Annu. Rev. Neurosci. 2003. 26:355–80doi: 10.1146/annurev.neuro.26.041002.131137Copyrightc 2003 by Annual Reviews. All rights reservedGENERATING TH E CEREBRAL CORTICALAREA MAPElizabeth A. Grove and Tomomi Fukuchi-ShimogoriDepartment of Neurobiology, Pharmacology, and Physiology, University of Chicago,947 East 58th Street, Chicago, Illinois 60637; email: [email protected],[email protected] Words cerebral cortex, patterning, signaling center, morphogen Abstract The view that the cortical primordium is initially patterned in similarways to the rest of the embryo has been a conceptual breakthrough. We now havea new starting point for understanding how the cortical area map is established andhow maps may change and evolve. Here we review findings that signaling moleculessecreted from distinct cortical signaling centers establish positional information inthe cortical primordium and regulate regional growth. In other embryonic systems,positional signals would regulate the patterned expression of transcription factors,leading, in a gene regulatory cascade, to the patterned differentiation of the tissue.We discuss candidate transcription factors with respect to such a model of corticalpatterning. Finally, embryonic structures interact to pattern one another. We reviewdata suggesting that the thalamus and cortex are patterned independently then interactto generate the final cortical area map.INTRODUCTIONA spectacular advance in biology has been to uncover many of the principles andmolecular mechanisms that underlie embryonic patterning of the vertebrate andinvertebrate body plans (Wolpert 1996). These discoveries have been applied tounderstanding the morphogenesis of tissues as diverse as the flywingandthevertebrate spinal cord (Wolpert 1996). However, not until recently has evidenceemerged that the mammalian cerebral cortex is patterned, at least in part, by thesame types of mechanisms (Bishop et al. 2000, Bishop et al. 2002, Ragsdale &Grove 2001, Garel et al. 2003, Mallamaci et al. 2000, Rubenstein et al. 1999).The pattern to be explained is the division of the cerebral cortex into anatomicallydistinct and functionally specialized areas, which form a species-specific areamap (Nauta & Feirtag 1986). The mechanisms that initiate map formation indevelopment have been elusive. Here, we review classic and recent studies thatopen up the problem in a new way.Other structures are patterned in the embryo by signaling centers that lieat the boundaries of the tissue to be patterned. These centers release signaling0147-006X/03/0721-0355$14.00 355356GROVEFUKUCHI-SHIMOGORIproteins that regulate regional growth and specify regional identity in the tissue.In an influential model, signaling proteins called morphogens diffuse through thetissue and establish a gradient that directly confers positional information (Wolpert1996). Not all secreted signaling molecules involved in patterning are morphogens,however. For example, in the developing spinal cord, Wingless-Int (WNT)1 and3a form a concentration gradient of protein but one that coordinates tissue growth,a different element of patterning (Megason & McMahon 2002).In some systems, such as the Drosophila embryo or the embryonic vertebratespinal cord, it is well established that cells respond to different levels of a mor-phogen by expressing specific transcription factors (Wolpert 1996). These region-ally expressed transcription factors in turn control regional expression of down-stream genes that regulate local differentiation of the tissue (Briscoe & Ericson2001). Here we review new findings that begin to fit development of the corticalarea map to such a model.Development of the area map, however, cannot be understood by focusingon the cortical primordium in isolation. In the adult, the cerebral cortex can beviewed, both by connections and function, to be at the apex of a hierarchy of brainstructures (Nauta & Feirtag 1986). Because embryonic tissues interact to patternone another, other developing brain structures would be expected to influencecortical pattern. In particular, a key feature of the cortical area map is that differentareas receive distinct sets of projections from different thalamic nuclei, relayinginformation from the periphery and other parts of brain (Nauta & Feirtag 1986).We survey selected recent studies of whether thalamic axons and their activityinitiate, maintain, or regulate select features of cortical area identity. Meanwhile,differential thalamocortical innervation across the cortex is itself a major partof cortical patterns. We summarize findings that molecular cues in the cortexguide thalamic axons but that patterning in the thalamus itself and cues along thethalamocortical pathway are also critical to initiating this component of corticalpattern. Taken as a whole, available data suggest that the cerebral cortex andthalamus are first patterned independently and then coordinate and interact togenerate the mature cortical area map.CLASSIC MODELS OF CORTICAL PATTER NINGOver the past several years, two classic models have dominated research into thedevelopment of the cortical area map (Rakic 1988, O’Leary 1989) (Figure 1).The P rotomap ModelIn the protomap model of area map formation (Rakic 1988), the cortical pri-mordium is patterned as it is generated. Although this model was proposed beforerecent advances in understanding molecular mechanisms of embryonic patterning,it already implied that the cortex is like other parts of the embryo, patterned ascells are dividing. Intrinsic area differences, specified by molecular determinants,PATTERNING THE CEREBRAL CORTEX 357are set up in the ventricular zone (VZ), the germinal cell layer of the cortical pri-mordium. As newborn neurons migrate out of the VZ in radial arrays they carrythe area protomap with them to form the cortical plate (CP), the incipient greymatter of the cortex.The P rotoco rtex ModelThe cytoarchitectonic features classically used to define areas appear relatively latein corticogenesis; moreover, some early cortical transplant experiments suggestedprolonged plasticity of area identity (O’Leary 1989, Schlaggar & O’Leary 1991).These observations support the protocortex model (O’Leary 1989), in which thecortical primordium is essentially homogeneous as it is generated and is patternedinto areas later by cues from axons growing in from the thalamus. Patterningmechanisms in c ortex are thus somewhat distinct from those in the