| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Kolling Institute of Medical Research, University of Sydney, Royal North Shore Hospital, St. Leonards, New South Wales 2065, Australia
Address all correspondence and requests for reprints to: Janet Martin, Kolling Institute of Medical Research, University of Sydney, Royal North Shore Hospital, St. Leonards, New South Wales 2065. E-mail: janetlm{at}med.su.oz.au
 |
Abstract |
|---|
 Top Abstract IntroductionMaterial and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsDiscussionReferences |
|---|
The actions of the IGFs are modulated by members of a family of six binding proteins (IGFBP-1 to -6) with high affinity for the IGFs (6). The IGFBPs differ in core protein structure and degree or type of posttranslational modification, with proteolyzed, glycosylated and/or phosphorylated variants occurring for each protein (7, 8). In addition, there is evidence that IGFBP-1, -2, -3, and -5 associate with components on the cell surface or within the extracellular matrix (9, 10, 11). The IGFBPs modulate the proliferative and mitogenic effects of the IGFs by either inhibiting IGF action through limiting growth factor access to specific cell surface receptors, or by enhancing IGF activity (12, 13).
The development and progression of malignant diseases may be associated with abnormal secretion of proteases (14), some of which have now been shown to have activity against one or more of the IGFBPs. For example, IGFBP-3 can act as a substrate for cathepsin D and prostate specific antigen (PSA), proteases associated with malignant disease of the breast and prostate, respectively (15, 16). Limited proteolysis of the IGFBPs results in modification of their IGF-binding properties, usually decreasing the affinity of the IGFBP for its ligand (17, 18, 19). Such a change may lead to release of the IGF from inhibition with subsequent restoration of its mitogenic activity, and has also been suggested to be the mechanism whereby enhancement of IGF mitogenic activity by IGFBP-3 occurs (20). Therefore, IGFBP proteolysis may be a significant contributing factor in uncontrolled cell proliferation, and the factors regulating this process – production, activation, or inhibition of the proteases or their inhibitors – are potentially important elements in the IGF regulatory system. In this study we investigated the presence of IGFBP protease activity secreted by MCF-7 breast cancer cells, and describe a novel mechanism by which IGFs may regulate IGFBP protease activity.
|
Material and Methods |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsDiscussionReferences |
|---|
Cell culture
Human breast carcinoma cells (MCF-7) were obtained from American
Type Culture Collection (Rockville, MD). Cells were maintained in 75
cm2 flasks (Corning, NY) in RPMI medium supplemented with
10% FCS, 4 mM glutamine, 10 mg/L bovine insulin, and 25
mM HEPES. Cultures were grown to confluency and split 1:3
weekly. The medium was changed 24 h after passaging, then once
every other day. Cell-conditioned medium was prepared by incubating
confluent cultures with serum-free medium (SFM, phenol red-free RPMI
supplemented with 1 g/L BSA, 60 mg/L penicillin, 100 mg/L streptomycin
sulfate, 4 mM glutamine, and 25 mM HEPES) for
24 h, discarding the medium, and changing to fresh SFM for a
further 48 h. Pooled conditioned media were centrifuged to remove
cell debris and stored at -20 C until required for analysis.
For IGF-I-stimulation experiments, 24-place multiwell plates (Corning, NY) were seeded with approximately 50,000 cells/cm2 and grown in medium containing 10% FCS, 4 mM glutamine, and 10 mg/L bovine insulin for 5–7 days (with medium changes) until confluent. The cultures were then changed to SFM and maintained for 24 h before the addition of IGF-I in SFM. Cultures were incubated for a further 72 h, then the medium was collected and frozen at -20 C until analyzed.
RIAs
IGFBP-3 in MCF-7 cell-conditioned medium and gel chromatography
eluates was determined by RIA as described previously (22), using
antiserum R100 at 1:100,000 final dilution (23).
IGFBP-3 proteolysis
Medium samples were prepared for protease analysis by
equilibration with sodium acetate buffer (0.1 M, pH 5.5 or
as indicated for individual experiments). Five milliliters of
conditioned MCF-7 medium was concentrated to approximately 100 µl by
centrifugation through a Centricon-10 ultrafiltration unit (Amicon,
Beverley, MA), flushed with a total of 10 ml sodium acetate buffer, and
diluted to starting volume (5 ml) in the same buffer. To detect
protease activity, pure human plasma-derived IGFBP-3 (50 ng) was
incubated with 50 µl prepared medium in a final volume of 100 µl in
sodium acetate buffer at 37 C for 24 h. Protease inhibitors were
diluted in this same buffer for addition as indicated for individual
experiments. Proteolyzed samples were analyzed by SDS-PAGE and
immunoblot or fast performance liquid chromatography as indicated
below.
Superose-12 column fractionation
Size fractionation was performed by FPLC on a Superose-12 10/30
column (Pharmacia, Uppsala, Sweden). Before each run, the column was
equilibrated for 30 min at 1 ml/min with 50 mM sodium
phosphate buffer containing 0.1 M sodium chloride and 0.2
g/L sodium azide, pH 6.5. Samples (200 µl) were applied to the column
using a Pharmacia injection valve V-7, and the eluate was monitored for
absorbance at 280 nm through a Pharmacia UV-1 single path monitor.
Fractions of 0.5 ml were collected and stored at -20 C before
analysis. The column was calibrated by monitoring elution positions of
human IGF-I (7.6 kDa), human GH (22 kDa), human IGFBP-2 (32 kDa), and
BSA (67 kDa) under the same conditions.
SDS-PAGE and western blotting
Protein separation was carried out on 12% SDS-polyacrylamide
gels with 4% stacking gel as described previously (23). Before
electrophoresis, samples and prestained low molecular weight
electrophoresis standards (Pharmacia) were prepared by the addition of
concentrated sample buffer to give a final concentration of 0.0125
M Tris-HCl (pH 6.8) containing 3% SDS, 10% glycerol.
Samples were then heated at 90 C for 3 min immediately before loading
onto gels. Electrophoretic separation was carried out at 45 V for
2 h, then at 100 V for 16–17 h.
For ligand and immunoblot analysis, proteins were electrophoretically transferred from the polyacrylamide gels onto Hybond C nitrocellulose membrane (Amersham, Bucks, UK) as previously described (23). After the transfer, the membranes were incubated in Tris-buffered saline (TBS, 10 mM Tris, 50 mM sodium chloride, pH 7.4) containing 10 g/L BSA and 0.2 g/L sodium azide (blocking buffer) for 3 h at 37 C. The blocked membrane was then incubated with [125I]IGF-I (1 x 106 cpm/50 ml) or with rabbit antihuman IGFBP-3 antiserum (R100) at a final dilution of 1:5000 in incubation buffer (TBS containing 0.2 g/L sodium azide, 10 g/L BSA, and 0.05% Nonidet P-40) overnight at 22 C. The membranes were then washed to remove unbound tracer or antibody, and immunoblots were incubated for a further 2 h with radioiodinated protein A (1 x 106 cpm/50 ml) and washed again as before. IGFBP-3 and related protein fragments were visualized and quantified by phosphorimage analysis (Molecular Devices, Sunnyvale, CA).
Binding assays
Competitive binding assays were carried out as previously
described (21) to compare binding of IGF-I, [des(1, 2, 3)]IGF-I, and
[LR3]IGF-I to IGFBP-3. Briefly, IGF-I (0.01–2.5
ng/tube), [des(1, 2, 3)]IGF-I (0.1–25ng/tube), and
[LR3]IGF-I (1–250 ng/tube) were incubated for 2 h
at 22 C with 0.5 ng IGFBP-3 and [125I]IGF-I (1 x
104 cpm) in a final volume of 300 µl binding buffer
containing 0.1 M sodium phosphate, 0.2 g/L sodium azide,
2.5 g/L BSA, at pH 6.5. Anti-IGFBP-3 antibody (R100, 0.5 µl) was then
added in 25 µl, and tubes were incubated for a further 1 h at 22
C. Complexes were precipitated by the addition of goat antirabbit
-globulin (2.5 µl) and 1 ml cold 60 g/L polyethylene glycol 6000
in 0.15 M NaCl, and centrifugation for 20 min at 4000 rpm.
Supernatants were decanted and pellets counted for 2 min.
|
Results |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsDiscussionReferences |
|---|
). In some experiments a
minor band of 21 kDa was also observed. When the pH of the media was
adjusted to 4.5–5.5, IGFBP-3 degradation was detected by the
simultaneous loss of the 40- to 45-kDa doublet and the 30-kDa IGFBP-3
and a significant accumulation of the 21-kDa species. No degradation of
IGFBP-3 occurred during incubation in unconditioned media at any pH
value (not shown).
|
). Both the 40- to 45-kDa doublet
and the 30-kDa IGFBP-3 species disappeared with increasing time, and an
increase in the 21-kDa fragment was apparent over the 24-h incubation
period. This was verified by phosphorimage analysis of the data (Fig. 2B); within 9 h of incubation 90% of the 40- to 45-kDa IGFBP-3
had disappeared, with complete loss of intact IGFBP-3 apparent by
24 h.
|
). Fifty microliters
of each eluted fraction was incubated with 50 ng IGFBP-3 at pH 5.5 for
24 h, and the presence of IGFBP-3 protease activity was detected
by immunoblot. Fractions 27–28 of the MCF-7 medium were most active in
causing a degradation of 40- to 45-kDa IGFBP-3. This corresponds to an
approximate molecular mass for the protease activity of 25–30 kDa and
may represent more than one enzyme.
|
). Proteolysis of the 40- to 45-kDa IGFBP-3 doublet was
substantially blocked by incubation of medium with the cysteine and
serine protease inhibitor leupeptin (250 µM) or the
divalent cation chelator EDTA (25 mM). However, inhibitors
of serine proteases (benzamidine and aprotinin) did not prevent the
breakdown of the 40- to 45-kDa IGFBP-3. These data suggest that the
IGFBP-3 degrading protease in MCF-7 cells is a cysteine protease
with cation dependence. The pH optimum, size, and inhibitor profile of
this protease activity are consistent with the presence of a lysosomal
enzyme such as cathepsin B. However,
trans-epoxysuccinyl-L-leucylamido(4-guanidino)-butane
(E-64), an inhibitor of a number of cysteine proteases including
cathepsin B, did not inhibit proteolysis of the 40- to 45-kDa IGFBP-3
(data not shown). In addition, concentrations of pure cathepsin B up to
50 µg/100 µl only caused partial proteolysis of IGFBP-3, whereas
immunoblotting for cathepsin B in 100 µl conditioned medium gave a
barely detectable signal under conditions in which 1 µg of the pure
enzyme was easily detectable (data not shown). Finally, immunodepletion
of conditioned medium using anticathepsin B antibody did not
consistently prevent IGFBP-3 proteolysis (data not shown).
|
30 kDa) by immunoblot analysis but elutes in the
position expected for intact IGFBP-3 on FPLC, suggesting that the
30-kDa form is generated during the electrophoresis (24). In view of
this, we examined whether the MCF-7 medium-proteolyzed IGFBP-3 fragment
estimated to be 21 kDa by SDS-PAGE appears smaller than intact IGFBP-3
on FPLC. IGFBP-3 incubated with conditioned medium at neutral pH or at
pH 5.5 for 24 h was subjected to Superose-12 chromatography, and
the resulting eluted fractions analyzed for immunoreactive IGFBP-3 by
RIA. As shown in Fig. 5, IGFBP-3 incubated with medium
at neutral pH eluted in fractions 26–27, consistent with a molecular
mass of 40–45 kDa. By contrast, IGFBP-3 incubated with medium at pH
5.5 eluted in fractions 29–31, corresponding to a molecular mass of
approximately 20 kDa. These results confirm our IGFBP-3 immunoblot
findings, indicating that the 21-kDa product of IGFBP-3 proteolysis
exists at that size in medium regardless of the method used to identify
it.
|
). Like the
30-kDa band, the 21-kDa IGFBP-3 fragment detectable by immunoblot (Fig. 6, left panel, lane 2) could not be detected by an IGF-I
ligand blot (Fig. 6, right panel, lane 2), indicating that
the proteolyzed 21-kDa IGFBP-3 fragment no longer retained IGF-I
binding activity. Similarly, IGF-II ligand blotting failed to reveal
this protein (not shown).
|
, lane 3); phosphorimage analysis
indicated that IGFBP-3 proteolysis was reduced by 90% in the
presence of medium from IGF-treated MCF-7 cells.
To test whether the inhibitory effect of IGF-I was due to inhibition of
protease production by MCF-7 cells or inhibition of protease activity,
we examined the degradation of 500 ng/ml IGFBP-3 by conditioned medium
from untreated MCF-7 cells in the presence of exogenous IGF-I in a
cell-free system. As shown by the immunoblot in Fig. 7, the addition of 800 ng/ml IGF-I resulted in apparent stabilization of
the 40- to 45-kDa IGFBP-3, completely inhibiting its disappearance. At
the same time, however, the 30-kDa IGFBP-3 apparent in the control
incubation with unconditioned medium (Fig. 7, lane 1) was no longer
visible, and the 21-kDa fragment band was only slightly reduced in
intensity (Fig. 7, lane 3). This result indicates that although the
proteolysis of 40- to 45-kDa IGFBP-3 is prevented in the presence of
IGF-I, the 30-kDa IGFBP-3 still undergoes cleavage. This may indicate
the activity of more than one protease, differential activity of a
single enzyme towards the 30- and 40- to 45-kDa forms of IGFBP-3, or
that IGF-binding to intact IGFBP-3 inhibits its proteolysis.
|
Mechanism of IGF inhibition of IGFBP-3 proteolysis
Changes in IGFBP-3 proteolysis in the presence of IGF-I have been
thought to involve IGF binding to IGFBP-3. We examined whether this was
the mechanism of IGF inhibition of the MCF-7-derived IGFBP-3 protease
using [LR3]IGF-I and [des(1, 2, 3)]IGF-I, two analogs of
IGF-I that have reduced binding to IGFBPs (29). The markedly reduced
binding affinity of [LR3]IGF-I and [des(1, 2, 3)]IGF-I to
natural human IGFBP-3 was confirmed by a competitive binding assay
using increasing concentrations of unlabeled IGF-I,
[LR3]IGF-I, and [des(1, 2, 3)]IGF-I to displace
radioiodinated IGF-I from pure IGFBP-3. As shown in Fig. 8A, [des(1, 2, 3)]IGF-I and [LR3]IGF-I had
approximately 5% and 0.2% the potency of IGF-I for binding to
IGFBP-3, respectively. When [LR3]IGF-I was tested for
inhibition of 40- to 45-kDa IGFBP-3 proteolysis in the absence of
cells, immunoblot analysis indicated that this IGF analog, which bound
IGFBP-3 with markedly reduced affinity compared with IGF-I, was a
potent inhibitor of proteolysis of IGFBP-3 by MCF-7-conditioned medium
(Fig. 8B), with 50 ng/ml able to prevent degradation of 500 ng/ml
IGFBP-3.
|
. IGF-I was the most potent inhibitor of
IGFBP-3 proteloysis, with half-maximal effect (EC50)
apparent at 15–30 ng/ml IGF-I (Fig. 8C). IGF-II, which shows similar
IGF-binding to IGFBP-3 as IGF-I (21), showed a similar inhibitory
effect on IGFBP-3 proteolysis as IGF-I (data not shown). In contrast
with its 500-fold reduced affinity for IGFBP-3 however,
[LR3]IGF-I showed only a 4- to 5-fold reduced potency
compared with IGF-I at inhibiting IGFBP-3 fragmentation, with an
EC50 of approximately 80 ng/ml in four experiments, whereas
[des(1, 2, 3)]IGF-I had an approximately 10-fold reduced potency
(EC50 300 ng/ml from two experiments).
[LR3]IGF-I had a similar inhibitory effect on the
proteolysis of [125I]-labeled IGFBP-3, showing almost the
same potency as IGF-I (data not shown). These results indicate that the
inhibitory effects of IGFs on the 40- to 45-kDa IGFBP-3 protease
activity have a pattern of specificity different from their specificity
for the IGF-binding site of IGFBP-3, and therefore occur independently
of interaction between the IGF and IGFBP-3. |
Discussion |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsDiscussionReferences |
|---|
In the present study, incubation of plasma-derived IGFBP-3 with MCF-7-conditioned medium at pH 5.5 yielded an immunoreactive product of 21 kDa. A wide range of fragments generated by IGFBP-3 proteolysis has been reported. Although most of this diversity could be due to the different cleavage sites of these proteases, some may arise from inaccuracies inherent to the method of size estimation or the use of different IGFBP-3 as substrate. Thus, recombinant E coli-derived IGFBP-3 will yield nonglycosylated fragments of lower molecular mass than the recombinant Chinese hamster ovary (CHO) cell-derived or plasma-purified proteins. For example, cathepsin D hydrolysis of CHO IGFBP-3 yields major products of 24 and 11 kDa (15), the major matrix metalloproteinases 1–3 give rise to 10.8- and 19-kDa species from nonglycosylated IGFBP-3 (33), plasmin proteolysis of human osteosarcoma-derived IGFBP-3 generates fragments of 30, 20, and 16 kDa (34), and PSA degradation of nonglycosylated IGFBP-3 yields predominantly 19.5- and 16-kDa species (16). Because the two components of 40- to 45-kDa plasma IGFBP-3 are believed to be due to variable glycosylation of Asn172 (35), neither the 30-kDa component of plasma IGFBP-3 nor the 21-kDa product of MCF-7 proteolysis of likely to contain this glycosylation site, because both appear as single bands by SDS-PAGE. Indeed, the 30-kDa form must be cleaved just amino-terminally of Asn172 to account for its size. Furthermore, the detection of the 21-kDa fragment by antiserum R100, which appears to have predominantly amino-terminal specificity (36), suggests that the 21-kDa fragment contains the amino-terminus of IGFBP-3. The exact composition of this fragment, and its relationship to fragments generated by the actions of other proteases on IGFBP-3, await purification and amino acid analysis.
The 21-kDa fragment resulting from IGFBP-3 proteolysis was not seen on an IGF-ligand blot, indicating a loss of IGF binding affinity. Changes in IGF affinity have been reported for other IGFBP fragments generated by the action of a variety of different proteases; pregnancy-associated protease was first identified on the basis of its degradation of IGFBP-3 to a form not detectable by ligand blot (17, 18). PSA cleaves IGFBP-3 to yield products with markedly reduced affinity for IGF-I and slightly reduced affinity for IGF-II (26, 37), whereas limited proteolysis of IGFBP-2 (28, 38), IGFBP-4 (27, 39, 40), and IGFBP-5 (25) will also alter their apparent IGF-binding affinity by IGF-ligand blot or solution-binding analysis. The physiological consequences of reduced IGF binding affinity have been difficult to determine. The inhibitory effects of IGFBP-3 on IGF-stimulated mitogenesis in prostate epithelial cells are reversed in the presence of IGFBP-3-degrading PSA (37), whereas inhibition of IGFBP-3 proteolysis in the PC-3 prostate carcinoma cell line correlates with increased IGF-mediated cell proliferation (41). In addition, there is some evidence that the reported potentiating effects of IGFBPs on IGF bioactivity may be attributed to alterations in IGF affinity resulting from proteolytic modification of the binding protein (20, 42); however, the mechanism by which decreased IGF binding activity may lead to an enhancement, rather than merely loss of inhibition, of IGF activity is as yet unexplained. Moreover, a recent study has shown that of two IGFBP-3 fragments derived from proteolysis of E coli-derived recombinant human IGFBP-3 by plasmin, one retained weak IGF-binding activity and antagonized IGF-stimulated mitogenesis, whereas the other non-IGF-binding fragment inhibited both IGF-I- and insulin-stimulated mitogenesis in chick embryo fibroblasts, consistent with intrinsic inhibitory activity of this IGFBP fragment (34). This argues against the generally accepted view that fragmentation of an IGFBP leads to increased mitogenesis through increased IGF bioactivity, and raises the intriguing possibility that proteolyzed IGFBPs have IGF-independent effects on cell proliferation. It remains to be determined whether the 21-kDa IGFBP-3 we identified in the present study has inhibitory or stimulatory IGF-independent effects on MCF-7 cell proliferation.
In vitro secretion of IGFBP-3 proteases has been described for a wide variety of normal and malignant cell types, including fibroblasts, osteoblasts, and osteosarcomas (15); prostate cells (37); and ovarian granulosa cells (43). Although some of these include known proteases such as matrix metalloproteases (33) or cathepsins (15), characterization of the protease described in the present study (i.e. for pH optimum, size, and class of protease) indicates lack of identity with other enzymes known to have activity against IGFBP-3. Its properties are similar to those of cathepsin B: optimal activity at pH 4.5–5.5, molecular mass of approximately 30 kDa, and inhibition by leupeptin and EDTA (44). However, degradation of IGFBP-3 was not inhibited by trans-epoxysuccinyl-L-leucylamido(4-guanidino)-butane (E-64), and immunodepletion of medium with cathepsin B antibodies did not consistently remove its protease activity. Further, pure cathepsin B appears to be a poor IGFBP-3 protease in vitro. Studies are currently underway to purify and identify this protease. In preliminary experiments, we detected similar IGFBP-3 proteolytic activity at pH 5.5 in medium derived from HEP G2 human hepatoma cells but not in Hs578-T breast cancer cells or neonatal fibroblast cells (our unpublished observations) suggesting either lack of production of the protease in these cell lines or the presence of endogenous inhibitors.
Detection of a protease in MCF-7 medium with optimal activity at pH 5.5 is in contrast with the findings of Conover and DeLeon (15) who described IGFBP-3 protease activity in MCF-7 cell-conditioned medium (later ascribed to cathepsin D) at pH 3. This inconsistency is difficult to interpret due to differences in materials used and assaying techniques employed in the two studies. The source of the IGFBP-3 preparations differed in that our protease assays measured degradation of human plasma IGFBP-3 rather than recombinant CHO-derived IGFBP-3, and much higher concentrations (500 ng/ml) of IGFBP-3 were digested in our protease assays compared with the tracer amounts used in the earlier study. It is possible that there may be differences in susceptibility to protease action between recombinant and plasma-derived IGFBP-3 at pH 5.5, or alternatively, the concentration of cathepsin D present in MCF-7-conditioned medium may be insufficient to degrade this quantity of IGFBP-3 at pH 3.
Although maximal activity of the enzyme described in the present study occurs at pH values below those seen in the extracellular environment of normal cells, it is well known that tumor cells acidify their microenvironments by the actions of a variety of mechanisms including Na+/H+ exchange and H+-ATPase activity (45). pH values as low as 5.5 have been reported in human tumors, although not specifically demonstrated in breast carcinomas (46). It may be speculated, therefore, that proteases such as those described in this report may be active in the extracellular space surrounding breast cancer cells. Low pH values such as these may also be found in intracellular organelles such as endosomes, and it is possible that IGFBP-3 proteolysis occurs primarily within the cell, perhaps yielding forms with as yet unrecognized intracellular actions.
Regulation of IGFBP protease activity by IGFs is now a well-recognized phenomenon; however, inhibition of protease activity by IGFs in a cell-free system, independent of interaction between growth factor and binding protein, is a novel finding. Previous studies have demonstrated that the effects of IGFs on IGFBP proteolysis may derive from occupancy of the IGFBP increasing or decreasing its susceptibility to proteolytic fragmentation (presumably through changes to the tertiary structure of the IGFBP in the presence of its ligand) or through receptor-mediated changes in the production of proteases or protease inhibitors. For example, IGFBP-2 proteolysis in smooth muscle cells and IGFBP-4 proteolysis in fibroblasts is enhanced by IGF-I and IGF-II apparently through increased protease production (27, 28), whereas IGFBP-3 and IGFBP-5 proteolysis in fibroblasts is inhibited by IGFs and IGF analogs that bind the IGFBPs but not by analogs that do not (25), thereby implicating direct IGF-IGFBP interaction in the process of inhibition. In the present study, we found that not only was IGF-I able to inhibit IGFBP-3 proteolysis by the MCF-7-conditioned medium, but that [LR3]IGF-I, an IGF analog that does not bind IGFBP-3, had the same effect when added to reaction mixtures containing IGFBP-3 and medium from untreated cells. The use of a cell-free system ruled out the possibility of altered protease/protease inhibitor production, whereas the efficacy of the nonbinding IGF analog indicated that the mechanism by which IGFs were inhibiting protease activity was unrelated to their binding with IGFBP-3. Grimes and Hammond (43) also noted reduced IGFBP-3 protease activity in medium from porcine ovarian granulosa cells treated with IGF-I, IGF-II, or [LR3]IGF-I, but did not distinguish between an IGF effect on protease production, presumably receptor mediated, or activity. Our findings suggest that there may be direct interaction between the protease that generates the 21-kDa IGFBP-3 fragment and IGF-I.
The key difference between our observation and previous reports of IGF inhibition of IGFBP-3 proteolysis is that we demonstrated inhibition by free rather than IGFBP-bound IGFs, because the nonbinding analog [LR3]IGF-I is inhibitory. If IGFBP-3 proteolysis increases IGF bioavailability by releasing bound IGFs, the inhibition of proteolysis by free IGFs would act to limit the further release of IGFs from IGFBP-3; i.e. our data support a negative feedback loop regulating free IGFs. Although this would appear to limit the potential mitogenic activity of IGFs on breast cancer cells, this assumes that only free IGFs are growth stimulatory for these cells, i.e. that IGF binding to IGFBP-3 is inhibitory. Although incubation of MCF-7 cells with exogenous recombinant human IGFBP-3 has been shown to inhibit estradiol-stimulated cell proliferation (47), the opposite has also been shown: transfection of MCF-7 cells with cDNA encoding IGFBP-3 enhances IGF-stimulated MCF-7-cell proliferation (48). It is also likely that the IGFBP-3 thus protected from degradation will have effects on other components of the IGF system, both in the epithelial cells and in stromal cells such as fibroblasts: IGFBP-3 attenuates IGF-stimulated IGFBP-5 production by, or release from, fibroblasts (49), and inhibits proteolysis of IGFBP-4 by fibroblast conditioned medium, apparently through direct interaction with the protease (50, 51). Taken together, these observations point to an extremely complex regulatory system, involving both positive and negative feedback mechanisms. Dissecting the roles and physiological significance of the individual players of this system promises to be a challenging task, and one that is essential to a complete understanding of the IGF axis in malignant disease.
|
Footnotes |
|---|
Received September 16, 1996.
|
References |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  P. W. Parodi Dairy Product Consumption and the Risk of Breast Cancer J. Am. Coll. Nutr., December 1, 2005; 24(suppl_6): 556S - 568S. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Grohmann, E. Foulstone, G. Welsh, J. Holly, J. Shield, E. Crowne, and C. Stewart Isolation and validation of human prepubertal skeletal muscle cells: maturation and metabolic effects of IGF-I, IGFBP-3 and TNF{alpha} J. Physiol., October 1, 2005; 568(1): 229 - 242. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. F. Singer, M. Mogg, W. Koestler, M. Pacher, E. Marton, E. Kubista, and M. Schreiber Insulin-Like Growth Factor (IGF)-I and IGF-II Serum Concentrations in Patients with Benign and Malignant Breast Lesions: Free IGF-II Is Correlated with Breast Cancer Size Clin. Cancer Res., June 15, 2004; 10(12): 4003 - 4009. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. M. Firth and R. C. Baxter Cellular Actions of the Insulin-Like Growth Factor Binding Proteins Endocr. Rev., December 1, 2002; 23(6): 824 - 854. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Matsui and T. Takahashi Mouse Testicular Leydig Cells Express Klk21, a Tissue Kallikrein That Cleaves Fibronectin and IGF-Binding Protein-3 Endocrinology, November 1, 2001; 142(11): 4918 - 4929. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Salahifar, R. C. Baxter, and J. L. Martin Differential Regulation of Insulin-Like Growth Factor-Binding Protein-3 Protease Activity in MCF-7 Breast Cancer Cells by Estrogen and Transforming Growth Factor-{beta}1 Endocrinology, September 1, 2000; 141(9): 3104 - 3110. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. C. Baxter Insulin-like growth factor (IGF)-binding proteins: interactions with IGFs and intrinsic bioactivities Am J Physiol Endocrinol Metab, June 1, 2000; 278(6): E967 - E976. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. A. Coverley, J. L. Martin, and R. C. Baxter The Effect of Phosphorylation by Casein Kinase 2 on the Activity of Insulin-Like Growth Factor-Binding Protein-3 Endocrinology, February 1, 2000; 141(2): 564 - 570. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. L. Martin and R. C. Baxter Oncogenic ras Causes Resistance to the Growth Inhibitor Insulin-like Growth Factor Binding Protein-3 (IGFBP-3) in Breast Cancer Cells J. Biol. Chem., June 4, 1999; 274(23): 16407 - 16411. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Lopez-Otin and E. P. Diamandis Breast and Prostate Cancer: An Analysis of Common Epidemiological, Genetic, and Biochemical Features Endocr. Rev., August 1, 1998; 19(4): 365 - 396. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Z. Shi, W. Xu, F. Loechel, U. M. Wewer, and L. J. Murphy ADAM 12, a Disintegrin Metalloprotease, Interacts with Insulin-like Growth Factor-binding Protein-3 J. Biol. Chem., June 9, 2000; 275(24): 18574 - 18580. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||
) for 24 h at 37 C, then fractionated by
Superose-12 chromatography at pH 6.5 as described in Materials
and Methods. Eluted fractions were assayed for IGFBP-3 by RIA.
Elution positions of calibration proteins (as in Fig. 3
) or [[LR3]IGF-I
(
) to IGFBP-3 was determined by solution binding assay as detailed
in Materials and Methods. Results are expressed as
percentage bound of total added IGF-I tracer after correction for
nonspecific tracer binding. B, Inhibition of IGFBP-3 proteolysis by
[LR3]IGF-I in a cell-free system. Proteolysis mixtures
contained 500 ng/ml IGFBP-3; MCF-7-conditioned medium from untreated
cells; and [LR3]IGF-I at 0, 10, 50, 250, 500, and 1000
ng/ml in lanes 1–6, respectively. Proteolyzed samples were analyzed by
immunoblot and visualized by phosphorimaging. C, Phosphorimage analysis
of immunoblots from protease mixtures containing IGFBP-3; conditioned
medium from untreated cells; and IGF-I (•), [des(1–3)]IGF-I (