material is available:
Fig. S1. Nucleotide sequence of codon-optimized she-
wasin A gene and deduced amino acid sequence.
This supplementary material can be found in the
online version of this article.
Please note: As a service to our authors and readers,
this journal provides supporting information supplied
by the authors. Such materials are peer-reviewed
and may be re-organized for online delivery, but are
not copy-edited or typeset. Technical support issues
arising from supporting information (other than miss-
ing files) should be addressed to the authors.
Characterization of recombinant shewasin A I. Simo˜es et al.
3186 FEBS Journal 278 (2011) 3177–3186 ª 2011 The Authors Journal compilation ª 2011 FEBS
. Shewasin A, an active pepsin homolog from the bacterium
Shewanella amazonensis
Isaura Simo
˜
es
1,2
, Rosa
´
rio Faro
1
, Daniel Bur
3
, John Kay
4
and. pro-
duction of recombinant shewasin A, the pepsin- like
homolog from the bacterium S. amazonensis, and dem-
onstrate that it displays all of the enzymatic properties
characteristic