IGF Binding Proteins Proteases Research Articles

Acid-activated insulin-like growth factor binding protein protease activity of Cathepsin D in normal and malignant prostatic epithelial cells and seminal plasma

In this study, we demonstrate insulin-like growth factor binding protein (IGFBP) acid proteolysis in conditioned media (CM) from normal and malignant primary cultures of prostatic epithelial cells, prostatic cell lines, and in seminal plasma. We further demonstrate the absence of such activity in CM from prostatic stromal cells. Radio-labeled IGFBPs (1-6) were incubated with various acidified CM and seminal plasma. None of these media showed IGFBP proteolytic activity at neutral pH, but all CM from prostatic epithelial cells (PC-E) demonstrated strong IGFBP proteolysis at acidic pH. No acid-activated proteolysis was observed in the CM from stromal cell cultures. In order to ascertain the role of cathepsin D, anti-cathepsin antibodies were used to immunodeplete the media of the selected enzymes prior to incubation with IGFBPs. Depletion of cathepsin D greatly reduced the proteolytic activity of the PC-E CM. Additionally, purified cathepsin D yielded a digestion pattern identical to that produced by prostatic cell CM and seminal plasma, following acidic incubation with IGFBP-3. Remarkably, the proteolytic pattern generated by seminal plasma, when incubated with IGFBP-3 at neutral pH, corresponded to that produced by prostate-specific antigen (PSA), demonstrating the interpolation of both neutral and acid proteases from prostate cells into seminal plasma. In conclusion, prostatic epithelial cells secrete acid-specific IGFBP protease(s) related to cathepsin D. Although no significant statistical difference was observed in the degree of acid-specific proteolysis in the media from normal versus malignant primary epithelial cell cultures, physiological characteristics of the malignant state might facilitate increased cathepsin D activity. We suspect this proteolysis may play a role in prostatic cell proliferation and invasive tumor growth.

Acid‐activated insulin‐like growth factor binding protein protease activity of Cathepsin D in normal and malignant prostatic epithelial cells and seminal plasma

In this study, we demonstrate insulin-like growth factor binding protein (IGFBP) acid proteolysis in conditioned media (CM) from normal and malignant primary cultures of prostatic epithelial cells, prostatic cell lines, and in seminal plasma. We further demonstrate the absence of such activity in CM from prostatic stromal cells. Radio-labeled IGFBPs (1–6) were incubated with various acidified CM and seminal plasma. None of these media showed IGFBP proteolytic activity at neutral pH, but all CM from prostatic epithelial cells (PC-E) demonstrated strong IGFBP proteolysis at acidic pH. No acid-activated proteolysis was observed in the CM from stromal cell cultures. In order to ascertain the role of cathepsin D, anti-cathepsin antibodies were used to immunodeplete the media of the selected enzymes prior to incubation with IGFBPs. Depletion of cathepsin D greatly reduced the proteolytic activity of the PC-E CM. Additionally, purified cathepsin D yielded a digestion pattern identical to that produced by prostatic cell CM and seminal plasma, following acidic incubation with IGFBP-3. Remarkably, the proteolytic pattern generated by seminal plasma, when incubated with IGFBP-3 at neutral pH, corresponded to that produced by prostate-specific antigen (PSA), demonstrating the interpolation of both neutral and acid proteases from prostate cells into seminal plasma. In conclusion, prostatic epithelial cells secrete acid-specific IGFBP protease(s) related to cathepsin D. Although no significant statistical difference was observed in the degree of acid-specific proteolysis in the media from normal versus malignant primary epithelial cell cultures, physiologicalcharacteristics of the malignant state might facilitate increased cathepsin D activity. We suspect this proteolysis may play a role in prostatic cell proliferationand invasive tumor growth. J. Cell. Physiol. 171:196–204, 1997. © 1997 Wiley-Liss, Inc.

Regulation of insulin-like growth factor-binding protein-4 protease activity by estrogen and parathyroid hormone in SaOS-2 cells: implications for the pathogenesis of postmenopausal osteoporosis.

The cellular mechanisms involved in the accelerated bone loss occurring in association with estrogen deprivation as seen following the menopause are not fully understood. Insulin-like growth factor-I (IGF-I) is the local regulator of osteoblasts and one of its binding proteins, insulin-like growth factor-binding protein-4 (IGFBP-4), binds to IGF-I and suppresses biological activity. Previous studies have shown that the binding activity of IGFBP-4 in the conditioned medium of parathyroid hormone (PTH)-treated SaOS-2 osteoblastic-like cells is enhanced twofold and that this PTH-enhanced IGFBP-4 binding activity is abolished by 17 beta-estradiol. Levels of IGFBP-4 in the conditioned medium have been reported to be regulated not only at the level of production but also at the level of degradation which is catalyzed by a protease that specifically cleaves IGFBP-4. We have, therefore, studied the effects of 17 beta-estradiol and PTH on IGFBP-4 protease activity using SaOS-2 cells. SaOS-2 cells produce a protease that specifically cleaves IGFBP-4 into two fragments of approximately 18 and 14 kilodaltons. IGFBP-4 protease activity in the conditioned medium from PTH-treated cells was suppressed, while this PTH-induced suppression of protease activity was reversed by the addition of 17 beta-estradiol to the cultures. IGFBP-4 proteolytic activity was stimulated by IGF-I or IGF-II added exogenously and was inhibited by EDTA or protease inhibitors. IGFBP-4 proteolyzed in the conditioned medium from cells treated with PTH and 17 beta-estradiol was less effective at inhibiting IGF-I-stimulated [3H]thymidine incorporation into DNA compared with that proteolyzed in the conditioned medium from PTH-treated cells. The simplest explanation is that 17 beta-estradiol suppressed the inhibitory effect of PTH on osteoblastic activity by inhibiting the PTH-induced suppression of IGFBP-4 protease activity.

7S nerve growth factor is an insulin-like growth factor-binding protein protease

7S nerve growth factor is an insulin-like growth factor-binding protein protease

Proteolysis of insulin-like growth factor (IGF)-binding protein-3 (IGFBP-3) in 150-kilodalton IGFBP complexes by a cation-dependent protease activity in adult rat serum promotes the release of bound IGF-I.

Insulin-like growth factor I (IGF-I) circulates in plasma predominantly in a 150-kDa complex together with IGF-binding protein-3 (IGFBP-3), an approximately 40-kDa glycoprotein that binds IGF-I with high affinity, and an 85-kDa acid-labile subunit that does not bind IGFs. The 150-kDa complex serves as a potential reservoir of IGF-1 by sequestering the growth factor in the vascular compartment. Before IGF-I can reach the tissues, however, it must be released from the complex in a form that can cross the capillary-endothelial barrier. The present study demonstrates that proteolysis of IGFBP-3 in the 150-kDa complex occurs in vitro and results in the release of IGF-I. Specific IGFBP-3 protease activity in adult rat serum was demonstrated by incubating the serum at 37 C in the presence and absence of various protease inhibitors, followed by ligand blotting. Intact and truncated (30-kDa) IGFBP-3 were almost completely proteolyzed after 12 h of incubation, whereas IGFBP-2 and IGFBP-4 in the serum were unchanged even after 48-h incubation. The IGFBP-3 protease activity was inhibited by EDTA, indicating cation dependence, and by 4-(2-aminoethyl)-benzenesulfonyl fluoride, a serine protease inhibitor. Fractionation of the serum after incubation indicated that IGFBP-3 was proteolyzed while it was part of the 150-kDa complex. Proteolysis of human IGFBP-3 (hIGFBP-3) also occurred within 150-kDa complexes when ternary complexes that had been reconstituted from recombinant hIGFBP-3, rat acid-labile subunit, and IGF-I were incubated with rat serum. Release of IGF-I from ternary complexes after proteolysis of hIGFBP-3 was indicated by the observation that some of the [125I]IGF-I initially associated with reconstituted 150-kDa complexes was transferred to endogenous rat IGFBPs during the incubation. Similar proteolysis of IGFBP-3 within 150-kDa complexes in vivo would provide a mechanism for mobilizing IGF-I from the circulating reservoir in plasma as well as for the turnover of IGFBP-3.

Serum concentrations of free and total insulin-like growth factor-I, IGF binding proteins -1 and -3 and IGFBP-3 protease activity in boys with normal or precocious puberty.

Circulating IGF-I and IGF binding protein-3 (IGFBP-3) levels both increase in puberty where growth velocity is high. The amount of free IGF-I is dependent on the IGF-I level and on the concentrations of the specific IGFBPs. Furthermore, IGFBP-3 proteolysis regulates the bioavailability of IGF-I. However, the concentration of free IGF-I and possible IGFBP-3 proteolytic activity in puberty has not previously been studied. We investigated serum levels of easily dissociable IGF-I concentrations and ultrafiltrated free IGF-I levels by specific assays in 60 healthy boys and in 5 boys with precocious puberty before and during GnRH agonist treatment. In addition, total serum IGF-I, IGFBP-1 and IGFBP-3 levels as well as IGFBP-3 protease activity were determined. Free (dissociable and ultrafiltrated) IGF-I concentrations were significantly higher in pubertal boys than in prepubertal children and correlated significantly with the molar ratio between IGF-I and IGFBP-3 (r = 0.69, P < 0.0001 and r = 0.54, P = 0.0008, respectively) and inversely with IGFBP-1 (r = -0.47, P < 0.0001 and r = -0.43, P = 0.0003, respectively). Multiple regression analysis suggested that IGFBP-3 level, and not IGFBP-1, was the major determinant of the free IGF-I serum level in normal boys. Free IGF-I levels were elevated in boys with precocious puberty and decreased during GnRH treatment. IGFBP-3 proteolysis was constant throughout puberty (mean 20%). We conclude that easily dissociable and ultrafiltrated free IGF-I serum levels are increased in boys with normal and precocious puberty and suggest that the increased free IGF-I serum concentration in puberty primarily reflects changes in total concentrations of IGF-I and IGFBPs secondary to increased GH secretion, but that it is not influenced by changes in IGFBP-3 proteolysis.

Identification and characterization of insulin-like growth factors (IGFs), IGF-binding proteins (IGFBPs), and IGFBP proteases in human synovial fluid

Identification and characterization of insulin-like growth factors (IGFs), IGF-binding proteins (IGFBPs), and IGFBP proteases in human synovial fluid

Identification and characterization of insulin-like growth factors (IGFs), IGF-binding proteins (IGFBPs), and IGFBP proteases in human synovial fluid.

The insulin-like growth factors (IGFs) and IGF-binding proteins (IGFBPs) in human synovial fluid play an important role in maintaining articular cartilage metabolism. In this study we measured the concentrations of IGF-I, IGF-II, and IGFBP-3 in normal human synovial fluid by RIA and characterized the IGFBPs by Western ligand blot (WLB), Western immunoblot, and immunoprecipitation. We also extended the study and compared normal synovial fluid to synovial fluids from patients with osteoarthritis (OA) and rheumatoid arthritis (RA). The concentrations of IGF-I, IGF-II, and IGFBP-3 in normal synovial fluid were 19 +/- 3 (mean +/- SE), 194 +/- 14, and 349 +/- 65 ng/mL, respectively. In synovial fluid of patients with OA, IGF-I levels were elevated, whereas IGF-II was decreased, and the IGFBP-3 level was similar to the control value. In patients with RA, both IGF-I and IGFBP-3 were elevated, whereas IGF-II remained unchanged. WLB and immunoprecipitation of normal synovial fluid revealed IGFBP-1 (26-29 kDa), IGFBP-2 (32 kDa), IGFBP-3 (42- to 39-kDa doublet), and IGFBP-4 (24 kDa); the IGFBP-3 doublet was very faint. In RA synovial fluid, all IGFBPs were dramatically increased, whereas little change was seen in the synovial fluid of OA. Western immunoblot against IGFBP-3 revealed a prominent 30-kDa immunoreactive fragment of IGFBP-3 in synovial fluids of normal adults as well as in those of RA and OA patients. This was concurrent with detectable IGFBP-3 protease activity, which was characterized to be of the metallo- and serine protease family. Thus, in normal synovial fluid, there is a balance of circulating IGF, IGFBP, and proteases to modulate the bioactivity of IGF. In pathological states, the increased IGF-I concentrations were accompanied by an increase in IGFBP-I concentrations were accompanied by an increase in IGFBP-3 levels in synovial fluid. These findings suggest that alteration of the IGF and IGFBP axis in pathological states may be important for understanding the underlying pathophysiology of disordered articular growth and metabolism.

Sequential changes in insulin-like growth factor I (IGF-I) and IGF-binding proteins in children with end-stage liver disease before and after successful orthotopic liver transplantation.

Pediatric end-stage liver disease (ESLD) leads to poor linear growth and wasting. After orthotopic liver transplantation (OLT), catch-up growth occurs unpredictably and with a delay. The bulk of circulating insulin-like growth factor I (IGF-I) and its major circulating binding protein, IGF-binding protein-3 (IGFBP-3), is derived from the liver. We hypothesized that growth failure in ESLD, both before and after OLT, may result from abnormalities in the IGF-IGFBP axis. Serum IGF-I, IGFBP-1, and insulin were measured by RIA, and IGFBP-3 was determined by immunoradiometric assay in 26 children with ESLD (mean of 3.7 samples pre-OLT and 4.2 samples post-OLT per patient) and 30 age-matched controls. In addition, serum IGFBPs were visualized by Western ligand blotting. IGFBP-3 and IGFBP-2 were also observed by immunoblotting with specific antisera. IGFBP-3 protease activity was determined by protease gels using recombinant human IGFBP-3 label as substrate. Anthropometric measurements were performed according to standard techniques. Pre-OLT, IGF-I (32.7 +/- 4.8 micrograms/L), and IGFBP-3 (1.11 +/- 0.10 mg/L) were significantly lower than control values [IGF-I, 168.3 +/- 16.5 micrograms/L (P = 0.0001); IGFBP-3, 2.57 +/- 0.17 mg/L (P = 0.0001)]. Post-OLT, IGF-I (179.2 +/- 19.7 micrograms/L; P = NS) rose to control levels, whereas IGFBP-3 (3.49 +/- 0.14 mg/L; P = 0.002) became significantly greater than the control value. IGFBP-1 was significantly higher pre-OLT (78.9 +/- 9.6 micrograms/L; P = 0.0001) than post-OLT (45.7 +/- 6.9 micrograms/L), and both were significantly higher than control values (18.5 +/- 2.5 micrograms/L; P = 0.0001 vs. pre-OLT and P = 0.0002 vs. post-OLT). There was a trend toward higher insulin levels both pre-OLT (15.5 +/- 1.8 mU/L) and post-OLT (12.5 +/- 1.4 mU/L) compared with control values (9.7 +/- 1.1 mU/L; P = 0.06 vs. pre-OLT). IGFBP-1 was negatively correlated with serum insulin post-OLT (P = 0.008), but there was no correlation pre-OLT. Western ligand blotting confirmed the changes in IGFBP-3 pre- and post-OLT. Immunoblotting demonstrated a reduction in all mol wt forms of IGFVBP-3 pre-OLT. Protease assays demonstrated the appearance of IGFBP-3 proteolysis only at a time coincidental with the operative stress of OLT; overall, there was no difference in protease activity pre- and post-OLT. IGFBP-2 was unchanged post-OLT compared with pre-OLT, although levels were higher than control values. Mid-upper arm circumference and triceps skin fold thickness SD score 3 months post-OLT and weight SD score 1 yr post-OLT were significantly higher than those at OLT. In conclusion, IGF-I and IGFBP-3 are reduced, and IGFBP-1 and IGFBP-2 are increased in children with ESLD. After OLT, IGF-I levels return to normal, but marked abnormalities in IGFBPs remain. These changes may help to explain at least in part the growth failure seen in pediatric ESLD both before and after successful OLT.

Altered Insulin-like Growth Factor–II (IGF-II) Level and IGF-Binding Protein–3 (IGFBP-3) Protease Activity in Interstitial Fluid Taken from the Skin Lesion of Psoriasis

In the present study, we have investigated insulin-like growth factors (IGFs) and their binding proteins (IGFBPs) in serum and artificially raised blister fluid from uninvolved and involved areas of nine patients with psoriasis. Both levels of IGFs and IGFBP-3, and profiles of IGFBP in serum and fluid from the uninvolved areas of these patients were comparable to those seen in normal subjects. In fluid from the involved areas, the IGF-II but not IGF-I level was significantly elevated. Among five molecular forms of IGFBP, the density of 41.5- and 38.5-kDa forms of IGFBP-3 were apparently increased in fluid from the involved areas, shown by Western ligand blotting. Radioimmunoassay further showed that the IGFBP-3 concentration in the involved areas was significantly raised. Immunoblotting revealed that the predominant form of IGFBP-3 in fluid from the uninvolved areas was a 29-kDa proteolytically modified product. In contrast, intact doublet IGFBP-3 was the main form of IGFBP-3 in fluid from the involved areas. Fluid from the involved areas but not the matched serum concentration-dependently inhibited the degradation of 125I-labeled nonglycosylated IGFBP-3 (ngIGFBP-3) caused by fluid from the uninvolved areas, suggesting the presence of an IGFBP-3 protease inhibitor(s) in psoriatic skin lesion. These findings suggest that the alterations in IGF/IGFBP system may contribute to the pathogenesis of psoriasis.

Serum levels of insulin-like growth factor-I and insulin-like growth factor binding protein-3 in children with insulin-dependent diabetes mellitus.

Pubertal development has recently been evaluated from the standpoint of changes in insulin-like growth factor (IGF)-I and IGF-binding protein-3 (IGFBP-3) levels in healthy children. We studied puberty related changes in serum IGF-I and IGFBP-3 levels in 24 patients (11 prepubertal) with insulin dependent diabetes mellitus (IDDM) and 26 healthy subjects (14 prepubertal). Serum IGF-I and IGFBP-3 levels were assayed using immunoradiometric assays and radioimmunoassays, respectively. Serum IGF-I and IGFBP-3 levels in diabetics did not increase during puberty, as opposed to those in healthy children. Serum IGF-I and IGFBP-3 levels of diabetic patients were found to be lower than those of control subjects during puberty (p < 0.0001 and p < 0.05, respectively). Proteolysis is believed to be a general mechanism to increase IGF bioavailability in the presence of IGFBPs. Increased IGFBP-3 protease activity has been shown in sera of children with IDDM as well as a decrease in this activity in response to insulin therapy. Our data displaying low IGFBP-3 levels in diabetic children may be due to increased proteolysis, which also causes a shift in IGF-I to its lower molecular weight forms. Higher rate of clearance of the latter may be the reason for the low IGF-I levels we observed in children with IDDM. The moderate correlation between insulin dose and IGFBP-3 levels (r = 0.5, p < 0.01) may suggest insulin to be a contributing factor in the regulation of IGFBP-3 levels. We conclude that regulation of IGF-I and IGFBP-3 concentrations is disturbed in children with IDDM, in particular during adolescence.

Estrogen- but not androgen-dominant human ovarian follicular fluid contains an insulin-like growth factor binding protein-4 protease.

Estrogen-dominant follicular fluid (FFe) and granulosa-luteal cell conditioned media, in contrast to androgen-dominant FF (FFa), contain barely detectable levels of insulin-like growth factor binding protein-4 (IGFBP-4) by ligand binding techniques. The current study was designed to evaluate the possibility of an IGFBP-4 protease in FFe, which may alter the affinity of IGFBP-4 for insulin-like growth factors (IGFs), rendering IGFBP-4 undetectable by ligand binding techniques. FFe and FFa were obtained from regularly menstruating women, and FFe was also obtained during in vitro fertilization procedures. Mixing experiments were performed by using human recombinant IGFBP-4 or IGFBP-4 in nonpregnancy serum (NPS) as substrate and FF as the source of the putative protease. Incubation of NPS at 37C for 5 h in the presence of FFe resulted in the reduction of IGFBP-4 to barely detectable levels when analyzed by Western ligand blotting, with no change occurring in the levels of the other binding proteins present in NPS. In contrast, incubation of FFa with NPS under similar conditions had no effect on the levels of IGFBP-4. The disappearance of IGFBP-4 when NPS was mixed with FFe exhibited optimal pH-dependence at pH 7-9. Inhibition of the putative protease by aprotinin, ethylenediaminetetraacetic acid, and 1,10-phenanthroline supports its identification as a metalloserine protease. Western immunoblot analysis detected the presence of a proteolytic fragment of approximately 17-18 kDa after incubation of recombinant IGFBP-4 in the presence of FFe but not in the presence of FFa. Similar incubation of other recombinant human IGFBPs did not reveal their degradation, further suggesting that the protease in FFe is specific for IGFBP-4. These data demonstrate the presence of an IGFBP-4-specific metalloserine protease in FFe but not in FFa, and they suggest that proteolytic cleavage may be responsible for effectively decreasing levels of inhibitory IGFBP-4 and thus increasing bioavailability of IGF peptides within estrogen-dominant follicles. The importance of this mechanism may lie in providing the dominant follicle with more available IGFs to synergize with gonadotropins in stimulating estradiol production and in inhibiting this synergy in androgen-dominant and atretic follicles.

Properties of an insulin-like growth factor-binding protein-4 protease that is secreted by smooth muscle cells.

Smooth muscle cells (SMC) secrete insulin-like growth factor (IGF)-binding protein-4 (IGFBP-4) and an IGFBP-4 protease. The purpose of this study was to determine the characteristics of this IGFBP-4 protease and to compare its inhibitor profile to those of IGFBP-5 and IGFBP-2 proteases, which are also present in SMC-conditioned medium. Cultured SMC were exposed to serum-free medium for periods of 24-72 h, and the amount of proteolytic activity in the conditioned medium was assessed by its capacity to degrade pure IGFBP-4. Minimal activity (e.g. < 20% of IGFBP-4 degraded in 24 h at 37 C) was present in conditioned medium unless IGF-I or IGF-II was added. This resulted in more than 60% of the intact IGFBP-4 being degraded in 14 h. The activity was a calcium-dependent serine protease and was inhibited by EDTA or 3,4-dicloroisocoumarin. Calcium, but not zinc, could restore proteolytic activity. Heparin alone inhibited IGFBP-4 proteolysis by more than 60%. When heparin cofactor-II and antithrombin-III (AT-III) were added alone, they each had an effect. The combination of heparin plus AT-III was no more active than heparin alone, but the combination of heparin cofactor-II and heparin resulted in near complete inhibition. Peptides that contained the active sites of AT-III or alpha 1-antichymotrypsin were potent inhibitors of the IGFBP-4 protease. The medium also contained proteolytic activities for IGFBP-2 and IGFBP-5. Comparison of the inhibitor profiles for the IGFBP-4 and IGFBP-5 proteolytic activities revealed major differences, but the IGFBP-2 proteolytic activity was very similar to that of the IGFBP-4 protease. IGFBP-4 zymography showed a band with a molecular mass estimate of 48 kilodaltons. In contrast, when IGFBP-2 was used as the substrate, a single band at 36 kilodaltons was visualized. These data taken together with the protease inhibitor results suggest that the IGFBP-2, IGFBP-4, and IGFBP-5 proteases are members of a similar family of calcium-dependent serine proteases, but they are distinct proteases. As IGFBP-4 is a potent inhibitor of IGF action, and the activity of this protease is regulated by IGF exposure, the protease represents a novel system for regulating the actions of IGF-I in this cell type.

Properties of an insulin-like growth factor-binding protein-4 protease that is secreted by smooth muscle cells

Properties of an insulin-like growth factor-binding protein-4 protease that is secreted by smooth muscle cells

Alterations in insulin-like growth factor (IGF)-dependent IGF-binding protein-4 proteolysis in transformed osteoblastic cells.

Insulin-like growth factor (IGF)-binding protein-4 (IGFBP-4) is secreted by a variety of osteoblastic cells and appears to be an integral component of bone cell physiology. We have previously reported that normal human osteoblast-like (hOB) cells secrete IGFBP-4 as well as a novel IGFBP-4 protease, which requires IGF for functional activity. In this study we assessed the IGFBP-4/IGFBP-4 protease system in transformed osteoblastic cells by Western ligand blotting and cell-free IGFBP-4 protease assays. Simian virus-40-immortalized hOB cells (HOBIT), human osteosarcoma cells (TE-85), and rat osteosarcoma cells (UMR 106-01, ROS 17/2.8) secrete IGFBP-4. In contrast to the rapid and dramatic proteolysis in hOB medium, medium conditioned by these cells had no apparent IGFBP-4 protease activity when assayed with exogenous IGF-II in culture or under cell-free conditions. Assayed in the presence of exogenous protease. HOBIT cells, but not the osteosarcoma cell lines, appeared to produce a cycloheximide-sensitive inhibitor of the IGFBP-4 proteolytic reaction. Transient cell transformation induced by incubating human osteoblasts transfected with a temperature-sensitive mutant of simian virus-40 T-antigen at the permissive temperature or by treating hOB cells with phorbol ester tumor promoters also resulted in inhibition of IGF-dependent IGFBP-4 proteolysis. Inhibition was observed if phorbol ester was added to the cultures at the time of medium change or after the protease had been expressed and secreted. Differences in IGFBP-4 proteolysis could not be accounted for by changes in IGFBP-4 messenger RNA expression or substrate levels. These data suggest that transformation is associated with alterations in the IGFBP-4/IGFBP-4 protease system in osteoblastic cells. Normal human osteoblasts secrete an IGF-dependent IGFBP-4 protease. The induction of an inhibitor of the IGF-dependent IGFBP-4 proteolytic reaction may be associated with early transformation processes. Fully tumorigenic bone cells expressed neither IGFBP-4 protease nor protease inhibitor activity.

Endogenous cathepsin D-mediated hydrolysis of insulin-like growth factor-binding proteins in cultured human prostatic carcinoma cells.

Insulin-like growth factor I (IGF-I) seems to play an important role in prostate cell growth and its actions may be modified by IGF-binding proteins (IGFBPs) secreted by prostate epithelial cells. The IGFBP system was studied in two human prostate carcinoma cell lines, PC3 and LNCaP. Androgen receptor-negative PC3 cells secrete IGFBP-3, IGFBP-4, and IGFBP-5, as determined by immunoprecipitation of the serum-free conditioned medium with specific IGFBP antibodies. Androgen receptor-positive LNCaP cells secrete IGFBP-2 and IGFBP-3. At neutral pH, there was little or no effect of a 24-h, 37 C cell-free incubation of PC3 and LNCaP conditioned media on IGFBP. On the other hand, when media was incubated at pH 3 for 24 h, [125I]IGFBP-3 hydrolysis and the virtual elimination of endogenous IGFBP detected by Western ligand blotting were observed. This loss was not due to the acid treatment, per se, since IGFBPs remained intact if the incubation at pH 3 was carried out at 4 C. The acid-activated IGFBP protease in LNCaP and PC3 cell-conditioned media was identified as cathepsin D based on acidic pH optimum and immunoblotting. Furthermore, immunoadsorption of cathepsin D from the media attenuated the acid-activated IGFBP hydrolysis [125I]IGF-I binding to prostate cancer cells was reduced in the presence of LNCaP conditioned media that had been incubated at neutral pH for 24 h (i.e. containing intact IGFBP) but not by acid-incubated conditioned media (i.e. cathepsin D-mediated hydrolyzed IGFBP). These data indicate that prostate carcinoma cells secrete specific IGFBPs, as well as a general IGFBP protease, cathepsin D. In the proper environment, cathepsin D is capable of hydrolyzing all endogenous IGFBP and, thus, modifying IGF-I action in prostatic cells.

Cleavage Analysis of Insulin-like Growth Factor (IGF)-dependent IGF-binding Protein-4 Proteolysis and Expression of Protease-resistant IGF-binding Protein-4 Mutants

Cultured human fibroblasts and osteoblast-like cells secrete an insulin-like growth factor (IGF)-dependent protease that cleaves IGF-binding protein-4 (IGFBP-4) into two fragments of approximately 18 and 14 kDa. Edman degradation of the isolated proteins established the amino termini of the reaction products. Sequence analysis of the 14-kDa carboxyl-terminal half of IGFBP-4 suggested cleavage after methionine at position 135 of the mature protein. Four variant IGFBP-4 molecules with single amino acid substitutions around this cleavage site were constructed and expressed. Wild-type and mutant IG-FBPs-4 bound IGF-I and IGF-II with equivalent affinities and, in the intact state, were equally effective inhibitors of IGF-I action. However, the IGFBP-4 mutants were relatively resistant to IGF-dependent proteolysis. A 5-6-h incubation in human fibroblast conditioned medium in the presence of IGF-II was sufficient for near total hydrolysis of wild-type IGFBP-4, whereas the mutant IGFBPs-4 were only minimally affected at this time. After a 24-h incubation with IGF-II, all mutant IGFBPs-4 showed extensive proteolysis, generating 18- and 14-kDa fragments. Pre-exposure of human fibroblasts in serum-free conditioned medium to IGF-II for 5 h potentiated subsequent IGF-I stimulation of DNA synthesis. When added with IGF-II, the protease-resistant mutant IG-FBPs-4, but not wild-type IGFBP-4, suppressed IGF-II enhancement of IGF-I-stimulated DNA synthesis. These biological studies suggest that the IGFBP-4/IGFBP-4 protease system may play a role modulating local cellular response to IGF-I.

Insulin-like growth factor binding protein (IGFBP) proteases: Functional regulators of cell growth

The IGFBP proteases were first described in pregnancy serum as a proteolytic activity against IGFBP-3. Since then, IGFBP proteases have been described in many other clinical situations, in various body fluids, and have been shown to cleave IGFBP-2 to -6 with varying specificity. The molecular nature of some of these proteases is being unraveled and three classes of IGFBP proteases have been recognized. These include kallikreins, cathepsins and matrix metalloproteinases (MMPs). We utilized two cellular systems to demonstrate the significance of IGFBP proteases in cellular growth regulation. In primary cultures of prostatic cells, we have shown that prostate-specific antigen (PSA) has the ability to enhance IGF mitogenic action by reducing the effects of IGFBPs. Similar kallikreins such as gamma nerve growth factor (NGF) share this activity. Within the prostatic milieu, we have also demonstrated epithelial production of the acid-activated IGFBP protease, cathepsin D, and its secretion into seminal plasma, as well as the serum of patients with prostate malignancy. We have also identified MMPs in prostatic cells and fluids. Using cultured airway smooth muscle (ASM) cells, we have demonstrated the synergism between IGFs and inflammatory agents in mediating ASM cell proliferation. Examination of this phenomenon revealed that these agents (e.g. leukotriene D 4 and interleukin1-β) induce the secretion of an IGFBP protease which cleaves the IGFBPs secreted by ASM cells, allowing IGFs to stimulate proliferation. Using several methods, including immunoblotting and immunodepletion techniques, we have identified this protease as MMP-1. These two pathophysiological systems demonstrate the importance of IGFBP proteases as autocrine paracrine growth regulators. Furthermore, IGFBP proteases may be critical elements in malignant and benign proliferative diseases, including prostate cancer and the ASM hyperplasia of long-standing asthma.

Serum insulin-like growth factor-binding protein-3 (IGFBP-3) levels and IGFBP-3 protease activity in normal, abnormal, and multiple human pregnancy

Serum insulin-like growth factor-binding protein-3 (IGFBP-3) levels and IGFBP-3 protease activity in normal, abnormal, and multiple human pregnancy

Serum insulin-like growth factor-binding protein-3 (IGFBP-3) levels and IGFBP-3 protease activity in normal, abnormal, and multiple human pregnancy.

Proteolytic activity directed against insulin-like growth factor (IGF)-binding protein-3 (IGFBP-3) is found in human pregnancy serum. We have investigated changes in this protease activity in pregnancies in which the fetus is small for gestational age and in multiple pregnancies. Maternal serum was obtained from 18 singleton pregnancies at 27 weeks gestational age (GA), and matched fetal serum was collected by cordocentesis. Fetuses were appropriate for GA (AGA; n = 6), small for gestational age (SGA) with evidence of uteroplacental insufficiency (UPI; starved SGA; n = 6), or SGA without UPI (nonstarved SGA; n = 6). In a second study, serum was obtained from women with singleton (n = 10), twin (n = 10), and higher multiple pregnancies (n = 10) at 9 weeks GA. All women with more than three fetuses underwent embryo reduction to 2 fetuses before 15 weeks gestation, when a second serum sample was obtained. Circulating IGF-I and IGF-II were measured by RIA, and IGFBP-3 was measured by both RIA and immunoradiometric assay. IGFBP-3 protease activity was assessed by Western ligand blotting after incubation with a normal nonpregnancy sera pool, immunoblotting, and specific protease assay. In the growth study, circulating maternal IGF-I and IGFBP-3 levels were not different in the three groups, but fetal IGF-I and IGFBP-3 levels were significantly lower in the UPI fetuses (IGF-I, 6.9 +/- 0.5 micrograms/L; IGFBP-3, 547 +/- 70 micrograms/L) than in either the nonstarved SGA fetuses (IGF-I, 27.8 +/- 6.3; IGFBP-3, 769 +/- 41 micrograms/L; P < 0.01) or the AGA fetuses (IGF-I, 39.4 +/- 3.4; IGFBP-3, 872 +/- 91 micrograms/L; P < 0.01). Maternal serum IGFBP-3 protease activity, measured by protease using [125I]IGFBP-3 as substrate, was increased in pregnancies complicated by UPI compared with GA-matched pregnancies in which the fetus was AGA or nonstarved SGA. No significant fetal serum protease activity was demonstrated. In the multiple pregnancies, IGFBP-3 rose significantly from 9-15 weeks GA in singleton (P = 0.005), twin (P = 0.004), and multiple (P = 0.007) pregnancies, and levels were higher in mothers of multiple pregnancies than in those of twin (P < 0.05) or singleton (P < 0.01) pregnancies at both 9 and 15 weeks GA. IGF-I levels were not different in the three groups and did not significantly increase between 9-15 weeks GA.(ABSTRACT TRUNCATED AT 400 WORDS)