Molecular cloning of a new human insulin-like growth factor binding protein.

We have cloned a new insulin-like growth factor's binding protein (IGFBP) from a human osteosarcoma cDNA library. Two conserved regions in the COOH-terminal third of the five known human IGFBPs were used to design primers and to perform polymerase chain reaction (PCR) with osteosarcoma cDNA as a template. One of the eight PCR products encoded a unique IGFBP sequence. The DNA sequence was used to synthesize probes to screen an osteosarcoma cDNA library and isolate full length cDNA clones. The amino acid sequence was deduced from one of them. It contains two possible signal peptidase cleavage sites yielding a mature molecule of 257 or 252 amino acids, and 18 cysteines in identical positions to the other IGFBPs. The most pronounced homology exists with human IGFBP-3 (50% in the NH2- and 45% in the COOH-terminal region).     

Identification of insulin-like growth factor binding proteins from cultured human epidermal keratinocytes

The role and mechanisms of action of insulin-like growth factors (IGFs) in skin remain unclear. Epidermal keratinocytes possess IGF-I receptors and are responsive to IGF-I, which is primarily derived from underlying dermal fibroblasts. IGF binding proteins (IGFBPs), also synthesized by fibroblasts, may be involved in paracrine targeting of IGF-I to its receptors. We therefore examined whether human keratinocytes synthesize IGFBPs and their mRNAs. Following culture in complete medium (containing bovine pituitary extract and epidermal growth factor) Western ligand blotting (WLB) of cell conditioned medium revealed a major band of 32 kD, a less abundant IGFBP of 24 kD at all passages, and a 37–42 kD IGFBP which increased in abundance in late passage. Immunoprecipitation followed by WLB confirmed that the predominant 32 kD band was IGFBP-2. Radioimmunoassay of IGFBP-1, -3, and -6 revealed detectable levels of IGFBP-3 and significant levels of IGFBP-6, but not IGFBP-1. Northern analysis following culture in complete medium revealed that at early passage IGFBP-1, -2, -4, and -6 mRNAs were detectable. IGFBP-3 and -5 mRNAs were not detectable. Following culture in growth factor-free medium a 37–42 kD band, consistent with IGFBP-3, was predominant and a 24 kD band consistent with IGFBP-4 was also present. These data demonstrate the expression of a distinct pattern of IGFBPs by cultured human keratinocytes dependent on culture conditions. Keratinocyte-derived IGFBPs are likely to play a role in the transport and targeting of IGF-I from dermally derived fibroblasts to the epidermis. © 1995 Wiley-Liss, Inc.     

Isolation from adult human serum of four insulin-like growth factor (IGF) binding proteins and molecular cloning of one of them that is increased by IGF I administration and in extrapancreatic tumor hypoglycemia.

We have isolated four insulin-like growth factor binding proteins (IGFBPs) from adult human serum by insulin-like growth factor (IGF) I affinity chromatography and high performance liquid chromatography. A 36-kDa binding protein (BP), not digestible with N-glycanase, is increased in patients with extrapancreatic tumor hypoglycemia and during IGF I administration in healthy adults. Its 38 NH2-terminal amino acids are identical to those of an IGFBP sequence derived from a human cDNA that cross-hybridizes with the rat IGFBP-2 cDNA. With probes encoding a NH2-terminal, COOH-terminal, and a middle region of this protein we have obtained three cDNA clones from a Hep G2 cDNA library; one encodes human IGFBP-2, and the other two presumably represent unspliced heteronuclear and alternatively spliced mRNA, respectively. A 28-30-kDa IGFBP represents a novel BP species in human serum. Its 30 NH2-terminal amino acids are not homologous to IGFBP-1, -2, or -3. It is not digestible with N-glycanase and does not bind 125I-IGF I. The NH2-terminal sequences of a 42/45- and a 31-kDa IGFBP are identical to that of human IGFBP-3. The 42/45-kDa proteins are two glycosylation variants of BP-3. The 31-kDa protein presumably is a degradation product of BP-3 that lacks the COOH terminus. It is likely that the different IGFBPs modulate auto-/paracrine and endocrine effects of IGFs on growth and metabolism in a different and specific manner.     

Binding proteins for insulin-like growth factors in adult rat serum. Comparison with other human and rat binding proteins

Insulin-like growth factor (IGF) binding protein has been purified from adult rat serum by affinity chromatography on agarose-IGF-II and high performance reverse-phase chromatography. The final preparation contains two components, of apparent molecular mass 50 and 56 kDa nonreduced, or 44 and 48 kDa reduced, both of which specifically bind IGF-I and IGF-II. Competitive binding data indicate association constants of 5-10 X 10(10) l/mol for both IGFs, with a slightly higher affinity for IGF-II than IGF-I. Amino-terminal sequence analysis yields a unique sequence, identical in 11 of the first 15 amino acids with that of a human plasma IGF binding protein (Martin, J. L., and Baxter, R. C. (1986) J. Biol Chem. 261, 8754-8760), and with slight homology to other human and rat IGF binding proteins characterized to date. By analogy with the binding protein from human plasma, it is likely that the rat protein is part of the growth-hormone dependent complex which appears to carry most or all of the circulating IGFs.     

Characterization of insulin-like growth factor binding proteins from human breast cancer cells

The insulin-like growth factor binding proteins (IGF-BPs) are structurally and immunologically distinct from the IGF type 1 or type 2 receptors and are characterized by two major forms: a large, GH-dependent BP found in human plasma (Mr = 150 k) and a small GH-independent BP (Mr = 28-42 k) present in human plasma, amniotic fluid, and HEP G2 cells. Using affinity cross-linking techniques, we have identified several binding proteins secreted by human breast cancer cell lines (Hs578T, MDA-231, T-47D, and MCF-7). Under nonreducing conditions these proteins migrated at an apparent Mr = 35, 28, 27, and 24 k, while reducing conditions revealed bands of apparent Mr = 35, 32, 27, and 24 k. Competitive binding studies in T-47D-conditioned media demonstrated that these BPs bound more IGF-II than IGF-I, and that IGF-II potently inhibited binding of either IGF-I or -II. Immunological studies using a polyclonal antibody against the HEP G2 small BP revealed no immunoreactive BP in conditioned media from MCF-7 and T-47D and only slight immunoreactivity in conditioned media from Hs578T and MDA 231. Analysis by Northern blot, using a probe from the cDNA sequence of the HEP G2 BP, demonstrated that Hs578T and MDA-231 cell lines contained small amounts of the 1.65 kilobase mRNA characteristic of the HEP G2 BP, while MCF-7 and T-47D tested negative.(ABSTRACT TRUNCATED AT 250 WORDS)     

High molecular weight insulin-like growth factor binding protein complex. Purification and properties of the acid-labile subunit from human serum

In the human circulation, the insulin-like growth factors (IGFs) circulate as part of a growth hormone-dependent 125- to 150-kDa complex. This complex has been postulated to contain, in addition to IGFs and one or more IGF-binding proteins, an acid-labile subunit (ALS) which does not itself bind IGFs. In this study, the ALS has been purified 1600-fold from human serum, and its binding properties have been examined. Fresh serum was fractionated on DEAE-Sephadex, and active fractions (determined by radioimmunoassay) were purified by affinity chromatography on an IGF-agarose column saturated with the plasma IGF-binding protein BP-53. After further high performance anion exchange chromatography, an ALS preparation was obtained which contained only an 84-86-kDa protein doublet, converting to a single 70-kDa band on N-glycanase treatment, and having an amino-terminal sequence unrelated to IGF-binding proteins or receptors. Pure ALS formed a complex with BP-53 (Ka approximately 5 x 10(8) M-1), immunoprecipitable by anti-BP-53 antiserum, only in the presence of IGF-I or IGF-II. This complex appeared at approximately 150 kDa on high performance gel chromatography. Pure ALS had no intrinsic IGF-binding activity and no effect on the binding of IGF-I or IGF-II to BP-53. These studies suggest that formation of the high molecular weight IGF-binding protein complex requires ALS, BP-53, and IGF.     

Cellular actions of insulin-like growth factor binding proteins.

The insulin-like growth factors (IGFs), insulin-like growth factor binding proteins (IGFBPs), and the IGFBP proteases are involved in the regulation of somatic growth and cellular proliferation both in vivo and in vitro. IGFs are potent mitogenic agents whose actions are determined by the availability of free IGFs to interact with the IGF receptors. IGFBPs comprise a family of proteins that bind IGFs with high affinity and specificity and thereby regulate IGF-dependent actions. IGFBPs have recently emerged as IGF-independent regulators of cell growth. Various IGFBP association proteins as well as cleavage of IGFBPs by specific proteases modulate levels of free IGFs and IGFBPs. The ubiquity and complexity of the IGF axis promise exciting discoveries and applications for the future.     

Structural analogs of human insulin-like growth factor I with reduced affinity for serum binding proteins and the type 2 insulin-like growth factor receptor

Four structural analogs of human insulin-like growth factor I (hIGF-I) have been prepared by site-directed mutagenesis of a synthetic IGF-I gene and subsequent expression and purification of the mutant protein from the conditioned media of transformed yeast. [Phe-1,Val1,Asn2, Gln3,His4,Ser8, His9,Glu12,Tyr15,Leu16]IGF-I (B-chain mutant), in which the first 16 amino acids of hIGF-I were replaced with the first 17 amino acids of the B-chain of insulin, has greater than 1,000-, 100-, and 2-fold reduced potency for human serum binding proteins, the rat liver type 2 IGF receptor, and the human placental type 1 IGF receptor, respectively. The B-chain mutant also has 4-fold increased affinity for the human placental insulin receptor. [Gln3,Ala4]IGF-I has 4-fold reduced affinity for human serum binding proteins, but is equipotent to hIGF-I at the types 1 and 2 IGF and insulin receptors. [Tyr15,Leu16]IGF-I has 4-fold reduced affinity for human serum binding proteins and 10-fold increased affinity for the insulin receptor. This peptide is also equipotent to hIGF-I at the types 1 and 2 IGF receptors. The peptide in which these four-point mutations are combined, [Gln3,Ala4,Tyr15,Leu16]IGF-I, has 600-fold reduced affinity for the serum binding proteins. This peptide has 10-fold increased potency for the insulin receptor, but is equipotent to hIGF-I at the types 1 and 2 IGF receptors. All four of these mutants stimulate DNA synthesis in the rat vascular smooth muscle cell line A10 with potencies reflecting their potency at the type 1 IGF receptor. These studies identify some of the domains of hIGF-I which are responsible for maintaining high affinity binding with the serum binding protein and the type 2 IGF receptor. In addition, these peptides will be useful in defining the role of the type 2 IGF receptor and serum binding proteins in the physiological actions of hIGF-I.     

Secretion and biological actions of insulin-like growth factor binding proteins in two human tumor-derived cell lines in vitro.

The insulin-like growth factors (IGFs) I and II are present in extracellular fluids associated with specific binding proteins (IGFBPs) that can modify their biologic actions. These studies were undertaken to determine which forms of IGFBP are secreted by endometrial carcinoma (HEC-1B) and breast carcinoma (MDA-231) cells, to characterize variables that control IGFBP secretion, and to study the effect of IGFBP-1 and IGFBP-2 on IGF-I stimulated cell proliferation. Secreted IGFBPs were identified by ligand blotting and IGFBP-1 was quantified using a specific radioimmunoassay (RIA). MDA-231 cell conditioned media (CM) contained four (43,000, 39,000, 30,000 and 24,000 Mr) forms of IGFBP, and HEC-1B cell CM contained three forms (39,000, 34,000 and 30,000 Mr). Immunoblotting showed that the 30,000 Mr form secreted by both cell types was IGFBP-1. Likewise the 34,000 Mr band in HEC-1B media reacted with IGFBP-2 antiserum and the 39,000 and 43,000 Mr bands reacted with IGFBP-3 antiserum. IGF-I stimulated the secretion of IGFBP-3 from both cell types and IGFBP-2 from HEC-1B cells but either decreased or caused no change in secretion of IGFBP-1 and a 24,000 Mr form. In contrast, insulin inhibited the secretion of IGFBP-1 but increased the secretion of the 24,000 Mr form. Compounds that elevate intracellular cAMP levels increased the secretion of IGFBP-3, IGFBP-1, and the 24,000 Mr form from both MDA-231 and HEC-1B cells. When sparse cultures of MDA-231 cells were used, addition of IGF-I caused a 24% increase in cell number after 48 hr. This mitogenic response was enhanced by the presence of recombinant human IGFBP-1 (45% increase in cell number, P less than 0.001). Bovine IGFBP-2 did not potentiate IGF-I stimulated cell proliferation. These findings show that two tumor cell lines secrete distinct forms of IGFBPs and that there is differential regulation of IGFBP secretion. At least one form secreted by both tumors may act as a positive autocrine modulator of IGF-I's growth stimulating actions.     

Circulating levels of insulin-like growth factors increase and molecular forms of their serum binding proteins change with human pregnancy.

Insulin-like growth factors (IGF) and binding proteins were measured in serum from pregnant and nonpregnant women. IGF-I determined by immunoassay after acid-ethanol extraction was increased by pregnancy (p less than 0.005) and was highest in the third trimester (p less than 0.01). Size exclusion chromatography of serum in acid before assay (i) gave a very similar IGF-I pattern, (ii) showed that IGF-II was much higher than IGF-I and (iii) revealed less serum IGF-binding protein activity in pregnancy and lactation. All IGF-binding proteins except binding protein-1 were markedly reduced by pregnancy. This indicates a major change in the main carrier protein for IGFs in the circulation and suggests that tissue targetting of IGFs may be altered during pregnancy.     

Expression and release of insulin-like growth factor binding proteins in isolated epiphyseal growth plate chondrocytes from the ovine fetus

Insulin-like growth factor-II (IGF-II) is an autocrine modulator of epiphyseal chondrogenesis in the fetus. The cellular availability of IGFs are influenced by the IGF-binding proteins (IGFBPs). In this study, we investigated the control of expression and release of IGFBPs from isolated epiphyseal growth plate chondrocytes from the ovine fetus by hormones and growth factors implicated in the chondrogenic process. Chondrocytes were isolated from the proliferative zone of the fetal ovine proximal tibial growth plate and maintained in monolayer culture at early passage number. Culture media conditioned by chondrocytes under basal conditions released IGFBPs of 24, 34, and 29 kDa, and a less abundant species of 39-43 kDa that were identified immunologically as IGFBP-4, IGFBP-2, IGFBP-5, and IGFBP-3, respectively. Messenger RNAs encoding each species were identified by Northern blot analysis within chondrocytes, as was mRNA encoding IGFBP-6. Exposure to IGF-I or IGF-II (13 or 26 nM) caused an increase in expression and release of IGFBP-3. The release of IGFBP-2 and IGFBP-5 were also potentiated without changes to steady state mRNA, and for IGFBP-5 this was due in part to a release from the cell membrane in the presence of IGF-II. Insulin (16.7 or 167 nM) selectively increased mRNA and the release of IGFBP-3, while cortisol (1 or 5 microM) inhibited both mRNA and release of IGFBP-2 and IGFBP-5. Transforming growth factor-beta1 (TGF-beta1) (0.1 or 0.2 nM) increased the expression and release of IGFBP-3, and caused an increase in mRNAs encoding IGFBP-2 and IGFBP-5. Neither growth hormone (GH), fibroblast growth factor-2, nor thyroxine (T(4)) had any effect on IGFBP expression or release. The results suggest that IGFBP expression and release within the developing growth plate can be modulated by IGF-II and other trophic factors, thus controlling IGF availability and action.     

The expression of insulin-like growth factor binding proteins in human hepatocellular carcinoma

Insulin-like growth factors (IGF), IGF receptors and IGF binding proteins (IGFBPs) play an important role in cell growth and differentiation. The liver is the major source of IGF-1 and at least two IGFBPs (IGFBP-1 and IGFBP-3). IGFBPs most often serve to attenuate the effects of IGF at the receptor level and thereby limit IGF-induced cell growth and differentiation. Although changes in IGFBP expression have been described during controlled liver growth such as hepatic regeneration following partial hepatectomy, there is limited knowledge of IGFBPs gene expression in uncontrolled growth or hepatocellular carcinoma. In the present study, we employed Northern blotting techniques to document the expression of IGFBP-1, 3 and 4 in normal human livers, cirrhotic and hepatocellular carcinoma tissues. The results revealed no differences in IGFBP-1, 3 and 4 mRNA levels between normal and cirrhotic tissues. However, the expression of all three IGFBPs mRNA were significantly down regulated in hepatocellular carcinoma tissues. These findings are in keeping with IGFBPs playing an important inhibitory role in the development and/or growth of hepatocellular carcinoma in humans.     

Growth hormone-dependent insulin-like growth factor (IGF) binding protein from human plasma differs from other human IGF binding proteins

A growth hormone-dependent binding protein for insulin-like growth factors (IGF-I and IGF-II) has been isolated from human plasma. Analyzed on SDS gels, the preparation contained a major protein band of 53 kDa, and a minor band of 47 kDa. After transfer to nitrocellulose, both species bound iodinated IGF-I, and could be detected using an antibody raised against the purified preparation. In contrast, an IGF binding protein purified from human amniotic fluid bound IGF-I but was not detectable immunologically. The amino acid comparison of the plasma binding protein preparation was different from that reported for amniotic fluid and HEP G2 hepatoma proteins, and the unique amino-terminal sequence, Gly-Ala-Ser-Ser-Ala-Gly-Leu-Gly-Pro-Val-, was different from that of the amniotic fluid and hepatoma proteins. This study indicates that the growth hormone-dependent IGF binding protein of human plasma is structurally and immunologically distinct from other IGF binding proteins.     

Matrix metalloproteinase-7 degrades all insulin-like growth factor binding proteins and facilitates insulin-like growth factor bioavailability

Proteolytic modification of insulin-like growth factor binding proteins (IGFBPs) plays an important physiological role in regulating insulin-like growth factor (IGF) bioavailability. Recently, we demonstrated that matrix metalloproteinase-7 (MMP-7)/Matrilysin produced by various cancer cells catalyzes the proteolysis of IGFBP-3 in vitro and regulates IGF bioavailability, resulting in an anti-apoptotic effect against anchorage-independent culture. In the present study, we investigated whether MMP-7 contributes to proteolysis of the other five IGFBPs, IGFBP-1, IGFBP-2, IGFBP-4, IGFBP-5, and IGFBP-6, and whether this results in phosphorylation of the IGF type 1 receptor (IGF-1R). MMP-7 cleaved all six IGFBPs, resulting in IGF-mediated IGF-1R phosphorylation, which was inhibited by EDTA treatment. These results suggest that MMP-7 derived from cancer cells can regulate IGF bioavailability in the microenvironment surrounding the tumor, where various kinds of IGF/IGFBP complexes are found, thereby favoring cancer cell growth and survival during the processes of invasion and metastasis.     

Insulin-like growth factor binding proteins of equine serum.

Ligand blotting analysis of serum from the horse using radiolabelled IGF-I revealed a protein at 96 kDa which was not present in serum from goat, cow, sheep, deer or donkey. These latter species all displayed five labelled bands in the range 24 to 41 kDa. Conversely, these were only weakly labelled in serum from the horse. Size exclusion chromatography of horse serum pre-incubated with radiolabelled IGF-I revealed reduced binding in the 130-kDa peak compared with goat plasma, and ligand blotting analysis indicated the 96-kDa protein was present in this peak. The 96-kDa protein from horse serum binds IGF-I and IGF-II specifically and appears to be unique to this species. The nature of this protein is at present unknown.     

Influence of exercise duration on serum insulin-like growth factor and its binding proteins in athletes

The changes in circulating concentrations of insulin-like growth factors during exercise have to date remained incomplete in their documentation. Therefore, we examined in 25 healthy athletes the effects of three different durations of three types of exercise – incremental ergometer cycling exercise (ICE), long-distance Nordic ski race (NSR) and a treadmill-simulated soccer game (TSG) lasting 20 min, 3 h, and 2 × 45 min separated by a 15-min half-time rest respectively, on plasma concentrations of growth hormone ([GH]), insulin-like growth factor-1 ([IGF-I]) and its binding proteins 1 and 3 ([IGFBP-1], [IGFBP-3]). Compared to baseline, serum [GH] increased by 15.2-fold after ICE (P < 0.001), 2.9-fold after NSR (P < 0.01) and 4.6-fold after TSG. Serum [IGF-I] rose by 11.9% after ICE (P < 0.001), while it decreased by −14.6% after NSR (P < 0.001) and was unchanged after TSG. Serum [IGFBP-1] was slightly increased (1.7-fold) after ICE (P < 0.01), but increased markedly (11.8-fold) after NSR (P < 0.001) and by 6.3-fold after the second session of TSG (P < 0.01) (it remained unchanged at the end of the first period of TSG, i.e. after 45-min exercise). The [IGFBP-3] increased by 14.7% after ICE (P < 0.001) and by 6% after TSG (P < 0.05) while it did not change after NSR. From our results it would appear that [IGFBP-1] increase to bind free IGF and hinder their insulin-like action during long-term exercise (lasting beyond 45 min). It is suggested that IGFBP-1 might thus contribute both to preventing hypoglycaemic action of IGF and to facilitating glucose uptake by muscle cells when muscle glycogen stores become deplete. Accepted: 27 May 1998     

Characterization and regulation of insulin-like growth factor binding proteins in human hepatic stellate cells

Cultured hepatic stellate cells (HSCs), the cell type primarily involved in the progression of liver fibrosis, secrete insulin-like growth factor-I (IGF-I) and IGF binding protein (IGFBP) activity. IGF-I exerts a mitogenic effect on HSCs, thus potentially contributing to the fibrogenic process in an autocrine fashion. However, IGF-I action is modulated by the presence of specific IGFBPs that may inhibit and/or enhance its biologic effects. Therefore, we examined IGFBP-1 through IGFBP-6 mRNA and protein expression in HSCs isolated from human liver and activated in culture. Regulation of IGFBPs in response to IGF-I and other polypeptide growth factors involved in the hepatic fibrogenic process was also assessed. RNase protection assays and ligand blot analysis demonstrated that HSCs express IGFBP-2 through IGFBP-6 mRNAs and release detectable levels of IGFBP-2 through IGFBP-5. Because IGF-I, platelet-derived growth factor-BB (PDGF-BB), and transforming growth factor-β (TGF-β) stimulate HSC proliferation and/or matrix production, we tested their effect on IGFBPs released by HSCs. IGF-I induced IGFBP-3 and IGFBP-5 proteins in a time-dependent manner without an increase in the corresponding mRNAs. IGFBP-4 protein levels decreased in response to IGF-I. TGF-β stimulated IGFBP-3 mRNA and protein but decreased IGFBP-5 mRNA and protein. In contrast, PDGF-BB failed to regulate IGFBPs compared with controls. Recombinant human IGFBP-3 (rhIGFBP-3) was then tested for its effect on IGF-I-induced mitogenesis in HSCs. rhIGFBP-3 inhibited IGF-I-stimulated DNA synthesis in a dose-dependent manner, with a peak effect observed at 25 nM IGFBP-3. Because TGF-β is highly expressed in cirrhotic liver tissue, we determined whether IGFBP-3 mRNA expression is increased in liver biopsies obtained from patients with an active fibroproliferative response due to viral-induced chronic active hepatitis. In the majority of these samples, IGFBP-3 mRNA was increased compared with normal controls. These findings indicate that human HSCs, in their activated phenotype, constitutively produce IGFBPs. IGF-I and TGF-β differentially regulate IGFBP-3, IGFBP-4, and IGFBP-5 expression, which, in turn, may modulate the in vitro and in vivo action of IGF-I. J. Cell. Physiol. 174:240–250, 1998. © 1998 Wiley-Liss, Inc.     

Insulin-like growth factor binding proteins: new proteins, new functions.

The insulin-like growth factors (IGFs), IGF binding proteins (IGFBPs), and IGFBP proteases regulate somatic growth and cellular proliferation both in vivo and in vitro. IGFs are potent mitogens whose actions are determined by the availability of free IGFs to interact with IGF receptors. IGFBPs comprise a family of six proteins that bind IGFs with high affinity and specificity and thereby regulate IGF-dependent actions. IGFBPs have recently emerged as IGF-independent regulators of cell growth. Cleavage of IGFBPs by specific proteases modulate levels of free IGFs and IGFBPs and thereby their actions. IGFBP-related proteins (IGFBP-rPs) bind IGFs with low affinity and also play important roles in cell growth and differentiation. The GH-IGF-IGFBP axis is complex and powerful. Future research on its physiology promises exciting insights into cell biology as well as therapies for diseases such as cancer and diabetes mellitus.     

Solution structure of human insulin-like growth factor II; recognition sites for receptors and binding proteins.

The three-dimensional structure of human insulin-like growth factor II was determined at high resolution in aqueous solution by NMR and simulated annealing based calculations. The structure is quite similar to those of insulin and insulin-like growth factor I, which consists of an alpha-helix followed by a turn and a strand in the B-region and two antiparallel alpha-helices in the A-region. However, the regions of Ala1-Glu6, Pro31-Arg40 and Thr62-Glu67 are not well-defined for lack of distance constraints, possibly due to motional flexibility. Based on the resultant structure and the results of structure-activity relationships, we propose the interaction sites of insulin-like growth factor II with the type 2 insulin-like growth factor receptor and the insulin-like growth factor binding proteins. These sites partially overlap with each other at the opposite side of the putative binding surface to the insulin receptor and the type 1 insulin-like growth factor receptor. We also discuss the interaction modes of insulin-like growth factor II with the insulin receptor and the type 1 insulin-like growth factor receptor.