By homologous recombination of an internal ribosome entry site and Cre recombinase coding region into the 3'-untranslated region of the mouse Emx1 gene, we have generated a strain of mice, Emx1(IRES)cre, that expresses the Cre recombinase in a spatial and temporal pattern like that observed for Emx1. When mated to reporter strains, these mice are a sensitive means to fate-map the Emx1-expressing cells of the developing forebrain. Our results demonstrate that radial glia, Cajal-Retzius cells, glutamatergic neurons, astrocytes, and oligodendrocytes of most pallial structures originate from an Emx1-expressing lineage. On the other hand, most of the pallial GABAergic neurons arise outside the Emx1-expressing lineage. Structures that are located near the basal ganglia (e.g., the amygdala and endopiriform nuclei) are not uniformly derived from Emx1-expressing cells.

The striatum has a central role in many neurobiological processes, yet little is known about the molecular control of its development. Inroads to this subject have been made, due to the discovery of transcription factors, such as the Dlx genes, whose expression patterns suggest that they have a role in striatal development. We report that mice lacking both Dlx-1 and Dlx-2 have a time-dependent block in striatal differentiation. In these mutants, early born neurons migrate into a striatum-like region, which is enriched for markers of the striosome (patch) compartment. However, later born neurons accumulate within the proliferative zone. Several lines of evidence suggest that mutations in Dlx-1 and Dlx-2 produce abnormalities in the development of the striatal subventricular zone and in the differentiation of striatal matrix neurons.

The Dlx homeobox gene family is expressed in a complex pattern within the embryonic craniofacial ectoderm and ectomesenchyme. A previous study established that Dlx-2 is essential for development of proximal regions of the murine first and second branchial arches. Here we describe the craniofacial phenotype of mice with mutations in Dlx-1 and Dlx-1 and -2. The skeletal and soft tissue analyses of mice with Dlx-1 and Dlx-1 and -2 mutations provide additional evidence that the Dlx genes regulate proximodistal patterning of the branchial arches. This analysis also elucidates distinct and overlapping roles for Dlx-1 and Dlx-2 in craniofacial development. Furthermore, mice lacking both Dlx-1 and -2 have unique abnormalities, including the absence of maxillary molars. Dlx-1 and -2 are expressed in the proximal and distal first and second arches, yet only the proximal regions are abnormal. The nested expression patterns of Dlx-1, -2, -3, -5, and -6 provide evidence for a model that predicts the region-specific requirements for each gene. Finally, the Dlx-2 and Dlx-1 and -2 mutants have ectopic skull components that resemble bones and cartilages found in phylogenetically more primitive vertebrates.

In the developing spinal cord, early progenitor cells of the oligodendrocyte lineage are induced in the motor neuron progenitor (pMN) domain of the ventral neuroepithelium by the ventral midline signal Sonic hedgehog (Shh). The ventral generation of oligodendrocytes requires Nkx6-regulated expression of the bHLH gene Olig2 in this domain. In the absence of Nkx6 genes or Shh signaling, the initial expression of Olig2 in the pMN domain is completely abolished. In this study, we provide the in vivo evidence for a late phase of Olig gene expression independent of Nkx6 and Shh gene activities and reveal a brief second wave of oligodendrogenesis in the dorsal spinal cord. In addition, we provide genetic evidence that oligodendrogenesis can occur in the absence of hedgehog receptor Smoothened, which is essential for all hedgehog signaling.

Genetic analysis of the development and evolution of the vertebrate head is at a primitive stage. Many homeo box genes, including the Distal-less family, are potential regulators of head development. To determine the function of Dlx-2, we generated a null mutation in mice using gene targeting. In homozygous mutants, differentiation within the forebrain is abnormal and the fate of a subset of cranial neural crest cells is respecified. The latter causes abnormal morphogenesis of the skeletal elements derived from the proximal parts of the first and second branchial arches. We hypothesize that the affected skull bones from the first arch have undergone a transformation into structures similar to those found in reptiles. These results show that Dlx-2 controls development of the branchial arches and the forebrain and suggests its role in craniofacial evolution.

Serotonergic (5-HT) neurons in the brainstem modulate a wide range of physiological processes and behaviors. Two transcription factor genes, Pet-1 and Nkx2.2, are necessary but not sufficient to specify the 5-HT transmitter phenotype. Here we show that the Lim class homeobox gene Lmx1b is required for proper formation of the entire 5-HT system in the hindbrain, as indicated by the loss of expression of genes necessary for serotonin synthesis and transport in Lmx1b null mice. Lmx1b and Pet1 act downstream of Nkx2.2, and their expression is independently regulated at the time when 5-HT transmitter phenotype is specified. Ectopic expression of Lmx1b plus Pet-1 is able to induce formation of 5-HT cells in the most ventral spinal cord, where Nkx2.2 is normally expressed. Combined expression of all three genes, Lmx1b, Pet-1, and Nkx2.2, drives 5-HT differentiation in the dorsal spinal cord. Our studies therefore define a molecular pathway necessary and sufficient to specify the serotonergic neurotransmitter phenotype.

Activation of the unfolded protein response (UPR) is involved in the pathogenesis of numerous CNS myelin abnormalities; yet, its direct role in traumatic spinal cord injury (SCI)-induced demyelination is not known. The UPR is an evolutionarily conserved cell defense mechanism initiated to restore endoplasmic reticulum homeostasis in response to various cellular stresses including infection, trauma, and oxidative damage. However, if uncompensated, the UPR triggers apoptotic cell death. We demonstrate that the three signaling branches of UPR including the PERK, ATF6, and IRE1α are rapidly initiated in a mouse model of contusive SCI specifically at the injury epicenter. Immunohistochemical analyses of the various UPR markers revealed that in neurons, the UPR appeared at 6 and 24-h post-SCI. In contrast, in oligodendrocytes and astroglia, UPR persisted at least for up to 3 days post-SCI. The UPR-associated proapoptotic transcriptional regulator CHOP was among the UPR markers upregulated in neurons and oligodendrocytes, but not in astrocytes, of traumatized mouse spinal cords. To directly analyze its role in SCI, WT and CHOP null mice received a moderate T9 contusive injury. Deletion of CHOP led to an overall attenuation of the UPR after contusive SCI. Furthermore, analyses of hindlimb locomotion demonstrated a significant functional recovery that correlated with an increase in white-matter sparing, transcript levels of myelin basic protein, and Claudin 11 and decreased oligodendrocyte apoptosis in CHOP null mice in contrast to WT animals. Thus, our study provides evidence that the UPR contributes to oligodendrocyte loss after traumatic SCI.

No abstract