0013-7227/90/1272-078l$02.00/0 Endocrinology Copyright © 1990 by The Endocrine Society
Vol. 127, No. 2 Printed in U.S.A.
Production of an Insulin-Like Growth Factor (IGF) Inducible IGF-Binding Protein by Human Skin Fibroblasts* JANET L. MARTIN AND ROBERT C. BAXTER Department of Medicine, University of Sydney, New South Wales 2006; and the Department of Endocrinology, Royal Prince Alfred Hospital, Camperdown, New South Wales 2050, Australia
ABSTRACT. Neonatal human skin fibroblasts produce insulin-like growth factor-binding proteins (IGFBPs) that have the potential to modulate the actions of the growth factors. We have examined the IGFBPs secreted by monolayer cultures of neonatal fibroblasts by ligand blotting with [125I]IGF-II and immunoblotting with antisera raised against three well characterized IGFBPs: IGFBP-1, IGFBP-2, and IGFBP-3. As detected by ligand blotting, medium from untreated fibroblasts contained IGFBP-3, a second IGFBP which appeared as a doublet of 2931K, and a smaller protein of 22K. Within 10 h of the addition of 50 ng/ml IGF-I, a markedly increased level of production of the 29-31K IGFBP doublet was detectable, with levels increasing
8- to 9-fold by 24 h compared to that in untreated samples. IGFI was approximately twice as potent as IGF-II at inducing 2931K IGFBP, with a half-maximal response at 15.4 ± 2.7 ng/ml IGF-I and 26.6 ± 1.6 ng/ml IGF-II (n = 3). Insulin tested at 1 Mg/ml did not induce 29-31K IGFBP. Neither GH nor the acidlabile subunit of the circulating high mol wt IGFBP complex induced 29-31K IGFBP or affected its induction by IGF-I or IGF-II. Immunoblotting demonstrated that IGF-inducible IGFBP did not react with antibodies to IGFBP-1, IGFBP-2, or IGFBP-3. These results indicate that IGF-I and IGF-II induce an IGFBP that is different from previously characterized human IGFBPs. (Endocrinology 127: 781-788, 1990)
I
T HAS been acknowledged for many years that the binding proteins (BPs) for the insulin-like growth factors (IGFs) have the potential to play an important role in the actions of the growth factors themselves. As recently reviewed, three distinct classes of IGFBPs have been identified (1). IGFBP-1 was first identified in amniotic fluid (2) and purified from this source (3) and human decidua (4); the cDNA for this protein has now been cloned and characterized (5). IGFBP-1 is regulated by glucose and insulin, as shown in both clinical studies (6, 7) and a tissue culture model of IGFBP-1 production (8). The second type of IGFBP, IGFBP-2, was originally purified from medium conditioned by the rat liver-derived cell line BRL-3A (9) and has recently been cloned from this source (10); the bovine homolog has been purified from the MDBK kidney cell line (11), while the primary structure for human IGFBP-2 has been determined from cDNA cloned from a human fetal liver library (12). The GH-dependent IGFBPs make up the third class of BPs. IGFBP-3, also known as BP-53, is the IGF-binding component that, in the presence of an
85-kilodalton (85K) acid-labile subunit and IGF-I or IGF-II, forms the high mol wt complex in which most of the IGFs are found in the circulation (13). First isolated from human plasma (14), homologous proteins have been isolated from the sera of rats (15) and pigs (16). IGFBP3 is glycosylated, appears as two IGF-binding bands on sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), and binds IGF-I and IGF-II with similar high affinity. Recent evidence indicates that IGFs are able to modulate the production of some of their BPs both in cell culture and in vivo. In sparse monolayer culture of fetal fibroblasts, IGF-I and IGF-II have been reported to induce immunoreactive IGFBP-1 production (17). In vivo, recombinant human IGF-I stimulates the production of 42K, 45K, and 49K glycosylated IGFBPs in diabetic or hypophysectomized rats (18, 19); since these proteins are able to form part of the GH-dependent complex, they are the rat homologs of the human IGFBP3 doublet. In contrast, Walton and Etherton (20) have shown no effect of infusion of IGF-I on immunoreactive IGFBP-3 levels in pigs, while demonstrating that circulating levels of both IGF-I and IGFBP-3 in these animals were increased in response to GH. Normal and transformed human fibroblasts produce a variety of IGFBPs (21,22). In this report we demonstrate
Received January 26, 1990. Address requests for reprints to: Janet Martin, Department of Medicine, Blackburn Building (D-06), University of Sydney, New South Wales 2006, Australia. * This work was supported by the National Health and Medical Research Council of Australia.
781
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 November 2015. at 15:01 For personal use only. No other uses without permission. . All rights reserved.
IGF-INDUCIBLE IGFBP FROM FIBROBLASTS
782
that in neonatal human fibroblasts, IGF-I and IGF-II, but not insulin, specifically induce the production of an IGFBP different from previously described human IGFBPs.
Materials and Methods Tissue culture plasticware, culture media, and fetal calf serum were purchased from Cytosystems (North Ryde, New South Wales, Australia); antibiotics, BSA (fraction V, RIA grade), bovine insulin, and protein-A were obtained from Sigma Chemical Co. (St. Louis, MO). Disuccinimidyl suberate was purchased from Pierce Chemical Co. (Rockford, IL). Electrophoresis consumables were obtained from Pharmacia (Uppsala, Sweden), and Bio-Rad (Richmond, CA). Autoradiographic film was purchased from
Amersham
medium (250 (A) were removed from each well, frozen, and replaced with fresh medium containing the test substance. All media were collected at 72 h and frozen until assayed. IGFBP3 was measured in conditioned medium samples by RIA (25), using antibody R7 (21). Sample preparation for SDS-PAGE analysis
Reagents
(Hyperfilm-MP)
E n d o • 1990 Vol 127 • No 2
(Bucks,
United Kingdom). IGF-I and IGF-II (23) and IGFBP-3 (14) were purified from Cohn fraction IV of human plasma, and IGFBP-1 (24) from amniotic fluid, as previously reported. Polyclonal antisera raised against IGFBP-3 and IGFBP-1 have been described previously (24,25); antiserum to bovine IGFBP2, raised against IGFBP purified from medium conditioned by MDBK cells, was the generous gift of Dr. John Ballard, CSIRO Division of Human Nutrition (Adelaide, South Australia), and Dr. John Wallace, University of Adelaide, South Australia. The a-subunit (acid-labile subunit) of the 140K GH-dependent IGF complex was purified as previously described (26). IGFBP secreted by SV-40-transformed IMR-90 fibroblasts (line AG 2804) was isolated from conditioned culture medium as previously reported (22). Biosynthetic methionyl human GH (hGH) was a generous gift from Kabi (Stockholm, Sweden). Tracers IGF-I and IGF-II tracers were prepared, as described in previous studies (23, 27), to specific activities of about 200 Ci/ g. Protein-A tracer was radioiodinated with Na125I (Amersham), using chloramine-T. Cross-linked tracer for the IGFBP-3 RIA (IGFBP-3 affinity-labeled with IGF-I) was prepared as previously described (25). Cell culture Neonatal fibroblasts were isolated as outgrowths from foreskin explants and maintained in monolayer culture at 37 C in Ham's F-12 medium containing 20 mM HEPES, 2 mM glutamine, and 10% fetal calf serum. Cultures were passaged by trypsin-EDTA treatment when confluent (every 7-10 days, 1:3 split) and used for experiments between the 3rd and 10th generations. For stimulation experiments, cells were seeded into 12-place multiwells at a density of 4 x 104 cells/well in medium containing 10% fetal calf serum and grown to confluence. Fibroblast monolayers were then rinsed with serum-free medium and incubated for a minimum of 48 h in medium without serum. This medium was replaced with 1.5 ml medium containing 2 g/ liter BSA with or without test substances as indicated. Twentyfour and 48 h after the initial additions, samples of conditioned
Conditioned media from stimulation experiments were acidified, ultrafiltered to remove IGFs, and concentrated before SDS-PAGE. Generally, 1-ml pools of media were acidified with an equal volume of 2 M acetic acid, incubated at room temperature for 1 h, then concentrated to about 200 ^1 by centrifugation through a Centricon-10 membrane (Amicon Corp., Lexington, MA). Samples were flushed through with a further 5 ml 1 M acetic acid, and the final concentrate (~100 /A) was lyophilized. Freeze-dried samples were reconstituted in 200 ^l 50 mM sodium phosphate, pH 6.5. Preliminary experiments using radiolabeled IGF-I indicated that this procedure removed >95% of the IGFs from the samples. Affinity labeling, SDS-PAGE, Western blotting, and densitometry To affinity label medium concentrates, 50-^1 aliquots were incubated overnight at 4 C with [125I]IGF-I or [125I]IGF-II in a final volume of 150 ii\ 50 mM sodium phosphate, pH 6.5. Disuccinimidyl suberate (7.5 mM in dimethylsulfoxide) was added to a final concentration of 0.25 mM, and the reaction was stopped 30 min later by the addition of 25 ^11 M Tris-HCl, pH 8. Pure proteins (IGFBP-1, IGFBP-3, and IGFBP purified from transformed fibroblast medium) were affinity labeled in buffer containing 0.5 g/liter BSA using the same protocol. Immunoprecipitation of affinity-labeled proteins was carried out as previously described (21). Medium concentrates, affinity-labeled samples, and calibration standards (Pharmacia) were prepared for electrophoresis by the addition of sample buffer to give final concentrations of 0.0125 M Tris (pH 6.8), 3% SDS, and 10% glycerol and boiled for 5 min; separation of proteins was achieved on 12% polyacrylamide gels. Gels containing affinity-labeled samples were then fixed and stained with 2.5 g/liter Coomassie blue in 25% isopropanol-10% acetic acid, destained in 25% methanol-10% acetic acid, then dried and exposed to Hyperfilm-MP autoradiography film for 3-7 days. For blotting studies, proteins were transferred to nitrocellulose (0.45 /*m; Schleicher and Schuell, Dassel, West Germany), using a Novablot electrophoretic transfer unit (LKB-Pharmacia Biotechnology, Bromma, Sweden). After electrophoresis, gels were placed in transfer buffer containing 39 mM glycine, 48 mM Tris, and 20% methanol (pH unadjusted) and shaken gently in this buffer for a minimum of 1 h. Transfer was completed over 1.5-2 h at 250 mamp, then the nitrocellulose sheets were blocked in 30 g/liter BSA at 37 C for a minimum of 6 h. For ligand blotting, sheets were then incubated with [125I]IGF-II (~1 X 106 cpm) at 4 C for 16 h, then washed three times with cold 50 mM sodium phosphate, pH 6.5, once with the same buffer containing 0.05% Nonidet P-40, then twice more in detergent-free buffer. For immunoblotting, blocked
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 November 2015. at 15:01 For personal use only. No other uses without permission. . All rights reserved.
IGF-INDUCIBLE IGFBP FROM FIBROBLASTS nitrocellulose sheets with transferred proteins were incubated with anti-hIGFBP-3 antibody (R7 antiserum; 1:500 final dilution) or anti-bIGFBP-2 antiserum (1:250 dilution) at 22 C or with anti-hIGFBP-1 antibody (A2 antiserum; 1:1000 final dilution) at 4 C. Incubation was continued for 16 h, then blots were washed as described above and incubated at 22 C for 1-2 h with 125I-labeled protein-A (~1 x 106 cpm). After washing again, Hyperfilm-MP was exposed to dried blots for 1-2 days at -70 C. Quantitation of developed autoradiographs was carried out using a Bio-Rad model 620 Video Densitometer attached to a chart recorder. The technique was validated by ligand blotting dilutions of medium with IGF-II tracer, scanning the autoradiograph, and plotting the absorbance peak height, expressed in arbitrary units, against concentration. The relationship obtained using this technique was linear between 0.05-50 ^1 medium and 4-80 arbitrary absorbance units (r2 = 0.992). The coefficient of variation of duplicate scans on a single sample, determined by analysis of variance on 12 duplicates, was 4.1%. All experiments were performed at least 3 times.
Results IGFBPs produced by neonatal skin fibroblasts over 3 days were detected by ligand blotting conditioned medium with [125I]IGF-II after fractionation by SDSPAGE. Figure 1 compares these IGFBPs with a pure plasma IGFBP-3 standard, which appeared as a charac94 • •
67 • •
43 • •
30 " •
20 • • BP-3
24h 48h 72h
24h 48h 72h
untreated
IGF-I treated
FIG. 1. [126I]IGF-II ligand blotting of conditioned medium from untreated and IGF-I-treated fibroblasts. Twenty-four-, 48- and 72-h conditioned media from untreated fibroblasts and those treated with 50 ng/ml IGF-I were electrophoresed unreduced on 12% polyacrylamide gels and transferred to nitrocellulose, as described in Materials and Methods. Blots were incubated with [125I]IGF-II, washed, and exposed to Hyperfilm-MP for 2 days. The positions of mol wt markers are indicated on the left by arrows. IGFBP-3 (500 ng), purified from human plasma as previously described (14) and indicated in the figure as BP3, is also shown for size comparison.
783
teristic doublet of approximately 50K (15). Cells incubated with or without 50 ng/ml IGF-I produced an IGFbinding doublet that comigrated with plasma IGFBP-3 and another doublet of approximately 29-31K. A minor IGF-binding species of 22K was also observed. In cells grown without IGF-I, the IGFBP-3 doublet was the predominant binding species, with the 22K and 29-31K bands barely detectable. IGF-I markedly induced the 293IK doublet, with a relatively small effect on the other IGFBPs. To quantitate these changes, autoradiographs were scanned with a densitometer, and the resulting peak heights for each radioactive species were compared (Fig. 2a). When quantitated in this way in three experiments, fibroblast IGFBP-3 and the 22K IGFBP were increased 1.5-2 fold by exposure of the cells to 50 ng/ml IGF-I, although this stimulation of production was not significant in the 24-h samples. RIA for IGFBP-3 in these samples was consistent with this finding (not shown); over four experiments, incubation with 50 ng/ml IGF-I over a 72-h period increased IGFBP-3 levels by 73 ± 10% (mean ± SE). A much more dramatic increase was observed with the 29-31K doublet. Within 24 h of IGF-I treatment, IGFBP activity was 15- to 20-fold higher in these media than in the controls, and this increased rate of production was maintained over 3 days. To examine the time course of induction of 29-3IK IGFBP in more detail, samples of media were taken at 2-h intervals for 24 h after the addition of IGF-I, concentrated, then analyzed by ligand blotting with IGF-II tracer and densitometry as before. As seen in Fig. 2b, 29-31K IGFBP was detectable in the medium within 10 h of IGF addition, and an increased rate of production was observed over the following 12 h, reaching a maximum between 22-24 h. To determine whether both IGF-I and IGF-II were able to induce the 29-31K IGFBP, fibroblasts were incubated with increasing concentrations of peptides in the range 2.5-50 ng/ml, and the 72-h conditioned medium was analyzed by IGF-II ligand blotting (Fig. 3). Both peptides showed dose-dependent stimulation of production of the 29-31K species, with the response showing greater sensitivity to IGF-I (Fig. 3a). Quantitation of these data by densitometry confirmed that IGF-I was approximately 2-fold more potent than IGF-II in its stimulation of IGFBP (Fig. 3b), with a half-maximal response at 15.4 ± 2.7 ng/ml for IGF-I and 26.6 ± 1.6 ng/ml for IGF-II (three experiments). When tested in the range 1-1000 ng, insulin was not able to induce production of the 29-31K IGFBP. The relationship between cell density and production of 29-31K IGFBP was also determined (not shown). Fibroblasts plated at densities of 1, 2, 4, or 8 X 104 cells/well and treated with either IGF-I or IGF-II showed production of 29-31K
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 November 2015. at 15:01 For personal use only. No other uses without permission. . All rights reserved.
IGF-INDUCIBLE IGFBP FROM FIBROBLASTS
784
22K IGFBP
29-31K IGFBP
IGFBP-3
E n d o • 1990 Vol 127-No 2
100
O80
o «60 I 40 20 "
20
3
1
2
3
1
4
Incubation time (days)
8
12
16
20
24
Incubation time (hours)
FIG. 2. IGF-stimulated production of IGFBPs by human fibroblasts. a, Induction of three IGFBPs over 3 days. Autoradiographs similar to those shown in Fig. 1 were scanned using a densitometer; peak absorbance heights for the IGFBP-3, 29-31K IGFBP, and 22K IGFBP species from unstimulated (•) and 50 ng/ml IGF-stimulated (M) samples are expressed in arbitrary units, where a 100% response represents the optical density of 29-31K IGFBP bands after 3-day induction by IGF-I. Statistical analysis was performed separately for each IGFBP by analysis of variance and Duncan's multiple range test; error bars represent the SD for three experiments. Significance of IGF-I effect at each time point: *, P < 0.005; **, P < 0.001. b, Twenty-four-hour time course of 29-31K IGFBP production. Fibroblasts were incubated with 50 ng/ml IGF-I, and medium samples were collected at 2-h intervals for 24 h and analyzed by SDS-PAGE, ligand blotting, and densitometry, as described in Materials and Methods. FIG. 3. Stimulation of 29-31K IGFBP production by IGF-I and IGF-II. a, Fibroblast cultures were incubated for 72 h in medium containing IGF-I (upper panel) or IGF-II (lower panel) in the range of 2.5-50 ng/ml, and the conditioned media were analyzed by IGF-II ligand blotting as described in Materials and Methods. Blots were exposed to Hyperfilm-MP for 2 days, b, Dose-response curves generated by densitometry after autoradiography. • , IGF-I; • , IGF-II. Data are the mean ± SD for three experiments. Also shown is an insulin dosecurve (O) derived from an identical experiment; no effect of insulin was seen up to 1000 ng/ml.
a IGF-I stimulation
2.5 5
100 •
10 25 ng/ml
50
IGF-II stimulation
2.5 5
10 25 ng/ml
IGFBP in proportion to cell number, with no difference in the response to IGF-I or IGF-II. In view of the fact that IGF-I and IGF-II circulate as part of a GH-dependent high mol wt complex, we next determined whether GH or the a (acid-labile)-subunit of the circulating complex (13) was able to elicit the same response as the IGFs, i.e. induction of 29-3IK IGFBP. Fibroblasts were treated with hGH (1 /ig/ml), a-subunit (1 /ug/ml), and IGF-I (50 ng/ml), either alone or in combination, and the conditioned medium was analyzed by IGF-II ligand blotting and autoradiography. As seen in Fig. 4, hGH, a-subunit, or a combination of the two caused no change in the intensity of the 29-31K doublet compared with that in untreated fibroblasts. However, in all samples in which IGF-I was included in the incubation, 29-3IK IGFBP was much more intensely labeled, indicating increased production of this protein. Com-
50
Peptide concentration (ng/ml)
• untr
GH
1
fey
•J IGF-I
a
IGF-I + GH
GH +a
IGF-I +a
IGF-I + GH +a
FIG. 4. Effect of treatment with IGF-I, GH, and the a-subunit of the GH-dependent IGFBP complex on 29-31K IGFBP induction. Fibroblasts were left untreated (untr) or treated with GH (1 fig/ml), otsubunit (1 Mg/ml), or IGF-I (50 ng/ml), alone or in combination as indicated, for 72 h. Ligand blotting of conditioned media was performed as described in Materials and Methods using [125I]IGF-II. The migration position on SDS-PAGE of a 30K mol wt marker is indicated by the arrow.
pared to the effect of IGF-I alone, the addition of either hGH or a-subunit had no significant effect, but a 52%
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 November 2015. at 15:01 For personal use only. No other uses without permission. . All rights reserved.
IGF-INDUCIBLE IGFBP FROM FIBROBLASTS
increase in labeling intensity was seen when both hGH and a-subunit were added (P = 0.047; n = 4). A direct size comparison between the IGF-induced IGFBP and other IGFBPs (determined by SDS-PAGE, ligand blotting, and autoradiography) is shown in Fig. 5. The IGF-inducible IGFBP is distinct from the IGFBP-3 doublet, the IGFBP-1 purified from amniotic fluid (28K), and the IGFBP purified from transformed human fibroblast medium (34K). The immunological relationship of induced 29-31K IGFBP to other IGFBPs was examined by immunoblotting with antisera to IGFBP-1, IGFBP-2, and IGFBP-3. As shown in Fig. 6, antiserum to IGFBP1 detects this protein purified from amniotic fluid, but does not react with the 29-3IK IGFBP or any other IGFBPs in medium conditioned by IGF-stimulated neonatal skin fibroblasts. Similarly, antiserum raised against bovine IGFBP-2 detected IGFBP purified from human transformed fibroblast-conditioned medium, but did not react with 29-3IK IGFBP from IGF-treated fibroblasts. Also shown in this figure and as reported previously (21), neonatal fibroblasts produce immunoreactive IGFBP-3 identical in size to human plasma IGFBP-3; however, IGF-I LIGAND BLOTTING
67 • •
43 • •
30 • •
g|i t i l l m •
20*IGFBP-1
Transformed fibroblast IGFBP
IMMUNOBLOTTING
67
43
30
20 IGFBP-1 medium IGFBP-1 ANTISERUM
Transformed fibroblast medium
IGFBP-3 medium
IGFBP-2 ANTISERUM IGFBP-3 ANTISERUM
FIG. 6. Immunoreactivity of 29-31K IGFBP compared to other human IGFBPs. IGFBP-1 (500 ng), transformed fibroblast IGFBP-1 (~1 ng), IGFBP-3 (500 ng), and IGF-I-treated medium (50 nl 5-fold concentrate) were electrophoresed and transferred to nitrocellulose. Blots were incubated with polyclonal antisera to human amniotic fluid IGFBP-1 (1:1000 final dilution), bovine IGFBP-2 (1:250), and human plasma IGFBP-3 (1:500) as indicated. Immune complexes were then detected by incubation with 125I-labeled protein-A (~1 X 106 cpm) and autoradiography, as described in Materials and Methods.
antiserum to this protein does not react with 29-3IK IGFBP doublet. These findings were confirmed using the same antisera to precipitate [125I] IGF-I affinity-labeled medium (not shown); none of the sera precipitated crosslinked species corresponding to the IGF-inducible IGFBP. Taken together, these results indicate that the IGF-induced IGFBP is immunologically distinct from IGFBP-1, IGFBP-2, and IGFBP-3.
lliiii
IGFBP-3
785
Medium
FIG. 5. Size comparison of human IGFBPs. Human plasma IGFBP-3, human amniotic fluid-derived IGFBP-1, IGFBP purified from transformed human fibroblasts (each at 500 ng/lane), and 50 [A 5-fold concentrated IGF-stimulated medium were electrophoresed on 12% polyacrylamide gels, transferred to nitrocellulose, then probed with [128I]IGF-II as described in Materials and Methods. Ligand blots were exposed for 4 h (for IGFBP-1 and IGFBP-3) and 16 h (for transformed fibroblast IGFBP and IGF-induced IGFBP).
Discussion In this study we demonstrated that neonatal human fibroblasts in monolayer culture produce an IGFBP of 29-31K in response to IGF-I or IGF-II, but not insulin. The observed induction is rapid and sustained, with the 29-31K species being detectable within 10 h of exposure to IGFs and showing accumulation in the medium over at least 72 h. IGF-I has been reported to stimulate the secretion of IGFBPs in various nonprimate and human systems. In porcine, rat, and mouse myocyte cultures exposed to IGF-I (28), an increase in total IGF-binding activity in the medium was used as a measure of increased IGFBP production, without a particular IGFinducible protein being identified. It was also demonstrated that porcine insulin was equipotent with IGF-I
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 November 2015. at 15:01 For personal use only. No other uses without permission. . All rights reserved.
786
IGF-INDUCIBLE IGFBP FROM FIBROBLASTS
in stimulating IGFBP (28), in contrast with our findings which showed no response to insulin at a concentration 20 times that of IGF-I or IGF-II. A rapid response to IGF-I has also been reported in cultures of rat osteoblasts (29), in which 10 nM (equivalent to 75 ng/ml) recombinant human IGF-I induced a 32K IGFBP within 8 h of treatment. In vivo, Zapf and co-workers (18) have shown that infusion of IGF-I in hypophysectomized rats induces an IGFBP, which appears as a doublet of 33-34K, after 8 h. The size similarities and rapid inducibility of these proteins suggest that they represent the rat equivalent of the IGFBP we have described in this report. In studies of human IGFBPs, the relationship between IGF stimulation of IGFBPs and cell density has been investigated by Hill et al. (17), who reported that IGF-I and IGF-II stimulate the production of immunoreactive IGFBP-1 by human fetal fibroblasts, and that the degree of stimulation shows an inverse correlation with cell density. This is in contrast to our findings in neonatal fibroblasts, where ligand blotting of concentrated culture media from sparse and confluent fibroblasts, with and without stimulation by IGFs, showed IGFBP-3 and 2931K IGFBP production directly proportional to cell number and no evidence of IGFBP-1. Immunoblotting of medium proteins and immunoprecipitation of affinitylabeled medium samples confirmed the absence of IGFBP-1. Whether the difference is one of developmental age of the tissue from which the cells were derived or cell culture conditions remains to be determined. In vivo, infusion of recombinant human IGF-I into normal adult subjects over 3-6 days has been reported to result in an increase in the circulating level of a ~36K IGFBP (30). However, induction of the 36K IGFBP was prevented by simultaneous infusion of GH, whereas in our study coincubation of fibroblasts with GH had no effect on the inducibility of 29-3IK IGFBP by IGF-I. This difference in regulation by GH may reflect differences between the in vitro and in vivo systems used in the two studies or may indicate the IGF inducibility of two distinct human IGFBPs. Consistent with the latter possibility, the protein induced by IGF-I infusion in vivo showed N-terminal sequence homology with IGFBP-2 (30), whereas the protein induced in fibroblasts did not react with an IGFBP-2 antiserum that detected a human IGFBP purified from a transformed fibroblast line. The IGF-inducible fibroblast IGFBP also failed to react with antiserum to either human IGFBP-1 or IGFBP-3. Thus, since this protein does not appear to contain epitopes recognized by any of these antisera, definitive classification cannot be made on the basis of these immunological studies. Mol wt has been a commonly used criterion for assigning newly identified IGFBPs to one or another of the recognized classes; however, given the technical difficul-
Endo • 1990 Vol 127 • No 2
ties associated with determination of accurate mol wt by SDS-PAGE, the similarity in core protein size of the predicted sequences of IGFBP-1 (25.3K) (5), IGFBP-2 (31.4K) (16), or IGFBP-3 (28.7K) (31) makes it difficult to distinguish between these proteins and the 29-3IK IGFBP doublet on the basis of size alone. In addition, glycosylation variants and breakdown species of lower mol wt have been identified for IGFBP-3 and may well exist for the other IGFBPs; thus, a doublet protein of 29-3IK could conceivably represent a protein with a core structure belonging to any of the three structurally distinct IGFBP classes. Final classification will, therefore, not be possible until sequence data become available. Both IGF-I and IGF-II were able to stimulate the production of 29-3IK IGFBP, with half-maximal responses at approximately 15 and 27 ng/ml, respectively. The observed IGF-I effect occurs at a concentration similar to that required for half-maximal stimulation of DNA synthesis in neonatal and adult fibroblasts (32, 33). It is difficult to determine to what extent endogenous IGFs will effect induction of the 29-3IK IGFBP, given that estimates of IGF production by human fibroblasts have ranged from 0.08 U/ml (-15-20 ng/ml) (34) to 0.2 ng/ml or less (35). It is possible that human fibroblasts do not make 7.5K IGFs, but, instead, secrete high mol wt precursors, IGF-lA or IGF-IB, with mol wt of 17.5K and 22K, respectively, which are the products of alternative splicing of the primary IGF-I RNA transcript (36). Conover et al. (37) reported no detectable 7.5K IGFI immunoreactivity in fetal and postnatal fibroblastconditioned medium, but found, instead, an IGF-IA prohormone of 9-17K mol wt after acid-gel chromatography. Clemmons and Shaw (38) purified high mol wt (21.5K) immunoreactive IGF from fibroblast medium with the same amino acid composition as IGF-IB (39). This peptide was shown by Clemmons and Shaw (38) to have enhanced biological activity in stimulating [3H]thymidine uptake by human fibroblasts, but not BALBs/c 3T3 fibroblasts, compared to IGF-I. Whether the precursor proteins can function in the same way as mature 7.5K IGF to induce the production of IGFBPs is not known; if, in fact, the high mol wt forms of the growth factors are stimulatory, and endogenous levels are high, it remains to be explained why the 29-31K IGFBP is barely detectable in medium to which no IGFs have been added. The precise role of the IGFBPs in modulating the anabolic and mitogenic activities of IGFs is still unclear; pure and impure preparations of the three classified types of IGFBPs have been shown to inhibit IGF activity (32, 40-42), while IGFBP-1 and IGFBP-3 have been reported to potentiate the actions of IGFs (32, 43, 44). However, the existence of multiple types of IGFBP suggests that the members of the different IGFBP classes may have more specific roles. Such specificity may be reflected in
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 November 2015. at 15:01 For personal use only. No other uses without permission. . All rights reserved.
IGF-INDUCIBLE IGFBP FROM FIBROBLASTS the quite marked differences in regulation of production of the IGFBPs; IGFBP-1 is modulated by insulin and glucose, IGFBP-3 in the circulation is GH dependent, while 29-31K IGFBP is induced by IGF-I and IGF-II. Elucidation of the functional significance of these differences in regulation and activity will be essential for a full understanding of the role of the IGFBPs in regulating the biological activity of IGF-I and IGF-II.
18.
19.
20.
References 1. Baxter RC, Martin JL 1989 Binding proteins for the insulin-like growth factors: structure, regulation and function. Prog Growth Factor Res 1:49 2. Chochinov RH, Mariz IK, Hajek AS, Daughaday WH 1977 Characterization of a protein in mid-term human amniotic fluid which reacts in the somatomedin-C radioreceptor assay. J Clin Endocrinol Metab 44:902 3. Povoa G, Enberg G, Jornvall H, Hall K 1984 Isolation and characterization of a somatomedin-binding protein from mid-term human amniotic fluid. Eur J Biochem 144:199 4. Koistinen R, Kalkkinen N, Huhtala M-L, Seppala M, Bohn H, Rutanen E-M 1986 Placental protein 12 is a decidual protein that binds somatomedin and has an identical N-terminal amino acid sequence with somatomedin-binding protein from human amniotic fluid. Endocrinology 118:1375 5. Brinkman A, Groffen C, Kortleve DJ, Geurts Van Kessel A, Drop SLS 1988 Isolation and characterization of a cDNA encoding the lower molecular weight insulin-like growth factor binding protein (IBP-1). EMBOJ 7:2417 6. Suikkari A-M, Koivisto VA, Rutanen E-M, Yki-Jarvinen H, Karonen S-L, Seppala M 1988 Insulin regulates the serum levels of low molecular weight insulin-like growth factor binding protein. J Clin Endocrinol Metab 66:266 7. Yeoh SI, Baxter RC 1988 Metabolic regulation of the growth hormone independent insulin-like growth factor binding protein in human plasma. Acta Endocrinol (Copenh) 119:465 8. Lewitt MS, Baxter RC 1990 Inhibitors of glucose uptake stimulate the production of insulin-like growth factor-binding protein (IGFBP-1) by human fetal liver. Endocrinology 126:1527 9. Knauer DJ, Wagner FW, Smith GL 1981 Purification and characterization of multiplication-stimulating activity (MSA) carrier protein. J Supramol Struct Cell Biochem 15:177 10. Brown AL, Chiarotti L, Orlowski CC, Mehlman T, Burgess WH, Ackerman EJ, Bruni CB, Rechler MM 1989 Nucleotide sequence and expression of a cDNA clone encoding a fetal rat binding protein for insulin-like growth factors. J Biol Chem 264:5148 11. Szabo L, Mottershead DG, Ballard FJ, Wallace JC 1988 The bovine insulin-like growth factor (IGF) binding protein purified from conditioned medium requires the N-terminal tripeptide in IGF-I for binding. Biochem Biophys Res Commun 151:207 12. Binkert C, Landwehr J, Mary J-L, Schwander J, Heinrich G 1989 Cloning, sequence analysis and expression of a cDNA encoding a novel insulin-like growth factor binding protein (IGFBP-2). EMBO J 8:2497 13. Baxter RC, Martin JL 1989 Structure of the Mr 140,000 growth hormone-dependent insulin-like growth factor binding protein complex: determination by reconstitution and affinity-labeling. Proc Natl Acad Sci USA 86:6898 14. Martin JL, Baxter RC 1986 Insulin-like growth factor binding protein from human plasma: purification and characterization. J Biol Chem 261:8754 15. Baxter RC, Martin JL 1987 Binding proteins for the insulin-like growth factors in adult rat serum: comparison with other human and rat binding proteins. Biochem Biophys Res Commun 147:408 16. Walton PE, Baxter RC, Burleigh BD, Etherton TD 1989 Purification of the serum acid-stable insulin-like growth factor binding protein from the pig {Sus scrofa). Comp Biochem Physiol 92B:561 17. Hill DJ, Camacho-Hubner C, Rashid P, Strain AJ, Clemmons DR
21.
22. 23. 24.
25. 26.
27. 28. 29. 30.
31.
32.
33. 34. 35. 36. 37.
787
1989 Insulin-like growth factor (IGF)-binding protein release by human fetal fibroblasts: dependency on cell density and IGF peptides. J Endocrinol 122:87 Zapf J, Hauri C, Waldvogel M, Futo E, Hasler H, Binz K, Guler HP, Schmid C, Froesch ER 1989 Recombinant human insulin-like growth factor I induces its own specific carrier protein in hypophysectomized and diabetic rats. Proc Natl Acad Sci USA 86:3813 Clemmons DR, Thissen JP, Maes N, Ketelslegers JM, Underwood LE 1989 Insulin-like growth factor-I (IGF-I) infusion into hypophysectomized or protein-deprived rat induces specific IGF-binding proteins in serum. Endocrinology 125:2967 Walton PE, Etherton TD 1989 Effects of porcine growth hormone and insulin-like growth factor-I (IGF-I) on immunoreactive IGFbinding protein concentration in pigs. J Endocrinol 120:153 Martin JL, Baxter RC 1988 Insulin-like growth factor binding proteins (IGF-BPs) produced by human skin fibroblasts: immunological relationship to other human IGF-BPs. Endocrinology 123:1907 Martin JL, Willetts KA, Baxter RC 1990 Purification and properties of a novel insulin-like growth factor-II binding protein from transformed human fibroblasts. J Biol Chem 265:4124 Baxter RC, De Mellow JSM 1986 Measurement of insulin-like growth factor-II by radioreceptor assay using ovine placental membranes. Clin Endocrinol (Oxf) 24:26719 Baxter RC, Martin JL, Wood MH 1987 Two immunoreactive binding proteins for insulin-like growth factors in human amniotic fluid: relationship to fetal maturity. J Clin Endocrinol Metab 65:423 Baxter RC, Martin JL 1986 Radioimmunoassay of growth-hormone dependent insulin-like growth factor binding protein in human plasma. J Clin Invest 78:1504 Baxter RC, Martin JL, Beniac VA 1989 High molecular weight insulin-like growth factor binding protein complex. Purification and properties of the acid-labile subunit from human serum. J Biol Chem 264:11843 Baxter RC, Brown AS 1982 Purification of tracer for somatomedinC radioimmunoassay by hydrophobic interaction chromatography. Clin Chem 28:485 McCusker RH, Clemmons DR 1988 Insulin-like growth factor binding protein secretion by muscle cells: effect of cellular differentiation and proliferation. J Cell Physiol 137:505 Schmid C, Zapf J, Froesch ER 1989 Production of carrier proteins for insulin-like growth factors by rat osteoblastic cells. Regulation by IGF-I and cortisol. FEBS Lett 244:328 Zapf J, Guler HP, Schmid C, Froesch ER, Binoux M, Hossenlopp P, Born W, Fischer JA, Kiefer M, Masiarz F, Merryweather J, NTerminal amino acid sequence and cloning of a novel human insulin-like growth factor (IGF) carrier protein (CP) induced by IGF-I and suppressed by growth hormone (GH). 71st Annual Meeting of The Endocrine Society, Seattle WA, 1989, p 274 (Abstract) Wood WI, Cachianes G, Henzel WJ, Winslow GA, Spencer SA, Hellmiss R, Martin JL, Baxter RC 1988 Cloning and expression of the growth hormone-dependent insulin-like growth factor-binding protein. Mol Endocrinol 2:1176 De Mellow JSM, Baxter RC 1988 Growth hormone-dependent insulin-like growth factor (IGF) binding protein both inhibits and potentiates IGF-I stimulated DNA synthesis in human skin fibroblasts. Biochem Biophys Res Commun 156:199 Conover CA, Rosenfeld RG, Hintz RL 1987 Insulin-like growth factor-II binding and action in human fetal fibroblasts. J Cell Physiol 133:560 Clemmons DR, Underwood LE, Van Wyk JJ 1981 Hormonal control of immunoreactive somatomedin production by cultured human fibroblasts. J Clin Invest 67:10 Adams SO, Kapadia M, Mills B, Daughaday WH 1984 Release of insulin-like growth factors and binding protein activity into serumfree medium of cultured human fibroblasts. Endocrinology 115:520 Rotwein P 1986 Two insulin-like growth factor I messenger RNAs are expressed in human liver. Proc Natl Acad Sci USA 83:77 Conover CA, Baker BK, Hintz RL 1989 Cultured human fibroblasts secrete insulin-like growth factor 1A prohormone. J Clin Endocri-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 November 2015. at 15:01 For personal use only. No other uses without permission. . All rights reserved.
788
IGF-INDUCIBLE IGFBP FROM FIBROBLASTS
nol Metab 69:25 38. Clemmons DR, Shaw DS 1986 Purification and biological properties of fibroblast somatomedin. J Biol Chem 261:10293 39. Rotwein P, Folz RJ, Gordon JI1987 Biosynthesis of human insulinlike growth factor I (IGF-I). J Biol Chem 262:11807 40. Drop SLS, Valiquette G, Guyda HJ, Corvol MT, Posner BI 1979 Partial purification and characterization of a binding protein for insulin-like activity. Acta Endocrinol (Copenh) 90:505 41. Ritvos O, Ranta T, Jalkanen J, Suikkari A-M, Voutilainen R, Bohn H, Rutanen E-M 1988 Insulin-like growth factor (IGF) binding protein from human decidua inhibits the binding and biological
Endo • 1990 Vol 127 • No 2
action of IGF-I in cultured choriocarcinoma cells. Endocrinology 122:2150 42. Knauer DJ, Smith GL 1980 Inhibition of biological activity of multiplication-stimulating activity by binding to its carrier protein. Proc Natl Acad Sci USA 77:7252 43. Blum WF, Jenne EW, Reppin F, Kietzmann K, Ranke MB, Bierich JR 1989 Insulin-like growth factor (IGF-I)-binding protein complex is a better mitogen than free IGF-I. Endocrinology 125:766 44. Elgin RG, Busby Jr WH, Clemmons DR 1987 An insulin-like growth factor (IGF) binding protein enhances the biologic response to IGF-I. Proc Natl Acad Sci USA 84:3254
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 November 2015. at 15:01 For personal use only. No other uses without permission. . All rights reserved.
Vol. 127, No. 2 Printed in U.S.A.
Production of an Insulin-Like Growth Factor (IGF) Inducible IGF-Binding Protein by Human Skin Fibroblasts* JANET L. MARTIN AND ROBERT C. BAXTER Department of Medicine, University of Sydney, New South Wales 2006; and the Department of Endocrinology, Royal Prince Alfred Hospital, Camperdown, New South Wales 2050, Australia
ABSTRACT. Neonatal human skin fibroblasts produce insulin-like growth factor-binding proteins (IGFBPs) that have the potential to modulate the actions of the growth factors. We have examined the IGFBPs secreted by monolayer cultures of neonatal fibroblasts by ligand blotting with [125I]IGF-II and immunoblotting with antisera raised against three well characterized IGFBPs: IGFBP-1, IGFBP-2, and IGFBP-3. As detected by ligand blotting, medium from untreated fibroblasts contained IGFBP-3, a second IGFBP which appeared as a doublet of 2931K, and a smaller protein of 22K. Within 10 h of the addition of 50 ng/ml IGF-I, a markedly increased level of production of the 29-31K IGFBP doublet was detectable, with levels increasing
8- to 9-fold by 24 h compared to that in untreated samples. IGFI was approximately twice as potent as IGF-II at inducing 2931K IGFBP, with a half-maximal response at 15.4 ± 2.7 ng/ml IGF-I and 26.6 ± 1.6 ng/ml IGF-II (n = 3). Insulin tested at 1 Mg/ml did not induce 29-31K IGFBP. Neither GH nor the acidlabile subunit of the circulating high mol wt IGFBP complex induced 29-31K IGFBP or affected its induction by IGF-I or IGF-II. Immunoblotting demonstrated that IGF-inducible IGFBP did not react with antibodies to IGFBP-1, IGFBP-2, or IGFBP-3. These results indicate that IGF-I and IGF-II induce an IGFBP that is different from previously characterized human IGFBPs. (Endocrinology 127: 781-788, 1990)
I
T HAS been acknowledged for many years that the binding proteins (BPs) for the insulin-like growth factors (IGFs) have the potential to play an important role in the actions of the growth factors themselves. As recently reviewed, three distinct classes of IGFBPs have been identified (1). IGFBP-1 was first identified in amniotic fluid (2) and purified from this source (3) and human decidua (4); the cDNA for this protein has now been cloned and characterized (5). IGFBP-1 is regulated by glucose and insulin, as shown in both clinical studies (6, 7) and a tissue culture model of IGFBP-1 production (8). The second type of IGFBP, IGFBP-2, was originally purified from medium conditioned by the rat liver-derived cell line BRL-3A (9) and has recently been cloned from this source (10); the bovine homolog has been purified from the MDBK kidney cell line (11), while the primary structure for human IGFBP-2 has been determined from cDNA cloned from a human fetal liver library (12). The GH-dependent IGFBPs make up the third class of BPs. IGFBP-3, also known as BP-53, is the IGF-binding component that, in the presence of an
85-kilodalton (85K) acid-labile subunit and IGF-I or IGF-II, forms the high mol wt complex in which most of the IGFs are found in the circulation (13). First isolated from human plasma (14), homologous proteins have been isolated from the sera of rats (15) and pigs (16). IGFBP3 is glycosylated, appears as two IGF-binding bands on sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), and binds IGF-I and IGF-II with similar high affinity. Recent evidence indicates that IGFs are able to modulate the production of some of their BPs both in cell culture and in vivo. In sparse monolayer culture of fetal fibroblasts, IGF-I and IGF-II have been reported to induce immunoreactive IGFBP-1 production (17). In vivo, recombinant human IGF-I stimulates the production of 42K, 45K, and 49K glycosylated IGFBPs in diabetic or hypophysectomized rats (18, 19); since these proteins are able to form part of the GH-dependent complex, they are the rat homologs of the human IGFBP3 doublet. In contrast, Walton and Etherton (20) have shown no effect of infusion of IGF-I on immunoreactive IGFBP-3 levels in pigs, while demonstrating that circulating levels of both IGF-I and IGFBP-3 in these animals were increased in response to GH. Normal and transformed human fibroblasts produce a variety of IGFBPs (21,22). In this report we demonstrate
Received January 26, 1990. Address requests for reprints to: Janet Martin, Department of Medicine, Blackburn Building (D-06), University of Sydney, New South Wales 2006, Australia. * This work was supported by the National Health and Medical Research Council of Australia.
781
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 November 2015. at 15:01 For personal use only. No other uses without permission. . All rights reserved.
IGF-INDUCIBLE IGFBP FROM FIBROBLASTS
782
that in neonatal human fibroblasts, IGF-I and IGF-II, but not insulin, specifically induce the production of an IGFBP different from previously described human IGFBPs.
Materials and Methods Tissue culture plasticware, culture media, and fetal calf serum were purchased from Cytosystems (North Ryde, New South Wales, Australia); antibiotics, BSA (fraction V, RIA grade), bovine insulin, and protein-A were obtained from Sigma Chemical Co. (St. Louis, MO). Disuccinimidyl suberate was purchased from Pierce Chemical Co. (Rockford, IL). Electrophoresis consumables were obtained from Pharmacia (Uppsala, Sweden), and Bio-Rad (Richmond, CA). Autoradiographic film was purchased from
Amersham
medium (250 (A) were removed from each well, frozen, and replaced with fresh medium containing the test substance. All media were collected at 72 h and frozen until assayed. IGFBP3 was measured in conditioned medium samples by RIA (25), using antibody R7 (21). Sample preparation for SDS-PAGE analysis
Reagents
(Hyperfilm-MP)
E n d o • 1990 Vol 127 • No 2
(Bucks,
United Kingdom). IGF-I and IGF-II (23) and IGFBP-3 (14) were purified from Cohn fraction IV of human plasma, and IGFBP-1 (24) from amniotic fluid, as previously reported. Polyclonal antisera raised against IGFBP-3 and IGFBP-1 have been described previously (24,25); antiserum to bovine IGFBP2, raised against IGFBP purified from medium conditioned by MDBK cells, was the generous gift of Dr. John Ballard, CSIRO Division of Human Nutrition (Adelaide, South Australia), and Dr. John Wallace, University of Adelaide, South Australia. The a-subunit (acid-labile subunit) of the 140K GH-dependent IGF complex was purified as previously described (26). IGFBP secreted by SV-40-transformed IMR-90 fibroblasts (line AG 2804) was isolated from conditioned culture medium as previously reported (22). Biosynthetic methionyl human GH (hGH) was a generous gift from Kabi (Stockholm, Sweden). Tracers IGF-I and IGF-II tracers were prepared, as described in previous studies (23, 27), to specific activities of about 200 Ci/ g. Protein-A tracer was radioiodinated with Na125I (Amersham), using chloramine-T. Cross-linked tracer for the IGFBP-3 RIA (IGFBP-3 affinity-labeled with IGF-I) was prepared as previously described (25). Cell culture Neonatal fibroblasts were isolated as outgrowths from foreskin explants and maintained in monolayer culture at 37 C in Ham's F-12 medium containing 20 mM HEPES, 2 mM glutamine, and 10% fetal calf serum. Cultures were passaged by trypsin-EDTA treatment when confluent (every 7-10 days, 1:3 split) and used for experiments between the 3rd and 10th generations. For stimulation experiments, cells were seeded into 12-place multiwells at a density of 4 x 104 cells/well in medium containing 10% fetal calf serum and grown to confluence. Fibroblast monolayers were then rinsed with serum-free medium and incubated for a minimum of 48 h in medium without serum. This medium was replaced with 1.5 ml medium containing 2 g/ liter BSA with or without test substances as indicated. Twentyfour and 48 h after the initial additions, samples of conditioned
Conditioned media from stimulation experiments were acidified, ultrafiltered to remove IGFs, and concentrated before SDS-PAGE. Generally, 1-ml pools of media were acidified with an equal volume of 2 M acetic acid, incubated at room temperature for 1 h, then concentrated to about 200 ^1 by centrifugation through a Centricon-10 membrane (Amicon Corp., Lexington, MA). Samples were flushed through with a further 5 ml 1 M acetic acid, and the final concentrate (~100 /A) was lyophilized. Freeze-dried samples were reconstituted in 200 ^l 50 mM sodium phosphate, pH 6.5. Preliminary experiments using radiolabeled IGF-I indicated that this procedure removed >95% of the IGFs from the samples. Affinity labeling, SDS-PAGE, Western blotting, and densitometry To affinity label medium concentrates, 50-^1 aliquots were incubated overnight at 4 C with [125I]IGF-I or [125I]IGF-II in a final volume of 150 ii\ 50 mM sodium phosphate, pH 6.5. Disuccinimidyl suberate (7.5 mM in dimethylsulfoxide) was added to a final concentration of 0.25 mM, and the reaction was stopped 30 min later by the addition of 25 ^11 M Tris-HCl, pH 8. Pure proteins (IGFBP-1, IGFBP-3, and IGFBP purified from transformed fibroblast medium) were affinity labeled in buffer containing 0.5 g/liter BSA using the same protocol. Immunoprecipitation of affinity-labeled proteins was carried out as previously described (21). Medium concentrates, affinity-labeled samples, and calibration standards (Pharmacia) were prepared for electrophoresis by the addition of sample buffer to give final concentrations of 0.0125 M Tris (pH 6.8), 3% SDS, and 10% glycerol and boiled for 5 min; separation of proteins was achieved on 12% polyacrylamide gels. Gels containing affinity-labeled samples were then fixed and stained with 2.5 g/liter Coomassie blue in 25% isopropanol-10% acetic acid, destained in 25% methanol-10% acetic acid, then dried and exposed to Hyperfilm-MP autoradiography film for 3-7 days. For blotting studies, proteins were transferred to nitrocellulose (0.45 /*m; Schleicher and Schuell, Dassel, West Germany), using a Novablot electrophoretic transfer unit (LKB-Pharmacia Biotechnology, Bromma, Sweden). After electrophoresis, gels were placed in transfer buffer containing 39 mM glycine, 48 mM Tris, and 20% methanol (pH unadjusted) and shaken gently in this buffer for a minimum of 1 h. Transfer was completed over 1.5-2 h at 250 mamp, then the nitrocellulose sheets were blocked in 30 g/liter BSA at 37 C for a minimum of 6 h. For ligand blotting, sheets were then incubated with [125I]IGF-II (~1 X 106 cpm) at 4 C for 16 h, then washed three times with cold 50 mM sodium phosphate, pH 6.5, once with the same buffer containing 0.05% Nonidet P-40, then twice more in detergent-free buffer. For immunoblotting, blocked
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 November 2015. at 15:01 For personal use only. No other uses without permission. . All rights reserved.
IGF-INDUCIBLE IGFBP FROM FIBROBLASTS nitrocellulose sheets with transferred proteins were incubated with anti-hIGFBP-3 antibody (R7 antiserum; 1:500 final dilution) or anti-bIGFBP-2 antiserum (1:250 dilution) at 22 C or with anti-hIGFBP-1 antibody (A2 antiserum; 1:1000 final dilution) at 4 C. Incubation was continued for 16 h, then blots were washed as described above and incubated at 22 C for 1-2 h with 125I-labeled protein-A (~1 x 106 cpm). After washing again, Hyperfilm-MP was exposed to dried blots for 1-2 days at -70 C. Quantitation of developed autoradiographs was carried out using a Bio-Rad model 620 Video Densitometer attached to a chart recorder. The technique was validated by ligand blotting dilutions of medium with IGF-II tracer, scanning the autoradiograph, and plotting the absorbance peak height, expressed in arbitrary units, against concentration. The relationship obtained using this technique was linear between 0.05-50 ^1 medium and 4-80 arbitrary absorbance units (r2 = 0.992). The coefficient of variation of duplicate scans on a single sample, determined by analysis of variance on 12 duplicates, was 4.1%. All experiments were performed at least 3 times.
Results IGFBPs produced by neonatal skin fibroblasts over 3 days were detected by ligand blotting conditioned medium with [125I]IGF-II after fractionation by SDSPAGE. Figure 1 compares these IGFBPs with a pure plasma IGFBP-3 standard, which appeared as a charac94 • •
67 • •
43 • •
30 " •
20 • • BP-3
24h 48h 72h
24h 48h 72h
untreated
IGF-I treated
FIG. 1. [126I]IGF-II ligand blotting of conditioned medium from untreated and IGF-I-treated fibroblasts. Twenty-four-, 48- and 72-h conditioned media from untreated fibroblasts and those treated with 50 ng/ml IGF-I were electrophoresed unreduced on 12% polyacrylamide gels and transferred to nitrocellulose, as described in Materials and Methods. Blots were incubated with [125I]IGF-II, washed, and exposed to Hyperfilm-MP for 2 days. The positions of mol wt markers are indicated on the left by arrows. IGFBP-3 (500 ng), purified from human plasma as previously described (14) and indicated in the figure as BP3, is also shown for size comparison.
783
teristic doublet of approximately 50K (15). Cells incubated with or without 50 ng/ml IGF-I produced an IGFbinding doublet that comigrated with plasma IGFBP-3 and another doublet of approximately 29-31K. A minor IGF-binding species of 22K was also observed. In cells grown without IGF-I, the IGFBP-3 doublet was the predominant binding species, with the 22K and 29-31K bands barely detectable. IGF-I markedly induced the 293IK doublet, with a relatively small effect on the other IGFBPs. To quantitate these changes, autoradiographs were scanned with a densitometer, and the resulting peak heights for each radioactive species were compared (Fig. 2a). When quantitated in this way in three experiments, fibroblast IGFBP-3 and the 22K IGFBP were increased 1.5-2 fold by exposure of the cells to 50 ng/ml IGF-I, although this stimulation of production was not significant in the 24-h samples. RIA for IGFBP-3 in these samples was consistent with this finding (not shown); over four experiments, incubation with 50 ng/ml IGF-I over a 72-h period increased IGFBP-3 levels by 73 ± 10% (mean ± SE). A much more dramatic increase was observed with the 29-31K doublet. Within 24 h of IGF-I treatment, IGFBP activity was 15- to 20-fold higher in these media than in the controls, and this increased rate of production was maintained over 3 days. To examine the time course of induction of 29-3IK IGFBP in more detail, samples of media were taken at 2-h intervals for 24 h after the addition of IGF-I, concentrated, then analyzed by ligand blotting with IGF-II tracer and densitometry as before. As seen in Fig. 2b, 29-31K IGFBP was detectable in the medium within 10 h of IGF addition, and an increased rate of production was observed over the following 12 h, reaching a maximum between 22-24 h. To determine whether both IGF-I and IGF-II were able to induce the 29-31K IGFBP, fibroblasts were incubated with increasing concentrations of peptides in the range 2.5-50 ng/ml, and the 72-h conditioned medium was analyzed by IGF-II ligand blotting (Fig. 3). Both peptides showed dose-dependent stimulation of production of the 29-31K species, with the response showing greater sensitivity to IGF-I (Fig. 3a). Quantitation of these data by densitometry confirmed that IGF-I was approximately 2-fold more potent than IGF-II in its stimulation of IGFBP (Fig. 3b), with a half-maximal response at 15.4 ± 2.7 ng/ml for IGF-I and 26.6 ± 1.6 ng/ml for IGF-II (three experiments). When tested in the range 1-1000 ng, insulin was not able to induce production of the 29-31K IGFBP. The relationship between cell density and production of 29-31K IGFBP was also determined (not shown). Fibroblasts plated at densities of 1, 2, 4, or 8 X 104 cells/well and treated with either IGF-I or IGF-II showed production of 29-31K
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 November 2015. at 15:01 For personal use only. No other uses without permission. . All rights reserved.
IGF-INDUCIBLE IGFBP FROM FIBROBLASTS
784
22K IGFBP
29-31K IGFBP
IGFBP-3
E n d o • 1990 Vol 127-No 2
100
O80
o «60 I 40 20 "
20
3
1
2
3
1
4
Incubation time (days)
8
12
16
20
24
Incubation time (hours)
FIG. 2. IGF-stimulated production of IGFBPs by human fibroblasts. a, Induction of three IGFBPs over 3 days. Autoradiographs similar to those shown in Fig. 1 were scanned using a densitometer; peak absorbance heights for the IGFBP-3, 29-31K IGFBP, and 22K IGFBP species from unstimulated (•) and 50 ng/ml IGF-stimulated (M) samples are expressed in arbitrary units, where a 100% response represents the optical density of 29-31K IGFBP bands after 3-day induction by IGF-I. Statistical analysis was performed separately for each IGFBP by analysis of variance and Duncan's multiple range test; error bars represent the SD for three experiments. Significance of IGF-I effect at each time point: *, P < 0.005; **, P < 0.001. b, Twenty-four-hour time course of 29-31K IGFBP production. Fibroblasts were incubated with 50 ng/ml IGF-I, and medium samples were collected at 2-h intervals for 24 h and analyzed by SDS-PAGE, ligand blotting, and densitometry, as described in Materials and Methods. FIG. 3. Stimulation of 29-31K IGFBP production by IGF-I and IGF-II. a, Fibroblast cultures were incubated for 72 h in medium containing IGF-I (upper panel) or IGF-II (lower panel) in the range of 2.5-50 ng/ml, and the conditioned media were analyzed by IGF-II ligand blotting as described in Materials and Methods. Blots were exposed to Hyperfilm-MP for 2 days, b, Dose-response curves generated by densitometry after autoradiography. • , IGF-I; • , IGF-II. Data are the mean ± SD for three experiments. Also shown is an insulin dosecurve (O) derived from an identical experiment; no effect of insulin was seen up to 1000 ng/ml.
a IGF-I stimulation
2.5 5
100 •
10 25 ng/ml
50
IGF-II stimulation
2.5 5
10 25 ng/ml
IGFBP in proportion to cell number, with no difference in the response to IGF-I or IGF-II. In view of the fact that IGF-I and IGF-II circulate as part of a GH-dependent high mol wt complex, we next determined whether GH or the a (acid-labile)-subunit of the circulating complex (13) was able to elicit the same response as the IGFs, i.e. induction of 29-3IK IGFBP. Fibroblasts were treated with hGH (1 /ig/ml), a-subunit (1 /ug/ml), and IGF-I (50 ng/ml), either alone or in combination, and the conditioned medium was analyzed by IGF-II ligand blotting and autoradiography. As seen in Fig. 4, hGH, a-subunit, or a combination of the two caused no change in the intensity of the 29-31K doublet compared with that in untreated fibroblasts. However, in all samples in which IGF-I was included in the incubation, 29-3IK IGFBP was much more intensely labeled, indicating increased production of this protein. Com-
50
Peptide concentration (ng/ml)
• untr
GH
1
fey
•J IGF-I
a
IGF-I + GH
GH +a
IGF-I +a
IGF-I + GH +a
FIG. 4. Effect of treatment with IGF-I, GH, and the a-subunit of the GH-dependent IGFBP complex on 29-31K IGFBP induction. Fibroblasts were left untreated (untr) or treated with GH (1 fig/ml), otsubunit (1 Mg/ml), or IGF-I (50 ng/ml), alone or in combination as indicated, for 72 h. Ligand blotting of conditioned media was performed as described in Materials and Methods using [125I]IGF-II. The migration position on SDS-PAGE of a 30K mol wt marker is indicated by the arrow.
pared to the effect of IGF-I alone, the addition of either hGH or a-subunit had no significant effect, but a 52%
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 November 2015. at 15:01 For personal use only. No other uses without permission. . All rights reserved.
IGF-INDUCIBLE IGFBP FROM FIBROBLASTS
increase in labeling intensity was seen when both hGH and a-subunit were added (P = 0.047; n = 4). A direct size comparison between the IGF-induced IGFBP and other IGFBPs (determined by SDS-PAGE, ligand blotting, and autoradiography) is shown in Fig. 5. The IGF-inducible IGFBP is distinct from the IGFBP-3 doublet, the IGFBP-1 purified from amniotic fluid (28K), and the IGFBP purified from transformed human fibroblast medium (34K). The immunological relationship of induced 29-31K IGFBP to other IGFBPs was examined by immunoblotting with antisera to IGFBP-1, IGFBP-2, and IGFBP-3. As shown in Fig. 6, antiserum to IGFBP1 detects this protein purified from amniotic fluid, but does not react with the 29-3IK IGFBP or any other IGFBPs in medium conditioned by IGF-stimulated neonatal skin fibroblasts. Similarly, antiserum raised against bovine IGFBP-2 detected IGFBP purified from human transformed fibroblast-conditioned medium, but did not react with 29-3IK IGFBP from IGF-treated fibroblasts. Also shown in this figure and as reported previously (21), neonatal fibroblasts produce immunoreactive IGFBP-3 identical in size to human plasma IGFBP-3; however, IGF-I LIGAND BLOTTING
67 • •
43 • •
30 • •
g|i t i l l m •
20*IGFBP-1
Transformed fibroblast IGFBP
IMMUNOBLOTTING
67
43
30
20 IGFBP-1 medium IGFBP-1 ANTISERUM
Transformed fibroblast medium
IGFBP-3 medium
IGFBP-2 ANTISERUM IGFBP-3 ANTISERUM
FIG. 6. Immunoreactivity of 29-31K IGFBP compared to other human IGFBPs. IGFBP-1 (500 ng), transformed fibroblast IGFBP-1 (~1 ng), IGFBP-3 (500 ng), and IGF-I-treated medium (50 nl 5-fold concentrate) were electrophoresed and transferred to nitrocellulose. Blots were incubated with polyclonal antisera to human amniotic fluid IGFBP-1 (1:1000 final dilution), bovine IGFBP-2 (1:250), and human plasma IGFBP-3 (1:500) as indicated. Immune complexes were then detected by incubation with 125I-labeled protein-A (~1 X 106 cpm) and autoradiography, as described in Materials and Methods.
antiserum to this protein does not react with 29-3IK IGFBP doublet. These findings were confirmed using the same antisera to precipitate [125I] IGF-I affinity-labeled medium (not shown); none of the sera precipitated crosslinked species corresponding to the IGF-inducible IGFBP. Taken together, these results indicate that the IGF-induced IGFBP is immunologically distinct from IGFBP-1, IGFBP-2, and IGFBP-3.
lliiii
IGFBP-3
785
Medium
FIG. 5. Size comparison of human IGFBPs. Human plasma IGFBP-3, human amniotic fluid-derived IGFBP-1, IGFBP purified from transformed human fibroblasts (each at 500 ng/lane), and 50 [A 5-fold concentrated IGF-stimulated medium were electrophoresed on 12% polyacrylamide gels, transferred to nitrocellulose, then probed with [128I]IGF-II as described in Materials and Methods. Ligand blots were exposed for 4 h (for IGFBP-1 and IGFBP-3) and 16 h (for transformed fibroblast IGFBP and IGF-induced IGFBP).
Discussion In this study we demonstrated that neonatal human fibroblasts in monolayer culture produce an IGFBP of 29-31K in response to IGF-I or IGF-II, but not insulin. The observed induction is rapid and sustained, with the 29-31K species being detectable within 10 h of exposure to IGFs and showing accumulation in the medium over at least 72 h. IGF-I has been reported to stimulate the secretion of IGFBPs in various nonprimate and human systems. In porcine, rat, and mouse myocyte cultures exposed to IGF-I (28), an increase in total IGF-binding activity in the medium was used as a measure of increased IGFBP production, without a particular IGFinducible protein being identified. It was also demonstrated that porcine insulin was equipotent with IGF-I
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 November 2015. at 15:01 For personal use only. No other uses without permission. . All rights reserved.
786
IGF-INDUCIBLE IGFBP FROM FIBROBLASTS
in stimulating IGFBP (28), in contrast with our findings which showed no response to insulin at a concentration 20 times that of IGF-I or IGF-II. A rapid response to IGF-I has also been reported in cultures of rat osteoblasts (29), in which 10 nM (equivalent to 75 ng/ml) recombinant human IGF-I induced a 32K IGFBP within 8 h of treatment. In vivo, Zapf and co-workers (18) have shown that infusion of IGF-I in hypophysectomized rats induces an IGFBP, which appears as a doublet of 33-34K, after 8 h. The size similarities and rapid inducibility of these proteins suggest that they represent the rat equivalent of the IGFBP we have described in this report. In studies of human IGFBPs, the relationship between IGF stimulation of IGFBPs and cell density has been investigated by Hill et al. (17), who reported that IGF-I and IGF-II stimulate the production of immunoreactive IGFBP-1 by human fetal fibroblasts, and that the degree of stimulation shows an inverse correlation with cell density. This is in contrast to our findings in neonatal fibroblasts, where ligand blotting of concentrated culture media from sparse and confluent fibroblasts, with and without stimulation by IGFs, showed IGFBP-3 and 2931K IGFBP production directly proportional to cell number and no evidence of IGFBP-1. Immunoblotting of medium proteins and immunoprecipitation of affinitylabeled medium samples confirmed the absence of IGFBP-1. Whether the difference is one of developmental age of the tissue from which the cells were derived or cell culture conditions remains to be determined. In vivo, infusion of recombinant human IGF-I into normal adult subjects over 3-6 days has been reported to result in an increase in the circulating level of a ~36K IGFBP (30). However, induction of the 36K IGFBP was prevented by simultaneous infusion of GH, whereas in our study coincubation of fibroblasts with GH had no effect on the inducibility of 29-3IK IGFBP by IGF-I. This difference in regulation by GH may reflect differences between the in vitro and in vivo systems used in the two studies or may indicate the IGF inducibility of two distinct human IGFBPs. Consistent with the latter possibility, the protein induced by IGF-I infusion in vivo showed N-terminal sequence homology with IGFBP-2 (30), whereas the protein induced in fibroblasts did not react with an IGFBP-2 antiserum that detected a human IGFBP purified from a transformed fibroblast line. The IGF-inducible fibroblast IGFBP also failed to react with antiserum to either human IGFBP-1 or IGFBP-3. Thus, since this protein does not appear to contain epitopes recognized by any of these antisera, definitive classification cannot be made on the basis of these immunological studies. Mol wt has been a commonly used criterion for assigning newly identified IGFBPs to one or another of the recognized classes; however, given the technical difficul-
Endo • 1990 Vol 127 • No 2
ties associated with determination of accurate mol wt by SDS-PAGE, the similarity in core protein size of the predicted sequences of IGFBP-1 (25.3K) (5), IGFBP-2 (31.4K) (16), or IGFBP-3 (28.7K) (31) makes it difficult to distinguish between these proteins and the 29-3IK IGFBP doublet on the basis of size alone. In addition, glycosylation variants and breakdown species of lower mol wt have been identified for IGFBP-3 and may well exist for the other IGFBPs; thus, a doublet protein of 29-3IK could conceivably represent a protein with a core structure belonging to any of the three structurally distinct IGFBP classes. Final classification will, therefore, not be possible until sequence data become available. Both IGF-I and IGF-II were able to stimulate the production of 29-3IK IGFBP, with half-maximal responses at approximately 15 and 27 ng/ml, respectively. The observed IGF-I effect occurs at a concentration similar to that required for half-maximal stimulation of DNA synthesis in neonatal and adult fibroblasts (32, 33). It is difficult to determine to what extent endogenous IGFs will effect induction of the 29-3IK IGFBP, given that estimates of IGF production by human fibroblasts have ranged from 0.08 U/ml (-15-20 ng/ml) (34) to 0.2 ng/ml or less (35). It is possible that human fibroblasts do not make 7.5K IGFs, but, instead, secrete high mol wt precursors, IGF-lA or IGF-IB, with mol wt of 17.5K and 22K, respectively, which are the products of alternative splicing of the primary IGF-I RNA transcript (36). Conover et al. (37) reported no detectable 7.5K IGFI immunoreactivity in fetal and postnatal fibroblastconditioned medium, but found, instead, an IGF-IA prohormone of 9-17K mol wt after acid-gel chromatography. Clemmons and Shaw (38) purified high mol wt (21.5K) immunoreactive IGF from fibroblast medium with the same amino acid composition as IGF-IB (39). This peptide was shown by Clemmons and Shaw (38) to have enhanced biological activity in stimulating [3H]thymidine uptake by human fibroblasts, but not BALBs/c 3T3 fibroblasts, compared to IGF-I. Whether the precursor proteins can function in the same way as mature 7.5K IGF to induce the production of IGFBPs is not known; if, in fact, the high mol wt forms of the growth factors are stimulatory, and endogenous levels are high, it remains to be explained why the 29-31K IGFBP is barely detectable in medium to which no IGFs have been added. The precise role of the IGFBPs in modulating the anabolic and mitogenic activities of IGFs is still unclear; pure and impure preparations of the three classified types of IGFBPs have been shown to inhibit IGF activity (32, 40-42), while IGFBP-1 and IGFBP-3 have been reported to potentiate the actions of IGFs (32, 43, 44). However, the existence of multiple types of IGFBP suggests that the members of the different IGFBP classes may have more specific roles. Such specificity may be reflected in
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 November 2015. at 15:01 For personal use only. No other uses without permission. . All rights reserved.
IGF-INDUCIBLE IGFBP FROM FIBROBLASTS the quite marked differences in regulation of production of the IGFBPs; IGFBP-1 is modulated by insulin and glucose, IGFBP-3 in the circulation is GH dependent, while 29-31K IGFBP is induced by IGF-I and IGF-II. Elucidation of the functional significance of these differences in regulation and activity will be essential for a full understanding of the role of the IGFBPs in regulating the biological activity of IGF-I and IGF-II.
18.
19.
20.
References 1. Baxter RC, Martin JL 1989 Binding proteins for the insulin-like growth factors: structure, regulation and function. Prog Growth Factor Res 1:49 2. Chochinov RH, Mariz IK, Hajek AS, Daughaday WH 1977 Characterization of a protein in mid-term human amniotic fluid which reacts in the somatomedin-C radioreceptor assay. J Clin Endocrinol Metab 44:902 3. Povoa G, Enberg G, Jornvall H, Hall K 1984 Isolation and characterization of a somatomedin-binding protein from mid-term human amniotic fluid. Eur J Biochem 144:199 4. Koistinen R, Kalkkinen N, Huhtala M-L, Seppala M, Bohn H, Rutanen E-M 1986 Placental protein 12 is a decidual protein that binds somatomedin and has an identical N-terminal amino acid sequence with somatomedin-binding protein from human amniotic fluid. Endocrinology 118:1375 5. Brinkman A, Groffen C, Kortleve DJ, Geurts Van Kessel A, Drop SLS 1988 Isolation and characterization of a cDNA encoding the lower molecular weight insulin-like growth factor binding protein (IBP-1). EMBOJ 7:2417 6. Suikkari A-M, Koivisto VA, Rutanen E-M, Yki-Jarvinen H, Karonen S-L, Seppala M 1988 Insulin regulates the serum levels of low molecular weight insulin-like growth factor binding protein. J Clin Endocrinol Metab 66:266 7. Yeoh SI, Baxter RC 1988 Metabolic regulation of the growth hormone independent insulin-like growth factor binding protein in human plasma. Acta Endocrinol (Copenh) 119:465 8. Lewitt MS, Baxter RC 1990 Inhibitors of glucose uptake stimulate the production of insulin-like growth factor-binding protein (IGFBP-1) by human fetal liver. Endocrinology 126:1527 9. Knauer DJ, Wagner FW, Smith GL 1981 Purification and characterization of multiplication-stimulating activity (MSA) carrier protein. J Supramol Struct Cell Biochem 15:177 10. Brown AL, Chiarotti L, Orlowski CC, Mehlman T, Burgess WH, Ackerman EJ, Bruni CB, Rechler MM 1989 Nucleotide sequence and expression of a cDNA clone encoding a fetal rat binding protein for insulin-like growth factors. J Biol Chem 264:5148 11. Szabo L, Mottershead DG, Ballard FJ, Wallace JC 1988 The bovine insulin-like growth factor (IGF) binding protein purified from conditioned medium requires the N-terminal tripeptide in IGF-I for binding. Biochem Biophys Res Commun 151:207 12. Binkert C, Landwehr J, Mary J-L, Schwander J, Heinrich G 1989 Cloning, sequence analysis and expression of a cDNA encoding a novel insulin-like growth factor binding protein (IGFBP-2). EMBO J 8:2497 13. Baxter RC, Martin JL 1989 Structure of the Mr 140,000 growth hormone-dependent insulin-like growth factor binding protein complex: determination by reconstitution and affinity-labeling. Proc Natl Acad Sci USA 86:6898 14. Martin JL, Baxter RC 1986 Insulin-like growth factor binding protein from human plasma: purification and characterization. J Biol Chem 261:8754 15. Baxter RC, Martin JL 1987 Binding proteins for the insulin-like growth factors in adult rat serum: comparison with other human and rat binding proteins. Biochem Biophys Res Commun 147:408 16. Walton PE, Baxter RC, Burleigh BD, Etherton TD 1989 Purification of the serum acid-stable insulin-like growth factor binding protein from the pig {Sus scrofa). Comp Biochem Physiol 92B:561 17. Hill DJ, Camacho-Hubner C, Rashid P, Strain AJ, Clemmons DR
21.
22. 23. 24.
25. 26.
27. 28. 29. 30.
31.
32.
33. 34. 35. 36. 37.
787
1989 Insulin-like growth factor (IGF)-binding protein release by human fetal fibroblasts: dependency on cell density and IGF peptides. J Endocrinol 122:87 Zapf J, Hauri C, Waldvogel M, Futo E, Hasler H, Binz K, Guler HP, Schmid C, Froesch ER 1989 Recombinant human insulin-like growth factor I induces its own specific carrier protein in hypophysectomized and diabetic rats. Proc Natl Acad Sci USA 86:3813 Clemmons DR, Thissen JP, Maes N, Ketelslegers JM, Underwood LE 1989 Insulin-like growth factor-I (IGF-I) infusion into hypophysectomized or protein-deprived rat induces specific IGF-binding proteins in serum. Endocrinology 125:2967 Walton PE, Etherton TD 1989 Effects of porcine growth hormone and insulin-like growth factor-I (IGF-I) on immunoreactive IGFbinding protein concentration in pigs. J Endocrinol 120:153 Martin JL, Baxter RC 1988 Insulin-like growth factor binding proteins (IGF-BPs) produced by human skin fibroblasts: immunological relationship to other human IGF-BPs. Endocrinology 123:1907 Martin JL, Willetts KA, Baxter RC 1990 Purification and properties of a novel insulin-like growth factor-II binding protein from transformed human fibroblasts. J Biol Chem 265:4124 Baxter RC, De Mellow JSM 1986 Measurement of insulin-like growth factor-II by radioreceptor assay using ovine placental membranes. Clin Endocrinol (Oxf) 24:26719 Baxter RC, Martin JL, Wood MH 1987 Two immunoreactive binding proteins for insulin-like growth factors in human amniotic fluid: relationship to fetal maturity. J Clin Endocrinol Metab 65:423 Baxter RC, Martin JL 1986 Radioimmunoassay of growth-hormone dependent insulin-like growth factor binding protein in human plasma. J Clin Invest 78:1504 Baxter RC, Martin JL, Beniac VA 1989 High molecular weight insulin-like growth factor binding protein complex. Purification and properties of the acid-labile subunit from human serum. J Biol Chem 264:11843 Baxter RC, Brown AS 1982 Purification of tracer for somatomedinC radioimmunoassay by hydrophobic interaction chromatography. Clin Chem 28:485 McCusker RH, Clemmons DR 1988 Insulin-like growth factor binding protein secretion by muscle cells: effect of cellular differentiation and proliferation. J Cell Physiol 137:505 Schmid C, Zapf J, Froesch ER 1989 Production of carrier proteins for insulin-like growth factors by rat osteoblastic cells. Regulation by IGF-I and cortisol. FEBS Lett 244:328 Zapf J, Guler HP, Schmid C, Froesch ER, Binoux M, Hossenlopp P, Born W, Fischer JA, Kiefer M, Masiarz F, Merryweather J, NTerminal amino acid sequence and cloning of a novel human insulin-like growth factor (IGF) carrier protein (CP) induced by IGF-I and suppressed by growth hormone (GH). 71st Annual Meeting of The Endocrine Society, Seattle WA, 1989, p 274 (Abstract) Wood WI, Cachianes G, Henzel WJ, Winslow GA, Spencer SA, Hellmiss R, Martin JL, Baxter RC 1988 Cloning and expression of the growth hormone-dependent insulin-like growth factor-binding protein. Mol Endocrinol 2:1176 De Mellow JSM, Baxter RC 1988 Growth hormone-dependent insulin-like growth factor (IGF) binding protein both inhibits and potentiates IGF-I stimulated DNA synthesis in human skin fibroblasts. Biochem Biophys Res Commun 156:199 Conover CA, Rosenfeld RG, Hintz RL 1987 Insulin-like growth factor-II binding and action in human fetal fibroblasts. J Cell Physiol 133:560 Clemmons DR, Underwood LE, Van Wyk JJ 1981 Hormonal control of immunoreactive somatomedin production by cultured human fibroblasts. J Clin Invest 67:10 Adams SO, Kapadia M, Mills B, Daughaday WH 1984 Release of insulin-like growth factors and binding protein activity into serumfree medium of cultured human fibroblasts. Endocrinology 115:520 Rotwein P 1986 Two insulin-like growth factor I messenger RNAs are expressed in human liver. Proc Natl Acad Sci USA 83:77 Conover CA, Baker BK, Hintz RL 1989 Cultured human fibroblasts secrete insulin-like growth factor 1A prohormone. J Clin Endocri-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 November 2015. at 15:01 For personal use only. No other uses without permission. . All rights reserved.
788
IGF-INDUCIBLE IGFBP FROM FIBROBLASTS
nol Metab 69:25 38. Clemmons DR, Shaw DS 1986 Purification and biological properties of fibroblast somatomedin. J Biol Chem 261:10293 39. Rotwein P, Folz RJ, Gordon JI1987 Biosynthesis of human insulinlike growth factor I (IGF-I). J Biol Chem 262:11807 40. Drop SLS, Valiquette G, Guyda HJ, Corvol MT, Posner BI 1979 Partial purification and characterization of a binding protein for insulin-like activity. Acta Endocrinol (Copenh) 90:505 41. Ritvos O, Ranta T, Jalkanen J, Suikkari A-M, Voutilainen R, Bohn H, Rutanen E-M 1988 Insulin-like growth factor (IGF) binding protein from human decidua inhibits the binding and biological
Endo • 1990 Vol 127 • No 2
action of IGF-I in cultured choriocarcinoma cells. Endocrinology 122:2150 42. Knauer DJ, Smith GL 1980 Inhibition of biological activity of multiplication-stimulating activity by binding to its carrier protein. Proc Natl Acad Sci USA 77:7252 43. Blum WF, Jenne EW, Reppin F, Kietzmann K, Ranke MB, Bierich JR 1989 Insulin-like growth factor (IGF-I)-binding protein complex is a better mitogen than free IGF-I. Endocrinology 125:766 44. Elgin RG, Busby Jr WH, Clemmons DR 1987 An insulin-like growth factor (IGF) binding protein enhances the biologic response to IGF-I. Proc Natl Acad Sci USA 84:3254
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 November 2015. at 15:01 For personal use only. No other uses without permission. . All rights reserved.
