253
Binding characteristics of pro-insulin-like growth factor-II from cancer patients: binary and ternary complex formation with IGF binding proteins-1 to -6 J J Bond1, S Meka and R C Baxter1 Kolling Institute of Medical Research, University of Sydney, Royal North Shore Hospital, St Leonards, New South Wales 2065, Australia 1
Cooperative Research Centre for Diagnostic Technologies, Brisbane, Queensland 4001, Australia
(Requests for offprints should be addressed to R C Baxter, Kolling Institute of Medical Research, Royal North Shore Hospital, St Leonards, New South Wales 2065, Australia; Email:
[email protected])
Abstract Many tumours secrete IGF-II in incompletely processed precursor forms. The ability of these pro-IGF-II forms to complex with the six IGF binding proteins (IGFBPs) is poorly understood. In this study, pro-IGF-II has been extracted from the serum and tumour tissue of two patients with non-islet cell tumour hypoglycaemia. These samples were used to study binary complex formation with IGFBPs-1 to -6 using competitive IGF-II binding assays and ternary complex formation with IGFBP-3 and IGFBP-5. In each case, IGFBPs-1 to -6 showed little difference in their ability to form binary complexes with recombinant IGF-II or tumour-derived pro-IGF-II forms, when the preparations were standardised according to IGF-II
Introduction Mature insulin-like growth factor (IGF)-II is a 7·5 kDa peptide secreted by cells into the blood after enzymatic cleavage from a larger precursor pro-IGF-II. It is produced throughout the body tissues as shown by mRNA distribution and immunohistochemistry (Sussenbach et al. 1992). Pro-IGF-II consists of the mature growth factor of 67 residues with a carboxy-terminal extension of 88 amino acids known as the E-peptide. Various forms of this peptide have been reported as a result of incomplete cleavage of the carboxy-terminal end. Zumstein et al. (1985) found a 10 kDa molecule from human serum that contained a 21 residue E-domain extension. Gowan et al. (1987) purified a 15 kDa IGF-II molecule from human serum. An increased proportion of these higher molecular weight IGF-II forms has been reported in plasma and tissue extracts from patients with non-islet cell tumour hypoglycaemia (NICTH), a syndrome resulting from IGF-II overproduction by mesenchymal tumours of nonpancreatic origin (Daughaday et al. 1988, Daughaday & Kapadia 1989). More recent reports have shown that
immunoreactivity. As previously described, ternary complex formation by acid-labile subunit (ALS) with IGFBP-3 and pro-IGF-II was greatly decreased compared with complex formation with mature IGF-II. In contrast, ALS bound similarly to IGFBP-5 in the presence of pro-IGF-II and mature IGF-II. These studies suggest that pro-IGF-II preferentially forms binary complexes with IGFBPs, and ternary complexes with IGFBP-5, rather than ternary complexes with IGFBP-3 as seen predominantly in normal serum. This may increase the tissue availability of serum pro-IGF-II, allowing its insulin-like potential to be realised. Journal of Endocrinology (2000) 165, 253–260
IGF-II overexpression and pro-IGF-II forms are more widespread than first thought in various cancers (Horiuchi et al. 1994, Singer et al. 1995, Hodzic et al. 1997, Van der Ven et al. 1997). IGF-II circulates in the blood and competes with IGF-I for high affinity interactions with six IGF-binding proteins (IGFBPs-1 to -6). The IGFBPs have core molecular masses in the range of 23 000–31 000 daltons, although this may be increased by glycosylation, particularly in the case of IGFBP-3, with glycoforms of 40–45 kDa. In addition, the binary complexes which IGF-I and IGF-II form with IGFBP-3 and IGFBP-5, may also form a ternary complex with the glycoprotein called acid-labile subunit (ALS) (Baxter et al. 1989, Twigg & Baxter 1998). These complexes are involved in maintaining the homeostasis of glucose metabolism and availability of IGFs to perform other important cellular functions in the body (Baxter 1991, Jones & Clemmons 1995). Little is known about how the E-domain of pro-IGF-II affects its affinity for IGFBPs and the ability to form ternary complexes. IGFBPs that do not form ternary complexes with ALS have the potential to carry IGF-II from the circulation to
Journal of Endocrinology (2000) 165, 253–260 0022–0795/00/0165–253 2000 Society for Endocrinology Printed in Great Britain
Online version via http://www.endocrinology.org
254
J J BOND
and others ·
IGFBP binding of pro-IGF-II
target tissues. Levels of IGFBP-2 and IGFBP-6 are markedly elevated in the serum of NICTH patients (Baxter 1996). In the serum of these patients, the occurrence of the 150 kDa ternary complex is reported to be reduced (Daughaday & Kapadia 1989) and 40–50 kDa IGF–IGFBP complexes are increased. It has been suggested that forms of pro-IGF-II disrupt the affinity of the ternary IGFBP-3 complexes, leading to hypoglycaemia (Baxter et al. 1995). Knowledge of the ability of pro-IGF-II to form binary and ternary complexes in serum of these patients is important in understanding the significance of elevated secretion of pro-IGF-II by certain cancers. This report describes the characterisation of the forms of pro-IGF-II in NICTH tumour and serum, and the use of these preparations of pro-IGF-II to test their binding characteristics to the six IGFBPs and ALS.
Materials and Methods Peptides Recombinant human (rh) IGF-II was a gift from Kabi Peptide Hormones, Stockholm, Sweden. IGFBP-1 was purified from human amniotic fluid (Baxter et al. 1987); rhIGFBP-2 was a gift from Sandoz (Basel, Switzerland); IGFBP-3 was purified from human plasma (Martin & Baxter 1986); rhIGFBP-4 was purchased from Austral Biologicals, San Ramon, CA, USA. rhIGFBP-5 was expressed in our laboratory in 911 human embryonic retina cells by Dr S Firth, using an adenoviral vector, and purified essentially as described for IGFBP-3 (Firth et al. 1999). Its amino-terminus was confirmed by sequencing as that of authentic hIGFBP-5 and its concentration was determined by quantitative amino acid analysis (unpublished data). IGFBP-6 was purified from human fibroblastconditioned medium (Martin et al. 1990). ALS was purified from human serum as previously described (Baxter et al. 1989). IGF-II was iodinated with chloramine T and purified as previously reported (Baxter 1990). IGF-II tracer was cross-linked to IGFBP-5 and purified essentially as described for the cross-linking of IGF-I tracer to IGFBP-3 (Baxter & Martin 1986). In brief, the iodo-IGF-II in 50 mM Na phosphate, 0·05% bovine albumin, pH 7·0, was incubated with 5 µg rhIGFBP-5 for 16 h at 2 C, then cross-linked for 30 min at 2 C with 0·25 mM final concentration of disuccinimidyl suberate (Pierce, Rockford, IL, USA). The reaction was stopped with an excess of Tris–HCl, pH 7·8, and the cross-linked IGFBP-5–IGF-II was separated from unreacted IGF-II at 22 C on a column of Sephadex G-100 (Pharmacia, Uppsala, Sweden) equilibrated and eluted in 0·5 M acetic acid, 0·1 M NaCl, 0·25% bovine albumin, pH 3·0. Journal of Endocrinology (2000) 165, 253–260
Sources of pro-IGF-II Serum from the same female patient with NICTH described by Baxter et al. (1995) was collected postoperatively. Pre-operative serum samples of an 87-yearold man with type II diabetes and a recurring NICTH tumour of the abdominal cavity were also collected. Subsequently, tumour samples from the male patient were collected in sterile Dulbecco’s modified Eagle’s medium (DMEM; Sigma, St Louis, MO, USA) during surgery for the removal of the tumour. Chromatographic fractionation of IGF-II Tumour samples were solubilised in 6 M guanidine hydrochloric acid (Progen, Darra, Australia) and 1 tablet of complete enzyme inhibitor (Boehringer Mannheim, Mannheim, Germany) per 50 ml buffer for 16 h at 4 C. The solubilised tissue was poured into 50 ml tubes (Falcon, Becton-Dickinson Labware, Lincoln Park, NJ, USA) and centrifuged at 4200 r.p.m. at 4 C for 20 min. The supernatant was dialysed overnight in Spectra/Por 3 membrane (MWCO: 3500; Spectrum Medical Industries, Houston, TX, USA) against 5 changes of 0·5 mol/l acetic acid and 75 mmol/l NaCl at room temperature. The diluent was then centrifuged as above and the supernatant fractionated on a 5100 cm column of Sephadex G-100 fine (Pharmacia Biotech, Uppsala, Sweden) and eluted at 2 ml/min in 0·5 mol/l acetic acid and 75 mmol/l NaCl, collecting 10 ml fractions. IGF-II was measured in column fractions using a conventional IGF-II RIA (Baxter 1990). The peak fractions containing IGF-II reactive material were pooled and concentrated against phosphate-buffered saline pH 6·5 in an ultrafiltration chamber (Amicon 3000 molecular mass cut-off filter, Amicon, Beverly, MA, USA). To remove further endogenous binding proteins and mature IGF-II, the G-100 column fractions and also the serum samples (1 ml) were fractionated on a 1·6100 cm column of Bio-Gel P-60 fine (Bio-Rad, Richmond, CA, USA), 0·5 M acetic acid and 75 mM NaCl as described by Baxter et al. (1995). Concentrations were determined by IGF-II RIA, and pro-IGF-II forms from the peaks off the column were analysed by SDS-PAGE and Western blotting. Western ligand blot To detect the pro-IGF-II forms, tissue extracts, serum samples and column fractions were prepared for SDSPAGE and detection on Western immunoblots. The acidified protein samples were freeze-dried in Corex tubes (Corning, NY, USA) and resuspended in 0·0625 M Tris– HCl buffer (pH 6·8) containing 4% sodium dodecylsulphate (SDS) and 2·5% 2-mercaptoethanol. The samples were heated in a boiling water bath for 3–5 min, centrifuged, and the supernatant subjected to SDS-PAGE. The www.endocrinology.org
IGFBP binding of pro-IGF-II ·
proteins were separated on a 16% acrylamide separating gel and overlaid by a 4% stacking gel for SDS-PAGE of the proteins, using a PROTEAN II electrophoresis unit (BioRad). Alternatively, column fractions were analysed on the Phastsystem (Pharmacia) by 20% homogeneous SDSPAGE. Proteins were visualised by subsequent staining with silver stain. For immunoblots, the proteins were transferred electrophoretically from SDS gels to polyvinyldifluoride (PVDF) membrane (BioRad) and incubated with a rabbit antiserum raised to natural human IGF-II (Baxter 1990). Control incubations were carried out with preimmune serum. Antibody binding was visualised by an incubation with a secondary antibody, biotinylated donkey anti-rabbit IgG (Amersham, Amersham, Bucks, UK) followed by streptavidin-horseradish peroxidase conjugate (Boehringer Mannheim).
J J BOND
and others
X-100 pH 6·0. Incubation for binary complex formation consisted of IGFBP-5 (5 ng/tube, approx. 175 pmol), increasing concentrations of recombinant IGF-II or proIGF-II sample, and 125 I-labelled IGF-II (10 000 c.p.m.) in a total volume of 250 µl. After 2 h incubation at 22 C, 250 µl 0·25% gamma globulin in assay buffer and 500 µl 25% polyethylene glycol (PEG) in ice-cold water were added for 15 min. The radioactive pellet was precipitated by centrifuging at 2200 r.p.m. for 15 min at 4 C. Supernatants were decanted and the pellet was washed with 1 ml 6·25% polyethylene glycol in ice-cold water, and centrifuged in a J6 centrifuge for 10 min at 4200 r.p.m. at 4 C. Supernatants were decanted and radioactivity in the pellet was measured by gamma counting. Ternary complex formation assays For IGFBP-3 ternary complex formation with IGF-II, I-labelled ALS tracer (10 000 c.p.m.) was incubated with 5 ng IGFBP-3 and increasing amounts of IGF-II or proIGF-II (0·01–10 ng/tube), in a total of 300 µl of the phosphate buffer as used in the binary complex formation assay. The conditions used for precipitation of the ternary complex were exactly the same as for the immunoprecipitation of binary IGFBP-3–IGF-II described in the section above. Since a precipitating IGFBP-5 antiserum was not available, the ability of IGFBP-5 to form ternary complexes was analysed using a competition assay with a constant amount of ALS (25 ng) per tube, crosslinked 125 I-labelled IGF-II–IGFBP-5 (10 000 c.p.m./tube), unlabelled IGFBP-5 (50 ng/tube) and an increasing amount of unlabelled IGF-II or pro-IGF-II (0·5–50 ng/tube). Controls without IGF-II were included in each assay. Incubations were in a total volume of 300 µl buffer containing 50 mM sodium phosphate, 0·25% bovine serum albumin, pH 6·5. A rabbit anti-human ALS antiserum, AL2/2, affinity purified on a column of Protein A-Sepharose (Pharmacia) was added at 25 µl of a 1:50 dilution to precipitate the complex. This antibody concentration had previously been shown to optimally bind the IGFBP-5 ternary complexes formed. After 1 h incubation at 22 C, precipitating antibody (goat anti-rabbit immunoglobulin) was added as 25 µl of a 1:10 dilution. After 30 min, 1·0 ml cold polyethylene glycol 6000 in 0·15 M NaCl was added, and after 15 min, each tube was centrifuged at 4000 r.p.m. at 4 C for 20 min, the supernatant decanted, and the pellet containing the precipitated complex counted in a gamma counter. 125
Competitive binding curves for binary complexes Competitive binding assays were used to measure the formation of binary complexes between IGFBPs and IGF-II. Except for IGFBP-5, assays were performed in a final volume of 0·3 ml buffer containing 0·1 M sodium phosphate, 0·02% sodium azide and 0·25% bovine serum albumin pH 6·5. Incubations consisted of IGFBP (IGFBP-1, -2, -4, -6: 1 ng, approx. 35 fmol; IGFBP-3: 2·5 ng, approx. 60 fmol), increasing concentrations of recombinant IGF-II or extracted NICTH pro-IGF-II, and 125I-labelled IGF-II (10 000 c.p.m.). For assays involving IGFBPs-1, -2 and -3, after 2 h at 22 C, 0·5 µl antiIGFBP-1, -2 or -3 antiserum was added in 25 µl buffer to the appropriate binary complex. After 1 h at 22 C, 2·5 µl goat anti-rabbit immunoglobulin in 25 µl buffer were added. After another hour at 22 C, bound radioactivity was precipitated by adding 1 ml cold 60 g/l polyethylene glycol 6000 in 0·15 M NaCl, waiting 15 min, then centrifuging for 20 min at 4200 r.p.m. at 2 C in a J6 centrifuge (Beckman, Palo Alto, CA, USA). Supernatants were decanted and radioactivity in the pellet was measured by gamma counting. It has previously been demonstrated for IGFBP-3 binding studies that the immunoprecipitation method of separating bound from free ligand gives identical results to the use of charcoal, or of concanavalin A (Martin & Baxter 1986). We have previously reported the use of immunoprecipitation for both IGFBP-1 (Baxter et al. 1987) and IGFBP-2 (Ho & Baxter 1997) binding assays, but have not compared it to alternative methods. For assays with IGFBP-4 and IGFBP-6, after 2 h incubation at room temperature, 1 ml 0·5% activated charcoal separation buffer was added to all incubations at 4 C, which were then centrifuged for 10 min at 4000 r.p.m. at 2 C. Half of the supernatant was then pipetted into a fresh tube and gamma counted for 2 min. IGFBP-5 assays were performed in 44 mM sodium phosphate, 100 mM HEPES, 0·1% BSA and 0·1% Triton www.endocrinology.org
Results Characterisation of chromatographic fractionation of IGF-II by radioimmunoassay and Western blots using polyclonal IGF-II antibodies Serum from the female patient with NICTH was fractionated on a column of Bio-Gel P-60 to separate Journal of Endocrinology (2000) 165, 253–260
255
256
J J BOND
and others
· IGFBP binding of pro-IGF-II
Figure 1 (a) IGF-II immunoreactivity in a 1 ml serum sample from a patient with an IGF-II-secreting tumour, fractionated by chromatography on a column of Bio-Gel P-60. The column, 1·6100 cm, was equilibrated and eluted at 0·2 ml/min in 0·5 mol/l acetic acid and 75 mmol/l NaCl. The profile shows three main immunoreactive peaks (arrows). (b) Immunoblot of NICTH patient serum and tissue (composite of three gels). Samples were separated by 16·5% SDS-PAGE and detected by immunoblot with IGF-II antiserum. Molecular mass standards (kDa) are indicated: alcohol dehydrogenase (45 kDa), carbonic anhydrase (34 kDa), myoglobin (23 kDa), lysozyme (16 kDa), and aprotinin (7 kDa). Lane 1: normal serum IGF-II (7·5 kDa) from panel a, peak 3 of a NICTH cancer patient. Lane 2: serum pro-IGF-II (10 and 20 kDa) from panel a, peak 2 of a NICTH cancer patient. Lane 3: recombinant IGF-II (200 ng 7 kDa ). Lane 4: NICTH tumour tissue solubilised in SDS sample buffer (10–15 kDa).
pro-IGF-II from IGF-II. The resulting profile, with three major IGF-II immunoreactive peaks, is shown in Fig. 1a. The first peak corresponds to IGFBPs (>30 kDa) which bind the radiolabelled IGF-II in the RIA. Fractions in peaks 2 and 3 were pooled and subjected to SDS-PAGE, then immunoblotted using polyclonal antibodies to IGF-II to detect the various forms of IGF-II. Peak 2 contained two immunoreactive bands of 10 kDa and 20 kDa (Fig. 1b, lane 2) and peak 3, mature IGF-II (7 kDa, Fig. 1b, lane 1). Recombinant IGF-II (200 ng) was used as a control and was detected at 7 kDa (Fig. 1b, lane 3). In contrast, using the polyclonal IGF-II antibody we detected several forms of pro-IGF-II from NICTH tumour tissue different from those in serum. A NICTH tissue extract prepared with SDS sample buffer contained a triplet of 13–15 kDa and a single band of 11 kDa (Fig. 1b, lane 4). In addition, NICTH tissue extracts were fractionated first on Sephadex G-100 and then on Bio-Gel P-60, yielding two immunoreactive peaks (Fig. 2a). In peak 2, we detected a faint band at 14 kDa and two bands at 12 and 10·5 kDa (Fig. 2b) by IGF-II immunoblot. The amount of pro-IGF-II in these preparations was quantified by IGF-II RIA. Journal of Endocrinology (2000) 165, 253–260
Binary complex formation between IGF-II and IGFBPs-1 to -6 Competitive binding assays were used to determine the ability of IGFBPs to form binary complexes with IGF-II. Figure 3 shows the binding curves of IGFBPs-1 to -6 of NICTH patient serum and tumour tissue extract of pro-IGF-II. Measured using an immunoprecipitation assay, radiolabelled IGF-II was displaced from IGFBP-1 slightly more efficiently by pro-IGF-II than by recombinant IGF-II in triplicate experiments (Fig. 3a), indicating a relative potency increase of up to twofold for pro-IGF-II compared with IGF-II. IGFBP-2 and IGFBP-3 interacted with recombinant IGF-II and pro-IGF-II with equal potency (Fig. 3b and 3c). Measured using a PEG precipitation assay, the ability of IGFBP-5 to form a binary complex with pro-IGF-II was the same compared with recombinant IGF-II (Fig. 3e). A competitive charcoal binding assay was used to test the binding affinity of IGFBP-4 and IGFBP-6 with recombinant IGF-II and pro-IGF-II. No difference was observed between the binary complex formation of proIGF-II and IGFBP-4 and IGFBP-6 compared with that determined using recombinant IGF-II (Fig. 3d and 3f). www.endocrinology.org
IGFBP binding of pro-IGF-II ·
Figure 2 (a) IGF-II immunoreactivity in IGF-II-secreting tumour tissue extracted with 8 M guanidine HCl. The extract was fractionated by chromatography on a column of Sephadex G100 and then Bio-Gel P-60. The same column conditions were used as described in Fig. 1a. The profile shows two main peaks of apparent IGF-II immunoreactivity (arrows): (1) interference by IGF-binding proteins in the IGF-II RIA, and (2) pro-IGF-II forms. (b) Immunoblot of tumour tissue extract from panel a, peak 2 showing isolation of large forms of IGF-II (10–14 kDa) and no normal IGF-II. Samples were electrophoresed on a 16·5% SDS-PAGE gel and blotted onto PVDF membrane. After blocking with 3% BSA, polyclonal antibodies to IGF-II were used to label immunoreactive bands which were visualised by horseradish peroxidase. Molecular mass standards (kDa) are indicated and the same markers were used as shown in Fig. 1b with the addition of the top standard of glutamic dehydrogenase (55 kDa) and the bottom standard of insulin B chain (4 kDa).
J J BOND
and others
Figure 3 Competitive binding assay of IGFBPs-1 to -6 in which radioiodinated IGF-II binding to the binding proteins was competed by recombinant IGF-II (M) or pro-IGF-II from NICTH serum () or tissue extract (). (a, b and c) Binding to IGFBP-1, -2 and -3 respectively, in which protein-bound tracer was separated from free by immunoprecipitation. (d and f) Binding to IGFBP-4 and IGFBP-6 respectively, in which protein-bound tracer was separated from free by charcoal adsorption of free tracer. (e) Binding to IGFBP-5, in which protein-bound tracer was separated from free by PEG precipitation. Specific binding is shown; non-specific binding was maximally 5%. One representative experiment of three independent studies, each with similar results, is shown in each panel.
Discussion
Ternary complex formation between ALS, IGFBP-3 and IGF-II
In this study, various forms of pro-IGF-II extracted from the serum and tumour of patients with NICTH have been described. We have also, for the first time, defined the binding characteristics of ALS and all six IGFBPs with pro-IGF-II as compared with IGF-II. Because the proIGF-II preparations used contain more than one species,
The ability of IGFBP-3 and pro-IGF-II in increasing concentrations to form ternary complexes was tested using radiolabelled ALS, precipitating complexes with IGFBP-3 antiserum. Figure 4a shows that, in triplicate experiments, IGFBP-3 (5 ng/tube) bound ALS tracer effectively in the presence of IGF-II, whereas pro-IGF-II from NICTH tumour extract was at least tenfold less effective than normal IGF-II in enabling ALS to bind to IGFBP-3. Since no precipitating IGFBP-5 antiserum was available, a different assay format was used to determine IGFBP-5 ternary complexes, involving their precipitation with ALS antiserum. Binding of covalent IGF-II– IGFBP-5 tracer to a fixed amount of ALS was competed by non-covalent binary complexes containing unlabelled IGFBP-5 and increasing concentrations of IGF-II or pro-IGF-II. Figure 4b shows that, in duplicate experiments, pro-IGF-II was, if anything, slightly more effective than recombinant IGF-II in displacing the crosslinked IGFBP-5–IGF-II tracer from ALS, demonstrating that, in contrast to IGFBP-3 complexes, ternary complexes with IGFBP-5 formed normally with pro-IGF-II.
Figure 4 (a) Ternary complex formation in which ALS tracer is bound by IGFBP-3 (5 ng) in the presence of increasing concentrations of recombinant IGF-II () or pro-IGF-II extracted from an IGF-II-secreting tumour (), as shown by the immunoblot in Fig. 2b. Complexes were precipitated with IGFBP-3 antiserum. (b) Ternary complex formation in which rhIGFBP-5–IGF-II crosslinked tracer is displaced from ALS by a fixed concentration of rhIGFBP-5 and increasing concentrations of recombinant IGF-II () or pro-IGF-II extracted from an IGF-II-secreting tumour (), as shown by the immunoblot in Fig. 2b. Complexes were precipitated with ALS antiserum.
www.endocrinology.org
Journal of Endocrinology (2000) 165, 253–260
257
258
J J BOND
and others ·
IGFBP binding of pro-IGF-II
our results can only indicate the ‘average’ activity of the forms present. As described previously (Liu et al. 1993), the pro-IGF-II bands detected by immunoreactivity from a NICTH patient were larger than normal IGF-II. It is not known to what extent the difference in size of the pro-IGF-II forms is related to the sequence length of the E-peptide or the degree of glycosylation of the various forms. The E-domain is known to be subject to O-linked glycosylation (Hudgins et al. 1992, Daughaday et al. 1993, Duguay et al. 1998, Hizuka et al. 1998). In this study we found bands of 13–15 kDa and single bands of 11 by SDS solubilisation of tumour tissue. In comparison, a faint band at 14 kDa and two bands at 10·5 and 12 kDa were detected in the same tumour tissue after size-fractionation on a P-60 gel permeation column. These forms of proIGF-II extracted from NICTH tumour tissue are similar to previous findings of IGF-II bands with molecular masses of 10 or 15 kDa in serum of NICTH patients (Gowan et al. 1987, Hudgins et al. 1992, Liu et al. 1993). Zumstein et al. (1985) showed a 10 kDa peptide from plasma protein fractions contained 21 amino acids of the E-domain of pro-IGF-II. The larger forms may be O-glycosylated at position 8 or 21 of the E peptide. We used this partially purified mixture of pro-IGF-II forms from NICTH tumour tissue to define their binding characteristics with the six IGFBPs. The ability of IGFBPs to carry the excess IGF-II that occurs in the serum of NICTH patients is probably crucial in maintaining glucose homeostasis. IGFBPs that do not form ternary complexes with ALS have the potential to carry IGFs from the circulation to their target tissues. As described by Baxter (1996), levels of IGFBP-2 and IGFBP-6 are markedly elevated in the serum of NICTH patients. Little is known of the levels of IGFBP-4 and IGFBP-5 in the serum of these patients due to the limited availability of specific immunoassays for these proteins. We have shown that pro-IGF-II binds to IGFBPs-1 to -6 with similar affinity to IGF-II. Presumably, these IGFBPs could act as carriers of excess IGF-II present in NICTH serum. In particular, those IGFBPs that are upregulated, such as IGFBP-2 and IGFBP-6 may have a significant role in IGF transport in NICTH. It has previously been shown that ternary complex formation is greatly reduced in the presence of pro-IGF-II compared with normal IGF-II (Baxter et al. 1995). The present study confirms this result for IGFBP-3, and shows that the reduced ternary complex is not a result of disrupted binary complex formation, but appears to be due to reduced binding of IGFBP-3–pro-IGF-II to ALS. Recently, we reported that IGFBP-5 also forms complexes with ALS (Twigg & Baxter 1998). We have now shown that pro-IGF-II is similar to IGF-II in forming binary complexes with IGFBP-5 and, in contrast to IGFBP-3, binary complexes with IGFBP-5 interact with ALS at least as well in the presence of pro-IGF-II as with IGF-II. This implies that, in patients with high circulating pro-IGF-II Journal of Endocrinology (2000) 165, 253–260
levels, there may be a tendency for the ternary complexed pro-peptide to be bound to IGFBP-5 rather than to IGFBP-3. The actual distribution of IGF-II species between the two IGFBPs will, however, depend upon their relative concentrations. It has been postulated that the reduced formation of ternary complex is a contributing factor in the tumour-associated hypoglycaemia in NICTH patients. It has been shown previously that this is associated with reduced growth hormone (GH) secretion, leading to a decrease in serum ALS and IGFBP-3 (Daughaday et al. 1995). Our data support the evidence to date that the excessive hypoglycaemic activity of NICTH tumour IGF-II corresponds with the inability of pro-IGF-II to form predominantly the larger 150 kDa ternary complexes with IGFBP-3. Since GH secretion is suppressed, it is possible that levels of IGFBP-5 in the serum are also depressed, since serum levels of this protein are GH dependent (Ono et al. 1996). Therefore, the importance of our findings that IGFBP-5 can preferentially form ternary complexes in NICTH is difficult to quantitate. Since all six IGFBPs form binary complexes well with proIGF-II, and the concentrations of several of these proteins are elevated in NICTH, a shift to binary complexes is likely to be the major contributor to increased IGF bioavailability. Similar basic carboxyl-terminal domains of IGFBP-3 (Firth et al. 1998) and IGFBP-5 (Twigg et al. 1998) are believed to be the preferential binding site for ALS. Pro-IGF-II does not prevent binary complex formation when compared with IGF-II, but the carboxy-terminal extension of the pro-IGF-II molecule is presumably responsible for the reduced ability to form ternary complex with IGFBP-3 and ALS. It is difficult to explain why such complexes still form normally with IGFBP-5. One possible explanation is that the heavy N-linked glycosylation of IGFBP-3, absent from IGFBP-5, may interact with O-linked carbohydrate on pro-IGF-II in a way which inhibits ALS binding. Consistent with this hypothesis, pro-IGF-II extracted from a NICTH patient serum, which was shown to lack the normal O-linked glycosylation, formed apparently normal ternary complex with IGFBP-3 (Daughaday et al. 1993). High levels of IGF-II expression, seen in several human embryonal tumours such as Wilm’s tumour, rhabdomyosarcoma and hepatoblastoma (Reeve et al. 1985, Scott et al. 1985, Gray et al. 1987), are not always associated with hypoglycaemia. Gelato and Vassalotti (1990) showed that 60% of IGF-II from 7 phaeochromocytomas was in large molecular mass forms (8·7–10 kDa). However, serum IGF-II levels suggested very little tumour-derived IGF-II is released into the circulation. The precise combination of factors leading from IGF-II overproduction to hypoglycaemia is thus still incompletely understood. Our demonstration that tumour-derived pro-IGF-II binds all IGFBPs normally, and only inhibits ternary complex formation with IGFBP-3, will allow a better understanding of the www.endocrinology.org
IGFBP binding of pro-IGF-II ·
significance of altered IGFBP levels in patients with tumours overexpressing IGF-II.
References Baxter RC 1990 Radioimmunoassay for insulin-like growth factor (IGF)-II: interference by pure IGF-binding proteins. Journal of Immunoassay 11 445–458. Baxter RC 1991 Insulin-like growth factor (IGF) binding proteins: the role of serum IGFBPs in regulating IGF availability. Acta Paediatrica (Suppl) 372 107–114. Baxter RC 1996 The role of insulin-like growth factors and their binding proteins in tumor hypoglycemia. Hormone Research 46 195–201. Baxter RC & Martin JL 1986 Radioimmunoassay of growth hormone-dependent insulin-like growth factor binding protein in human plasma. Journal of Clinical Investigation 78 1504–1512. Baxter RC, Martin JL & Wood MH 1987 Two immunoreactive binding proteins for insulin-like growth factors in human amniotic fluid: relationship to fetal maturity. Journal of Clinical Endocrinology and Metabolism 65 423–431. 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. Journal of Biological Chemistry 264 11843–11848. Baxter RC, Holman SR, Corbould A, Stranks S, Ho PJ & Braund W 1995 Regulation of the insulin-like growth factors and their binding proteins by glucocorticoid and growth hormone in non-islet cell tumor hypoglycemia. Journal of Clinical Endocrinology and Metabolism 80 2700–2708. Daughaday WH, Emanuele MA, Brooks MH, Barbato AL, Kapadia M & Rotwein P 1988 Synthesis and secretion of insulin-like growth factor II by a leiomyosarcoma with associated hypoglycemia. New England Journal of Medicine 319 1434–1440. Daughaday WH & Kapadia M 1989 Significance of abnormal serum binding of insulin-like growth factor-II in the development of hypoglycemia in patients with non-islet cell tumours. Proceedings of the National Academy of Sciences of the USA 86 6778–6782. Daughaday WH, Trivedi B & Baxter RC 1993 Serum ‘big insulin-like growth factor-II’ from patients with tumor hypoglycemia lacks normal E-domain O-linked glycosylation, a possible determinant of propeptide processing. Proceedings of the National Academy of Sciences of the USA 90 5823–5827. Daughaday WH, Trivedi B & Baxter RC 1995 Abnormal serum IGF-II transport in non-islet cell tumor hypoglycemia results from abnormalities of both IGF binding protein-3 and acid-labile subunit and leads to elevation of serum free IGF-II. Endocrine 3 425– 428. Duguay SJ, Jin Y, Stein J, Duguay AN, Gardener P & Steiner DF 1998 Post-translational processing of the insulin-like growth factor-II precursor. Journal of Biological Chemistry 273 18443– 18451. Firth SM, Ganeshprasad U & Baxter RC 1998 Structural determinants of the ligand and cell surface binding of insulin-like growth factor binding protein-3 Journal of Biological Chemistry 273 2631–2638. Firth SM, Ganeshprasad U, Poronnik P, Cook DI & Baxter RC 1999 Adenoviral-mediated expression of human insulin-like growth factor-binding protein-3. Protein Expression and Purification 16 202–211. Gelato MC & Vassalotti J 1990 Insulin-like growth factor-II: possible local growth factor in pheochromocytoma. Journal of Clinical Endocrinology and Metabolism 71 1168–1174. Gowan LK, Hampton B, Hill DJ, Schlueter RJ & Perdue JF 1987 Purification and characterisation of a unique high molecular www.endocrinology.org
J J BOND
and others
weight form of insulin-like growth factor-II. Endocrinology 121 449–458. Gray A, Tam AW, Dull TJ, Hayflick J, Pintar J, Cavenee WK, Koufos A & Ullrich A 1987 Tissue-specific and developmentally regulated transcription of the insulin-like growth factor 2 gene. DNA 6 283–295. Hizuka N, Fukuda I, Takano K, Asakawa-Yasumoto K, Okubo Y & Demura H 1998 Serum high molecular weight form of insulin-like growth factor-II from patients with non-islet cell tumor hypoglycemia is O-glycosylated. Journal of Clinical Endocrinology and Metabolism 83 2875–2877. Ho PJ & Baxter RC 1997 Characterization of truncated insulin-like growth factor-binding protein-2 in human milk. Endocrinology 138 3811–3818. Hodzic D, Delacroix L, Willemsen Ph, Bensbaho K, Collette J, Broux R, Lefebvre P, Legros JJ, Grooteclaes M & Winkler R 1997 Characterisation of the IGF system and analysis of the possible molecular mechanisms leading to IGF-II overexpression in mesothelioma. Hormone and Metabolic Research 29 549– 555. Horiuchi T, Shinohara Y, Sakamoto Y, Hizuka N, Watabe T, Mokuda O, Tanaka K, Shimizu N, Sugano I & Nagao K 1994 Expression of insulin-like growth factor-II by a gastric carcinoma associated with hypoglycemia. Virchows Archiv 424 449–452. Hudgins RW, Hampton B, Burgess WH & Perdue JF 1992 The identification of O-glycosylated precursors of insulin-like growth factor-II. Journal of Biological Chemistry 267 8153–8160. Jones JI & Clemmons DR 1995 Insulin-like growth factors and their binding proteins: biological actions. Endocrine Reviews 16 3–34. Liu F, Baker BK, Powell DR & Hintz RL 1993 Characterisation of proinsulin-like growth factor–II E-region immunoreactivity in serum and other biological fluids. Journal of Clinical Endocrinology and Metabolism 76 1095–1100. Martin JL & Baxter RC 1986 Insulin-like growth factor-binding protein from human plasma. Purification and characterization. Journal of Biological Chemistry 261 8754–8760. Martin JL, Willetts KE & Baxter RC 1990 Purification and properties of a novel insulin-like growth factor-II binding protein from transformed human fibroblasts. Journal of Biological Chemistry 265 4124–4130. Ono T, Kanzaki S, Seino Y, Baylink DJ & Mohan S 1996 Growth hormone (GH) treatment of GH-deficient children increases serum levels of insulin-like growth factors (IGFs), IGF-binding protein –3 and –5, and bone alkaline phosphatase isoenzyme. Journal of Clinical Endocrinology and Metabolism 81 2111–2116. Reeve AE, Eccles MR, Wilkins RJ, Bell GI & Millow LJ 1985 Expression of insulin-like growth factor-II transcripts in Wilm’s tumour. Nature 317 258–260. Scott J, Cowell J, Robertson ME, Priestley LM, Wadey R, Hopkins B, Pritchard J, Bell GI, Rall LB, Graham CF & Knott TJ 1985 Insulin-like growth factor-II gene expression in Wilm’s tumour and embryonic tissues. Nature 317 260–262. Singer C, Rasmussen A, Smith HS, Lippman ME, Lynch HT & Cullen KJ 1995 Malignant breast epithelium selects for insulin-like growth factor-II expression in breast stroma: evidence for paracrine function. Cancer Research 55 2448–2454. Sussenbach JS, Steenberg PH & de Pagter Holthuizen P 1992 Structure and expression of the human insulin-like growth factor genes. Growth Regulation 2 1–9. Twigg SM & Baxter RC 1998 Insulin-like growth factor (IGF)binding protein-5 forms an alternative ternary complex with IGFs and the acid labile subunit. Journal of Biological Chemistry 273 6074–6079. Twigg SM, Kiefer MC, Zapf J & Baxter RC 1998 Insulin-like growth factor binding protein-5 complexes with the acid labile Journal of Endocrinology (2000) 165, 253–260
259
260
J J BOND
and others ·
IGFBP binding of pro-IGF-II
subunit: role of the carboxyl-terminal domain. Journal of Biological Chemistry 273 28791–28798. Van der Ven LTM, Roholl PJM, Gloudemans T, Van Buul-Offers SC, Welters MJP, Bladergroen BA, Faber JAJ, Sussenbach JS & Den Otter W 1997 Expression of insulin-like growth factors (IGFs), their receptors and IGF binding protein-3 in normal, benign and malignant smooth muscle tissues. Journal of Cancer 75 1631–1640.
Journal of Endocrinology (2000) 165, 253–260
Zumstein PP, Luthi C & Humbel RE 1985 Amino acid sequence of a variant proform of insulin-like growth factor-II. Proceedings of the National Academy of Sciences of the USA 82 3169–3172.
Received 25 August 1999 Revised manuscript received 24 November 1999 Accepted 11 January 2000
www.endocrinology.org
