Ganglioside GD1a negatively regulates matrix metalloproteinase-9 expression in mouse FBJ cell lines at the transcriptional level.

Abstract

Mouse FBJ virus-induced osteosarcoma FBJ-S1 cells rich in GD1a are not readily metastatic, whereas FBJ-LL cells with low levels of GD1a are highly metastatic. GD1a was previously shown to suppress metastasis of mouse FBJ cells and to upregulate caveolin-1 and stromal interaction molecule 1 expression. The present study demonstrates that matrix metalloproteinase-9 (MMP-9) expression renders FBJ-LL cells invasive. MMP-9 is inversely regulated by GD1a, based upon four observations: MMP-9 mRNA content was 5 times higher in FBJ-LL cells than FBJ-S1 cells; a GD1a-re-expressing FBJ-LL cell variant produced through beta1,4GalNAcT-1 cDNA transfection expressed lower levels of MMP-9; exogenous addition of GD1a to FBJ-LL cells decreased MMP-9 production in a dose- and time-dependent manner; and treatment of GD1a-rich cells with D-PDMP or siRNA targeting St3gal2 decreased GD1a expression, but augmented MMP-9 expression. This is the first report demonstrating that GD1a negatively regulates expression of MMP-9 at the transcriptional level.

Full text

Page 1

Apparent suppression of MMP-9 activity by GD1a as determined
by gelatin zymography
Dan Hua, Xuan Tana, Toshinori Satob, Sadako Yamagataa, Tatsuya Yamagataa,*
aLaboratory of Tumor Biology and Glycobiology, Shenyang Pharmaceutical University, Shenyang 110016, People’s Republic of China
bDepartment of Biosciences and Informatics, Keio University, Hiyoshi, Yokohama 223-8522, Japan
Received 11 August 2006
Available online 22 August 2006
Abstract
Gelatin zymography is widely used to detect and evaluate matrix metalloproteinase-9 (MMP-9) activity. MMP-9 transcription was
previously shown to be negatively regulated by ganglioside GD1a. [D. Hu, Z. Man, T. Xuan, P. Wang, T. Takaku, S. Hyuga, X.S.
Yao, T. Sato, S. Yamagata, T. Yamagata, Ganglioside GD1a regulation of matrix metalloproteinase-9 (MMP-9) expression in mouse
FBJ cell Lines: GD1a suppression of MMP-9 expression stimulated by PI3K-Akt and p38 though not by the Erk signaling pathway,
2006, submitted for publication.] Zymography of MMP-9 of FBJ-M5 cells preincubated with GD1a indicated a greater decrease in activ-
ity than expected from mRNA suppression. Incubation of conditioned medium containing MMP-9 with GD1a caused MMP-9 activity
to decrease. Examination was thus made to confirm that MMP-9 activity is actually suppressed and/or MMP-9 protein undergoes deg-
radation by GD1a. GD1a was found to have no effect on MMP-9 activity and Western blots indicated GD1a not to diminish MMP-9
during electrophoresis under reducing conditions. GD1a appeared to mediate the binding of a portion of MMP-9 with certain molecules,
with consequently greater molecular mass on the gel, to cause decrease in the activity of MMP-9 at the site where it would normally
appear. Caution should be used in doing gelatin zymography since molecules other than GD1a may similarly work, causing decrease
in MMP-9 activity in zymography.
? 2006 Elsevier Inc. All rights reserved.
Keywords: Ganglioside; GD1a; Zymography; Metastasis; MMP-9; MMP-2
During the course of our study on ganglioside GD1a
regulation of metastasis of mouse osteosarcoma-derived
FBJ-LL cells, a matrix metalloproteinase-9 (MMP-9) but
not MMP-2 mRNA expression was found to be down-reg-
ulated by GD1a [5]. MMP-9 and MMP-2 are implicated in
metastatic potential of tumor cells [1–4]. MMP-9 was noted
to be high in FBJ-LL with less GD1a, but suppressed in
FBJ-S1 cells rich in GD1a [5]. The FBJ-LA5-30 cell, a
GD1a-reexpressing FBJ-LL cell variant through b1-
4GalNAcT-1 (GM2/GD2-synthase) cDNA transfection
[6], showed less MMP-9 production, compared to a
mock-transfectant M5 cell. GD1a regulation was con-
firmed by exogenous addition of GD1a to FBJ-M5 or
FBJ-LL cells which displayed decreased production of
MMP-9 in GD1a-dose- and time-dependent manners.
Depletion of GD1a from cells such as FBJ-S1 and FBJ-
5-30 by an inhibitor of glucosylceramide synthase D-
PDMP or inhibition of sialyltransferase by siRNA target-
ing St3gal2 brought about increase in MMP-9 mRNA pro-
duction. Assessment was subsequently made of MMP-9
activity by gelatin zymography using aliquots of culture
medium subsequent to cell incubation with GD1a. mRNA
production was suppressed by GD1a [5] and accordingly
enzyme activity as determined by gelatin zymography
should also be suppressed GD1a-dose and incubation time
dependently. But the decrease in activity was actually more
than expected. GD1a in the culture medium was kept in the
aliquot to be assayed by gelatin zymography. There is thus
0006-291X/$ - see front matter ? 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.bbrc.2006.08.062
*Corresponding author. Fax: +86 24 23986433.
E-mail address: tcyamagata@cool.odn.ne.jp (T. Yamagata).
www.elsevier.com/locate/ybbrc
Biochemical and Biophysical Research Communications 349 (2006) 426–431
BBRC

Page 2

the possibility that MMP-9 activity in the culture medium
may have been suppressed by GD1a.
Though MMP-9 activity decreased when conditioned
medium containing MMP-9 was pre-incubated with
GD1a in the absence of cells, incubation of the gel with
GD1a following electrophoresis of MMP-9 indicated
GD1a to have no effect on MMP-9 activity in the gel, sug-
gesting GD1a not to inhibit MMP-9 activity. From immu-
noblot analysis, following gel electrophoresis under native
conditions, MMP-9 of the GD1a-treated conditioned
medium was shown less compared to the control, this being
consistent with zymographical results. Following the
reducing PAGE, MMP-9 in the GD1a-treated conditioned
medium was exactly as much as that in the control. Thor-
ough scanning of the zymogram disclosed partial electro-
phoresis of MMP-9 as bands any of which had a
molecular mass of more than 92 kDa. GD1a would thus
appear to assist MMP-9 in its association with certain mol-
ecules in conditioned medium, so as to decrease the amount
of MMP-9 at the site where it would normally appear. This
was not seen to be the case with MMP-2. Caution should
be used in doing gelatin zymography, since certain mole-
cules other than GD1a may also function to decrease
apparent MMP-9 activity in zymography.
Materials and methods
Cell lines and culture. The highly metastatic mouse osteosarcoma cell
line, FBJ-LL, and poorly metastatic cell line, FBJ-S1, were produced from
a FBJ virus-induced osteosarcoma of the BALB/c mouse [7]. FBJ-S1 cells
expressed GM3 and GD1a, whereas FBJ-LL cells expressed GM3 and
slightly expressed GD1a. LA5-30 cells were obtained by transfection of
FBJ-LL cells with b1-4GalNAcT-1 (GM2/GD2-synthase), and mock-
transfectant M5 cells, as control [6]. GD1a expression in LA5-30 cells was
5-fold that of M5. Migration capacity of LA5-30 cells was about one-tenth
that of M5, comparable to the capacity of FBJ-S1 cells. When M5 cells
were inoculated into mice, metastatic nodules were observed in liver, lung,
kidney, and adrenal glands within 4–5 weeks, while LA5-22 cell trans-
plantation did not show any sign of metastasis [6]. The cells were main-
tainedinmediumcontaining
Corporation, NY, USA) supplemented with 10% fetal bovine serum
(TBD-TianJin HaoYang Biological Company, TianJin, China), 100 U/ml
penicillin, and 100 lg/ml streptomycin, and incubated in a humidified
(37 ?C, 5% CO2, and 95% air) incubator (Sanyo, Tokyo, Japan). The cells
were usually grown in a 60 mm culture dish (BD Falcon, CA, USA) and
passaged on reaching 75% confluency. To see the effects of GD1a on cells,
the cells were seeded, washed with serum-free RPMI-1640 at 24 h, and
incubated with or without 50 lM GD1a in the absence of serum for the
period of time indicated. In order to study the effects of GD1a on MMPs,
cells were incubated in the absence of serum for 24 h, and the medium was
collected and further incubated with or without 50 lM GD1a for the time
indicated.
Chemicals and antibodies. Ganglioside GD1a from bovine brain was
purchased from Wako (Osaka, Japan). The primers used in this study were
designed by Primer3 software and synthesized by Invitrogen Shanghai.
Rabbit anti-MMP-9 and anti-MMP-2 polyclonal antibodies were pur-
chased from Chemicon. Horseradish peroxidase (HRP)-linked anti-rabbit
IgG secondary antibody was from Cell Signaling (MA, USA).
RNA extraction and RT-PCR. About 1 · 106cells were harvested and
total RNA was extracted using the Qiagen RNeasy Kit according to
instructions of manufacturer. One microgram of RNA, taken as indicated
by absorption, was subjected to RT-PCR using the TaKaRa RT-PCR kit
(AMV) Ver. 3.0 with a PC707 Program Temp Cont System (ASTEC,
RPMI-1640 (Gibico,Invitrogen
Japan). The product obtained was analyzed by 2% agarose electrophore-
sis. Following ethidium bromide staining (0.05% ethidium bromide in
TAE buffer) for 30 min, the intensity of the stained band was assessed with
a Bio-profile Bio 1D image analyzer (Vilber Lourmat, Marne-la-Vallee,
France) at 312 nm. Primers were synthesized at the Invitrogen (Shanghai,
China) and primer sequences used for PCR in this study were as follows:
for b-actin, sense 50-ACACTGTGTGCCCATCTACGAGG-30and anti-
sense 50-AGGGGCCGGACTCGTCGTCATACT-30; for MMP-9, sense
50-CTGACTACGATAAGGACGGCAA-30and antisense 50-ATACTGG
ATGCCGTCTATGTCG-30; for MMP-2, sense 50-ACCTGGATGCCGT
CGTGGAC-30and antisense 50-TGTGGCAGCACCAGGGCAGC -30.
Western blotting. About 1 · 106cells were lysed in 1 ml sample buffer
(62.5 mM Tris–HCl, 2% SDS, 10% glycerine, 5% b-mercaptoethanol, and
0.03% bromophenol blue) and boiled at 100 ?C for 5 min. An aliquot of
the lysate was loaded onto a 10% SDS–polyacrylamide gel. Following
electrophoresis, the gel was blotted and subjected to Western blotting. The
blotted membrane was reacted with primary antibody at 1/2000 dilution,
followed by horseradish peroxidase (HRP)-conjugated anti-rabbit IgG
secondary antibody (1/3000 dilution). Western blots were visualized by
ECL Western blotting detection reagents (Amersham Biosciences) so as to
enhance chemiluminescence subsequently to be exposed to Fuji XR film.
Lanes were scanned and the optical density was determined by the Bio-
profile Bio 1D image analyzer.
Gelatin zymography. Gelatinase activity was determined according to
the method previously described [7]. In brief, 8% polyacrylamide gel of
1 mm thickness containing 0.3 mg/ml gelatin was used and the proteins
were separated by Lammli’s buffer system. Cells were inoculated into
60 mm culture dish with 4 ml culture medium. After overnight culture
medium was discarded, and cells were washed with RPMI-1640 con-
taining no serum and further incubated with 50 lM GD1a for the time
indicated in the medium without serum. At the indicated time, condi-
tioned medium was obtained to use for the measurement of gelatinase
activity. An aliquot of conditioned medium was mixed with the equal
amount of Lammli’s sample buffer with no reducing reagent and
without heating applied to electrophoresis. The gel following electro-
phoresis was rinsed with 2.5% Triton X-100 for 1 h at room tempera-
ture followed by incubation in the reaction buffer (10 mM CaCl2,
50 mM Tris–HCl, pH 7.4, and 0.02% NaN3) for 16 h at 37 ?C, fixed
with 50% methanol–10% acetic acid for 30 min, stained with 0.02%
CBB in 50% methanol–10% acetic acid for 1 h, and destained with 20%
methanol–10% acetic acid until clear white bands appeared on the blue
background. Gelatin zymography depicts MMPs as negatively stained
bands that were scanned, and the optical density was determined using
the Bio-profile Bio, ID image analyzer.
Results
Inverse regulation of MMP-9 by GD1a
MMP-9 transcription was found inversely regulated by
GD1a content; MMP-9 (95 kDa activity) was low in
FBJ-S1 cells rich in GD1a, while several times more in
FBJ-LL having less GD1a (Fig. 1A). MMP-9 but not
MMP-2 mRNA expression was seen to be down-regulated
by GD1a. Fig. 1A shows the FBJ-LA5-30 cell, a GD1a-
reexpressing FBJ-LL cell variant produced through b1-
4GalNAcT-1 (GM2/GD2-synthase) cDNA transfection
[6], to have caused decrease in MMP-9 production, com-
pared to mock-transfectant M5. Gelatin zymography indi-
cated MMP-9 activity to be inversely related to GD1a.
MMP-9 activity was low in GD1a-rich FBJ-S1 cells,
though by one order of magnitude higher in GD1a-less
FBJ-LL cells (Fig. 1B). FBJ-LA5-30 cells expressing
GD1a to the same extent as FBJ-S1 cells had far less
D. Hu et al. / Biochemical and Biophysical Research Communications 349 (2006) 426–431
427

Page 3

MMP-9 compared to the mock transfectant, FBJ-M5 cells.
This situation was not observed for MMP-2.
GD1a regulation of MMP-9 was previously confirmed
by treatment of FBJ-5-30 cells with D-PDMP, an inhibitor
of glucosylceramide synthesis, or siRNA targeting St3gal2,
which brought about increase in MMP-9 mRNA [5]. GD1a
regulation of MMP-9 was further confirmed by exogenous
addition of GD1a to M5 or LL cells, in both of which
MMP-9 mRNA production was seen to decrease dose
and time dependently. MMP-9 would thus appear nega-
tively regulated by GD1a. Fig. 1C shows GD1a to suppress
MMP-9 mRNA expression of FBJ-M5 cells and this sup-
pression was reflected in MMP-9 activity as determined
by gelatin zymography shown in Fig. 1D. The activity of
MMP-9 (40%) of the cells incubated with GD1a was less
than expected, judging from mRNA suppression (60%).
But there was no effect of GD1a on MMP-2 mRNA or
activity [5].
Possible degradation of MMP-9 by GD1a in conditioned
medium
Examination was made as to whether GD1a suppresses
not only MMP-9 gene transcription but also MMP-9 activ-
ity in culture medium. For this purpose, FBJ-M5 cells rich
in MMP-9 were incubated in serum-free medium for 24 h
to prepare conditioned medium containing MMP-9. Fol-
lowing centrifugation for cell removal, the conditioned
medium was incubated with 50 lM or 100 lM GD1a for
12 h. Each aliquot of incubation medium was assayed for
MMP activity by gelatin zymography. As shown in
Fig. 2A (FBJ-M5 cells) and B (FBJ-LL cells), MMP-9
activity decreased in conditioned medium with GD1a incu-
bation, while MMP-9 activity of conditioned medium incu-
bated without GD1a for 12 h was essentially the same as 0
time control (data not shown). MMP-2 activity remained
unchanged throughout incubation. Fig. 2C shows MMP-
9 activity in the medium incubated with GD1a to have
decreased time dependently.
Decrease in MMP-9 activity may possibly have been due
to its suppression or degradation during incubation in cul-
ture medium in the presence of GD1a. GD1a may bind to
MMP-9 to suppress activity of the latter toward the gelatin
substrate, to catalyze its degradation, or meditate some
proteolytic enzyme for this degradation. To determine
which of these is applicable, freshly obtained conditioned
medium recovered from FBJ-M5 cells following 24 h in
culture without GD1a was separated with a gelatin-con-
taining gel and its SDS removed. The gel was then incubat-
ed in GD1a-contained reaction buffer. But as shown in
Fig. 2D, enzyme activity on the gel was not affected by
incubation with GD1a, thus ruling out any possibility that
putative binding of GD1a to MMP-9 hinders the activity
and that GD1a catalyzes MMP-9 degradation. Rather,
GD1a would appear to assist and facilitate degradation
of MMP-9 in conditioned medium.
Fig. 1. Ganglioside GD1a negatively regulates metalloproteinase-9 (MMP-9) but not MMP-2. FBJ-S1, -LL, -M5, and -LA5-30 cells were cultured under
standard conditions followed by determination of MMP-9 and -2 expression by RT-PCR (A). On aliquots of culture medium of FBJ-S1, -LL, -M5, and -
LA5-30 cells gelatin zymography was carried out (B). For clarification of the effects of GD1a on MMP-9 expression, RNA was extracted from FBJ-M5
cells incubated with or without 50 lM GD1a in serum-free medium for 12 h and examined for the expression of MMP-9 mRNA (C). Gelatin zymography
was done using aliquots of culture medium (D). (C) RT-PCR measurement is normalized by the expression of b-actin in a way so that MMP-9 mRNA
content divided by that of b-actin of the control is expressed as 1 with which mRNA expression of the cell treated with GD1a was compared. MMP-9
activity of the control is taken as 1 in (D). Independent experiments were conducted several times with results essentially in agreement. Typical results are
shown here.
428
D. Hu et al. / Biochemical and Biophysical Research Communications 349 (2006) 426–431

Page 4

Total MMP-9 not affected by GD1a despite apparently
possible suppression
To find whether MMP-9 is degraded in conditioned
medium with GD1a, MMP-9 protein was determined using
an antibody. Gelatin zymography being much more sensi-
tive than immunoblotting, culture medium of GD1a-treat-
ed and control cells was concentrated by about 20 times
using an Amicon filtration tube (for MW 10,000). Aliquots
of condensed medium (Fig. 3A) exhibited essentially the
same activity as that of MMP-9 activity previously noted.
Each aliquot was separated by Laemmli’s electrophoresis
(procedure including boiling for 5 min in the presence of
2-mercaptoethanol),electroblotted
membrane, and immunostained with anti-MMP-9 or
anti-MMP-2 antibody. Fig. 3B shows Western blots of
onto nitrocellulose
MMP-9 and MMP-2 and, surprisingly, no differences in
MMP-9 protein content for culture medium incubated with
GD1a and the control. Any degradation of MMP-9 during
MMP-9 incubation in culture medium appears not to have
occurred through the action of GD1a. Gelatin zymography
shown in Fig. 3A was performed under native conditions
using SDS but not 2-mercaptoethanol, without boiling.
Electrophoresis was thus conducted under the same
(native) conditions as gelatin zymography followed by elec-
troblotting (Fig. 3C). The density of MMP-9 immunoblot-
ting with GD1a incubation was less than for the control,
this being consistent with the results of gelatin zymography
(Fig. 3A), indicating intact MMP-9 protein to be partially
absent from its site. Careful examination of gelatin zymog-
raphy of the GD1a-treated sample (Fig. 3A) disclosed faint
activity bands, each with molecular mass of 120 kDa or
Fig. 2. GD1a in culture medium decreases MMP-9 but not MMP-2 activity. To examine the effects of GD1a on MMP-9, conditioned medium of FBJ-M5
(A) or FBJ-LL (B) cells incubated without GD1a and serum for 24 h was obtained and incubated with 50 lM or 100 lM GD1a for 12 h. Aliquots of the
medium were subsequently assayed for MMPs. (C) Conditioned medium was incubated with 50 lM GD1a for the time indicated and MMPs were
determined by gelatin zymography. (D) Aliquots of conditioned medium were electrophoresed as in gelatin zymography and a portion of the gel
containing MMPs was cut and rinsed in 2.5% Triton X-100 for 1 h at room temperature followed by incubation in reaction buffer containing 50 lM GD1a
for 16 h at 37 ?C. The control gel was incubated with reaction buffer without GD1a. (A–C) Indicate experiments two times separately with similar results
obtained. Typical zymograms appear on the left of the figure. In (D) the control and sample have been electrophoresed in duplicate.
D. Hu et al. / Biochemical and Biophysical Research Communications 349 (2006) 426–431
429

Page 5

exceeding 190 kDa, possibly owing to the association of
MMP-9 with some molecules present in culture medium
in the presence of GD1a.
Discussion
Due to its simplicity and sensitivity and lack of any bet-
ter method, gelatin zymography is used for matrix metallo-
proteinase measurement in cells and tissue with or without
drug application [8–10], and the term zymography often
means gelatin zymography for assay of MMPs, though
zymography has been used other than for the detection
and measurement of MMPs, such as application to other
enzymes [11–13].
MMP-9 and MMP-2 are secreted by cells and thus,
for examination of its effects on MMPs, a drug is intro-
duced into the culture medium. Aliquots to be assessed
for effect on MMPs consequently contain the drug.
Should MMP activity be shown suppressed on a gelatin
zymogram, the suppression may possibly be explained
by the following: (1) The drug suppression of MMP
mRNA expression with consequent reduction in enzyme
protein content, (2) impairment of MMP protein synthe-
sis, (3) inhibition of MMP secretion into the medium, (4)
degradation of MMP secreted in the medium by GD1a
itself or degradation mediated by GD1a, or (5) denatur-
ing MMP in the medium via drug interaction during gel-
atin zymography. Any one or all these possibilities may
be applicable.
GD1a administered to FBJ-M5 cells inhibited MMP-9
mRNA (Ref. [5] and Fig. 1C), reduction in enzyme activ-
ity (Fig. 1D), this possibly being a reflection of decrease
in mRNA, shows at least possibility (1) to be applicable.
MMP-9 activity (40%) of FBJ-M5 cells incubated with
GD1a was less than expected, based on the degree of
mRNA suppression (60%), so that GD1a would appear
to suppress not only MMP-9 gene transcription but
MMP-9 activity as well, in culture medium. Pure
Fig. 3. MMP-9 activity is apparently suppressed by GD1a but total MMP-9 does not change. FBJ-M5 cells were incubated without GD1a to prepare
conditioned medium which was subsequently incubated with or without 50 lM GD1a for additional 6 h and then concentrated by 20 times using an
Amicon filtration tube (for MW 10,000); its MMP activity was then determined (A). Molecular mass markers appear in the right lane of (A). Western
blotting of MMP-9 and -2 was conducted subsequent to Laemmli’s electrophoresis (procedure including boiling for 5 min in the presence of 2-
mercaptoethanol) (B), or electrophoresis under native conditions as for gelatin zymography (C). Measurement of MMP activity and protein was each
expressed as the mean value of two independent experiments. Typical results for gelatin zymography or Western blots appear on the left of the figures.
430
D. Hu et al. / Biochemical and Biophysical Research Communications 349 (2006) 426–431

Page 6

MMP-9 being difficult to obtain, conditioned medium
containing MMP-9 secreted by FBJ-M5 cells was used
in this study to find whether additional incubation with
GD1awoulddiminish enzyme
zymography.
Incubation of conditioned medium containing MMP-9
with GD1a was noted to decrease MMP-9 activity (Figs.
2A–C), suggesting (4) and/or (5) to be applicable. But incu-
bation in GD1a-containing reaction buffer following elec-
trophoresis of conditioned medium in the gel containing
gelatin failed to diminish enzyme activity (Fig. 2D), indi-
cating GD1a itself not to degrade MMP-9 or lessen
MMP-9 activity. Still, GD1a may possibly mediate the deg-
radation of MMP-9 by the enzyme in culture medium.
Staining of MMP-9 with antibody following electrophore-
sis under reducing conditions (Fig. 3B) showed protein
content not to change during conditioned medium incuba-
tion with GD1a. GD1a thus clearly does not mediate
MMP-9 degradation in conditioned medium, and so (4)
and (5) would not be applicable.
With electrophoresis of conditioned medium incubated
with GD1a under native conditions, immunoblotting and
staining with antibody, decrease in MMP-9 subsequent
to GD1a treatment compared to the control (Fig. 3C)
was essentially the same as that indicated on the gelatin
zymogram (Fig. 3A), indicating MMP-9, once secreted
into the culture medium, to be affected by GD1a in such
a way as to cause MMP-9 to form a complex with
GD1a or with other molecule(s) assisted by GD1a. The
zymogram also showed several activity bands corre-
sponding topositions of
(Fig. 3D). There thus should also be the possibility (6)
that GD1a with micelles binds MMP-9 or GD1a assists
MMP-9 to bind to some molecules in culture medium to
form a higher molecular weight complex, thus decreasing
apparent activity at the site where MMP-9 would nor-
mally appear. Human neutrophils have been shown to
contain 25 kDa protein which, on binding to MMP-9,
forms an activity band at 135 kDa in zymography [14].
Synovial fluid from patients with rheumatoid arthritis
showed MMP-9 activity at 130 and 224 kDa as well as
92 kDa [15]. Analysis of mouse uterus indicated the pres-
ence of high molecular mass proteins that were immuno-
reactive toward the anti-MMP-9 antibody [16]. Pro-
MMP-9 binds to TIMP-1 through a hemopexin domain
and active MMP-9 binds to TIMP-1 and TIMP-2
[17,18], though either of which bound to proMMP-9 or
MMP-9 is liberated in an electric field in gelatin zymog-
raphy. MMP-9 has thus been shown to bind to other
molecules, but in this study was noted to bind to certain
molecules in culture medium in the presence of GD1a.
The mechanism for this and nature of the bound mole-
cules remain to be elucidated. The possibilities (2) and
(3) have yet to be demonstrated, but it is beyond the
scope of this work.
activity in gelatin
highermolecularmass
References
[1] S. Curran, G.I. Murray, Matrix metalloproteinases: molecular aspects
of their roles in tumour invasion and metastasis, Eur. J. Cancer 36
(2000) 1621–1630.
[2] G. Klein, E. Vellenga, M.W. Fraaije, W.A. Kamps, E.S.J.M. de Bont,
The possible role of matrix metalloproteinase (MMP)-2 and MMP-9
in cancer, e.g. acute leukemia, Crit. Rev. Oncol./Hematol. 50 (2004)
87–100.
[3] J.E. Rundhaug, Matrix metalloproteinases and angiogenesis, J. Cell.
Mol. Med. 9 (2005) 267–285.
[4] L.A. Liotta, W.G. Stetler-Stevenson, Tumor invasion and metastasis:
an importance of positive and negative regulation, Cancer Res. 51
(1991) 5054–5059.
[5] D. Hu, Z. Man, T. Xuan, P. Wang, T. Takaku, S. Hyuga, X.S. Yao,
T. Sato, S. Yamagata, T. Yamagata, Ganglioside GD1a regulation of
matrix metalloproteinase-9 (MMP-9) expression in mouse FBJ cell
Lines: GD1a suppression of MMP-9 expression stimulated by PI3K-
Akt and p38 though not by the Erk signaling pathway, 2006,
Submitted for publication.
[6] S. Hyuga, S. Yamagata, Y. Takatsu, M. Hyuga, H. Nakanishi, K.
Furukawa, T. Yamagata, Suppression of FBJ-LL cell adhesion to
vitronectin by ganglioside GD1a and loss of metastatic capacity,
International J. Cancer 83 (1999) 685–691.
[7] S. Yamagata, R. Tanaka, Y. Ito, S. Shimizu, Gelatinase of murine
metastatic tumor cells, Biochem. Biophys. Res. Commun. 158 (1989)
228–234.
[8] S. Yamagata, Y. Yoshii, J.G. Suh, R. Tanaka, S. Shimizu, Occur-
rence of an active form of gelatinase in human gastric and colorectal
carcinoma tissues, Cancer Lett. 59 (1991) 51–55.
[9] K. Tamakoshi, F. Kikkawa, A. Nawa, H. Ishikawa, K. Mizuno, A.
Tamakoshi, S. Yamagata, N. Suganuma, Y. Tomoda, Characteriza-
tion of extracellular matrix-degrading proteinase and its inhibitor in
gynecologic cancer tissues with clinically different metastatic form,
Cancer 76 (1995) 2565–2571.
[10] M. Hamaguchi, S. Yamagata, A.A. Thant, H. Xiao, H. Iwata, T.
Mazaki, H. Hanafusa, Augmentation of metalloproteinase (gelatin-
ase) activity secreted from Rous sarcoma virus-infected cells corre-
lates with transforming activity of src, Oncogene 10 (1995) 1037–
1043.
[11] R.O. Miura, S. Yamagata, Y. Miura, T. Harada, T. Yamagata,
Analysis of glycosaminoglycan-degrading enzymes by substrate gel
electrophoresis (zymography), Anal. Biochem. 225 (1995) 333–340.
[12] M. Cao, M. Sahmi, J.G. Lussier, C.A. Price, Plasminogen activator
and serine protease inhibitor-E2 (protease nexin-1) expression by
bovine granulosa cells in vitro, Biol. Reprod. 71 (2004) 887–893.
[13] S. Hattori, H. Fujisaki, T. Kiriyama, T. Yokoyama, S. Irie, Real-time
zymography and reverse zymography: a method for detecting
activities of matrix metalloproteinases and their inhibitors using
FITC-labeled collagen and casein as substrates, Anal. Biochem. 301
(2002) 27–34.
[14] L. Kjeldsen, A.H. Johnsen, H. Sengeløv, N. Borregaard, Isolation
and primary structure of NGAL, a novel protein associated with
human neutrophil gelatinase, J. Biol. Chem. 268 (1993) 10425–10432.
[15] G.S. Makowskia, M.L. Ramsby, Zymographic analysis of latent and
activated forms of matrix metalloproteinase-2 and -9 in synovial fluid:
correlation to polymorphonuclear leukocyte infiltration and in
response to infection, Clin. Chim. Acta 329 (2003) 77–81.
[16] J. Hu, X. Zhang, W.B. Nothnick, T.E. Spencer, Matrix metallopro-
teinases and their tissue inhibitors in the developing neonatal mouse
uterus, Biol. Reprod. 71 (2004) 1598–1604.
[17] I. Stamenkovic, Matrix metalloproteinases in tumor invasion and
metastasis, Cancer Biol. 10 (2000) 415–433.
[18] K. Brew, D. Dinakarpandian, H. Nagase, Tissue inhibitors of
metalloproteinases: evolution, structure and function, Biochim. Bio-
phys. Acta 1477 (2000) 267–283.
D. Hu et al. / Biochemical and Biophysical Research Communications 349 (2006) 426–431
431