Orly Reiner Department of Molecular Genetics, The Weizmann Institute of Science, 76100 Rehovot, Israel. e-mail: [email protected] Research on homeobox genes has shown them to have crucial roles in many developmental processes. A new study on the homeobox transcription factor, ARX, offers insights into neuronal migration and, surprisingly, testis development. Aristaless-related proteins (including Arx) belong to the Pax family of homeobox transcription factors1, several members of which are known to be involved in genetic diseases. On page 359 of this issue, Kunio Kitamura and colleagues2 use a gene-targeting approach to identify a pivotal role for the Arx transcription factor in brain and testis development. These findings prompted the investigators to ask whether any human diseases are caused by ARX inactivation. Their search led them to several individuals suffering from X-linked lissencephaly with abnormal genitalia3 (XLAG), who turn out to have mutations in ARX. Can't get there from here Lissencephaly is a severe human brain malformation that is associated with abnormal neuronal migration. The form of lissencephaly syndrome caused by mutant ARX is somewhat different from classical lissencephaly. It involves reduced cortical thickness and immature cerebral white matter. Small interneurons are notably missing in the XLAG neocortex as well. Unlike lissencephalies caused by mutations in MDCR4 (also known as LIS1), DCX5, 6, or RELN7, 8, the cortex in XLAG is organized in a simplified three-layer structure, and no layer inversions or duplications are involved. The generation of Arx knockout mice by Kitamura et al.2 allowed the developmental abnormalities of brain and testis associated with human XLAG to be elucidated. One of the difficulties the authors circumvented was the fact that hemizygous mutant male mice died within half a day after birth. To overcome this problem, delivery was delayed until E20.5, and the mice were then hand fed for three days. Their studies showed that the cortex of these mice is thin as a result of decreased proliferation of progenitor cells in the entire ventricular zone. Radial migration seemed to be normal, whereas migration of interneurons from the germinal eminences was markedly affected. More specifically, Arx deficiency resulted in the loss of one route of migration from the medial germinal eminences (MGE) to the cortical intermediate zone, whereas interneurons born in the same position migrated normally to the cortical subventricular zone. This was accompanied by abnormal expression (but not loss) of Titf1 (also known as Nkx2.1) and Dlx, markers of the MGE. Although loss of Dlx1/2 or Titf1 in the MGE results in a significant reduction in the number of neocortical interneurons9, the underlying mechanisms may be different from that observed in Arx mutant mice. Whereas at E18.5 the interneurons in wildtype animals were scattered throughout the cortical plate, the Arx mutant interneurons that were able to migrate were confined to the subplate. Later, at P3, the wildtype GABAergic interneurons were condensed in layer V, but in mutant mice these cells were scattered throughout the cortical plate. Thus, this study is the first to demonstrate clearly that migration of GABAergic interneurons occurs in two stages. First, 'tangential migration' occurs from the germinal eminences to the intermediate zone/subventricular zone. Then the cells move vertically to take up their positions in the cortical anlage10. ARX and mental retardation XLAG is not the only disease associated with mutations in ARX. In humans, a number of remarkably diverse mutations in this locus have been described2, 11-13. Interestingly, all individuals with ARX mutations suffer from mental retardation, but most have normal MRIs (magnetic resonance imaging). However, this is the first clearly documented example of multiple mutations within a single gene resulting in phenotypic presentations with normal and abnormal MRI patterns. This suggests that some mutations do not affect neuronal migration in an obvious manner. In fact, when a mutation is identified in ARX, it is difficult to predict the phenotype because the associated symptoms are variable to the point where patients with identical mutations have been diagnosed as suffering from distinct syndromes—Partington syndrome or West syndrome. Moreover, people with ARX mutations not only have differing physical maladies, but variable behavioral features as well; some exhibit language deficits, and others exhibit aggressive behavior. For example, two point mutations within the conserved paired-like homeodomain resulted in XLAG, whereas a different point mutation in an adjacent conserved proline residue caused mental retardation with myoclonic epilepsy11. Point mutations in regions outside of the homeodomain result in mental retardation with normal MRI findings in most affected individuals13. Mutations in ARX may result in a specific phenotypic presentation in females as well. Some female XLAG carriers have neurological abnormalities, such as agenesis of the corpus callosum with or without mild-to-moderate neurological symptoms that include attention deficit disorder, mental retardation or epilepsy14. In contrast, female mutant mice developed normally and were fertile, although some were obese. Abnormal differentiation of interneurons expressing neuropeptide-Y (NPY) was found in the present study2. As NPY-interneurons are known to be involved in excitability control, it is possible that the epileptic phenotype found in human patients and in the mouse model could be associated with this phenomenon. Testis development In addition to the brain phenotype, individuals with XLAG have abnormal testicular differentiation with altered external genitalia. Presentation varies from apparently female genitalia to moderately hypoplastic but clearly male genitalia. The testicular phenotypes of male Arx null mice and of genotypic XLAG human males are remarkably similar, with defects in Leydig cell differentiation. It is interesting to note, however, that ARX is not expressed in Leydig cells but rather in other interstitial cells. This is in contrast to the brain, where the phenotype occurs in the cells that express the protein. Through use of specific markers, Kitamura et al.2 have revealed several putative targets for Arx, such as Wnt8b and Lhx9. Products of both genes were completely missing from thalamic eminences in mutant mice. The identification of targets in other affected cell types should improve our understanding of the roles of ARX in the differentiation of NPY-interneurons and in the testis. Finally, the work of Kitamura et al.2 raises the question of whether the lissencephalic phenotype results from the combined defects in neuroblast proliferation, abnormal differentiation, abnormal nerve tracks, or migration. It is possible to distinguish between them. Combining the mouse model with cell-specific or inducible gene inactivation may answer this question, perhaps explaining why individuals with mutant ARX have such wide-ranging phenotypes. REFERENCES Meijlink, F., Beverdam, A., Brouwer, A., Oosterveen, T.C. & Berge, D.T. Int. J. Dev. Biol. 43, 651-663 (1999). | PubMed | Kitamura, K. et al. Nature Genet. 32 359-369 (2002). | Article | PubMed | Berry-Kravis, E. & Israel, J. Ann. Neurol. 36, 229-233 (1994). | PubMed | Reiner, O. et al. Nature 364, 717-721 (1993). | PubMed | des Portes, V. et al. Cell 92, 51-61 (1998). | PubMed | Gleeson, J.G. et al. Cell 92, 63-72 (1998). | PubMed | Rice, D.S. & Curran, T. Annu. Rev. Neurosci. 24, 1005-1039 (2001). | Article | PubMed | Hong, S.E. et al. Nature Genet. 26, 93-96 (2000). | Article | PubMed | Anderson, S.A., Marin, O., Horn, C., Jennings, K. & Rubenstein, J.L. Development 128, 353-363 (2001). | PubMed | Nadarajah, B. & Parnavelas, J.G. Nature Rev. Neurosci. 3, 423-432 (2002). | Article | PubMed | Stromme, P. et al. Nature Genet. 30, 441-445 (2002). | Article | PubMed | Stromme, P., Mangelsdorf, M.E., Scheffer, I.E. & Gecz, J. Brain Dev. 24, 266-268 (2002). | Article | PubMed | Bienvenu, T. et al. Hum. Mol. Genet. 11, 981-991 (2002). | Article | PubMed | Bonneau, D. et al. Ann. Neurol. 51, 340-349 (2002). | Article | PubMed |