Integrin-mediated action of insulin-like growth factor binding protein-2 in tumor cells.

The neoplastic production of the insulin-like growth factor binding protein (IGFBP)-2 often correlates with tumor malignancy and aggressiveness. Since IGFBP-2 contains an RGD motif in its C-terminus, it was hypothesized that this protein may act independently of IGF on tumor cells through integrins. To investigate this, integrin binding, intracellular signaling and the impact of IGFBP-2 on cell adhesion and proliferation were examined in two tumor cell lines. In tracer displacement studies, up to 30% of the added (125)I-hIGFBP-2 specifically bound to the cells. Bound (125)I-hIGFBP-2 was reversibly displaced by IGFBP-2, IGFBP-1 and RGD-(Gly-Arg-Asp)-containing peptides, but not by IGFBP-3, -4, -5, -6 and RGE-(Gly-Arg-Glu)-containing peptides. Blocking with antibodies directed against different integrins and with fibronectin demonstrated that IGFBP-2 cell surface binding is specific for alpha5beta1-integrin. Incubation of IGFBP-2 with equimolar quantities of IGF-I and IGF-II annihilated RGD-specific binding. IGFBP-2 binding at the cell surface led to dephosphorylation of the focal adhesion-kinase (FAK) of up to 37% (P<0.01), and of the p42/44 MAP-kinases of up to 40% (P<0.01). In addition, IGFBP-2 promoted de-adhesion of the cells dose-dependently by up to 30% (P<0.05), and reduced proliferation by 24% (P<0.01). Since one of the cell lines used does not express a functional IGF-I receptor, these data demonstrate that IGFBP-2 can act in an IGF-independent manner, at least in part by an interaction with alpha5beta1-integrin.

Correlation of type I insulin-like growth factor receptor (IGF-I-R) and insulin receptor-related receptor (IRR) messenger RNA levels in tumor cell lines from pediatric tumors of neuronal origin.

The insulin receptor-related receptor (IRR) is a member of the insulin receptor family. So far no ligand has yet been discovered for this receptor type (orphan receptor). IRR, insulin receptor (IR), and insulin-like growth factor-I receptor (IGF-I-R) are all tyrosine kinases. The cellular function of the IRR is not known. The expression of IRR mRNA is restricted to a few, e.g. neuronal tissues, and has also been found in neuroblastomas. Since tyrosine kinase receptors, including the IGF-I-R, may be involved in tumor genesis, we examined the expression of IRR mRNA and IGF-I-mRNA in 18 tumor cell lines using RT-PCR and the solution hybridization/RNAse protection assay. In particular, the mRNA levels of IRR and IGF-I-R were compared by semi-quantitative RT-PCR in seven neuroblastomas and 11 soft tissue sarcomas (STS), five of which were of neuronal origin. In all of the seven neuroblastoma cell lines and in five of the 11 STS cell lines, the IRR mRNA was detected. In addition, the IRR mRNA was expressed in rhabdomyosarcoma, in leiomyosarcoma, in one of the Ewing sarcoma and in the neurofibrosarcoma cell line. The last two tumor cell types are of neuronal origin. The levels of expression of IGF-I-R and IRR mRNA of the neuroblastoma cell lines were closely related (r = 0.82, P < 0.002). Furthermore, IRR mRNA was found only in cell lines that also expressed IGF-I-R mRNA. In conclusion, cell lines from pediatric tumors of neuronal origin express IRR mRNA simultaneously with a another tyrosine kinase receptor (IGF-I-R) mRNA. The tight coupling of their mRNA expression suggests a functional association of both receptors in the tumor cells.

Secretion of noncomplexed 'Big' (10-18 kD) forms of insulin-like growth factor-II by 12 soft tissue sarcoma cell lines.

The paraneoplastic production of pro-insulin-like growth factor-II (IGF-II) forms causes tumour hypoglycaemias and presumably also has an effect on tumour cell growth. We investigated the molecular weights of IGF-II forms and their ability to form complexes with IGF binding proteins (IGFBPs) in conditioned culture media (CM) from 12 paediatric soft tissue sarcoma (STS) cell lines and from two healthy fibroblast lines. Untreated CM were separated by size exclusion chromatography using biocompatible HPLC. Subsequently, IGF-II, IGFBP-2 and IGFBP-3 were determined in the HPLC fractions by specific RIAs. In the CM, IGF-II concentrations between 0.5 and 8.6 ng/10(6) cells were measured but no IGF-I was detectable. Parallel to this investigation, a high IGF-II mRNA level averaging 44.4 +/- 29.7% was measured by semi-quantitative RT-PCR. The STS cell lines secreted a higher proportion of big-IGF-II forms reaching 10-18 kD (10-33% of the total IGF-II secreted) compared to the healthy fibroblasts (2.5-5%). At the same time, the proportion of IGF-II bound with IGFBP in complexes of 35- 70 kD and 150 kD was reduced by up to 85% in CM from tumour cells. The tumour cell lines apparently secrete a different spectrum of IGF-II forms than healthy fibroblasts. The reduced ability to form complexes with IGFBP and the higher molecular weight of the IGF-II forms produced by the tumour cells indicate that these forms could in fact be the known tumour-associated pro-IGF-II forms. Due to these characteristics, the big-IGF-II forms probably have an altered biological effect on the tumour cells when compared to IGF-II.