Purification and characterization of a major collagenase from Streptomyces parvulus

Article abstract of DOI:10.1271/bbb.80357

A major collagenase was purified about 96-fold from a crude enzyme sample of Streptomyces parvulus by chromatography on Q-Sepharose, Sephacryl S-200, and butyl-Toyopearl. The purified enzyme showed a relative molecular mass of approximately 52,000 on SDS-

Full text of DOI:10.1271/bbb.80357

Biosci. Biotechnol. Biochem., 73 (1), 21–28, 2009
Purification and Characterization of a Major Collagenase
*
from Streptomyces parvulus
Yasuko SAKURAI,1 Hideshi INOUE,1;2 Wataru NISHII,2 Takayuki TAKAHASHI,1 Yuichi IINO,1
y
1;2;
Masayuki YAMAMOTO,1 and Kenji TAKAHASHI
1Department of Biophysics and Biochemistry, Graduate School of Science, University of Tokyo,
7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033, Japan
2School of Life Sciences, Tokyo University of Pharmacy and Life Sciences,
1432-1 Horinouchi, Hachioji, Tokyo 192-0392, Japan
Received May 27, 2008; Accepted August 21, 2008; Online Publication, January 7, 2009
[doi:10.1271/bbb.80357]
A major collagenase was purified about 96-fold from
a crude enzyme sample of Streptomyces parvulus by
chromatography on Q-Sepharose, Sephacryl S-200, and
butyl-Toyopearl. The purified enzyme showed a relative
molecular mass of approximately 52,000 on SDS–PAGE
and a pH optimum at about 9.0, and was strongly
inhibited by metal-chelating agents. It also cleaved
4-phenylazobenzyloxycarbonyl-Pro-Leu-Gly-Pro-D-Arg
specifically at the Leu-Gly bond, with a Km value of
0.60 mM at pH 9.0 at 37 ꢀC. Based on the amino acid
sequences of the N-terminal region and internal tryptic
peptides, the corresponding gene was cloned. The DNA
sequence of the cloned gene indicated that the enzyme is
produced as an 864-residue precursor protein with a
408-residue prepro sequence followed by a 456-residue
mature enzyme moiety. The enzyme is most homolo-
gous with the collagenase from S. coelicolor, the
identity being 73%, and it is thought to be a member
of the Vibrio collagenase subfamily.
the corresponding enzyme precursors predicted, but
collagenases that have been both enzymatically and
structurally well characterized are rather scarce.
Crude collagenase preparation from Streptomyces
parvulus subsp. citrinus (collagenase N-2, Nitta Zeratin,
Osaka, Japan) is commercially available and has been
used in various biochemical and biomedical studies, but
it has not yet been purified completely and fully
characterized. Hence it is important to purify the
enzyme to homogeneity and to clarify fully its enzy-
matic and structural characteristics. In the present study
we purified a major collagenase from S. parvulus and
investigated various characteristics of it including the
amino acid sequences of the precursor form (864
residues) and the mature enzyme (456 residues) by gene
cloning and nucleotide/protein sequencing.
Materials and Methods
Materials. A bacterial strain of S. parvulus subsp. citrinus (IFO
14435) was obtained from the Fermentation Institute (Tsukuba, Japan).
A crude enzyme sample (collagenase N-2) prepared from the same
bacterial strain was from Nitta Zeratin. It had been prepared from a
culture filtrate of the above bacteria by ammonium sulfate fractiona-
tion, followed by chromatographies on DEAE-Toyopearl and Sepha-
dex G-100.19) Azocasein, benzamidine, bovine pancreatic trypsin
inhibitor, diisopropylfluorophosphate, and Nꢀ-tosyl-phenylalanine
chloromethyl ketone were from Sigma (St. Louis, MO), and bestatin
and leupeptin from Peptide Institute (Osaka, Japan). 4-Phenylazoben-
Key words: amino acid sequence; characterization;
collagenase; purification; Streptomyces par-
vulus
Bacterial collagenases (IUBMB: EC 3.4.24.3) are
metallo-endopeptidases capable of digesting native,
triple-helical collagens, and are classified into two major
subfamilies, M9A including Vibrio collagenases1) and
M9B including Clostridium collagenases2) in the
MEROPS classification. They are useful in tissue
dissociation experiments, including the isolation of
hepatocytes, adipocytes, and other cells for research
purposes, as well as in medical procedures such as the
preparation of vascular endothelial cells for seeding
vascular prostheses and the isolation of pancreatic islets
for transplantation.2) To date bacterial collagenases
have been purified from various species, such as C.
histolyticum,3–6) C. perfringens,7) Pseudomonas marino-
glutinosa,8) V. alginolyticus,9–11) and Streptomyces
sp.,12–16) and partially characterized. The V. parahae-
molticus collagenase was obtained by expression of the
gene in Escherichia coli and partially characterized.17,18)
Moreover many more collagenase genes have been
cloned and sequenced and the amino acid sequences of
.
zyloxycarbonyl(PZ)-Pro-Leu-Gly-Pro-D-Arg-OH 2H2O (PZ-peptide)
was from Nova Biochem (Lauferfingen, Switzerland). Q-Sepharose
and Sephacryl S-200 were from Pharmacia Biotech (Piscataway, NJ),
and butyl-Toyopearl and a TSKgel ODS-120T column were from
Tosoh Co. (Tokyo). Reagents for protein and peptide sequencing were
from Applied Biosystems (Tokyo). Other reagents were of analytical
grade, largely obtained from Wako Pure Chemical (Tokyo).
Determination of enzyme activity. Azocoll was routinely used as a
substrate unless otherwise specified, essentially as described else-
where.20) Azocoll (4 mg) suspended in 890 ml of 50 mM Tris–HCl
buffer, pH 7.5, was mixed with 10 ml of an enzyme solution in the same
buffer, and the mixture was kept at 37 ꢀC for 10–30 min under shaking.
The reaction was stopped by the addition of 100 ml of 30% trichloro-
acetic acid, and the mixture was centrifuged at 40;000 ꢁ g for 10 min.
The absorbance of the supernatant was measured at 540 nm. The blank
sample was treated in the same manner without the enzyme. One unit
*
y
The nucleotide sequence data for the collagenase are available from the DDBJ data bank under accession no. AB429498.
To whom correspondence should be addressed. Fax: +81-49-297-8168; E-mail: kenjitak@ls.toyaku.ac.jp
Abbreviations: PZ, 4-phenylazobenzyloxycarbonyl; PCR, polymerase chain reaction

22
Y. SAKURAI et al.
of enzyme activity was defined as the amount causing an increase of
1.0 in the absorbance at 540 nm per min under the standard assay
conditions. When the volume of enzyme solution exceeded 10 ml, the
volume of the buffer used to suspend azocall was reduced accordingly
to adjust the final volume of the assay mixture to 900 ml.
Determination of N-terminal and partial internal amino acid
sequences. The N-terminal amino acid sequence of the purified protein
was determined using an Applied Biosystems pulse-liquid protein
sequencer model 477A. To determine the partial internal sequences of
the enzyme, the protein was heat-denatured by incubation at 100 ꢀC for
1 h. The denatured protein (about 1 nmol) was then digested at 37 ꢀC
with trypsin (Worthington, Lakewood, NJ) for 24 h in 83 ml of 70 mM
Tris–HCl buffer, pH 8.0, and 10 mM CaCl2. Trypsin (1.2 mg) was
added at the start, and after 3 h and 20 h. The reaction was stopped by
the addition of 10 ml of 1 N HCl. The resulting peptides were separated
by HPLC using a Hitachi (Tokyo) 655A-11 system on a column
(0:46 ꢁ 25 cm) of TSKgel ODS-120T (Tosoh). The peptides were
eluted with linear gradients of acetonitrile (0–50% over 30 min, then to
100% over the next 5 min) in 0.1% trifluoroacetic acid at a flow rate of
0.8 ml/min. The effluent was monitored by measuring the absorbance
at 215 and 280 nm, and the peptide peak fractions were collected and
lyophilized. The peptide fractions were numbered with a suffix, ‘‘T,’’ in
the order of elution. An aliquot of each peptide fraction dissolved in
water was submitted to automated amino acid sequencing using an
Applied Biosystems pulse-liquid protein sequencer model 477A.
Purification procedures. All purification steps were performed at
4 ꢀC. A crude enzyme sample (collagenase N-2) (50 mg) was dissolved
in 5 ml of 5 mM Tris–HCl buffer, pH 8.0, and 4 mM CaCl2, and
dialyzed against the same buffer (4;000 ml ꢁ 2). The dialyzed sample
was loaded on a column of Q-Sepharose (1:6 ꢁ 10 cm) equilibrated
with the same buffer, and then eluted with a linear gradient of 5 mM
(500 ml) to 1 M (500 ml) of Tris–HCl buffer, pH 8.0, and 4 mM CaCl2.
The active fractions were pooled and concentrated using Centricell
(Polyscience, Warrington, PA), and applied to a column (3:0 ꢁ 100
cm) of Sephacryl S-200 equilibrated and eluted with 50 mM Tris–HCl
buffer, pH 8.0, 4 mM CaCl2, and 0.2 M NaCl. The pooled active
fraction was concentrated using Centricell and applied to a column
(2:1 ꢁ 19:5 cm) of butyl-Toyopearl equilibrated with 50 mM Tris–HCl
buffer, pH 8.0, 4 mM CaCl2, and 20% ammonium sulfate. Elution
was performed by a linear gradient from 50 mM Tris–HCl buffer,
pH 8.0, 4 mM CaCl2, and 10% ammonium sulfate (1,000 ml) to
5 mM Tris–HCl buffer, pH 8.0, and 4 mM CaCl2 (1,000 ml). The
active fractions were pooled and stored frozen at ꢂ20 ꢀC. When
necessary, part of the pooled fraction was concentrated using
Centricell, desalted through a PD-10 column, and stored frozen at
ꢂ20 ꢀC. The protein concentration was determined from the absorb-
ance at 280 nm assuming A280 (0.1% solution, 1-cm light path) = 1.0
unless otherwise specified.
Cloning and nucleotide sequencing of the collagenase gene.
Cloning of the collagenase gene was performed essentially as described
previously.22) Based on the amino acid sequences Leu-Arg-Ile-Arg-Ala-
Gln-Glu-Met-Thr, Val-Leu-Val-Ile-Asn-His-Thr-Cys, and Phe-Gly-
Asp-Gly-Thr-Thr-Ser-Ala-Ala-Ala-Asn-Pro-Ala, the degenerated pri-
mers were designed as follows: Fwd-1, CTI (C/A)GI ATI (C/A)GI
GCI CA(G/A) GA(G/A) ATG AC; Fwd-2, GTI CTI GT(G/C) ATC
AA(C/T) CA(C/T) AC(G/C/A/T) T; and Rev, GC IGG (G/A)TT
IGC IGC IGC I(G/C)(A/T) IGT IGT ICC (G/A)TC ICC (G/A)AA.
The first polymerase chain reaction (PCR) was performed using a DNA
extract from S. parvulus as the template and a set of primers, Fwd-2
and Rev. The PCR conditions were as follows: an initial denaturation
of 94 ꢀC for 5 min; 25 cycles of 94 ꢀC for 30 s, 55 ꢀC for 2 min, and
72 ꢀC for 2 min; and a final extension of 72 ꢀC for 7 min. Using the
0.9-kb first PCR product as the template, the second PCR was
performed with a set of primers, Fwd-1 and Rev. The PCR product was
extracted from agarose gel after electrophoresis, labeled using a DIG
Labeling Kit (Boehringer Mannheim, Mannheim, Germany) according
to the manufacturer’s instructions, and used as the specific probe for
Southern and colony hybridizations. Since a 5-kb BamHI fragment was
detected as a major band by Southern hybridization (Fig. 10), the
library for colony hybridization was prepared as follows: the BamHI
fragments around 5-kb were extracted from agarose gel after electro-
phoresis, ligated into the BamHI site of plasmid BlueScript SK+, and
introduced into E. coli DH5 ꢀ. Detection in Southern and colony
hybridizations was performed using a DIG Luminescent Detection
Kit for Nucleic Acid (Boehringer Mannheim). Plasmid DNA was
prepared from a positive clone by the boiling lysis method. DNA
sequencing was performed using a DNA sequencer model 4000L
(LI-COR).
SDS–PAGE. To check the homogeneity of the purified enzyme and
to estimate its molecular mass, SDS–PAGE was performed under non-
reducing conditions in 10% polyacrylamide gel by the method of
Laemmli,21) followed by Coomasie Brilliant Blue staining.
Determination of pH/activity profile. The activity toward azocoll
was determined at 37 ꢀC at pH 4.0–6.0 in 50 mM sodium acetate
buffers, at pH 7.0–9.0 in 50 mM Tris–HCl buffers, at pH 9.5–11.0 in
0.1 M Na2CO3–NaHCO3 buffers, and at pH 11.5–12.0 in 0.1 M
Na2CO3–NaOH buffers. The experiments were performed 4 times.
Inhibition studies. The enzyme (0.008 mg in 10 ml of 7.5 mM Tris–
HCl buffer, pH 8.0, and 4 mM CaCl2) was mixed with the various
reagents dissolved in 2–20 ml of 50 mM Tris–HCl buffer, pH 9.0
(containing a small volume of ethanol where necessary), and kept at
37 ꢀC for 5 min, and then the remaining activity toward azocoll was
determined under standard assay conditions.
Stability studies. The enzyme (0.12 mg in 150 ml of 7.5 mM Tris–HCl
buffer, pH 8.0, and 4 mM CaCl2) was kept at different temperatures for
30 min or 1 h, and then the remaining activity toward azocoll was
determined under standard assay conditions. The experiments were
performed 4 times.
Sequence comparison. Sequence alignment and the construction of
a phylogenetic tree were performed using the program Clustal W.23)
Kinetic studies. The reaction mixtures contained various volumes of
PZ-peptide21) solution (0.34 nmol/ml in 50 mM Tris–HCl buffer,
pH 9.0), enzyme solution (5 ml, 0.004 mg), 50 mM Tris–HCl buffer,
pH 9.0 (80 ml), and distilled water in a total volume of 190 ml. The
mixtures were incubated at 37 ꢀC for 5 min unless otherwise specified.
The reaction was stopped by heating at 100 ꢀC for 2 min, and the digest
was analyzed by HPLC using a Hitachi (Tokyo) 655A-11 system on a
column (0:46 ꢁ 25 cm) of TSKgel ODS-120T (Tosoh, Tokyo). The
peptides were eluted with linear gradients of acetonitrile (0–40% over
0–20 min and 40–100% over 20–25 min) in 0.1% trifluoroacetic acid at
a flow rate of 0.8 ml/min. Under these conditions, PZ-peptide and its
hydrolysis products, PZ-Pro-Leu and Gly-Pro-D-Arg, were eluted at
22.1, 22.8, and 1.9 min respectively (see Fig. 5). The extent of
cleavage was estimated by amino acid analysis after acid hydrolysis
(6 N HCl, 110 ꢀC, 24 h) of the Gly-Pro-D-Arg fraction using an Applied
Biosystems automated derivatizer-analyzer (420A/130A) (Applied
Biosystems, Foster City, CA). Vmax and Km values were estimated from
a Lineweaver-Burk plot using the reaction rates at 10 different
substrate concentrations.
Prediction of the signal sequence cleavage site, secondary
structures, and the hydrophobicity profile. The signal sequence
cleavage site was predicted following Heijne.24) The secondary
structures were predicted by the GOR4 program25,26) and follow-
ing Chou and Fasman.27) A hydrophobicity plot was obtained by
the method of Kyte and Doolittle.28,29)
Results and Discussion
A major collagenase was purified from the crude
enzyme sample (collagenase N-2) from S. parvulus by
successive steps of chromatography on Q-Sepharose,
Sephacryl S200, and butyl-Toyopearl. The results are
shown in Fig. 1 and Table 1. The enzyme activity was
largely eluted from the Q-Sepharose column with the
equilibration buffer at the flow-through position, and
most of the non-enzyme protein was bound to the

Streptomyces parvulus Collagenase: Purification and Characterization
23
Fig. 2. SDS–PAGE of Purified S. parvulus Collagenase.
The S. parvulus collagenase finally purified by chromatography
on butyl-Toyopearl was submitted to SDS–PAGE. Electrophoresis
was performed under non-reducing conditions in a 10% gel,
followed by Coomasie Brilliant Blue staining. M, molecular mass
markers; S, sample.
Fig. 1. Purification of a Collagenase from S. parvulus.
a, Chromatography on a Sepharose S-200 column. The active
fraction, which had passed through the Q-Sepharose column, was
chromatographed on the column (3:0 ꢁ 100 cm) equilibrated and
eluted with 50 mM Tris–HCl buffer, pH 8.0, 4 mM CaCl2, and 0.2 M
NaCl at a flow rate of 28 ml/h. Fractions of 8.3 ml were collected,
and the active fractions under the bar were pooled. b, Chromatog-
raphy on a butyl-Toyopearl column. The active fraction from the
Sepharose S-200 column was applied to a butyl-Toyopearl column
(2:1 ꢁ 19:5 cm) equilibrated with 50 mM Tris–HCl buffer, pH 8.0,
4 mM CaCl2, and 20% ammonium sulfate. Elution was performed by
a linear gradient from 50 mM Tris–HCl buffer, pH 8.0, 4 mM CaCl2,
and 10% ammonium sulfate (1,000 ml) to 5 mM Tris–HCl buffer,
pH 8.0, and 4 mM CaCl2 (1,000 ml) at a flow rate of 50 ml/h.
Fractions of 10 ml were collected, and the active fractions under the
bar were pooled. The broken line in (b) shows the concentration of
ammonium sulfate (AS) in the eluting buffer.
Fig. 3. pH-Dependence of Activity of S. parvulus Collagenase
toward Azocoll.
Activity toward azocoll was measured at various pH values at
37 ꢀC. , sodium acetate buffers; , Tris–HCl buffers; , Na2CO3–
NaHCO3 buffers; , Na2CO3–NaOH buffers. Each point represents
an average of the values obtained in four independent experiments.
The vertical bars represent standard deviations.
Table 1. Purification of a Collagenase from S. parvulus
Total
Specific
activity
(units) (units/mg)
Proteina
(mg)
Purification Yield
(-fold)
activity
Step
(%)
of the major peak. These results might indicate the
presence of a minor collagenase component (see
Fig. 10). The possibility cannot be excluded that a small
amount of the minor enzyme contaminated the purified
enzyme fraction. However, since no additional band was
observed on SDS–PAGE, the amount might have been
extremely small or its molecular mass might have been
indistinguishable from that of the major enzyme.
Crude enzyme
Q-Sepharose
50
488
349
221
270
9.8
97
1
9.9
100
72
3.6
2.1
0.29b
Sephacryl S-200
Butyl-Toyopearl
105
938
10.7
95.7
45
55
A crude enzyme sample (collagenase N-2) (50 mg) was used as the starting
material.
aThe protein was estimated assuming A280 (0.1%, 1 cm) = 1.0, except for
the crude enzyme.
The purified major enzyme was used in the subse-
quent characterization studies. To date, collagenases
from Streptomyces species have been purified from
various sources, but have been characterized only
partially.12–16) Figure 3 shows the pH-dependence of
the activity toward azocoll; the enzyme is highly active
in a pH range of 8–10, with a maximum around pH 9.
This pH optimum is similar to that of 8–9 reported for
Streptomyces sp. C-51,14) but higher than the pH optima
in a range of 6.0–8.0, reported for collagenases from
Streptomyces sp. 138215) and 3B16) and other collage-
nases.4,9,10,18) As shown in Fig. 4, the enzyme was fairly
stable up to 50 ꢀC. Above 50 ꢀC, it became remarkably
unstable, and it was almost completely inactivated by
incubation at 60 ꢀC for 30 min (Fig. 4). The results
obtained by incubation for 1 h were nearly the same,
except that 45% of the original activity was lost at 55 ꢀC.
bThis value was corrected to 0.12 mg using a calculated A280 (0.1%,
1 cm) = 2.363.
column and eluted by the gradient elution (data not
shown). The enzyme in the flow-through fraction from
Q-Sepharose was further purified by chromatography on
Sepharose S-200 and butyl-Toyopearl. The total activity
was somewhat increased after butyl-Toyopearl chroma-
tography (Table 1), presumably due to the removal of an
inhibitory contaminant. Through these steps, the enzyme
was purified from the crude sample about 96-fold at
a yield of 55%. The purified enzyme gave a single
peak on SDS–PAGE, and its relative molecular mass
was estimated to be approximately 52,000 (Fig. 2).
Upon chromatography on Sepharose S-200 and butyl-
Toyopearl, a minor activity peak was observed in front

24
Y. SAKURAI et al.
Fig. 4. Effects of Temperature on the Stability of S. parvulus
Collagenase.
The enzyme was kept at various temperatures for 30 min (— —)
or 1 h (– – – –) at pH 8.0, and then the remaining activity toward
azocoll was determined under standard assay conditions. Each point
represents an average of the values obtained in four independent
experiments. The vertical bars represent standard deviations (solid
line, 30 min; dotted line, 1 h).
Fig. 5. HPLC Pattern of a Digest of PZ-Peptide by S. parvulus
Collagenase.
PZ-peptide (24 nmol) was hydrolyzed by the enzyme (0.004 mg)
for 5 min under standard assay conditions, and the digest was
submitted to HPLC. The small peak (retention time, 18.2 min) eluted
before the PZ-peptide is a non-peptide impurity.
Table 2. Effects of Various Reagents on the Activity of S. parvulus
Collagenase
The purified enzyme was mixed with each reagent in 50 mM Tris–
HCl buffer, pH 9.0, and kept at 37 ꢀC for 5 min, and then the remaining
activity toward azocoll was determined under standard assay con-
ditions.
(data not shown). This is consistent with the fact that the
crude enzyme (collagenase N-2) used in this study has
been reported to cleave bovine tendon collagen (Sigma,
St. Louis, MO) specifically at X-Gly bonds (X,
predominantly Hyp and Ala), but scarcely to hydrolyze
bovine serum albumin or casein.19) These results
indicate that the purified enzyme is a member of the
bacterial collagenase family. The kinetic parameters, Km
and Vmax values, were determined to be 0:60 ꢃ 0:03 mM
and 37 ꢃ 2 mmol/l ꢄ min respectively. The V. parahae-
molyticus and C. histolyticum collagenases have been
reported to have Km values of 1.06 mM and 14.82 mM
respectively toward 2-furanacryloyl-Leu-Gly-Pro-Ala at
pH 8.0 at 25 ꢀC.17) Thus the Km value of the present
enzyme toward PZ-peptide was apparently similar to
that of the V. parahaemolyticus collagenase, although
the substrates used were different.
Concentrationa Inhibition
Inhibitor
(mM)
(%)
EDTA
8-Quinolinol
1
1
89
73
63
53
26
16
10
10
10
7
o-Phenanthroline
2,3-Dimercapto-1-propanol
L-Cysteine
1
1
10
1
Bestatin
Leupeptin
1
Iodoacetic acid
Benzamidine
1
1
Bovine pancreatic trypsin inhibitor
Diisopropylfluorophosphate
Tosyl-L-phenylalanine chloromethyl
ketone
0.1
10
1
0
0
N-terminal sequencing of the purified enzyme yielded
a single sequence of 33 residues. In addition, internal
amino acid sequences of 135 residues in total were also
determined at the protein level by analyzing a tryptic
digest of the purified enzyme. These peptide sequences
were found to agree completely with the protein
sequence deduced from DNA sequencing (Fig. 6),
indicating the homogeneity of the purified enzyme.
The complete amino acid sequence of the prepro form of
the collagenase was deduced by cloning and nucleotide
sequencing of the gene clone obtained from S. parvulus,
as shown in Fig. 6. The total number of residues was
864, and the calculated relative molecular mass was
93,560. The N-terminal and some internal amino acid
sequences determined at the protein level are also
included in Fig. 6. Thus the prepro enzyme was
composed of 864 amino acid residues, including a
408-residue prepro peptide and a 456-residue mature
enzyme. In the prepro peptide, the N-terminal 33
residues and the remaining 375 residues were predicted
to be the signal peptide and the propeptide respectively.
The mature enzyme is produced by cleavage at the
Ser-Ser bond. S. parvulus therefore should produce an
endopeptidase capable of cleaving this peptide bond.
Purification and characterization of this processing
enzyme would be interesting, especially since no
aFinal concentration in the assay mixture.
The collagenase from Streptomyces sp. 3B has been
reported to be less stable, with an optimal temperature of
37 ꢀC.16) Other collagenases have been reported to
be stable up to 40 ꢀC,10,15) or near 50 ꢀC.8) Therefore
the temperature stability of the present enzyme appears
to be higher than those of most other collagenases. The
enzyme was strongly inhibited by metal-chelating
agents, such as EDTA, 8-quinolinol, o-phenanthroline,
and 2,3-dimercapto-1-propanol, but was rather insensi-
tive to inhibitors of serine and cysteine endopeptidases
and aminopeptidases, as shown in Table 2. This in-
dicates that the enzyme is a typical metalloendopepti-
dase, like many other bacterial collagenases containing
an essential Zn2þ. Strong inhibition by EDTA and
o-phenanthroline has also been reported for collagenases
from Streptomyces sp. A8 and 3B.13,16)
As shown in Fig. 5, PZ-peptide, a typical synthetic
peptide substrate for collagenases, was cleaved specif-
ically at the Leu-Gly bond. Under similar conditions,
approximately 70% hydrolysis occurred in 30 min, and
complete hydrolysis in several h at this peptide bond,
and no hydrolysis was observed at other peptide bonds

Streptomyces parvulus Collagenase: Purification and Characterization
25
Fig. 6. Nucleotide and the Deduced Amino Acid Sequences of S. parvulus Collagenase.
The initiation codon was predicted to be GTG at position 271, as in the case of the S. coelicolor ortholog. The deduced N-terminal amino acid
sequence and the amino acid sequences of the tryptic peptides determined are shown in one-letter code below the nucleotide sequence, and are
underlined. N-Term, the N-terminal sequence of the enzyme; T, tryptic peptide. Numbers stand for peptide peak numbers on HPLC. The peptide
number is shown before each peptide, except for T-6, T-8, and T-19 which are shown after the peptide. X, residue not unambiguously identified.
An asterisk indicates a stop codon.
endopeptidase with such a specificity is known. The
calculated relative molecular mass and pI value of the
mature enzyme were 49,368 and 5.21 respectively. This
relative molecular mass is roughly consistent with that
(approximately 52,000) for the purified enzyme as
determined by SDS–PAGE. The relative molecular
masses reported for other collagenases from Streptomyces
sp. are highly variable: 35,000,12) 30,000–40,000,15)
75,000,13) 100,000 (type I) and 90,000–110,000 (type
II),14) and 116,000 (type I) and 97,000 (type II).16) Thus
the molecular mass of the present enzyme was different
from any of these collagenases. These results indicate
that there are marked variations in structural and
enzymatic properties even among the Streptomyces
collagenases.
The enzyme had a peptidase domain with a metal-
lopeptidase active-site motif, His-Glu-X-X-His, at posi-
tions 527–531 (prepro protein numbering is used
throughout), including the putative catalytic Glu528
and two metal ligands, His527 and His531, and a third
putative metal ligand, Glu556. The enzyme is thought to
have a PKD (polycystic kidney disease) domain30)
(residues 678–737) and a PPC (bacterial pre-peptidase
C-terminal) domain31) (residues 781–851) in the C-
terminal region as judged by the domain structure of the
S. coelicolor collagenase (CAA16449).

26
Y. SAKURAI et al.
–
–
Fig. 7. Amino Acid Sequence Comparison of S. parvulus Collagenase with Homologous Collagenases.
The amino acid sequences of the prepro forms of the collagenases from S. parvulus (this study), S. coelicolor (CAA16449), M. xanthus
(YP 634336), and V. alginolyticus (EAS77672) are aligned. The active site residues are indicated by ꢅ, and the N-terminal residues of the
S. parvulus and V. alginolyticus enzymes are indicated by # and + respectively. The residues common to the S. parvulus collagenase are shaded.
Dashes indicate deletions.
The amino acid sequences of a number of bacterial
collagenases have been deduced from the nucleotide
sequences of their genes, but only the sequence of the
S. coelicolor collagenase has been determined among
Streptomyces sp. Among these, the collagenases from
S. coelicolor, Micrococcus xanthus (YP 634336), and
V. arginolyticus11) are among the most homologous to
the present enzyme, and their sequences are compared in
Fig. 7. Among the bacterial collagenases, the N-terminal
sequences of the mature forms have been determined
for the present enzyme (this study) and for the
collagenases from V. alginolyticus,11) V. anguillarum,32)
C. perfringens,7) and C. histolyticum.33) There is a
marked difference in the number of residues from the
N-terminus to the first His residue of the active-site
motif. These are 118, 147, 412, and 416 residues for the
present enzyme, V. anguillarum, V. alginolyticus, and
C. perfringens enzymes respectively, and 414 and 415
residues for the two enzymes from C. histolyticum.
Since the 456-residue enzyme is the only active form
identified so far for the present enzyme, we assume that
this form is the mature enzyme. The cleavage sites were
found to be Ser-Thr, His-Ala, and Arg-Ala bonds in the
collagenases from V. alginolyticus, V. anguillarum, and
C. perfringens respectively, and Ser-Ile, Arg-Ala, and
Arg-Val bonds in the collagenases G, A, and H from
C. histolyticum respectively. Thus, the cleavage sites
of Ser-Thr and Ser-Ile in the collagenases of V. algino-
lyticus and C. histolyticum (collagenase G) respectively
are similar to that of the present enzyme (viz., Ser-Ser).
On the other hand, the cleavage sites in the other
collagenases, His-Ala, Arg-Ala, Arg-Ala, and Arg-Val,
have a basic residue at the P1 position and a small
aliphatic residue at the P10 position. No marked
sequence homology was to be seen, however, beyond
the P1 and P10 positions. Hence a few types of
processing enzymes of different specificities might be
present. The differences in processing site, and hence in
the size of the pro part, are thought to be due to
differences in the mode of processing of proenzymes by

Streptomyces parvulus Collagenase: Purification and Characterization
27
Fig. 8. Phylogenetic Tree Based on the Amino Acid Sequences of the
Prepro Forms of Typical Bacterial Collagenases.
The accession numbers of the amino acid sequences of the
collagenases used are as follows: S. parvulus, AB429498; S. coeli-
color, CAA16449; M. xanthus, YP 634336; P. tunicata, EAR30787;
S. woodyi, EAV38041; V. alginolyticus, EAS77672; V. parahaemo-
lyticus, Q9AMB9; C. histolyticum, BAA86030; and C. perfringens,
BAB79879. The lengths of branches indicate the evolutionary
distances (substitutions/site) between the proteins. The bootstrap
value of each node is shown at that node.
Fig. 10. Southern-Blot Hybridization Analysis of Genomic DNA
from S. parvulus.
The genomic DNA was digested with restriction enzymes, as
follows: lane 1, SacI; lane 2, BamHI.
to be 73%, 49%, and 38% with the collagenases from
S. coelicolor, M. xanthus, and V. alginolyticus respec-
tively. On the other hand, the identity with the
C. histolyticum collagenase was 26%. Thus the present
collagenase was most closely related to the collagenase
from S. ceolicolor. Moreover, the S. coelicolor prepro
enzyme (865 residues) shows significant similarity to the
present enzyme in the size and sequence of the prepro
moiety as well, as shown in Fig. 7. The N-terminal
sequences of both prepro proteins start at the same
position. These results indicate that the present enzyme
is a member of the Vibrio collagenase subfamily
(MEROPS M9A) rather than the Clostridium collage-
nase subfamily (MEROPS M9B). This was further
confirmed by construction of a phylogenetic tree based
on the amino acid sequences of the prepro forms of
some typical bacterial collagenases, as shown in Fig. 8.
The active-site motif and the third metal ligand Glu are
all conserved in these enzymes.
The secondary structures of the prepro forms of the
present enzyme and the S. coelicolor enzyme as pre-
dicted by the GOR4 program are compared in Fig. 9a
and b. They are very much alike, but show some
differences, especially in the N-terminal regions (posi-
tions 420–450) of the mature enzyme forms. The
secondary structure contents of the mature forms of
both enzymes were estimated to be as follows: present
enzyme, ꢀ-helix, 20.2%, and ꢁ-strand, 24.3%; the
S. coelicolor enzyme, ꢀ-helix, 12.4%, and ꢁ-strand,
34.9%. Similar results were obtained by the Chou-
Fasman method (data not shown), and the secondary
structure contents were estimated to be as follows:
present enzyme, ꢀ-helix, 17.8%, and ꢁ-strand, 35.3%;
the S. coelicolor enzyme, ꢀ-helix, 11.3%, and ꢁ-strand,
35.9%. Thus the present enzyme appears to be richer in
ꢀ-helix than the S. coelicolor enzyme. Figure 9c and d
shows a comparison of the hydropathy plots of the two
enzymes. Again they are very similar, but show some
differences. In the mature forms, significant differences
are observed at positions 420–480 and 650–670. Hence
both enzymes presumably possess very similar tertiary
structures, but with some local differences. So far,
however, no crystal structure has been solved for
bacterial collagenases. Thus elucidation of the crystal
Fig. 9. Secondary Structure and Hydropathy Profile Comparison
between the Prepro Forms of S. parvulus and S. coelicolor Colla-
genases.
The secondary structures of the S. parvulus collagenase (a) and
S. coelicolor collagenase (b) were predicted by the GOR 4 program.
Long vertical bar, ꢀ-helix; short vertical bar, ꢁ-strand. The hydro-
pathy plots of the S. parvulus collagenase (c) and S. coelicolor
collagenase (d) were obtained by the method of Kyte and Doolittle.
The vertical dotted line indicates the N-terminal position (residue
409) of the mature S. parvulus collagenase and the corresponding
position in the S. coelicolor collagenase.
specific processing enzymes in the respective bacteria.
To our knowledge, the mechanism of activation of the
proenzymes of bacterial collagenases has not been
studied so far. It would be interesting to elucidate the
activation mechanism of the present enzyme, including
characterization of the processing enzyme.
When the sequence of the present enzyme in mature
form was compared with the corresponding parts of
other enzymes, the sequence identities were calculated

28
Y. SAKURAI et al.
Structural gene and complete amino acid sequence of Vibrio
alginolyticus collagenase. Biochem. J., 281, 703–708 (1992).
12) Rippon, J. W., Extracellular collagenase produced by Strepto-
myces madurae. Biochim. Biophys. Acta, 159, 147–152 (1968).
13) Chakraborty, R., and Chandra, A. L., Purification and character-
ization of a streptomycete collagenase. J. Appl. Bacteriol., 61,
331–337 (1986).
structures of the present enzyme and others is highly
desirable for further understanding of the structure/
function relationships and design of useful inhibitors of
bacterial collagenases.
The results of Southern-blot hybridization analysis of
the genomic DNA from S. parvulus are shown in
Fig. 10. Although the DNA region used in Southern
hybridizaion, nucleotides 1531–2423 in Fig. 6, involves
no restriction site of the enzymes used for digestion of
the genomic DNA, two or three hybridized bands are
present in every lane of Fig. 10. This suggests the
existence of another gene homologous to the cloned one.
This gene might correspond to the minor enzyme that
was not characterized, partly due to the paucity of the
enzyme sample. Although the action of the minor
enzyme on PZ-peptide was not analyzed, it is assumed
to have the same specificity as the major enzyme, as
judged by the specificity of the crude enzyme sample
toward collagen.19) Further studies are necessary to
identify the minor enzyme.
14) Endo, A., Murakawa, S., Shimizu, H., and Shiraishi, Y.,
Purification and properties of collagenase from a Streptomyces
species. J. Biochem., 102, 163–170 (1987).
15) Ivanko, O. V., and Varbanets’, L. D., Purification and physico-
chemical properties of Streptomyces sp. 1349 collagenase and
Streptomyces sp. 1382 keratinase. Mikrobiol. Z., 66, 11–24
(2004).
16) Petrova, D., Derekova, A., and Vlahov, S., Purification and
properties of individual collagenases from Streptomyces sp.
strain 3B. Folia Microbiol. (Prague), 51, 93–98 (2006).
17) Yu, M. S., and Lee, C. Y., Expression and characterization of
the prtV gene encoding a collagenase from Vibrio parahaemo-
lyticus in Escherichia coli. Microbiology, 145, 143–150 (1999).
18) Luan, X., Chen, J., Zhang, X. H., Li, Y., and Hu, G., Expression
and characterization of a metalloprotease from a Vibrio para-
haemolyticus isolate. Can. J. Microbiol., 53, 1168–1173 (2007).
19) Wada, K., Nishio, T., and Yasugi, S., Japan Kokai Tokkyo
Koho, 61-289885 (Dec. 19, 1986).
Acknowledgments
20) Wuensch, E., and Heidrich, H.-G., Zur quantitativen Bestim-
¨
We are grateful to Nitta Zeratin KK (Osaka, Japan)
for the kind gift of crude collagenase (collagenase N-2).
We also thank to Dr. S.B.P. Athauda for helpful
technical advice. This work was supported in part by
Grants-in-Aid for Scientific Research from the Ministry
of Education, Culture, Sports, Science, and Technology
of Japan.
mung der Kollagenase. Hoppe Seylers Z. Physiol. Chem., 333,
149–151 (1963).
21) Laemmli, U. K., Cleavage of structural proteins during the
assembly of the head of bacteriophage T4. Nature, 227, 680–
685 (1970).
22) Sambrook, J., and Russell, D. W., ‘‘Molecular Cloning, a
Laboratory Manual’’ 3rd ed., Vol. 1, Cold Spring Harbor
Laboratory Press, Cold Spring Harbor (2001).
23) Thompson, J. D., Higgins, D. G., and Gibson, T. J., CLUSTAL
W: improving the sensitivity of progressive multiple sequence
alignment through sequence weighting, position-specific gap
penalties and weight matrix choice. Nucleic Acids Res., 22,
4673–4680 (1994).
References
1) Fukushima, J., and Okuda, K., Vibrio collagenase. In ‘‘Hand-
book of Proteolytic Enzymes’’ 2nd ed., Vol. 1, eds. Barrett, A. J.,
Rawlings, N. D., and Woessner, J. F., Elsevier Academic Press,
London, pp. 414–415 (2004).
24) von Heijne, G., A new method for predicting signal sequence
cleavage sites. Nucleic Acids Res., 14, 4683–4690 (1986).
25) Garnier, J., Gibrat, J. F., and Robson, B., GOR method for
predicting protein secondary structure from amino acid se-
quence. Methods Enzymol., 266, 540–553 (1996).
2) Van Wart, H. E., Clostridium collagenases. In ‘‘Handbook of
Proteolytic Enzymes’’ 2nd ed., Vol. 1, eds. Barrett, A. J.,
Rawlings, N. D., and Woessner, J. F., Elsevier Academic Press,
London, pp. 416–419 (2004).
3) Yoshida, E., and Noda, H., Isolation and characterization of
collagenases I and II from Clostridium histolyticum. Biochim.
Biophys. Acta, 105, 562–574 (1965).
26) Combet, C., Blanchet, C., Geourjon, C., and Deleage, G.,
NPS@: Network protein sequence analysis. Trends Biochem.
Sci., 25, 147–150 (2000).
4) Kono, T., Purification and partial characterization of collage-
nolytic enzymes from Clostridium histolyticum. Biochemistry,
7, 1106–1114 (1968).
27) Chou, P. Y., and Fasman, G. D., Prediction of the secondary
structure of proteins from their amino acid sequence. Adv.
Enzymol. Relat. Areas Mol. Biol., 47, 45–148 (1978).
28) Kyte, J., and Doolittle, R. F., A simple method for displaying
the hydropathic character of a protein. J. Mol. Biol., 157, 105–
132 (1982).
5) Lwebuga-Mukasa, J. S., Harper, E., and Taylor, P., Collagenase
enzymes from Clostridium: characterization of individual
enzymes. Biochemistry, 15, 4736–4741 (1976).
6) Bond, M. D., and Van Wart, H. E., Characterization of the
individual collagenases from Clostridium histolyticum. Bio-
chemistry, 23, 3085–3091 (1984).
29) Gasteiger, E., Hoogland, C., Gattiker, A., Duvaud, S., Wilkins,
M. R., Appel, R. D., and Bairoch, A., Protein identification and
analysis tools on the ExPASy server. In ‘‘The Proteomics
Protocols Handbook,’’ ed. Walker, J. M., Humana Press, New
York, pp. 571–607 (2005).
7) Matsushita, O., Yoshihara, K., Katayama, S., Minami, J., and
Okabe, A., Purification and characterization of Clostridium
perfringens 120-kilodalton collagenase and nucleotide sequence
of the corresponding gene. J. Bacteriol., 176, 149–156 (1994).
8) Hanada, K., Mizutani, T., Yamagishi, M., Tsuji, H., Misaki, T.,
and Sawada, J., The isolation of collagenase and its enzymo-
logical and physico-chemical properties. Agric. Biol. Chem., 37,
1771–1781 (1973).
30) Hughes, J., Ward, C. J., Peral, B., Aspinwall, R., Clark, K., San
´
Millan, J. L., Gamble, V., and Harris, P. C., The polycystic
kidney disease 1 (PKD1) gene encodes a novel protein with
multiple cell recognition domains. Nat. Genet., 10, 151–160
(1995).
31) Yeats, C., Bentley, S., and Bateman, A., New knowledge from
old: in silico discovery of novel protein domains in Strepto-
myces coelicolor. BMC Microbiol., 3, 3–23 (2003).
32) Milton, D. L., Norqvist, A., and Wolf-Watz, H., Cloning of a
metalloprotease gene involved in the virulence mechanism of
Vibrio anguillarum. J. Bacteriol., 174, 7235–7244 (1992).
33) Matsushita, O., Jung, C. M., Katayama, S., Minami, J.,
Takahashi, Y., and Okabe, A., Gene duplication and multiplicity
of collagenases in Clostridium histolyticum. J. Bacteriol., 181,
923–933 (1999).
9) Lecroisey, A., Keil-Dlouha, V., Woods, D. R., Perrin, D., and
Keil, B., Purification, stability and inhibition of the collagenase
from Achromobacter iophagus. FEBS Lett., 59, 167–172 (1975).
10) Tong, N. T., Tsugita, A., and Keil-Dlouha, V., Purification and
characterization of two high-molecular-mass forms of Achro-
mobacter collagenase. Biochim. Biophys. Acta, 874, 296–304
(1986).
11) Takeuchi, H., Shibano, Y., Morihara, K., Fukushima, J., Inami,
S., Keil, B., Gilles, A. M., Kawamoto, S., and Okuda, K.,