The ability to generate expanded populations of individual cell types able to undergo normal differentiation in vitro and in vivo is of critical importance in the investigation of the mechanisms that underlie differentiation and in studies on the use of cell transplantation to repair damaged tissues. This review discusses two different approaches to the generation of expanded cell populations with phenotypes useful for either of these purposes. In one line of research, an analysis of the growth control properties of glial precursor cells of the CNS has revealed that cooperation between appropriate mitogens can promote extended precursor division in the absence of differentiation, thus allowing unprecedented expansion of a primary cell population without resort to the expression of activated oncogenes in the cells of interest. In a second line of research, H-2KbtsA58 transgenic mice have been developed in order to allow the direct derivation of conditionally immortal cell lines from many tissues of the body simply by dissection and growth of cells under permissive conditions. In both instances, cells grown for extended periods in vitro displayed normal patterns of differentiation when reintroduced in vivo. In addition, conditionally immortal astrocytes derived from H-2KbtsA58 mice appear to offer a simple cellular model for studying the ability of glial scar tissue to inhibit migration of glial precursor cells and extension of neurites from mature neurons.
Three clonal gonadotropin-releasing hormone (GnRH) neuronal cell lines were derived from a genetically induced tumor in transgenic mice. A transgene was constructed to target expression of simian virus 40 T antigen to GnRH neurons using the promoter/enhancer domains of the cell-specifically expressed GnRH gene. The resulting GT1 cells were characterized by morphology, the expression of neuron-specific genes, expression and processing of GnRH, pulsatile basal secretion of GnRH, release of GnRH in response to depolarization, and regulation of GnRH release by a variety of neurotransmitters and neuromodulators. By all of these criteria, GT1 cells are highly differentiated neuronal cell lines that provide valuable models for studying the cell biology of neuroendocrine cells.
A useful approach to molecular biologic studies of complex mammalian systems is the judicious use of clonal cell lines. For studies on neuronal populations of the central nervous system, useful cell lines are lacking. A general protocol that produces clonal lines by retroviral infection in vivo followed by in vitro infection with an additional oncogene-containing virus is detailed. The lines produced express several neuronal markers but not glial fibrillary acidic protein. The lines are phenotypically stable, freezable, and transfectable by standard methods. Retroviral combinations that have produced stable lines using this protocol are large T with v-src and v-myc, and large T with v-ras. Other oncogenic combinations may produce different, useful phenotypes for studies on the molecular development and function of CNS neurons.
We describe methods for grafting cultured cells into the newborn and the adult rat using as an example the immortalized hippocampal cell line HiB5. We discuss cell labeling strategies and animal preparation prior to as well as animal care after surgery. The principles of stereotaxic surgery are presented with a standard protocol for transplantation into the adult. A grafting technique that allows micro-volumes of cell suspension to be precisely placed in very small target sites with minimum trauma is described. Also described is a detailed protocol for stereotaxic placement in the newborn that utilizes a new instrument that allows safe, prolonged anesthesia and accurate, reproducible placement of cell suspensions.

