"Killer" impacts and life's origins.

(ii) hepatocyte growth factor/nerve growth factor/transforming growth factor–␤(TGF␤)/glial cell line–derived neurotrophic factor: 36 Ϯ 13 min –1 (n ϭ 12), (iii) TGF ␣/glial growth factor: 8 Ϯ 5 min –1 (n ϭ 8), and (iv) neurotrophin-3/plate-derived growth factor– AA: 16 Ϯ 6 min –1 (n ϭ 12). Each of these factors promoted cell survival, proliferation, or differentiation in other assays performed in our lab. To test the effect of extracellular matrix components, we cultured RGCs on Matrigel (Collaborative Biomedical Products) (mean EPSC frequency of 21 Ϯ 11 min –1 ; n ϭ 7) or in serum-free medium containing chon-droitin sulfate proteoglycan (5 ␮g/ml; Collaborative Biomedical Products) (mean frequency 2 Ϯ 2 min –1 ; n ϭ 6). Both components were active, as they fas-ciculated neurites of RGCs. 15. Astrocytes (7) and oligodendrocytes [B. A. Barres et al., Cell 70, 31 (1992)] were purified by immunopan-ning from postnatal rat optic nerves as described, because these glial cell types cannot be purified from other brain areas. Microglial cells were purified from the superior colliculus of P8 rats by immunopanning with a goat anti-rat immunoglobulin G antibody (Jackson ImmunoResearch Laboratories) and a rat anti-mouse Thy 1.2 antibody (Boehringer). RGCs were cultured with each glial cell type at a ratio of three to five glial cells per neuron in the defined serum free medium (8). In order to avoid contact between neurons and glial cells, we cultured RGCs on cover slips above glial feeding layers. and D-2-amino-5-phosphonopentanoic acid (50 ␮M, RBI) to the extracellular recording solution. 18. The average whole-cell resting membrane potential in RGCs cultured for 10 to 15 days without (– 48 Ϯ 1 mV, n ϭ 22) or with glial cells (– 49 Ϯ 1 mV, n ϭ 19) was not different. This, however, does not rule out that glial cells increased quantal release by depolar-izing nerve terminals. 19. For electron microscope analysis, cultures were fixed with glutaraldehyde (2.5% in phosphate-buff-ered saline for 30 min), treated with osmium tetroxide (1% for 30 min), dehydrated, and embedded in Epon 812 (Polysciences). For each culture, ultrathin sections were cut from at least two blocks with a glass knife, stained with lead citrate and uranyl acetate, and examined with a Phillips EM 410. Ultrastructural criteria to identify synapses were a pre-and a postsynaptic density, a synaptic cleft, and more than two synaptic vesicles in the presynaptic active zone. We counted synapses in photomicrographs (ϫ18,500 …