Differential mRNA Localization in Astroglial Cells in Culture SILVIA MEDRANO 1 AND OSWALD STEWARD 2 * 1 Department of Neuroscience, University of Virginia, Charlottesville, Virginia 22908 2 Reeve-Irvine Research Center and Departments of Anatomy & Neurobiology and Neurobiology & Behavior, University of California, Irvine, California 92697-4292 ABSTRACT Messenger RNA (mRNA) targeting to specific subcellular domains has been studied exten- sively in many cell types, and there is increasing evidence suggesting that mRNA sorting also occurs in astrocytes. As a step toward developing strategies to evaluate the signals that govern mRNA sorting in astrocytes, the authors studied the subcellular distribution of several repre- sentative mRNAs, poly(A) RNA and ribosomal RNA, in process-bearing (type-2) astroglial cells in culture. Nonradioactive in situ hybridization analysis revealed a gradual increase in the expres- sion of glial fibrillary acidic protein (GFAP) mRNA as type-2 astrocytes differentiated in culture. In mature cells, labeling was present in both cell bodies and processes. GFAP mRNA labeling was granular in nature and was particularly concentrated at branch points and at the tips of the processes. Unlike GFAP mRNA, vimentin, -tubulin, and - and -actin mRNAs were mainly confined to the cell bodies, with only occasional labeling seen in the processes. Nonradioactive and radioactive in situ hybridization analysis of poly(A) and ribosomal RNA, respectively, revealed labeling in cell bodies and processes of immature and differentiated astrocytes. Treatment with nocodazole, a microtubule depolymerizing agent, resulted in a substantial reduction of GFAP mRNA labeling in the processes, whereas treatment with cytochalasin D, a microfilament- disrupting agent, did not alter GFAP mRNA distribution. The results indicate that cultured type-2 astrocytes have the capacity to sort mRNAs to different subcellular domains and that the localization of GFAP mRNA to astrocyte processes requires intact microtubules. J. Comp. Neurol. 430:56 –71, 2001. © 2001 Wiley-Liss, Inc. Indexing terms: nonradioactive in situ hybridization; glial fibrillary acidic protein mRNA; cytoskeletal proteins The localization of messenger RNA (mRNA) to discrete subcellular domains is a common feature of gene expres- sion in polarized and morphologically complex cells. In the nervous system, mRNA sorting has been investigated ex- tensively in neurons (for reviews, see Steward, 1997; Steward and Singer, 1997) and oligodendrocytes (Amur- Umarjee et al., 1990; Barbarese, 1991; Ainger et al., 1993), and there is increasing evidence that sorting also occurs in astrocytes. For example, retinal Mu ¨ ller cells exhibit dif- ferential distribution of glial fibrillary acidic protein (GFAP) and -actin mRNAs (Sarthy et al., 1989). GFAP mRNA has been detected in the processes of reactive as- trocytes and radial glia in human brain and spinal cord (Landry et al., 1994) and in the processes of astrocytes in mixed neural cell cultures (Trimmer et al., 1991). Hence, mRNA sorting appears to be an important aspect of gene expression in all three of the cell types in the central nervous system that derive from the neuroectoderm. Studies on mRNA localization in other cell types have indicated certain features that suggest conserved mecha- nisms of mRNA transport and localization: 1) In polarized cells, most mRNAs are localized near the nucleus, whereas only a few mRNAs are localized at a distance from the nucleus. This differential localization suggests a mechanism for mRNA sorting within the cytoplasm (for review, see Steward, 1997). 2) The mRNAs that are present at a distance from the nucleus are often found in granules (Sundell and Singer, 1991; Ainger et al., 1993; Bassell et al., 1994; Ferrandon et al., 1994; Barbarese et al., 1995; Knowles et al., 1996; Tongiorgi et al., 1997). This granular distribution suggests that localized mRNAs are Grant sponsor: National Institutes of Health; Grant numbers: NS12333, NS29875, and 5T32-HD07323. *Correspondence to: Dr. Oswald Steward, Reeve-Irvine Research Cen- ter, University of California, Irvine, 1105 Gillespie Neuroscience Research Facility, Irvine, CA 92697-4292. E-mail: osteward@uci.edu Received 18 February 2000; Revised 12 September 2000; Accepted 11 October 2000 THE JOURNAL OF COMPARATIVE NEUROLOGY 430:56 –71 (2001) © 2001 WILEY-LISS, INC.