Cultured sympathetic neurons from neonatal rats release a Ca²⁺-dependent metalloproteinase capable of degrading native and denatured type I collagen. A large fraction of the metalloproteinase activity is released from growth cones of growing neurites, which suggests that this proteinase may be involved in neurite penetration into collagen-rich extracellular matrices during development. In the present study, we investigated the possibility that neurite penetration into 3-dimensional collagen gels requires the Ca²⁺-dependent metalloproteinase. Sympathetic neurons were plated on top of collagen gels and the neuronal metalloproteinase activity was inhibited with the collagenase inhibitor HSCH₂CH[CH₂CH(CH₃)₂]-CO-Phe-Ala-NH₂ (HS-LFA). HS-LFA inhibited the Ca²⁺-dependent metalloproteinase activity, penetration of neurites into collagen gels and release of ³H from ³H-collagen gels in a dose-dependent manner over very similar concentration ranges (EC₅₀ = 150–200 nM). Highly significant correlations existed between neurite penetration into collagen gels and metalloproteinase activity (r = 0.99), and metalloproteinase activity and release of ³H from ³H-collagen gels (r = 0.99). Although HS-LFA inhibited the growth of neurites within collagen gels, it had no effect on neurite outgrowth on collagen films. Thus, the neuronal Ca²⁺-dependent metalloproteinase appears to be required for neurite penetration into and growth within 3-dimensional collagen matrices, but not for growth along a 2-dimensional collagen substrate.

Guanine nucleotide-binding regulatory proteins, or G proteins, mediate the interaction of agonist receptors on the platelet surface with phospholipase C and adenylyl cyclase. To better understand this process, we have used several approaches to identify which G proteins are present in platelets, normal human megakaryocytes, and human erythroleukemia (HEL) cells, a leukemic cell line with megakaryocytic features. Because platelet and HEL cell responses to thrombin are inhibited by pertussis toxin, we have focused upon the members of the Gi family, whose alpha subunits can be ADP-ribosylated by that toxin. Western blots with antisera specific for Gi alpha demonstrated the presence in both platelets and HEL cells of the three best-described forms of this protein: Gi alpha 1, Gi alpha 2, and Gi alpha 3. Based upon immunoprecipitation studies with [35S]-methionine-labeled HEL cells, their relative abundance appears to be Gi alpha 2 much greater than Gi alpha 3 greater than Gi alpha 1. A HEL cell cDNA library screened with the Gi alpha antisera produced clones encoding Gi alpha 2 and Gi alpha 3 that had sequences similar to those reported from other sources. Gi alpha-specific probes created from these cDNA clones confirmed the presence of mRNA encoding Gi alpha 2 and Gi alpha 3 in both platelets (by Northern blotting) and megakaryocytes (by in situ hybridization). Thus the pertussis toxin substrates that have previously been detected in platelets and HEL cells are shown to be members of the Gi alpha family, all of which are candidates for interaction with receptors for thrombin and other agonists.

No abstract

1 RS-45041-190 (4-chloro-2-(imidazolin-2-yl)isoindoline) showed high affinity for 12 imidazoline receptors labelled by [3H]-idazoxan in rat (pKi = 8.66 ± 0.09), rabbit (pKi = 9.37 ± 0.07), dog (pKi = 9.32 ± 0.18) and baboon kidney (pKi = 8.85 ± 0.12), but had very low affinity for a2-adrenoceptors in rat cerebral cortex (pKi = 5.7 ± 0.09).2 RS-45041-190 showed low affinity for other adrenoceptors, dopamine, 5-hydroxytryptamine, and muscarinic receptors and dihydropyridine binding sites (selectivity ratio>1000). 3 RS-45041-190 showed moderate potency for the inhibition of monoamine oxidase A in vitro (pIC50= 6.12), but had much lower potency for monoamine oxidase B (pICs = 4.47), neither of which equated with its affinity for I2 receptors. 4 RS-45041-190 (0.001 to 3 mg kg-', i.v. and 1 ng-50 yg i.c.v.) had only small, transient effects on blood pressure and heart rate in anaesthetized rats. In conscious rats, RS-45041-190 had no effect on body core temperature or tail skin temperature (1 mg kg-', s.c.) or on activity or rotarod performance (10 mg kg-1, i.p.). There were also no effects on barbiturate sleeping time in mice after doses of 1-10 mg kg-, i.p.5 RS-45041-190 (10 and 25 mg kg-', i.p.) significantly increased food consumption in rats for up to 4 h after dosing, but unlike idazoxan (10 mg kg-, i.p.) did not increase water consumption. 6 RS-45041-190 is therefore a selective, high-affinity ligand at I2 imidazoline receptors and its hyperphagic effect may suggest a role for I2 imidazoline receptors in the modulation of appetite. However, in the absence of a selective agonist it is unclear whether this ligand is an agonist or an antagonist at I2 receptors.

No abstract