Impaired SDF1/CXCR4 Signaling in Glial Progenitors Derived From SOD1 G93A Mice Yongquan Luo, 1 Haipeng Xue, 1 Andrea C. Pardo, 2 Mark P. Mattson, 1 Mahendra S. Rao, 2,3 and Nicholas J. Maragakis 2 * 1 Laboratory of Neurosciences, National Institute on Aging Intramural Research Program, Baltimore, Maryland 2 Department of Neurology, Johns Hopkins University, Baltimore, Maryland 3 Invitrogen Corporation, Carlsbad, California Mutations in the superoxide dismutase 1 (SOD1) gene are associated with familial amyotrophic lateral sclerosis (ALS), and the SOD1 G93A transgenic mouse has been widely used as one animal model for studies of this neu- rodegenerative disorder. Recently, several reports have shown that abnormalities in neuronal development in other models of neurodegeneration occur much earlier than previously thought. To study the role of mutant SOD1 in glial progenitor biology, we immortalized glial restricted precursors (GRIPs) derived from mouse E11.5 neural tubes of wild-type and SOD1 G93A mutant mice. Immunocytochemistry using cell lineage markers shows that these cell lines can be maintained as glial progeni- tors, because they continue to express A2B5, with very low levels of glial fibrillary acidic protein (astrocyte), bIII- tubulin (neuron), and undetected GalC (oligodendrocyte) markers. RT-PCR and immunoblot analyses indicate that the chemokine receptor CXCR4 is reduced in SOD1 G93A GRIPs. Subsequently, SOD1 G93A GRIPs are unable to respond to SDF1a to activate ERK1/2 enzymes and the transcription factor CREB. This may be one pathway leading to a reduction in SOD1 G93A cell migration. These data indicate that the abnormalities in SOD1 G93A glial progenitor expression of CXCR4 and its mediated sig- naling and function occur during spinal cord develop- ment and highlight nonneuronal (glial) abnormalities in this ALS model. V V C 2007 Wiley-Liss, Inc. Key words: neural progenitors; migration; ERK activ- ation; neurodegeneration; chemotaxis Mutations in the gene for superoxide dismutase 1 (SOD1) have been associated with an autosomal domi- nant form of amyotrophic lateral sclerosis (ALS). Several transgenic mouse models of the mutant form of human SOD1 (G93A, i.e., glycine substituted to alanine at posi- tion 93, as well as G37R, G85R) resulted in the develop- ment of an animal model of the disease (Gurney et al., 1994; Bruijn et al., 1997). All of the mutant SOD1 mice develop a slowly progressive hindlimb paralysis, with early death. This is accompanied by motor neuron loss, astro- gliosis, and in most of the models cytosolic SOD1 inclu- sions (Wong et al., 1995; Bruijn et al., 1998). The SOD1 G93A mouse has been considered a model of a neu- rodegenerative disease, with early studies indicating that no significant pathology is present early in the course (Gurney et al., 1994). However, emerging data have sug- gested that many pathological and physiological changes precede the onset of hindlimb weakness. Chemokines are a family of small secreted proteins with diverse immune and neural functions, including control of leukocyte trafficking, organization of the hem- atopoetic/lymphopoetic system, and angiogenesis (Tran and Miller, 2003). Stromal-cell-derived factor (SDF1) belongs to the CXC subfamily and is the only known ligand for the membrane-bound G-protein-coupled re- ceptor CXCR4 (Oberlin et al., 1996; Rossi and Zlotnik, 2000). Three spliced variants of SDF1 have been found: SDF1a, SDF1b, and SDF1g (Gleichmann et al., 2000). The study of mouse models in which the SDF1/ CXCR4 pathway has been manipulated shows that CXCR4 mutant mice, like SDF1 null mutants, die at about the time of birth, with severe abnormalities in orga- nogenesis and effects on neural precursor migration in the cerebellum, an abnormal development of the dentate gyrus, and defects in interneuron migration in the cortex (Ma et al., 1998; Lu et al., 2002). These abnormalities have been attributed to the loss of an attractant to embry- onic neural progenitors (Bagri et al., 2002; Reiss et al., 2002; Stumm et al., 2003). SDF1 is strongly expressed by the meningeal cells in the pial layer of the cerebellum and is a chemoattractant for neural stem cells of the embryonic cerebellar external granular layer (Reiss et al., 2002; Zhu et al., 2002). In CXCR4 mutant mice, a defect in neuro- Contract grant sponsor: NIA Intramural Research Program; Contract grant sponsor: ALS Association; Contract grant number: R01NS041680; Contract grant sponsor: Packard Center for ALS Research at Johns Hop- kins. *Correspondence to: Nicholas J. Maragakis, MD, Department of Neurol- ogy, Johns Hopkins University, 600 N. Wolfe St., Meyer 6-119, Balti- more, MD 21287. E-mail: nmaragak@jhmi.edu Received 17 November 2006; Revised 20 March 2007; Accepted 9 April 2007 Published online 13 June 2007 in Wiley InterScience (www. interscience.wiley.com). DOI: 10.1002/jnr.21398 Journal of Neuroscience Research 85:2422–2432 (2007) ' 2007 Wiley-Liss, Inc.