| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
University of Bristol Division of Surgery (L.A.M., S.X., S.C.-H., J.K.F., J.M.P.H.), Department of Hospital Medicine, Bristol Royal Infirmary, Bristol BS2 8HW, United Kingdom; and Department of Cellular Physiology (J.M.P.), Babraham Institute, Babraham, Cambridge CB2 4AT, United Kingdom
Address all correspondence and requests for reprints to: Professor J. M. P. Holly, University of Bristol Division of Surgery, Department of Hospital Medicine, Bristol Royal Infirmary, Bristol BS2 8HW, United Kingdom. E-mail: jeff.holly{at}bris.ac.uk
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsMaterialsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsMaterialsResultsDiscussionReferences |
|---|
One such postulated mechanism for release is the proteolytic modification of IGFBP-3. This activity was first described in the circulation of pregnant women (2, 3) and then in various pathologies, such as diabetes (4), post surgery (5), severe illness (6), and cancer (7), indicating a more general mechanism. In all of the above examples, the pattern of fragmentation of IGFBP-3 has been the same. More recently, IGFBP-3 protease has also been described in several cell lines, both transformed and untransformed (8, 9); however, the pattern of fragmentation has differed from that seen in the circulation. Protease activity has also been described in normal physiological fluids outside the circulation. Xu et al. (10) described the presence of IGFBP-3 protease activity in interstitial fluid but the absence of such activity in P Inst F. Similarly Fernihough et al. (11) demonstrated protease activity in normal synovial fluid but significantly less activity in synovial fluid from joints affected by osteoarthritis (OA) and rheumatoid arthritis (RA). In both interstitial and synovial fluid, the pattern of proteolysis was the same as seen in pregnancy serum.
The IGFBP-3 protease has been classified as a metal-dependent serine protease that is active at neutral pH and at 37 C (2). Attempts to further characterize the identity of the in vivo IGFBP-3 protease have foundered because of the same general problems. There are many general extracellular proteases present throughout the body at relatively high levels, and IGFBP-3 seems particularly susceptible to their actions. The activity of these extracellular proteases is normally very tightly regulated because of the presence of even higher levels of general protease inhibitors; but any manipulation to fractionate samples may lead to separation of such inhibitors, revealing protease activity that was not active in the original intact sample. However, now that we have a broader experience of observing IGFBP-3 protease activity changing in different physiological fluids in a number of conditions, we have taken the simple approach of examining constituents of activity in such fluids after nondisruptive size fractionation. Knowing the difference in IGFBP-3 protease activity in the original samples, if, after fractionation, some components varied in line but others showed no relation, then we anticipated that we may get indications of which components could make important contributions to the in vivo activity and which are unrelated artifacts of the separation.
Various enzymes, such as plasmin (12), cathepsin D (13), matrix metalloproteinases (MMPs) (14), and prostate specific antigen (15), have been identified as IGFBP-3 proteases, but they all differ from the pregnancy-associated protease either in pattern of proteolysis, active pH, or inhibitor sensitivity. They do, however, seem to have a role as proteases found in the conditioned medium from cell lines and may be important within the pericellular environment in various conditions.
Various physiological regulators of IGFBP-3 protease activity have been demonstrated, such as insulin in insulin-dependent diabetes (4), and nutritional status in malnourished (16) and critically ill patients (8).
The mechanism by which the enzyme activity is switched on and off, however, is still unknown. It may be caused by the activation of latent enzymes, de novo production of enzymes, removal of enzyme inhibitors, or removal of an inhibitor that protects IGFBP-3. In this paper, we provide evidence that the enzymes are always present and regulation of the IGFBP-3 protease results from complex interaction of several enzymes and specific inhibitors.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsMaterialsResultsDiscussionReferences |
|---|
Serum from pregnant women (PS). A serum pool was made from 10 patients in the third trimester of pregnancy.
Interstitial fluid. Interstitial fluid was obtained by the application of 200 mHg of negative pressure to the skin for 1–2 h. Approximately 1.5 ml of N Inst F was taken from an area of normal skin and an area of plaque type psoriasis (P Inst F). The fluid was stored at -20 C.
Synovial fluid. Synovial fluid was obtained, by aspiration of the suprapatella pouch, from patients with OA (OA Syn F) or RA (RA Syn F) and normal healthy volunteers (N Syn F). The fluid was centrifuged and stored at -70 C.
Informed consent was obtained from all subjects, and the study was approved by the ethics committee of the United Bristol Healthcare Trust.
Size separation of serum and physiological fluids
NS and PS and synovial and interstitial fluids were
filtered through a 0.22-µm filter before size fractionation under
neutral conditions. One hundred microliters of sample was loaded onto a
Superose 12 column in a 0.05-M phosphate and
0.1-M NaCl buffer (pH 6.5) at a flow rate of 50 ml/h.
Two-hundred-fifty-microliter fractions were collected and stored at
-20 C.
Results shown, from the separation of the above fluids, are representative of at least two different pools and two different individual samples, to ensure that the pool used was not disproportionately influenced by a nonrepresentative sample.
|
Materials |
|---|
|
TopAbstractIntroductionMaterials and MethodsMaterialsResultsDiscussionReferences |
|---|
Assay for IGFBP-3 proteolytic activity
Intact samples and column fractions were assayed for their
ability to fragment radiolabeled ngIGFBP-3 (125I-ngIGFBP-3
iodinated using a Chloramine T method), according to the method of
Lamson et al. (17). Briefly, 5 µl of each intact sample
[in a total vol of 50 µl in 0.05 M phosphate and 0.1
M NaCl buffer, (pH 6.5)] and 50 µl of each column
fraction were incubated with 15,000 cpm of 125I-ngIGFBP-3
for 24 h at 37 C. The assay was stopped by boiling 20-µl
aliquots of the incubation with SDS loading buffer before loading onto
a 12.5% SDS polyacrylamide gel that was run overnight at 30 mA.
Fragmentation of the labeled substrate was visualized by fixing and
drying the gels and exposing them to x-ray film at -70 C for 2–3
days.
Inhibitor profile of IGFBP-3 proteolytic activity in fractions
after size separation
After identification of the fractions with IGFBP-3 proteolytic
activity, the effects of various inhibitors were tested for their
ability to inhibit fragmentation of ngIGFBP-3.
Fifty microliters of each fraction was incubated with 5 ng of
ngIGFBP-3 or 15,000 cpm of 125I-ngIGFBP-3 alone or with
EDTA (final concentration, 50 mM/liter); aprotinin (final
concentration, 1 mg/ml); 4-(2-aminoethyl)-benzene sulfonyl
(final concentration, 1 mM/liter); TIMP-1 (final
concentration, 113 µg/ml); zinc chloride (final concentration, 1
mM/liter); 1,10 phenanthroline (10 mM/liter);
-2 antiplasmin (final concentration, 5 µg/ml); and
antichymotrypsin (final concentration, 450 µg/ml).
Forty-microliter aliquots were removed, and the assay was stopped by
boiling with SDS loading buffer before loading the samples onto a
12.5% polyacrylamide gel. Samples with 125I-ngIGFBP-3 were
fixed and dried before exposure to x-ray film, and samples with
unlabeled ngIGFBP-3 were transferred by electrophoresis, at 70 mA
overnight, onto hybond C membranes for 4 h at 0.8 mA constant
current. Intact and fragmented ngIGFBP-3 were visualized by Western
immunoblotting. The membranes were blocked in tris-buffered saline and
3% nonfat milk before incubating with a specific monoclonal antibody
for IGFBP-3 (SCH 2/6 at 1:10,000) at room temperature overnight. The
membranes were then washed in tris-buffered saline to remove unbound
antibody before incubating with an antirabbit antibody conjugated to
peroxidase (1:10,000) for 1 h at room temperature. Binding of the
peroxidase-labeled antibody was visualized using enhanced
chemiluminesence with an ECL detection system (Amersham,
Buckinghamshire, UK) and exposure to x-ray film.
Detection of inhibition of proteolytic activity in column fractions
by intact samples and samples from the size separation of serum and
synovial fluid
Fifty microliters of the high and low molecular weight (MW)
areas of activity from the size separation of serum, interstitial
fluid, and synovial fluid were mixed with 5 µl of either NS, P Inst
F, or RA Syn F before a protease assay, as already described.
Additionally, 50 µl of the high and low MW areas of activity were
mixed with 50 µl of fractions from the separation of each fluid, with
little or no proteolytic activity before the protease assay.
Differential effect of IGFBP-3 from human serum and recombinant
IGFBP-3 on protease activity
Fifty microliters of the high and low MW areas of activity from
the size separation of NS were mixed with either 75 ng IGFBP-3
partially purified from NS, or increasing concentrations of recombinant
IGFBP-3 (both glycosylated and nonglycosylated; 50, 100, 200, and 500
ng) before a protease assay, as already described. Partial purification
of IGFBP-3 was achieved using an anti-IGFBP-3 affinity column.
Results shown are representative of assays performed at least three times using samples from different separations.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsMaterialsResultsDiscussionReferences |
|---|
).
|
). Similarly, the separation of serum from PS
reveals two distinct MW areas of activity corresponding to the MW of
the activity seen in NS (in fractions of more than 200K and in
fractions of about 60K) (Fig. 1C). A time-course experiment revealed
the same distribution of activity at 4, 12, and 24 h, with the
same resultant fragmentation pattern but with activity increasing with
time.
Size separation of normal interstitial fluid and P Inst F
We have previously reported (10) that, in contrast to the
very low IGFBP-3 protease activity found in the circulation of normal
adults, such activity is clearly seen in N Inst F (Fig. 2A); however, P Inst F has very little
activity (Fig. 2A). Incubation of column fractions from the size
separation of both interstitial fluids with 125I-ngIGFBP-3
shows that both samples have a distinct area of proteolytic activity in
fractions of about 60 kDa (Fig. 2, B and C), similar to that seen in
serum. The high MW region, seen after the separation of serum, is not
apparent.
|
).
Assessment of IGFBP-3 proteolytic activity in the fractions from the
size separation of N Syn F shows a low MW area of activity in fractions
of about 60K; and, like interstitial fluid, the high MW area is not
apparent (Fig. 3b, panel A). After separation of OA Syn F, there is
also a distinct area of proteolytic activity in fractions of about 60K
(Fig. 3b, panel B), whereas the separation of RA Syn F reveals a
distinct high and low MW area of activity comparable with serum (Fig. 3b, panel C).
|
). In addition, the high MW area was
also partially inhibited by 1,10 phenanthroline and zinc and completely
inhibited by 2 antiplasmin (data not shown: results summarized in
Table 1).
|
|
). The low MW area was not inhibited by 1,10
phenanthroline; but, similar to the effect of TIMP-1, there was an
increase in activity. Incubation with zinc had no effect on the amount
of intact IGFBP-3, but it altered the pattern of fragmentation,
suggesting that it was inhibiting an enzyme that fragments IGFBP-3
after initial cleavage. Whereas 2 antiplasmin completely inhibited
the low MW area of activity in NS, it had no effect on the low MW area
from PS (data not shown: results summarized in Table 1).
Inhibition of IGFBP-3 proteolytic activity in column fractions by
intact samples
Intact NS reduced the amount of IGFBP-3 proteolytic activity in
the high and low MW areas from serum (both NS and PS), the high MW area
from RA Syn F, and the low MW area from interstitial fluid and synovial
fluid, as seen by the increase in the amount of intact ngIGFBP-3
remaining in the presence of NS (Fig. 5, A and B). P Inst F also increased the amount of IGFBP-3 that remained
intact in a protease assay with the low MW area in serum and
interstitial fluid (Fig. 6). Intact RA
Syn F had the same inhibitory effect on the high MW from that synovial
fluid and the low MW areas from all three synovial fluids (see Fig. 8B).
|
|
|
) were found to be able to
reduce the amount of 125I-ngIGFBP-3 that was degraded by
the high and low MW areas of activity from both NS and PS (Fig. 7). The 150K fractions from the
separation of PS had no inhibitory effect (data not shown). The same
fractions from NS and corresponding fractions from RA Syn F were found
to reduce the amount of 125I-ngIGFBP-3 that was degraded by
both the high and low MW areas from RA Syn F and the low MW areas from
N Syn F and OA Syn F (Fig. 8). The 150K
fractions from the separation of OA synovial were able to reduce the
amount of degradation in the low MW area but to a lesser extend than RA
Syn F (Fig. 8). The same size fractions from the separation of N Syn F
had no inhibitory effect on the proteolytic activity in the low MW area
from the separation of normal synovial fluid.
|
).
To demonstrate that this inhibition could not be explained by substrate
competition, increasing concentrations of recombinant g- or ngIGFBP-3
were also added to the protease assay with the high and low MW areas of
activity. There was no reduction in proteolysis of the
125I-IGFBP-3 in the presence of the recombinant IGFBP-3 at
concentrations comparable with and greater than the amount of IGFBP-3
in the NS, intermediate inhibitory areas or in the purified IGFBP-3
(Fig. 9; results shown for ngIGFBP-3). At least 10-fold greater amounts
of recombinant IGFBP-3 were required to inhibit the activity in both
the high and low areas of activity.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsMaterialsResultsDiscussionReferences |
|---|
2 antiplasmin
suggests that plasmin or a plasmin-like enzyme is responsible for at
least some of the activity in the high MW area. The lower MW area was
partially inhibited by EDTA, aprotinin, and AEBSF; but there was an
increase in activity in the presence of TIMP-1 and 1,10 phenanthroline.
This suggests that the low MW area of activity is caused by at least
one metal-dependent serine protease, one or more of which can be
partially inactivated by a metalloproteinase. Inhibition of the low MW
area from NS, but not PS, by 2 antiplasmin suggests that, during the
separation of NS, plasmin may be activated. Although plasmin may not be
responsible for cleaving IGFBP-3 directly, it may activate latent
enzyme(s) involved in IGFBP-3 proteolysis. In PS, these enzymes seem to
already be active, and hence, blocking activated plasmin has no
effect. When the extravascular fluids were separated, both interstitial and synovial fluid gave similar results. The separation of both N Inst F and P Inst F resulted in a low MW area of activity of about 60K; likewise, the separation of N Syn F and OA Syn F also demonstrated the presence of IGFBP-3 proteolytic activity in fractions of about 60K. Like serum, IGFBP-3 proteolytic activity in RA Syn F was present in a high MW area of more than 200K and a lower MW area of about 60K. The high MW area from RA Syn F and the low MW areas of activity from both interstitial fluid and all three synovial fluids were inhibited in the same way as the activity in the two corresponding areas from the separation of serum, which suggests that the enzymes involved both in the circulation and outside the circulation are very similar.
Interstitial and synovial fluid have activity similar to that of PS when intact; the apparent absence of the high MW area in all but RA Syn F suggests that this activity is normally restricted to the vasculature. There is a large influx of proteins into synovial fluid in RA from the circulation caused by the inflammatory response and consequent increased capillary permeability, and this may be the source of this high MW area of activity in this condition.
The separation of serum, interstitial fluid, and synovial fluid reveals IGFBP-3 proteolytic activity after size separation, regardless of whether activity is present in the intact sample. The presence of activity in NS, P Inst F, RA, and OA Syn F after separation, but not in the intact sample, suggests the presence of an inhibitor. The ability of P Inst F to inhibit the activity in normal interstitial fluid and for RA Syn F to inhibit the activity in N Syn F has already been demonstrated (Ref. 11 ; and Maile and Whellams, unpublished observations). In this paper, we confirm the presence of an inhibitor in both interstitial fluid and synovial fluid by the ability of the intact samples to inhibit the high and low areas of activity from the separation of their respective fluids. We also demonstrate (for the first time) the presence of an inhibitor in NS, by the ability of NS to inhibit the activity in the high and low MW areas from the separation of both sera, the high and low areas from RA Syn F, and the low MW area from N Inst F, P Inst F, N Syn F, and OA Syn F.
The nature of this inhibition was further characterized by demonstrating that fractions of approximately 150 kDa from NS, but not from the separation of PS, could inhibit the high and low areas of activity from the separation of serum. The majority of IGFBP-3 in the circulation is found in 150-kDa ternary complex with IGF and acid-labile subunit. Inhibition by the 150-kDa fractions from the separation of NS suggests that the inhibitor of proteolytic activity may be associated with IGFBP-3 in NS or that IGFBP-3 is held in a specific confirmation that makes it inaccessible to the protease. This was further demonstrated using IGFBP-3, partially purified from NS, which also resulted in inhibition of activity. IGFBP-3 itself is not the inhibitor, however, because there was no inhibition of activity when either glycosylated or nonglycosylated recombinant IGFBP-3 was added to the protease assay. Protease activity in serum was first measured by mixing NS with PS and visualizing the loss of intact IGFBP-3 from NS, by Western immunoblotting; the lack of inhibition by NS in these experiments suggests that there must be a mechanism for overriding this in PS. PS may contain a component that either degrades or blocks this inhibitor. The inhibition by the 150-kDa fraction from RA Syn F, partial inhibition by OA Syn F, and lack of inhibition by the same size fractions from N Syn F also suggest the presence of an inhibitor associated with IGFBP-3. The lack of apparent inhibition by 150-kDa fractions from P Inst F suggest that the mechanism of regulation may be different in interstitial fluid. The data regarding the presence of an inhibitor associated with IGFBP-3 is, as yet, circumstantial; and further characterization will be required to confirm its existence.
The measurement of protease activity by incubation with radiolabeled substrate is, at best, a crude assessment, because the extent of proteolysis and the fragmentation pattern vary considerably, in relation to relative amounts of proteases and inhibitors present, the quality of tracer, the buffer used, and the length of incubation. We have found even greater variability when analyzing fractions after chromatographic separation. We therefore have made inferences only from qualitative differences and have not drawn conclusions from absolute amounts of various fragments produced, which vary between column runs.
We are still left with several pertinent questions, not the least of which is the identity of the protease responsible for the modification of IGFBP-3. We have provided evidence, however, that the activity in PS and the activity outside the circulation is caused by the same protease system; this argues against the decidua being the source of the PS-associated protease (18). Additionally, we have seen the same areas of activity after the separation of serum from catabolic cancer patients (Maile and Crown, unpublished observations); this also supports the idea that the proteases responsible are always present. Bang et al. (19) recently purified the fibrinolytic enzyme plasminogen from both PS and NS. Plasmin has been demonstrated to be capable of proteolyzing IGFBP-3 (12), and its abundance in serum means that it is not surprising that it was isolated from both PS and NS. As the authors imply, however, plasmin is an unlikely candidate for the circulating and extravascular IGFBP-3 protease, because it is very tightly regulated and only active in the circulation at sites of tissue injury. Although plasminogen activators increase in the circulation in pregnancy there is a concomitant rise in plasminogen activator inhibitors and, therefore, no overall increase in activity. MMPs have also been shown to play a role in the IGFBP-3 proteolytic activity in the serum from pregnant rats (14); our data support a role for MMPs in the IGFBP-3 protease system but suggest that they may be important regulators of serine proteases.
Our data confirms the study of Bang et al. (19), in which they demonstrated two distinct areas of IGFBP-3 proteolytic activity in PS of MWs similar to the activity described in this paper. In contrast to this study, however, no high MW area was apparent in NS; this may be because of the complexity of the system being studied and the limitations of the techniques employed. Our data, however, are consistent with another study that also reported high and low MW areas of activity of comparable size after size separation of NS (20).
This paper provides evidence that the IGFBP-3 protease system is caused by a complex interaction of various classes of enzymes and provides evidence that the enzymes responsible for the proteolysis of IGFBP-3, both in the circulation and outside, are always present. In NS, RA Syn F, and P Inst F, where no activity is apparent, the IGFBP-3 seems to be protected from proteolysis. The identity of the enzymes and inhibitors involved in the IGFBP-3 protease system and the factors regulating the presence or absence of the inhibitory IGFBP-3 remain to be elucidated.
Received January 26, 1998.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsMaterialsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  D. R. Clemmons, M. Sleevi, and W. H. Busby Jr. Recombinant, Nonglycosylated Human Insulin-Like Growth Factor-Binding Protein-3 (IGFBP-3) Is Degraded Preferentially after Administration to Type II Diabetics, Resulting in Increased Endogenous Glycosylated IGFBP-3 J. Clin. Endocrinol. Metab., December 1, 2005; 90(12): 6561 - 6568. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Oesterreicher, W. F. Blum, B. Schmidt, T. Braulke, and B. Kubler Interaction of Insulin-like Growth Factor II (IGF-II) with Multiple Plasma Proteins: HIGH AFFINITY BINDING OF PLASMINOGEN TO IGF-II AND IGF-BINDING PROTEIN-3 J. Biol. Chem., March 18, 2005; 280(11): 9994 - 10000. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 O. Fletcher, L. Gibson, N. Johnson, D. R. Altmann, J. M.P. Holly, A. Ashworth, J. Peto, and I. d. S. Silva Polymorphisms and Circulating Levels in the Insulin-Like Growth Factor System and Risk of Breast Cancer: A Systematic Review Cancer Epidemiol. Biomarkers Prev., January 1, 2005; 14(1): 2 - 19. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. R. Devi, D.-H. Yang, R. G. Rosenfeld, and Y. Oh Differential Effects of Insulin-Like Growth Factor (IGF)-Binding Protein-3 and Its Proteolytic Fragments on Ligand Binding, Cell Surface Association, and IGF-I Receptor Signaling Endocrinology, November 1, 2000; 141(11): 4171 - 4179. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Diamandi, J. Mistry, R. G. Krishna, and J. Khosravi Immunoassay of Insulin-Like Growth Factor-Binding Protein-3 (IGFBP-3): New Means to Quantifying IGFBP-3 Proteolysis J. Clin. Endocrinol. Metab., June 1, 2000; 85(6): 2327 - 2333. [Abstract] [Full Text]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
