Lack of Correspondence Between mRNA Expression for a Putative Cell Death Molecule (SGP-2) and Neuronal Cell Death in the Central Nervous System Gwenn A. Garden, zyxwvuts ' Mark Bothwell,' and Edwin W Rubel ',** ' Department of Physiology and Biophysics and Hearing Development Laboratories. and 'Department of Otolaryngology-Head and Neck Surgery, University of Washington, Seattle, Washington 98195 SUMMARY zyxwvutsrq Neuronal death during nervous system development, a widely observed phenomenon, occurs through unknown mechanisms. Recent evidence suggests an active, destruc- tive process requiring new gene expression. Sulfated gly- coprotein-2 (SGP-2), a secretory product of testicular Sertoli cells has been shown to up-regulate in several nonneural tissues undergoing programmed cell death and in several types of neuronal degeneration. In order to determine if this message up-regulates in neurons under- going developmentally determined cell death, we have studied the expression of SGP-2 mRNA in the develop- ing and adult rat central nervous system (CNS) with in zyxwvuts situ hybridization. We also report on the expression of this message in nonneural tissues from several regions of the developing embryo. The developing and adult rat cen- tral nervous system as well as widely varied tissues in the rat embryo express SGP-2 mRNA in a pattern that does not correlate with regions undergoing developmental cell death. In the nervous system, SGP-2 mRNA is ex- pressed in neuronal populations including motor neu- rons, cortical neurons, and hypothalamic neurons at ages when the period of developmental cell death has passed. In a nonneural tissue (palatal shelve epithelium) for which a developmental cell death period has been de- scribed, SGP-2 mRNA was not present in the region where cell death occurs. We conclude that SGP-2 mRNA expression cannot be correlated with programmed cell death in neural or nonneural tissues. The results of this study as well as recently reported SGP-2 homologies indi- cate a possible role for this protein in secretion and lipid transport. INTRODUCTION The developing nervous system gives rise to many more neurons than will remain in the mature ani- mal (Hamburger, 1982;Oppenheim, 1985). Regu- lation of neuronal population size occurs largely through the process of developmental neuronal death (Oppenheim, 1985). The cellular mecha- nisms underlying the death of neuronal subpopu- lations are not understood. Recent evidence sup- ports the concept of developmental neuronal death Received February 25, I99 1 ; accepted zyxwvutsrq April 30, 1991 Journal of Keurobiology, VoI. 22. No. 6, pp. 590-604 (1 99 1) zyxwvutsrqp C2 199 1 John Wiles. & Sons, Inc. CCC 0022-3034/9 1/060590- 15$04.00 * To whom correspondence should be addressed. zyxwvuts 590 as an active process requiring new gene expression within the dying cell. Inhibition of new mRNA or protein synthesis retards developmental neuronal death of chick motoneurons and dorsal root gan- glion cells, in vivo (Oppenheim, Prevette, Tytel, and Homma, 1990). and prevents cell death that results from deprivation of neurotrophic factors in cultured chick sensory and autonomic neurons (Martin et al., 1988: Scott and Davies, 1990). Two mutations in the nematode, Caenorhabditis ele- guns, ced-3, and ced-4, produce a phenotype lack- ing programmed neuronal death (Ellis and Hor- vitz, 1986). Genetic mosaic analysis has shown that loss of ced-3 or ced-4 function from the neu- rons destined to die prevents the process of pro- grammed neuronal death specifically in those cells that contain the mutation (Yuan and Horvitz,