No abstract

Insulin-like growth factor binding proteins (IGFBP) are thought to modulate the biological actions of the insulin-like growth factors (IGF), including possible regulatory roles in the growth and differentiation of the central nervous system. Extracellular fluids usually contain a mixture of IGFBPs, three of which have been cloned, sequenced, and designated IGFBP-1, -2, and -3. We used Western ligand blotting, immunoprecipitation, and competitive binding analysis to characterize IGFBPs found in fetal and adult rat cerebrospinal fluid (CSF) and IGFBPs produced by cultures of neonatal rat choroid plexus, astrocytes, and C6 glial cells. Pooled rat CSF contains primarily IGFBP-2 (a narrow band at Mr = 29,000), lesser quantities of IGFBP-3 (a multicomponent broad band at Mr = 37,500-43,000), and trace amounts of low mol wt IGFBPs. Conditioned medium from cultures of choroid plexus cells contained a single binding protein corresponding to IGFBP-2, whereas C6 cells made predominately an IGFBP corresponding to IGFBP-3. Astrocytes secreted two IGFBPs corresponding to IGFBP-2 and -3, primarily IGFBP-3. Neonatal CSF contained substantially more binding activity corresponding to IGFBP-2 than did adult CSF. In all samples showing Western ligand binding profiles corresponding to IGFBP-2, identification was established by immunoprecipitation. Competitive binding analysis performed on choroid plexus IGFBP showed preferential high affinity binding for IGF-II compared with that for IGF-I. In conclusion, CSF contains a mixture of distinct IGFBPs, primarily IGFBP-2. The other IGFBPs found in CSF are capable of being synthesized locally within the central nervous system by glial cells and neurons, suggesting that they are not derived from plasma by transport across the blood-brain barrier.

The description of the cellular localization of insulin-like growth factor (IGF) receptors in the central nervous system (CNS) remains incomplete, as do the descriptions of changes in their characteristics with respect to different developmental stages. We, therefore, performed affinity labeling studies in microsomal membrane preparations of adult and fetal rat brain and liver tissues with [125I]IGF-I and [125I]IGF-II. These studies demonstrated tissue- and developmental stage-specific structural variants of type I receptor alpha-subunits as well as type II receptors. The adult rat brain type I alpha-subunit had an apparent mol wt (Mr) of 127,000, whereas those of adult and fetal rat liver measured 140,000. Fetal rat brain microsomes, however, had two types of type I receptor alpha-subunits measuring 130,000 and 120,000 Mr. The larger subunit from fetal brain consistently migrated at an apparent Mr of 3,000, greater than subunits from adult brain. Both type I and II receptors were more abundant in fetal liver and brain than in adult tissues. Affinity labeling was also performed directly to monolayers of cultured fetal brain neurons and newborn astrocytes. These studies detected both type I and II receptors on the surfaces of both types of cells. However, only the high Mr (140,000) form of the type I alpha-subunit was detected in cultured CNS cells, suggesting that expression of low Mr variant receptors is altered in vitro. Type II receptors were demonstrated by immunohistochemistry in adult rat hypothalamic neurons. However, the majority of neurons did not react with type II receptor antibody. This finding implies that only a minority of hypothalamic neurons are capable of responding to IGF-II via type II receptors. On the other hand, all astrocytes had striking type II receptor immunoreactivity. This signifies a more general biological role for this receptor in astrocytes compared with neurons. These results suggest that different tissue-, developmental stage-, and cell-specific processes are mediated by IGF receptors and suggests new directions in which to explore potential biological actions for these receptor-ligand systems in the CNS.

The outcome of in utero cocaine exposure is unclear. To determine if cocaine affects neuronal growth and differentiation, we used PC-1 2 cells, which have a mitogenic response to IGF-I and differentiate into neurons on exposure to nerve growth factor. Differentiation was quantified as neurite extension after a 72-h exposure to 20 ng/ml nerve growth factor (dosage at 50% maximal effectiveness) and cocaine doses ranging from 0.01 to 10 gg/ml. The results were 49±2, 40±3, 29±2, 23±2, and 12±2% differentiation with respective cocaine concentrations of 0, 0.01, 0.1, 1, and 10 ,g/ml (P < 0.0001). Cocaine stability studies showed insignificant spontaneous hydrolysis under the conditions of this study. Cocaine did not affect cell viability or number, but had a relatively modest, statistically significant (P < 0.0001 ) inhibitory effect on IGF-I-stimulated thymidine incorporation. The dose-response curves for differentiation vs mitogenic response differed significantly (P = 0.021). Therefore, cocaine inhibition of these processes is probably mediated by different mechanisms, and not caused by generalized toxicity. To our knowledge, this is the first demonstration of cocaine effects on neuronal multiplication and differentiation in vitro. The results suggest in utero exposure may directly impair brain development. (J. Clin. Invest. 1994. 93:1179-1185