Báo cáo khoa học: Membrane type-1 matrix metalloprotease-independent activation of pro-matrix metalloprotease-2 by proprotein convertases docx

Membrane type-1 matrix metalloprotease-independent
activation of pro-matrix metalloprotease-2 by proprotein
convertases
Bon-Hun Koo, Hee-Hyun Kim, Michael Y. Park, Ok-Hee Jeon and Doo-Sik Kim
Department of Biochemistry, College of Life Science and Biotechnology, Yonsei University, Seoul, South Korea
Introduction
Matrix metalloproteases (MMPs) constitute a family
of 23 zinc-dependent endopeptidases. They are
involved in many biological processes and diseases,
and catalyze the proteolysis of extracellular and
non-extracellular matrix molecules [1]. For example,
MMP-2 is associated closely with organ growth,
endometrial cycling, wound healing, bone remodeling,
tumor invasion and metastasis [2]. Its functions are
executed through the degradation of components of
the basement membrane, including type IV collagen,
fibronectin, elastin, laminin, aggrecan and fibrillin
[1,3].
Keywords
furin; membrane type-1 matrix
metalloprotease (MT1-MMP); pro-matrix
metalloprotease-2 (pro-MMP-2); pro-MMP-2
activation; proprotein convertases
Correspondence
B H. Koo, Department of Biochemistry,
College of Life Science and Biotechnology,
Yonsei University, 134 Sinchon-Dong
Seodaemun-Gu, Seoul 120-749, South Korea
Fax: 82 2 312 6027
Tel: 82 2 313 2878
E-mail:

D S. Kim, Department of Biochemistry,
College of Life Science and Biotechnology,
Yonsei University, 134 Sinchon-Dong
Seodaemun-Gu, Seoul 120-749, South Korea
Fax: 82 2 312 6027
Tel: 82 2 2123 2700
E-mail:
(Received 14 July 2009, revised 26 August
2009, accepted 28 August 2009)
doi:10.1111/j.1742-4658.2009.07335.x
Matrix metalloprotease-2 is implicated in many biological processes and
degrades extracellular and non-extracellular matrix molecules. Matrix
metalloprotease-2 maintains a latent state through a cysteine–zinc ion
pairing which, when disrupted, results in full enzyme activation. This
pairing can be disrupted by a conformational change or cleavage within
the propeptide. The best known activation mechanism for pro-matrix
metalloprotease-2 occurs via cleavage of the propeptide by membrane
type-1 matrix metalloprotease. However, significant residual activation of
pro-matrix metalloprotease-2 is seen in membrane type-1 matrix meta-
lloprotease knockout mice and in fibroblasts treated with metalloprotease
inhibitors. These findings indicate the presence of a membrane type-1
matrix metalloprotease-independent activation mechanism for pro-matrix
metalloprotease-2 in vivo, which prompted us to explore an alternative
activation mechanism for pro-matrix metalloprotese-2. In this study, we
demonstrate membrane type-1 matrix metalloprotease-independent propep-
tide processing of matrix metalloprotease-2 in HEK293F and various
tumor cell lines, and show that proprotein convertases can mediate the
processing intracellularly as well as extracellularly. Furthermore, processed
matrix metalloprotease-2 exhibits enzymatic activity that is enhanced by
intermolecular autolytic cleavage. Thus, our experimental data, taken

together with the broad expression of proprotein convertases, suggest that
the proprotein convertase-mediated processing may be a general activation
mechanism for pro-matrix metalloprotease-2 in vivo.
Abbreviations
Con A, concanavalin A; dec-RVKR-cmk, dec-Arg-Val-Lys-Arg-chloromethyl ketone; FITC, fluorescein isothiocyanate; GAPDH, glyceraldehyde-3-
phosphate dehydrogenase; MMP, matrix metalloprotease; MT1-MMP, membrane type-1 matrix metalloprotease; MTT, 3-(4,5-dimethylthiazol-
2-yl)-2,5-diphenyl-tetrazolium bromide; PC, proprotein convertase; pro-MMP-2, pro-matrix metalloprotease-2; TGN, trans-Golgi network; TIMP,
tissue inhibitor of metalloprotease.
FEBS Journal 276 (2009) 6271–6284 ª 2009 The Authors Journal compilation ª 2009 FEBS 6271
Because of its potential for tissue destruction,
MMP-2 activity is regulated at multiple points, such
as gene expression, compartmentalization, zymogen
activation and enzyme inactivation by extracellular
inhibitors [e.g. tissue inhibitors of metalloproteases
(TIMPs)] [1]. Like other MMPs, pro-MMP-2 main-
tains a latent state via an interaction between a
thiol group of a propeptide cysteine residue and the
catalytic zinc ion in the active site [4]. Disruption of
this cysteine–zinc ion pairing, such as through confor-
mational changes [5] or proteolysis within the propep-
tide [e.g. by plasmin, thrombin or membrane
type-MMPs (MT-MMPs)] [6–12], is required for the
activation of the latent enzyme. The most studied acti-
vation mechanism for pro-MMP-2 is cleavage of the
propeptide by MT1-MMP, which requires cooperative
activity between MT1-MMP and TIMP-2 [7,13–15].
However, residual activation of pro-MMP-2 is
observed in MT1-MMP knockout mice, even though
the activation is reduced significantly [16,17]. Thus,
these data suggest that an MT1-MMP-independent

activation mechanism for pro-MMP-2 may also exist
in vivo.
Proprotein convertases (PCs), a family of Ca
2+
-
dependent serine proteases of which furin is the most
ubiquitous, have a major role in molecular maturation
[18–20]. Most PCs reside within the trans-Golgi net-
work (TGN), but some are present at the cell surface
via a transmembrane domain (e.g. furin) [21,22] or the
extracellular matrix (e.g. PACE4 and PC5A) [23]. Like
other PCs, furin cleaves its substrates immediately
downstream of the consensus sequence Arg-Xaa-
Arg ⁄ Lys-Arg (where Xaa is any amino acid) in the
TGN [18,19,24] or at the cell surface [25]. The signifi-
cance of furin in many biological processes is attri-
buted to its widespread expression [26] and the
developmental lethality of furin knockout mice [27].
Furthermore, elevated expression of PCs is frequently
observed in various human cancers and tumor cell
lines, implicating the importance of PCs in tumor
progression [28].
The presence of activated MMP-2 in MT1-MMP
knockout mice prompted us to explore an MT1-
MMP-independent activation mechanism for pro-
MMP-2. In this study, we demonstrate MT1-MMP-
independent propeptide processing of MMP-2 in
HEK293F and various tumor cell lines, where PCs
could mediate this processing. Furthermore, PC-pro-
cessed MMP-2 showed enzymatic activity, which was

enhanced following intermolecular autolytic cleavage.
Thus, these results strongly suggest a potential role
of PCs in pro-MMP-2 activation in PC-expressing
cells.
Results
MT-MMP-independent processing of pro-MMP-2
Previous studies have shown that MT1-MMP is a
major activator of pro-MMP-2 [29–31]. Furthermore,
TIMP-2 and integrin avb3 have been shown to play
important roles in MT1-MMP-mediated pro-MMP-2
activation [13–15,32]. Thus, prior to investigating the
role of MT1-MMP in pro-MMP-2 activation, expres-
sion of MT1-MMP, TIMP-2 and integrin avb3 was
characterized in COS-1, HCT116, HEK293F, MCF-7,
MDAH 2774, K-562, NCI-H460 and Hep G2 cells.
MT1-MMP protein was undetectable by immunoblot-
ting with anti-MT1-MMP rabbit polyclonal IgG in all
cell types (data not shown), whereas semi-quantitative
RT-PCR detected its mRNA in HCT116, K-562, NCI-
H460 and Hep G2 cells (Fig. 1A). Because other
MT-MMPs have also been demonstrated to process
pro-MMP-2 [8–11], their expression was also examined
in these cells. Although protein expression was not
determined, different levels of MT-MMP mRNA were
observed in the cells (Fig. 1A). Immunoblotting with
mouse anti-TIMP-2 IgG
2a
showed that COS-1 cells
possessed the highest levels of TIMP-2, whereas the
expression of this inhibitor was absent or much lower

in HCT116, HEK293F, MCF-7, MDAH 2774, K-562,
NCI-H460 and Hep G2 cells (Fig. 1B). Variable cell
surface expression of integrin avb3 was assessed by
flow cytometric analysis (Fig. 1C), although its expres-
sion was undetectable in HCT116, K-562 and Hep G2
cells (data not shown). These data demonstrate that
different cell types have different cellular components
for MT-MMP-mediated pro-MMP-2 activation.
To test the MT-MMP-independent processing of
pro-MMP-2, cells were incubated with metalloprotease
inhibitors (e.g. GM6001 and TIMP-2), and the condi-
tioned medium was analyzed by zymography. TIMP-2
is a strong inhibitor of MT1-MMP at high concentra-
tions [33], although it activates pro-MMP-2 with MT1-
MMP at low concentrations [34]. Thus, we used a high
concentration of TIMP-2 (5 lgÆmL
–1
) to completely
inhibit the enzymatic activities of MT1-MMP and
other TIMP-2-sensitive MMPs. The data showed that
the processed MMP-2 was clearly seen in the condi-
tioned medium of HEK293F, MDAH 2774 and MCF-
7 cells (Fig. 2A). However, processing was not affected
by incubation with the inhibitors in these cells
(Fig. 2A). Likewise, processed MMP-2 was observed
in the conditioned medium of K-562, NCI-H460 and
Hep G2 cells, and their incubation with the metallo-
protease inhibitors did not result in the prevention of
processing even though pro-MMP-2 accumulated
Activation of pro-matrix metalloprotease-2 by proprotein convertases B H. Koo et al.

6272 FEBS Journal 276 (2009) 6271–6284 ª 2009 The Authors Journal compilation ª 2009 FEBS
slightly in the conditioned medium of K-562 cells trea-
ted with the inhibitors (Fig. 2B). As a control, the con-
ditioned medium of concanavalin A (Con A)-treated
HT-1080 cells showed that MT1-MMP-mediated pro-
cessing of pro-MMP-2 was completely inhibited in the
presence of the metalloprotease inhibitors (Fig. 2C).
Unlike HT-1080 cells, endogenous MMP-2 expression
in all the cell lines used was low, so that lytic bands
could hardly be seen in the zymograms without con-
centrating the conditioned medium (Fig. 2A and not
shown in Fig. 2B).
MT-MMPs cleave pro-MMP-2 at the Asn66–Leu
[7–10] or Asn109–Tyr peptide bond [11]. Therefore, to
further investigate the role of MT-MMP in pro-MMP-
2 activation, pro-MMP-2 mutants incapable of cleav-
age by MT1-MMP and ⁄ or autolysis were generated
(N66I ⁄ L67V, N109I ⁄ Y110F and N66I ⁄ L67V ⁄ N109I ⁄
Y110F) (Fig. 3). In agreement with the data using
metalloprotease inhibitors, transient transfection of
these mutants into HEK293F, MCF-7 and
MDAH 2774 cells did not prevent cleavage, as pro-
cessing of the mutants persisted (Fig. 4A). By contrast,
processing was increased significantly in the cells
expressing the pro-MMP-2 N66I ⁄ L67V and
N66I ⁄ L67V ⁄ N109I ⁄ Y110F mutants by unknown
mechanisms (Fig. 4A). As a control, COS-1 cells
expressing the mutants with MT1-MMP showed that
the pro-MMP-2 N66I ⁄ L67V and N66I ⁄ L67V ⁄ N109I ⁄
Y110F mutants were not cleaved by MT1-MMP, and

that the pro-MMP-2 N109I⁄ Y110F mutant was not
processed autocatalytically following MT1-MMP
cleavage (Fig. 4B). Overall, these results suggest the
presence of an MT-MMP-independent activation
mechanism for pro-MMP-2 in some cell types.
PCs mediate propeptide processing of MMP-2
To further explore the MT-MMP-independent activa-
tion mechanism of pro-MMP-2, cells were incubated
with various inhibitors targeted towards serine (e.g.
aprotinin, chymostatin, and leupeptin) and aspartyl
(e.g. pepstatin) proteases, and the conditioned medium
was analyzed by zymography. As shown in Fig. 5A, this
processing was not affected by effective concentrations
of these protease inhibitors in HEK293F, MCF-7 and
MDAH 2774 cells. These data suggest that pro-MMP-2
processing may be mediated by proteases other than
serine and aspartyl proteases in these cells.
A
C
B
Fig. 1. Cellular expression of MT-MMPs,
TIMP-2 and integrin avb3. (A) Analysis of
MT-MMP mRNA in various cell lines. The
data represent semi-quantitative RT-PCR
analysis of MT-MMP mRNA in HCT116,
HEK293F, MCF-7, MDAH 2774, K-562,
NCI-H460 and Hep G2 cells. PCR products
(25 cycles) were resolved by agarose gel
electrophoresis and visualized by ethidium
bromide staining. Arrowheads indicate PCR

products of MT-MMPs or GAPDH. (B) Wes-
tern blotting of conditioned medium with
mouse anti-TIMP-2 IgG
2a
. The conditioned
medium from the cells on 12-well plates
was concentrated prior to SDS-PAGE. Arrow
indicates TIMP-2 protein. (C) Flow
cytometric analysis for cell surface
expression of integrin avb3. The percentage
changes in fluorescence intensity by the
presence of integrin avb3 are shown.
A sample lacking primary antibody was used
as a control (n = 3 representative experi-
ments).
B H. Koo et al. Activation of pro-matrix metalloprotease-2 by proprotein convertases
FEBS Journal 276 (2009) 6271–6284 ª 2009 The Authors Journal compilation ª 2009 FEBS 6273
A previous study has shown the furin processing
of pro-MMP-2 through cleavage of a consensus PC
cleavage site in the propeptide of MMP-2 (Arg98-Lys-
Pro-Arg101) [35]. We initially examined the enzymatic
activities of PCs in various cells using an enzymatic
activity assay. The results showed PC activities in the
cell lysates from HEK293F, MCF-7, MDAH 2774,
K-562, NCI-H460 and Hep G2, whereas negligible
amounts of PC activity were detected in COS-1 and
HCT116 cells (Fig. 5B). Next, we investigated whether
PCs play a crucial role in pro-MMP-2 processing by
incubating cells with increasing concentrations of the
membrane-permeable PC inhibitor dec-Arg-Val-Lys-

Arg-chloromethyl ketone (dec-RVKR-cmk). As shown
in Fig. 5C, D, pro-MMP-2 processing was reduced
significantly in the PC inhibitor-treated cells, suggesting
that PCs are major processing enzymes for pro-MMP-2
in these cell types. The cells did not undergo apoptosis
at the doses of inhibitor used, which was confirmed
by 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium
bromide (MTT) assay (data not shown). Next, experi-
ments were performed to identify which PCs associated
with the constitutive secretory pathway mediated this
processing. When pro-MMP-2 was expressed in COS-1
and HCT116 cells, mainly the intact MMP-2 zymogen
was detectable, whereas the processed form was negli-
gible in the conditioned medium (Fig. 5E). However,
processed MMP-2 was seen in the conditioned medium
of cells co-expressing pro-MMP-2 and PCs (Fig. 5E).
Parallel experiments performed in HEK293F and
MCF-7 cells showed endogenous processing of pro-
MMP-2, which was enhanced following the transfection
of PCs (Fig. 5E). The substitution of Arg101 with Ala
decreased PC processing significantly, as well as the
endogenous processing of pro-MMP-2. However, par-
tially and fully processed MMP-2 was marginally detect-
able in the conditioned medium of the cells expressing
the mutant (Fig. 5E). As such, PCs may be the major
processing enzymes for pro-MMP-2 in these cells.
PACE4- and PC5A-expressing cells exhibit intracel-
lular and extracellular processing of pro-MMP-2
Intracellular processing of pro-MMP-2 by furin has
been shown previously in COS-1 cells [35]. Despite

A
B
C
Fig. 2. MT-MMP-independent processing of pro-MMP-2 in various
cell types. (A) Effect of metalloprotease inhibitors on the propeptide
processing of MMP-2 in HEK293F, MDAH 2774 and MCF-7 cells.
Cells were treated with GM6001 (20 l
M) or TIMP-2 (5 lgÆmL
–1
)in
500 lL of 293 SFM-II medium on 24-well plates. After 14 h of incu-
bation, MMP-2 was captured from 500 lL of the conditioned med-
ium using gelatin–Sepharose according to the manufacturer’s
recommendations (Amersham Biosciences) and eluted in 30 lLof
SDS sample buffer, and the eluted sample was analyzed by zymog-
raphy; 10 lL of the conditioned medium (pre-column sample,
unconcentrated) was loaded on the first lane. Arrows indicate
pro- and processed MMP-2 (MMP-2). (B) Effect of metalloprotease
inhibitors on the processing of pro-MMP-2 in K-562, NCI-H460 and
Hep G2 cells. Cells were treated as described above. MMP-2 was
concentrated from 125 lL of the conditioned medium using gela-
tin–Sepharose and the eluted sample was used for zymographic
analysis. Conditioned medium from HT-1080 cells was used as a
positive control for pro-MMP-2. (C) Validation of the metallopro-
tease inhibitors to prevent MT1-MMP-mediated processing of
pro-MMP-2. HT-1080 cells were treated with Con A (50 lgÆmL
–1
)in
the presence and absence of the metalloprotease inhibitors for
14 h (n = 3 representative experiments).

Fig. 3. Structures of pro-MMP-2 mutants.
Activation of pro-matrix metalloprotease-2 by proprotein convertases B H. Koo et al.
6274 FEBS Journal 276 (2009) 6271–6284 ª 2009 The Authors Journal compilation ª 2009 FEBS
this, the cellular localization of pro-MMP-2 processing
by other PCs, including PACE4 and PC5A, remains
unknown. Therefore, whether PACE4- or PC5A-medi-
ated processing of pro-MMP-2 occurs intracellularly
was investigated using a variety of approaches. First,
we examined the enzymatic activities of PCs in condi-
tioned medium and cell lysates using the enzymatic
activity assay. The results showed PC activities in the
conditioned medium and cell lysates from COS-1 cells
expressing PCs (Fig. 6A). We further verified whether
the activitie cell lysates with the PC inhibitor (1 s
shown reflected exclusively the presence of PCs by
treating the conditioned medium and 00 lm). The
enzymatic activities were almost completely inhibited
(more than 98%) in the PC inhibitor-treated samples
(data not shown), excluding the enzymatic activities of
other proteases in the conditioned medium and cell
lysates. Next, pro-MMP-2-expressing COS-1 cells were
co-cultured with cells expressing furin, PACE4, PC5A
or MT1-MMP, and zymographic analysis of the condi-
tioned medium was performed. Although soluble furin
was detected in the conditioned medium (Fig. 6A) and
furin cell surface expression was also confirmed by
flow cytometry and cell surface biotinylation (Fig. 6C),
furin-expressing COS-1 cells did not show extracellular
processing of pro-MMP-2 (Fig. 6B). Moreover, when
purified pro-MMP-2 was incubated with purified furin

under cell-free conditions, processing did not occur
(data not shown). By contrast, the presence of PACE4,
PC5A or MT1-MMP resulted in extracellular cleavage
of pro-MMP-2 (Fig. 6B). Moreover, substitution of
Arg101 with Ala abolished the extracellular processing
of pro-MMP-2 by PACE4 and PC5A completely
(Fig. 6B), suggesting the direct cleavage of pro-MMP-2
by these enzymes.
To further compare PACE4 and PC5A processing
of pro-MMP-2 with that of furin, cell lysates from the
co-transfected cells were analyzed by zymography, in
which cells were pretreated with trypsin ⁄ EDTA to
remove MMP-2 from the cell surface (Fig. 6E). Cell
lysates from furin-expressing cells showed dramatically
larger amounts of processed MMP-2 than did those
from cells expressing PACE4 or PC5A (Fig. 6D).
Interestingly, processed MMP-2 was also seen in cell
lysates from HEK293F and MCF-7 cells expressing
pro-MMP-2 alone (Fig. 6D).
To further verify the intracellular localization of
PC-processed MMP-2, a FLAG-tagged pro-MMP-2
mutant, in which the FLAG epitope was inserted imme-
diately downstream of Arg98-Lys-Pro-Arg101, was gen-
erated. In these experiments, PC-processed MMP-2 with
a free N-terminal FLAG tag was detected by western
blotting with mouse anti-FLAG M1 IgG
2b
. The anti-
FLAG M1 IgG
2b

only recognizes proteins with a free
A
B
Fig. 4. (A) Zymograms of conditioned
medium from cells expressing pro-MMP-2
(WT), pro-MMP-2 N66I ⁄ L67V mutant
(N66I ⁄ L67V), pro-MMP-2 N109I ⁄ Y110F
mutant (N109I ⁄ Y110F) and pro-MMP-2
N66I ⁄ L67V ⁄ N109I ⁄ Y110F mutant
(N66I ⁄ L67V ⁄ N109I ⁄ Y110F). The conditioned
medium was analyzed without concentra-
tion. The bar graph shows the ratio of
processed MMP-2 to unprocessed
pro-MMP-2. Arrows indicate pro- and
processed MMP-2 (MMP-2). (B) Zymogram
of conditioned medium from COS-1 cells
expressing pro-MMP-2, pro-MMP-2
N66I ⁄ L67V mutant, pro-MMP-2
N109I ⁄ Y110F mutant and pro-MMP-2
N66I ⁄ L67V ⁄ N109I ⁄ Y110F mutant with or
without MT1-MMP. Note that COS-1 cells
expressing pro-MMP-2 N66I ⁄ L67V and
N66I ⁄ L67V ⁄ N109I ⁄ Y110F mutants show
prominent processing of the propeptide
without MT1-MMP expression. Arrows
indicate pro-, intermediate and processed
MMP-2 (MMP-2). Data are the means and
standard deviations of n = 3 experiments.
*P < 0.05 versus wild-type.
B H. Koo et al. Activation of pro-matrix metalloprotease-2 by proprotein convertases

FEBS Journal 276 (2009) 6271–6284 ª 2009 The Authors Journal compilation ª 2009 FEBS 6275
N-terminal FLAG tag. These results demonstrated that
furin-processed MMP-2 was recognized specifically by
the anti-FLAG M1 antibody (Fig. 7B), whereas pro-
MMP-2 and processed MMP-2 were detected by the
anti-MMP-2 antibody (Fig. 7A). Furthermore, muta-
tion of Arg101 to Ala resulted in the complete loss of
FLAG-tagged MMP-2, further supporting the specific-
ity of the anti-FLAG M1 antibody and validating this
experimental approach (Fig. 7B). Immunofluorescence
of fixed and permeabilized COS-1 cells stained with
anti-MMP-2 antibody revealed MMP-2 localization to
be mostly intracellular (Fig. 7C). Staining with anti-
FLAG M1 antibody also showed the intracellular local-
ization of PC-cleaved MMP-2, even though cells
expressing PACE4 or PC5A stained weakly (Fig. 7C).
Cells expressing MMP-2 alone exhibited little staining
with the anti-FLAG M1 antibody (Fig. 7C). Further-
more, no signal was observed in cells expressing the
pro-MMP-2-RKPA(101)-FLAG mutant with PCs,
providing additional evidence for the specificity of the
anti-FLAG M1 antibody (data not shown). Taken
together, these results demonstrate that furin-expressing
cells undergo exclusive intracellular processing of
pro-MMP-2, whereas cells expressing PACE4 or PC5A
exhibit both intracellular and extracellular processing
of pro-MMP-2.
Pro101 regulates excessive PC processing of
pro-MMP-2
Because Arg98-Lys-Pro-Arg101 in the propeptide of

MMP-2 is a minimal recognition motif for PC cleav-
A
C
E
D
B
Fig. 5. PC-dependent processing of
pro-MMP-2. (A) Effect of various inhibitors
on the propeptide processing of MMP-2
in HEK293F, MCF-7 and MDAH 2774
cells. Cells were treated with aprotinin
(20 lgÆmL
–1
), chymostatin (10 lgÆmL
–1
),
leupeptin (20 lgÆmL
–1
) and pepstatin (2 lM).
After 8 h of incubation, the conditioned
media were analyzed by zymography.
Arrows indicate pro- and processed MMP-2
(MMP-2). (B) Protease activities of PCs in
cell lysates from the cells indicated. (C)
Inhibition of pro-MMP-2 processing by a PC
inhibitor dec-RVKR-cmk in HEK293F, MCF-7
and MDAH 2774 cells. Cells were incubated
with 0–100 l
M of the inhibitor for 8 h, and
the conditioned medium was analyzed by

zymography. The bar graph shows the ratio
of processed MMP-2 to pro-MMP-2 in the
conditioned medium of cells treated with
100 l
M of the PC inhibitor. Data are the
means and standard deviations of n =3
experiments. *P < 0.05 versus untreated
control. (D) Inhibition of pro-MMP-2
processing by the PC inhibitor in K-562,
NCI-H460 and Hep G2 cells. Cells were
incubated with 100 l
M of the inhibitor. Note
that almost complete inhibition of pro-MMP-
2 processing is seen in the conditioned
medium of the inhibitor-treated cells.
(E) Zymograms of conditioned medium from
cells expressing pro-MMP-2 or pro-MMP-2
R101A mutant with furin, PACE4 or PC5A
(n = 3 representative experiments).
Activation of pro-matrix metalloprotease-2 by proprotein convertases B H. Koo et al.
6276 FEBS Journal 276 (2009) 6271–6284 ª 2009 The Authors Journal compilation ª 2009 FEBS
age [18,19,24], and Pro100 is highly conserved in the
MMP family, we investigated whether this amino acid
residue can play a role in regulating the PC cleavage
of pro-MMP-2. The conditioned medium of cells
expressing the pro-MMP-2 P100K mutant showed
highly increased propeptide processing by PCs
(PC5A > PACE4 > furin), whereas the PC-mediated
intracellular cleavage of the mutant was increased
slightly (Fig. 8A). Interestingly, the co-culture of COS-

1 cells expressing individual PCs and cells expressing
the mutant showed more robust processing of the
propeptide by extracellular PACE4 and PC5A, but not
by extracellular furin (Fig. 8B). These experimental
data suggest that Pro100 plays a critical role in regu-
lating excessive pro-MMP-2 processing, especially by
extracellular PACE4 and PC5A.
PC-cleaved MMP-2 is further processed to
achieve full activation
Because PCs cleave pro-MMP-2 immediately upstream
of the Cys102 residue that interacts with the catalytic
A
B
D E
C
Fig. 6. Cellular location of PC processing of
pro-MMP-2. (A) Protease activities of PCs in
conditioned medium and cell lysates from
COS-1 cells expressing PCs. (B) Co-culture
of COS-1 cells expressing pro-MMP-2 or
pro-MMP-2 R101A mutant with cells
expressing PCs or MT1-MMP. The condi-
tioned media were analyzed by zymography.
Note that processed MMP-2 is faintly seen
in the conditioned medium of cells express-
ing pro-MMP-2 alone. Arrows indicate
pro-, intermediate and processed MMP-2
(MMP-2). (C) Cell surface localization of furin
in the transfected COS-1 cells. Flow cytom-
etry and cell surface biotinylation

approaches were used to detect cell
surface furin. Arrowhead indicates furin.
(D) Zymographic analysis of cell lysates
from COS-1, HEK293F and MCF-7 cells
co-expressing pro-MMP-2 and PCs. Cells
were pretreated with 0.05% tryp-
sin ⁄ 0.53 m
M EDTA for 30 min on ice to
remove cell surface proteins. (E) Flow cyto-
metric analysis shows the cell surface locali-
zation of MMP-2 in COS-1, HEK293F and
MCF-7 transfected with pro-MMP-2. Cell
surface MMP-2 (dotted line) is absent in the
cells treated with trypsin ⁄ EDTA (T ⁄ E) (full
line) (n = 3 representative experiments).
B H. Koo et al. Activation of pro-matrix metalloprotease-2 by proprotein convertases
FEBS Journal 276 (2009) 6271–6284 ª 2009 The Authors Journal compilation ª 2009 FEBS 6277
domain zinc ion, MMP-2 processed in this manner
may not possess catalytic activity. Therefore, we inves-
tigated whether PC-cleaved MMP-2 exhibits enzymatic
activity or whether intermolecular autolytic cleavage at
the Asn109–Tyr peptide bond [36] is required for its
activity after initial processing. The degradation of
collagen IV and a fluorescein-conjugated gelatin by
processed MMP-2 from the conditioned medium of
HEK293F cells expressing pro-MMP-2, the pro-MMP-2
E404A mutant and the pro-MMP-2 N109I ⁄ Y110F
mutant in the presence and absence of furin was com-
pared. Because the catalytic glutamic acid residue
within the pro-MMP-2 active site is replaced in the

E404A mutant, a proteolytically inactive enzyme is
A
C
B
Fig. 7. Intracellular localization of
PC-cleaved MMP-2. Cell lysate was
obtained from COS-1 cells expressing
pro-MMP-2-RKPR(101)-FLAG or pro-MMP-2-
RKPA(101)-FLAG with or without furin.
Western blotting of cell lysate was
performed using anti-MMP-2 (A) and
anti-FLAG M1 (B) antibody. Arrowheads
indicate furin-processed MMP-2, and arrow
indicates uncleaved pro-MMP-2. (C) Confocal
microscope imaging of fixed and
permeabilized COS-1 cells stained with
anti-MMP-2 (total MMP-2 staining) or
anti-FLAG M1 (MMP-2 or FLAG M1 staining
is green). A negative control in which the
primary antibody was omitted showed no
signal (data not shown). Scale bar, 20 lm.
Activation of pro-matrix metalloprotease-2 by proprotein convertases B H. Koo et al.
6278 FEBS Journal 276 (2009) 6271–6284 ª 2009 The Authors Journal compilation ª 2009 FEBS
generated. Furthermore, the pro-MMP-2 N109I ⁄
Y110F mutant showed complete loss of intermolecular
autolytic cleavage following MT1-MMP cleavage
(Fig. 4B). Because of concerns about possible
unwanted structural effects of the processed MMP-2
during purification, these experiments were performed
using conditioned medium from cells expressing the

various constructs. The concentrations of MMP-2 were
measured by ELISA. Substrates were incubated with
an identical amount of MMP-2 and subsequently ana-
lyzed by SDS-PAGE or fluorometry. Although the
substrates were degraded proteolytically by MMP-2
that was processed via endogenous routes, furin-medi-
ated cleavage conferred increased proteolytic activity
on MMP-2 (Fig. 9). However, the pro-MMP-2
N109I ⁄ Y110F mutant showed less proteolytic activity
than the wild-type, even though its enzymatic activity
was also increased by furin-mediated cleavage (Fig. 9).
We also obtained similar results with the conditioned
medium from MCF-7 cells expressing the same
constructs (data not shown). However, furin-processed
MMP-2 from the conditioned medium of co-transfected
COS-1 cells showed much reduced levels of enzymatic
activity (e.g. 1 ⁄ 50th to 1 ⁄ 200th of the activity of
processed MMP-2 from HEK293F and MCF-7 cells)
by unknown mechanisms (data not shown). Taken
together, these data suggest that PC-cleaved MMP-2 is
further processed for full enzymatic activity.
Discussion
The role of MT1-MMP in pro-MMP-2 activation is
well established [7,14], but residual activation of pro-
MMP-2 can also be seen in MT1-MMP
) ⁄ )
mouse
fibroblasts cultured in collagen gel and lung extract
[16,17]. The residual activation observed might be
caused by the presence of other proteases, including

metalloproteases (e.g. other MT-MMPs) [8–11] and
A
B
Fig. 8. Regulation of excessive PC processing of pro-MMP-2 by
Pro100 residue. (A) PC processing of pro-MMP-2 and pro-MMP-2
P101K mutant. Conditioned medium (CM) and cell lysate from
co-transfected COS-1 cells were analyzed by zymography. Data are
the means and standard deviations of n = 3 experiments.
*P < 0.05 versus wild-type. (B) Co-culture of COS-1 cells express-
ing the pro-MMP-2 P100K mutant with PC-expressing cells. The
conditioned medium was analyzed by zymography (n = 3 represen-
tative experiments).
A
B
Fig. 9. Furin-processed MMP-2 gains its full activity by intermolec-
ular autolytic cleavage. Digestion assay of collagen IV (A) and a fluo-
rescein-conjugated gelatin (B). Conditioned medium was obtained
from HEK293F cells expressing pro-MMP-2, pro-MMP-2 E404A
mutant and pro-MMP-2 N109I ⁄ Y110F mutant with or without furin.
The digested substrates were analyzed as described in Materials
and methods. Zymogram and western blot using anti-MMP-2 show
equal amounts of MMP-2 used. Arrows indicate pro- and processed
MMP-2 (MMP-2). Data are the means and standard deviations of
n = 6 experiments.
ANOVA test indicates statistically significant dif-
ferences in furin versus control and in the N109I ⁄ Y110F mutant
versus the wild-type (P < 0.05).
B H. Koo et al. Activation of pro-matrix metalloprotease-2 by proprotein convertases
FEBS Journal 276 (2009) 6271–6284 ª 2009 The Authors Journal compilation ª 2009 FEBS 6279
serine proteases (e.g. urokinase ⁄ plasmin, thrombin,

chymase, cathepsin G and trypsin-2) [6,12,37,38].
Thus, we examined MT1-MMP-independent process-
ing of pro-MMP-2 using various approaches, all of
which provided evidence for non-MT1-MMP compo-
nents within cells. MT-MMPs have been shown to
process pro-MMP-2 through cleavage of pro-MMP-2
at the Asn66–Leu peptide bond by MT1-, MT2-,
MT3- and MT5-MMP, followed by intermolecular
autolytic cleavage at the Asn109–Tyr peptide bond [7–
10], or directly at the Asn109–Tyr peptide bond by
MT6-MMP [11]. Moreover, various MT-MMPs are
expressed in the cell types analyzed in this study. How-
ever, studies using metalloprotease inhibitors and pro-
MMP-2 mutants incapable of cleavage by MT-MMPs
excluded MT-MMPs as candidates for pro-MMP-2
processing in these cell lines. A previous study has also
shown residual MMP-2 activation in fibroblasts treated
with a general metalloprotease inhibitor [16]. These
data, taken together with ours, strongly suggest the
existence of non-MT-MMP components for pro-
MMP-2 processing in various cell types.
In this study, we found that PCs such as furin, PACE4
and PC5A mediate the MT-MMP-independent process-
ing of pro-MMP-2. Furin-mediated cleavage of pro-
MMP-2 immediately downstream of R(98)KPR(101)
has been reported previously [35], even though it has a
minimal PC recognition motif [18]. Moreover, thrombin-
mediated activation of pro-MMP-2 has been shown to
occur in human endothelial cells at a predicted PC
recognition motif for cleavage [12]. Because other

serine proteases, including factor Xa and plasmin, have
also been shown to activate pro-MMP-2 [6,39], it
would be interesting to investigate whether these prote-
ases can activate pro-MMP-2 by cleavage at the same
site. If this is the case, the PC recognition motif would
be a general cleavage site for pro-MMP-2 activation.
Although the conditioned medium of cells expressing
the PC-uncleavable mutant showed a significant
decrease in pro-MMP-2 processing, lower levels of
processed MMP-2 were also seen. MT-MMP expres-
sion may contribute to the residual processing of the
mutant in these cells as the mutant can be processed
more fully following MT1-MMP overexpression. PCs
are overexpressed in various cancers, including lung,
breast and skin, with a significant role in tumor
progression [40]. Moreover, our studies using metallo-
protease and PC inhibitors showed PCs to be major
enzymes for pro-MMP-2 processing in various tumor
cell lines. Thus, PC-mediated processing may be a
major mechanism for the activation of pro-MMP-2 in
PC-overexpressing cells (e.g. tumor cell lines), whereas
PCs are likely to act alongside MT-MMPs for
pro-MMP-2 activation in cells expressing high levels of
MT-MMPs.
Furin is localized predominantly in the TGN, but
some is present at the plasma membrane [21,22,25].
This enzyme is also secreted by cells and may be func-
tional in the extracellular space [41,42]. However,
despite numerous attempts, furin-mediated extracellu-
lar processing of pro-MMP-2 could not be detected,

leading to the conclusion that furin processing of pro-
MMP-2 is exclusively intracellular. This finding contra-
dicts that of Cao et al. [35], who reported that furin
processing was both intracellular and extracellular,
with intracellular processing as the predominant mech-
anism. Unlike furin, secreted PACE4 and PC5A are
anchored to the cell surface by heparan sulfate proteo-
glycan in the extracellular space despite their residence
in the TGN [23]. Likewise, MMP-2 has an affinity for
binding heparin via its hemopexin-like domain and
anchors to heparan sulfate proteoglycans [43,44].
Therefore, the binding property of heparan sulfate pro-
teoglycans may lead to their co-localization to the cell
surface, facilitating the extracellular processing of pro-
MMP-2. However, the inability of extracellular furin
to bind heparan sulfate proteoglycans may prevent its
co-localization with pro-MMP-2 at the cell surface and
its extracellular processing of pro-MMP-2. It is also
likely that, once exported from the cell, pro-MMP-2
undergoes a conformational change, thereby hamper-
ing furin processing. Although cells expressing PACE4
or PC5A exhibit both intracellular and extracellular
processing of pro-MMP-2, intracellular processing
is likely to be predominant because these PCs are
concentrated in the TGN to allow more efficient
processing of pro-MMP-2.
By contrast with previously published data [35],
furin-processed MMP-2 showed enzymatic activity.
However, when furin-processed MMP-2 was obtained
from the conditioned medium of COS-1 cells, its enzy-

matic activity was much lower than that obtained from
HEK293F and MCF-7 cells. Several reasons can
explain this observation. TIMP-2 was highly expressed
in COS-1 cells. Furthermore, TIMP-2 is a strong
inhibitor of MMP-2 at high concentrations [45],
although it activates pro-MMP-2 with MT1-MMP at
low concentrations [34]. Therefore, we propose that
the lack of enzymatic activity is a result of the presence
of a high concentration of TIMP-2 in the conditioned
medium of COS-1 cells. This was also supported by
our observation that autoproteolytic activation of
MT1-MMP-processed MMP-2 was detected marginally
in COS-1 cells (Fig. 4B). In this study, we found that
furin-processed MMP-2 was further processed to its
fully activated form, but the detailed mechanism
Activation of pro-matrix metalloprotease-2 by proprotein convertases B H. Koo et al.
6280 FEBS Journal 276 (2009) 6271–6284 ª 2009 The Authors Journal compilation ª 2009 FEBS
remains to be established. It is possible that TIMP-2
and integrin avb3 may mediate the full activation of
PC-processed MMP-2, as they have been suggested to
play a role in the maturation of MT1-MMP-cleaved
MMP-2 to its fully active form. Other plausible candi-
dates are fibronectin and heparan sulfate proteogly-
cans, which display an affinity for MMP-2 binding
[43,44]. MMP-2 can be localized and concentrated at
the cell surface via these molecules, which may pro-
mote intermolecular autolytic cleavage of PC-processed
MMP-2.
Materials and methods
Reagents

Mouse anti-MMP-2 IgG
1
(sc-13595) and anti-TIMP-2
IgG
2a
(sc-21735), as well as rabbit anti-MT1-MMP poly-
clonal IgG (sc-30074), were purchased from Santa Cruz
Biotechnology (Santa Cruz, CA, USA). Mouse anti-FLAG
M1 IgG
2b
, Con A, type IV collagen, aprotinin, trypsin
inhibitor, leupeptin and pepstatin were purchased from
Sigma-Aldrich (St. Louis, MO, USA). Monoclonal anti-
integrin avb3 antibody, GM6001 and human TIMP-2 were
purchased from Chemicon (Temecula, CA, USA), and dec-
RVKR-cmk was obtained from Calbiochem (San Diego,
CA, USA). Monoclonal anti-furin antibody (MON-148)
was purchased from Alexis Biochemicals (San Diego, CA,
USA) and purified furin from New England BioLabs
(Beverly, MA, USA) The EnzChek Gelatinase ⁄ Collagenase
Assay Kit was purchased from Molecular Probes (Eugene,
OR, USA) and MTT from Amresco (Solon, OH, USA).
Expression plasmids and site-directed
mutagenesis
blast programs from the National Center for Biotechnology
Information were used to search for expressed sequence tags.
Human testis cDNA (Marathon cDNA, Clontech, Palo Alto,
CA, USA) was used as a template to amplify the full-length
cDNA for MMP-2 (Accession no. NM_004530). Oligo-
nucleotide primers 5¢-GCTACGATG GAGGCGCTAATG

GCC-3¢ (start codon in italics) and 5¢-TCAGCAGCCTAGC
CAGTCGGATTTG-3¢ (stop codon in italics) were used for
PCR with Advantage 2 polymerase (Clontech). The 2-kb
PCR product was cloned into TOPO cloning vectors
(Invitrogen, Carlsbad, CA, USA) and sequenced completely.
For the full-length MMP-2 expression plasmid, its open read-
ing frame was digested with EcoRI and then re-cloned into
the pcDNA3.1 ⁄ myc-His(–) B vector (Invitrogen) digested
with EcoRI. For FLAG tag insertion between Arg101 and
Cys102 in the propeptide of MMP-2, site-directed muta-
genesis (Intron Biotechnology, Seongnam, Kyunggi, South
Korea) was performed using the forward primer 5¢-AGAC
CATGCGGAAGCCACGCGACTACAAAGACGATGACG
ACAAGTGCGGCAACCCAGATGTGGC-3¢ (FLAG enco-
ding site in italics) and the reverse primer 5¢-GCCACATC
TGGGTTGCCGCACTTGTCGTCATCGTCTTTGTAGTC
GCGTGGCTTCCGCATGGTCT-3¢ (FLAG encoding site
in italics). Site-directed mutagenesis was performed topre-
pare all the mutants used for this study. Expression plas-
mids for PACE4, PC5A and furin were kindly provided by
Nabil Seidah (Clinical Research Institute of Montreal,
Canada), and the human MT1-MMP expression plasmid
was kindly provided by Suneel Apte (Lerner Research
Institute, Cleveland, OH, USA).
Cell culture, transfection, cell treatments and cell
viability assay
HEK293F, COS-1, HCT116 (human colorectal carcinoma)
(ATCC no. CCL-247), MDAH 2774 (human ovarian ade-
nocarcinoma) (ATCC no. CRL-10303), NCI-H460 (human
lung carcinoma) (ATCC no. HTB-177) and HT-1080

(human fibrosarcoma) (ATCC no. CCL-121) cells were
maintained in Dulbecco’s modified Eagle’s medium contain-
ing 10% fetal bovine serum, and MCF-7 (human breast
adenocarcinoma) (ATCC no. HTB-22), K-562 (human
chronic myelogenous leukemia) (ATCC no. CCL-243) and
Hep G2 (human hepatocellular carcinoma) (ATCC no.
HB-8065) cells were maintained in RPMI-1640 medium
containing 10% fetal bovine serum. Transient transfections
with various plasmids were performed using Lipofecta-
mine 2000 according to the manufacturer’s recommenda-
tions (Invitrogen). For the analysis of secreted protein,
transfected cells were cultured and then changed to
293 SFM-II medium (Invitrogen). For comparison of cell
viability of the PC inhibitor-treated cells with control cells,
cells were cultured confluently in 96-well plates and further
incubated in the presence and absence of dec-RVKR-cmk
(100 lm) for 8 h. The numbers of viable cells were assessed
using the MTT assay.
Semi-quantitative RT-PCR
Total RNA was prepared using an RNeasy Mini Kit
(Qiagen, Chatsworth, CA, USA). Five micrograms of
mRNA was used for the synthesis of first-strand cDNA
employing the Superscript III First-Strand Kit (Invitrogen).
One microliter of cDNA was used as template for PCR.
Primer sets were as follows: for MT1-MMP, forward pri-
mer 5¢-CGAAGCCTGGCTACAGCAAT-3¢ and reverse
primer 5¢-CTCGTATGTGGCATACTCGC-3¢; for MT2-
MMP, forward primer 5¢-CCTGGACAACTATCCCA
TGC-3¢ and reverse primer 5¢-GCCAGACACTGATGGGC
TTG-3¢; for MT3-MMP, forward primer 5¢-GACTGACC

CCAGAAT GTCAG-3¢ and reverse primer 5¢-CTGCCAC
ACATCAAA GGCAC-3¢; for MT5-MMP, forward primer
5¢-GCCGGG CAGAACTGGTTAAA-3¢ and reverse
B H. Koo et al. Activation of pro-matrix metalloprotease-2 by proprotein convertases
FEBS Journal 276 (2009) 6271–6284 ª 2009 The Authors Journal compilation ª 2009 FEBS 6281
primer 5¢-CGAAA GCCTGGCGAATAGCT-3¢; for MT6-
MMP, forward primer 5¢-GGCTGACTCGCTATGGTTA
C-3¢ and reverse primer 5¢-GCCATCAGGGCATAGCT
CAT-3¢; for glyceraldehyde-3-phosphate dehydrogenase
(GAPDH), forward primer 5 ¢-GAAGCTCACTGGCATG
GCCTT-3¢ and reverse primer 5 ¢-CTCTCTTGCTCAGTG
TCCTTGCT-3¢.
Flow cytometric analysis and
immunofluorescence
Prior to the analysis of cell surface MMP-2, cells were fixed
with 4% paraformaldehyde for 30 min. For the analysis of
cell surface integrin avb3 and furin, the fixation step was
omitted. Then, detached cells were incubated with monoclo-
nal antibodies against integrin avb3, furin or MMP-2 at
4 °C for 2 h. The cells were washed and further incubated
with fluorescein isothiocyanate (FITC)-conjugated goat
anti-mouse IgG (Chemicon) for 1 h. Flow cytometry was
performed on a FACSCalibur (Becton Dickinson, San Jose,
CA, USA), and data were analyzed using WinMDI soft-
ware version 2.8 (The Scripps Research Institute, La Jolla,
CA, USA). For immunocytochemical staining, transfected
COS-1 cells were fixed and permeabilized with methanol at
)20 °C for 10 min. After incubation with monoclonal anti-
MMP-2 or anti-FLAG M1 antibody, and subsequently
with FITC-conjugated goat anti-mouse antibody, optical

images were acquired using an LSM 510 META confocal
microscope (Carl Zeiss, Thornwood, NY, USA).
ELISA, zymography, cell surface biotinylation and
western blotting
MMP-2 was quantified in the conditioned medium of trans-
fected cells using an MMP-2 ELISA kit according to the
manufacturer’s recommendations (Calbiochem). The condi-
tioned medium and cell lysate were analyzed for proteins
with gelatinolytic activity by identification of substrate lysis
by 8% SDS–PAGE containing 2 mgÆmL
–1
gelatin. Gels
were washed with 1% Triton X-100 for 1 h and incubated
for 14–20 h at 37 °Cin50mm Tris ⁄ HCl, pH 7.5, contain-
ing 20 mm CaCl
2
. Gels were stained with 0.2% Coomassie
Brilliant Blue R-250 in 40% methanol and 10% acetic acid.
Biotinylation of cells was performed on ice to prevent the
labeling of intracellular proteins, according to the manufac-
turer’s recommendations (Pierce, Rockford, IL, USA).
Western blotting was performed using reduced or non-
reduced cell lysates on SDS–PAGE, and then electroblotted
to poly(vinylidene difluoride), followed by detection of
bound antibody using enhanced chemiluminescence (Amer-
sham Biosciences, Pittsburgh, PA, USA). The signal inten-
sity of the relevant bands from western blotting was
quantified using ImageJ software (National Institutes of
Health, Bethesda, MD, USA).
Protease activity of PCs

The protease activities of PCs were assayed using a modified
method from a previous report [46]. Briefly, samples were
incubated with 100 lm pyroglutamyl-Arg-Thr-Lys-Arg-
4-methylcoumaryl 7-amide (pyr-RTKR-MCA, Peptide Insti-
tute, Osaka, Japan), 2 mm CaCl
2
,10lgÆmL
–1
each of
leupeptin, 4-[(2S,3S)-3-carboxyoxiran-2-ylcarbonyl-l-leucyl-
amido]butylguanidine, bestatin and pepstatin, 1 lgÆmL
–1
chymostatin and 30 ngÆmL
–1
TIMP-2 in 100 mm Tris ⁄ HCl
(pH 7.5) at 37 °C for 1 h. The amount of MCA released from
the substrate was measured in a fluorescence spectrophoto-
meter with a fluorescence microplate reader set for excitation
at 380 nm and emission detection at 460 nm. One unit of
activity is defined as the amount of furin that can release
1 nmol of MCA from the substrate in 1 mL of enzyme.
Collagen IV and fluorescein-conjugated gelatin
digestion assays
Substrates were incubated with conditioned medium con-
taining MMP-2 (5 lgÆmL
–1
MMP-2 per reaction) at 37 °C
for 6 h (5 lg of collagen IV per reaction) or fluorescein-
conjugated gelatin for 1 h according to the manufacturer’s
recommendations. The digested substrates were analyzed by

SDS–PAGE or a fluorescence microplate reader set for
excitation at 495 nm and emission detection at 515 nm.
Statistical analysis
Data are the means and standard deviations of n experi-
ments. Statistical analysis was performed using an unpaired
t-test or ANOVA test. A P value of less than 0.05 was
considered to be statistically significant.
Acknowledgements
We thank Dr Suneel S. Apte (Lerner Research
Institute, Cleveland, OH, USA) for helpful discussions
and critical reading of the manuscript. This work was
supported by a Korea Science and Engineering Foun-
dation (KOSEF) grant funded by the South Korean
government (MOST; nos ROA-2004-000-10297-0 and
2009-0081759). O H. Jeon and H H. Kim are fellow-
ship awardees by the Brain Korea 21 (BK21) program.
References
1 Somerville RP, Oblander SA & Apte SS (2003) Matrix
metalloproteinases: old dogs with new tricks. Genome
Biol 4, 216.
2 Woessner JF Jr (1994) The family of matrix metallopro-
teinases. Ann NY Acad Sci 732, 11–21.
Activation of pro-matrix metalloprotease-2 by proprotein convertases B H. Koo et al.
6282 FEBS Journal 276 (2009) 6271–6284 ª 2009 The Authors Journal compilation ª 2009 FEBS
3 Okada Y, Morodomi T, Enghild JJ, Suzuki K, Yasui
A, Nakanishi I, Salvesen G & Nagase H (1990)
Matrix metalloproteinase 2 from human rheumatoid
synovial fibroblasts. Purification and activation of the
precursor and enzymic properties. Eur J Biochem 194,
721–730.

4 Ra HJ & Parks WC (2007) Control of matrix metallo-
proteinase catalytic activity. Matrix Biol 26, 587–596.
5 Bescond A, Augier T, Chareyre C, Garcon D, Horne-
beck W & Charpiot P (1999) Influence of homocysteine
on matrix metalloproteinase-2: activation and activity.
Biochem Biophys Res Commun 263 , 498–503.
6 Baramova EN, Bajou K, Remacle A, L’Hoir C, Krell
HW, Weidle UH, Noel A & Foidart JM (1997)
Involvement of PA ⁄ plasmin system in the processing of
pro-MMP-9 and in the second step of pro-MMP-2
activation. FEBS Lett 405, 157–162.
7 Strongin AY, Collier I, Bannikov G, Marmer BL,
Grant GA & Goldberg GI (1995) Mechanism of cell
surface activation of 72-kDa type IV collagenase.
Isolation of the activated form of the membrane
metalloprotease. J Biol Chem 270 , 5331–5338.
8 Morrison CJ, Butler GS, Bigg HF, Roberts CR,
Soloway PD & Overall CM (2001) Cellular activation
of MMP-2 (gelatinase A) by MT2-MMP occurs via a
TIMP-2-independent pathway. J Biol Chem 276,
47402–47410.
9 Nakada M, Yamada A, Takino T, Miyamori H,
Takahashi T, Yamashita J & Sato H (2001) Suppression
of membrane-type 1 matrix metalloproteinase
(MMP)-mediated MMP-2 activation and tumor
invasion by testican 3 and its splicing variant gene
product, N-Tes. Cancer Res 61, 8896–8902.
10 Pei D (1999) Identification and characterization of the
fifth membrane-type matrix metalloproteinase
MT5-MMP. J Biol Chem 274, 8925–8932.

11 Nie J & Pei D (2003) Direct activation of pro-matrix
metalloproteinase-2 by leukolysin ⁄ membrane-type 6
matrix metalloproteinase ⁄ matrix metalloproteinase 25
at the asn(109)–Tyr bond. Cancer Res 63, 6758–6762.
12 Lafleur MA, Hollenberg MD, Atkinson SJ, Knauper V,
Murphy G & Edwards DR (2001) Activation of pro-
(matrix metalloproteinase-2) (pro-MMP-2) by thrombin
is membrane-type-MMP-dependent in human umbilical
vein endothelial cells and generates a distinct 63 kDa
active species. Biochem J 357, 107–115.
13 Caterina JJ, Yamada S, Caterina NC, Longenecker G,
Holmback K, Shi J, Yermovsky AE, Engler JA &
Birkedal-Hansen H (2000) Inactivating mutation of the
mouse tissue inhibitor of metalloproteinases-2 (Timp-2)
gene alters proMMP-2 activation. J Biol Chem 275,
26416–26422.
14 Hernandez-Barrantes S, Toth M, Bernardo MM, Yurk-
ova M, Gervasi DC, Raz Y, Sang QA & Fridman R
(2000) Binding of active (57 kDa) membrane type
1-matrix metalloproteinase (MT1-MMP) to tissue inhib-
itor of metalloproteinase (TIMP)-2 regulates MT1-
MMP processing and pro-MMP-2 activation. J Biol
Chem 275, 12080–12089.
15 Wang Z, Juttermann R & Soloway PD (2000) TIMP-2
is required for efficient activation of proMMP-2 in vivo.
J Biol Chem 275, 26411–26415.
16 Ruangpanit N, Price JT, Holmbeck K, Birkedal-Hansen
H, Guenzler V, Huang X, Chan D, Bateman JF &
Thompson EW (2002) MT1-MMP-dependent and
-independent regulation of gelatinase A activation in

long-term, ascorbate-treated fibroblast cultures: regula-
tion by fibrillar collagen. Exp Cell Res 272, 109–118.
17 Oblander SA, Zhou Z, Galvez BG, Starcher B, Shan-
non JM, Durbeej M, Arroyo AG, Tryggvason K &
Apte SS (2005) Distinctive functions of membrane
type 1 matrix-metalloprotease (MT1-MMP or MMP-
14) in lung and submandibular gland development are
independent of its role in pro-MMP-2 activation. Dev
Biol 277, 255–269.
18 Bergeron F, Leduc R & Day R (2000) Subtilase-like
pro-protein convertases: from molecular specificity to
therapeutic applications. J Mol Endocrinol 24, 1–22.
19 Nakayama K (1997) Furin: a mammalian subtilisin
⁄
Kex2p-like endoprotease involved in processing of a
wide variety of precursor proteins. Biochem J 327
(Pt 3), 625–635.
20 Thomas G (2002) Furin at the cutting edge: from
protein traffic to embryogenesis and disease. Nat Rev
Mol Cell Biol 3, 753–766.
21 Mayer G, Boileau G & Bendayan M (2003) Furin
interacts with proMT1-MMP and integrin alphaV at
specialized domains of renal cell plasma membrane.
J Cell Sci 116, 1763–1773.
22 Mayer G, Boileau G & Bendayan M (2004) Sorting of
furin in polarized epithelial and endothelial cells:
expression beyond the Golgi apparatus. J Histochem
Cytochem 52, 567–579.
23 Tsuji A, Sakurai K, Kiyokage E, Yamazaki T, Koide S,
Toida K, Ishimura K & Matsuda Y (2003) Secretory

proprotein convertases PACE4 and PC6A are heparin-
binding proteins which are localized in the extracellular
matrix. Potential role of PACE4 in the activation of
proproteins in the extracellular matrix. Biochim Biophys
Acta 1645, 95–104.
24 Seidah NG & Prat A (2002) Precursor convertases in
the secretory pathway, cytosol and extracellular milieu.
Essays Biochem 38, 79–94.
25 Koo BH, Longpre JM, Somerville RP, Alexander JP,
Leduc R & Apte SS (2006) Cell-surface processing of
pro-ADAMTS9 by furin. J Biol Chem 281, 12485–
12494.
26 Zheng M, Streck RD, Scott RE, Seidah NG & Pintar
JE (1994) The developmental expression in rat of prote-
ases furin, PC1, PC2, and carboxypeptidase E: implica-
B H. Koo et al. Activation of pro-matrix metalloprotease-2 by proprotein convertases
FEBS Journal 276 (2009) 6271–6284 ª 2009 The Authors Journal compilation ª 2009 FEBS 6283
tions for early maturation of proteolytic processing
capacity. J Neurosci 14, 4656–4673.
27 Roebroek AJ, Umans L, Pauli IG, Robertson EJ,
Van LF, Van DV & Constam DB (1998) Failure of
ventral closure and axial rotation in embryos lacking
the proprotein convertase furin. Development 125,
4863–4876.
28 Khatib AM, Siegfried G, Chretien M, Metrakos P &
Seidah NG (2002) Proprotein convertases in tumor
progression and malignancy: novel targets in cancer
therapy. Am J Pathol 160, 1921–1935.
29 Gilles C, Polette M, Piette J, Munaut C, Thompson
EW, Birembaut P & Foidart JM (1996) High level of

MT-MMP expression is associated with invasiveness of
cervical cancer cells. Int J Cancer 65, 209–213.
30 Ellerbroek SM, Fishman DA, Kearns AS, Bafetti LM
& Stack MS (1999) Ovarian carcinoma regulation of
matrix metalloproteinase-2 and membrane type 1 matrix
metalloproteinase through beta1 integrin. Cancer Res
59, 1635–1641.
31 Hotary K, Allen E, Punturieri A, Yana I & Weiss SJ
(2000) Regulation of cell invasion and morphogenesis in
a three-dimensional type I collagen matrix by mem-
brane-type matrix metalloproteinases 1, 2, and 3. J Cell
Biol 149, 1309–1323.
32 Deryugina EI, Ratnikov B, Monosov E, Postnova TI,
DiScipio R, Smith JW & Strongin AY (2001) MT1-
MMP initiates activation of pro-MMP-2 and integrin
alphavbeta3 promotes maturation of MMP-2 in breast
carcinoma cells. Exp Cell Res 263, 209–223.
33 Will H, Atkinson SJ, Butler GS, Smith B & Murphy G
(1996) The soluble catalytic domain of membrane
type 1 matrix metalloproteinase cleaves the propeptide
of progelatinase A and initiates autoproteolytic activa-
tion. Regulation by TIMP-2 and TIMP-3. J Biol Chem
271, 17119–17123.
34 Cao J, Rehemtulla A, Bahou W & Zucker S (1996)
Membrane type matrix metalloproteinase 1 activates
pro-gelatinase A without furin cleavage of the N-termi-
nal domain. J Biol Chem 271, 30174–30180.
35 Cao J, Rehemtulla A, Pavlaki M, Kozarekar P &
Chiarelli C (2005) Furin directly cleaves proMMP-2 in
the trans-Golgi network resulting in a nonfunctioning

proteinase. J Biol Chem 280, 10974–10980.
36 Strongin AY, Marmer BL, Grant GA & Goldberg GI
(1993) Plasma membrane-dependent activation of the
72-kDa type IV collagenase is prevented by complex
formation with TIMP-2. J Biol Chem 268, 14033–14039.
37 Shamamian P, Schwartz JD, Pocock BJ, Monea S,
Whiting D, Marcus SG & Mignatti P (2001) Activation
of progelatinase A (MMP-2) by neutrophil elastase,
cathepsin G, and proteinase-3: a role for inflammatory
cells in tumor invasion and angiogenesis. J Cell Physiol
189, 197–206.
38 Sorsa T, Salo T, Koivunen E, Tyynela J, Konttinen
YT, Bergmann U, Tuuttila A, Niemi E, Teronen O,
Heikkila P et al. (1997) Activation of type IV procolla-
genases by human tumor-associated trypsin-2. J Biol
Chem 272, 21067–21074.
39 Rauch BH, Bretschneider E, Braun M & Schror K
(2002) Factor Xa releases matrix metalloproteinase-2
(MMP-2) from human vascular smooth muscle cells
and stimulates the conversion of pro-MMP-2 to MMP-
2: role of MMP-2 in factor Xa-induced DNA synthesis
and matrix invasion. Circ Res 90, 1122–1127.
40 Bassi DE, Mahloogi H & Klein-Szanto AJ (2000) The
proprotein convertases furin and PACE4 play a signifi-
cant role in tumor progression. Mol Carcinog 28, 63–69.
41 Tsuneoka M, Nakayama K, Hatsuzawa K, Komada M,
Kitamura N & Mekada E (1993) Evidence for involve-
ment of furin in cleavage and activation of diphtheria
toxin. J Biol Chem 268, 26461–26465.
42 Thimon V, Belghazi M, Dacheux JL & Gatti JL (2006)

Analysis of furin ectodomain shedding in epididymal
fluid of mammals: demonstration that shedding of furin
occurs in vivo. Reproduction 132, 899–908.
43 Wallon UM & Overall CM (1997) The hemopexin-like
domain (C domain) of human gelatinase A (matrix
metalloproteinase-2) requires Ca
2+
for fibronectin and
heparin binding. Binding properties of recombinant
gelatinase A C domain to extracellular matrix and
basement membrane components. J Biol Chem 272,
7473–7481.
44 Yu WH & Woessner JF Jr (2000) Heparan sulfate
proteoglycans as extracellular docking molecules for
matrilysin (matrix metalloproteinase 7). J Biol Chem
275, 4183–4191.
45 Cao J, Sato H, Takino T & Seiki M (1995) The
C-terminal region of membrane type matrix metallo-
proteinase is a functional transmembrane domain
required for pro-gelatinase A activation. J Biol Chem
270, 801–805.
46 Tsuji A, Hashimoto E, Ikoma T, Taniguchi T, Mori K,
Nagahama M & Matsuda Y (1999) Inactivation of
proprotein convertase, PACE4, by alpha1-antitrypsin
Portland (alpha1-PDX), a blocker of proteolytic
activation of bone morphogenetic protein during
embryogenesis: evidence that PACE4 is able to form
an SDS-stable acyl intermediate with alpha1-PDX.
J Biochem 126, 591–603.
Activation of pro-matrix metalloprotease-2 by proprotein convertases B H. Koo et al.

6284 FEBS Journal 276 (2009) 6271–6284 ª 2009 The Authors Journal compilation ª 2009 FEBS