Notch, neural stem cells, and brain tumors.

The Notch pathway has a fundamental role during cell-fate specification in the developing mammalian nervous system. During neocortical development, Notch signaling inhibits neuronal differentiation and maintains the neural stem/progenitor cell pool to permit successive waves of neurogenesis, which are followed by gliogenesis. In addition, recent evidence suggests that Notch signaling is not uniformly used among distinct proliferative neural cells types, with the canonical cascade functional in neural stem cells but attenuated in neurogenic progenitors. Although the role of Notch in neural development is increasingly well understood, it has recently become evident that Notch also has a role in brain tumor biology. Notch receptors are overexpressed in many different brain tumor types, and they may have an initiating role in some. Stem-like cells in brain tumors share many similarities with neural stem/progenitor cells and may require Notch for their survival and growth. Understanding the role of Notch signaling in neoplastic and non-neoplastic stem/progenitor populations will advance our understanding of basic principles regulating developmental and stem cell biology and may also lead to more effective therapies for brain tumors.

Notch3 signaling initiates choroid plexus tumor formation.

Notch3 has been studied in the context of brain development, but whether it plays a role in the formation of brain tumors is unclear. We demonstrate that the introduction of constitutively active Notch3 into periventricular cells of embryonic day 9.5 mice causes the formation of choroid plexus tumors (CPTs). Tumors arose in the fourth ventricles in 83% of animals and were associated with hydrocephalus. They were microscopically highly similar to choroid plexus papillomas in humans, with an ongoing proliferation rate of 4-6%. Signs of Notch pathway activity were also present in human choroid plexus lesions, and receptor mRNA levels in papillomas were elevated over those in non-neoplastic choroid plexus. Notch2 was overexpressed approximately 500-fold in one case, suggesting that the role of this pathway in CPTs may not be specific to Notch3. Our findings indicate that activated Notch3 can function as an oncogene in the developing brain, and link the Notch pathway to human CPT pathogenesis.

N-terminal fatty-acylation of sonic hedgehog enhances the induction of rodent ventral forebrain neurons.

The adult basal ganglia arise from the medial and lateral ganglionic eminences, morphologically distinct structures found in the embryonic telencephalon. We have previously shown that temporal changes in sonic hedgehog (Shh) responsiveness determine the sequential induction of embryonic neurons that populate the medial and lateral ganglionic eminences. In this report, we show that Shh-mediated differentiation of neurons that populate the lateral ganglionic eminence express different combinations of the homeobox-containing transcription factors Dlx, Mash1 and Islet 1/2. Furthermore, we show that N-terminal fatty-acylation of Shh significantly enhances its ability to induce the differentiation of rat E11 telencephalic neurons expressing Dlx, Islet 1/2 or Mash1. Recent evidence indicates that in utero injection of the E9.5 mouse forebrain with retroviruses encoding wild-type Shh induces the ectopic expression of Dlx2 and severe deformities in the brain. In this report, we show that Shh containing a mutation at the site of acylation prevents either of these phenotypes. These results suggest that N-terminal fatty-acylation of Shh may play an important role in Shh-dependent signaling during rodent ventral forebrain formation.

An acylatable residue of Hedgehog is differentially required in Drosophila and mouse limb development.

The Drosophila Hedgehog protein and its vertebrate counterpart Sonic hedgehog are required for a wide variety of patterning events throughout development. Hedgehog proteins are secreted from cells and undergo autocatalytic cleavage and cholesterol modification to produce a mature signaling domain. This domain of Sonic hedgehog has recently been shown to acquire an N-terminal acyl group in cell culture. We have investigated the in vivo role that such acylation might play in appendage patterning in mouse and Drosophila; in both species Hedgehog proteins define a posterior domain of the limb or wing. A mutant form of Sonic hedgehog that cannot undergo acylation retains significant ability to repattern the mouse limb. However, the corresponding mutation in Drosophila Hedgehog renders it inactive in vivo, although it is normally processed. Furthermore, overexpression of the mutant form has dominant negative effects on Hedgehog signaling. These data suggest that the importance of the N-terminal cysteine of mature Hedgehog in patterning appendages differs between species.

Spatiotemporal selectivity of response to Notch1 signals in mammalian forebrain precursors.

The olfactory bulb, neocortex and archicortex arise from a common pool of progenitors in the dorsal telencephalon. We studied the consequences of supplying excess Notch1 signal in vivo on the cellular and regional destinies of telencephalic precursors using bicistronic replication defective retroviruses. After ventricular injections mid-neurogenesis (E14.5), activated Notch1 retrovirus markedly inhibited the generation of neurons from telencephalic precursors, delayed the emergence of cells from the subventricular zone (SVZ), and produced an augmentation of glial progeny in the neo- and archicortex. However, activated Notch1 had a distinct effect on the progenitors of the olfactory bulb, markedly reducing the numbers of cells of any type that migrated there. To elucidate the mechanism of the cell fate changes elicited by Notch1 signals in the cortical regions, short- and long-term cultures of E14.5 telencephalic progenitors were examined. These studies reveal that activated Notch1 elicits a cessation of proliferation that coincides with an inhibition of the generation of neurons. Later, during gliogenesis, activated Notch1 triggers a rapid cellular proliferation with a significant increase in the generation of cells expressing GFAP. To examine the generation of cells destined for the olfactory bulb, we used stereotaxic injections into the early postnatal anterior subventricular zone (SVZa). We observed that precursors of the olfactory bulb responded to Notch signals by remaining quiescent and failing to give rise to differentiated progeny of any type, unlike cortical precursor cells, which generated glia instead of neurons. These data show that forebrain precursors vary in their response to Notch signals according to spatial and temporal cues, and that Notch signals influence the composition of forebrain regions by modulating the rate of proliferation of neural precursor cells.

The Gsh2 homeodomain gene controls multiple aspects of telencephalic development.

Homeobox genes have recently been demonstrated to be important for the proper patterning of the mammalian telencephalon. One of these genes is Gsh2, whose expression in the forebrain is restricted to the ventral domain. In this study, we demonstrate that Gsh2 is a downstream target of sonic hedgehog and that lack of Gsh2 results in profound defects in telencephalic development. Gsh2 mutants have a significant decrease in the expression of numerous genes that mark early development of the lateral ganglionic eminence, the striatal anlage. Accompanying this early loss of patterning genes is an initial expansion of dorsal telencephalic markers across the cortical-striatal boundary into the lateral ganglionic eminence. Interestingly, as development proceeds, there is compensation for this early loss of markers that is coincident with a molecular re-establishment of the cortical-striatal boundary. Despite this compensation, there is a defect in the development of distinct subpopulations of striatal neurons. Moreover, while our analysis suggests that the migration of the ventrally derived interneurons to the developing cerebral cortex is not significantly affected in Gsh2 mutants, there is a distinct delay in the appearance of GABAergic interneurons in the olfactory bulb. Taken together, our data support a model in which Gsh2, in response to sonic hedgehog signaling, plays a crucial role in multiple aspects of telencephalic development.

Radial glial identity is promoted by Notch1 signaling in the murine forebrain.

In vertebrates, Notch signaling is generally thought to inhibit neural differentiation. However, whether Notch can also promote specific early cell fates in this context is unknown. We introduced activated Notch1 (NIC) into the mouse forebrain, before the onset of neurogenesis, using a retroviral vector and ultrasound imaging. During embryogenesis, NIC-infected cells became radial glia, the first specialized cell type evident in the forebrain. Thus, rather than simply inhibiting differentiation, Notch1 signaling promoted the acquisition of an early cellular phenotype. Postnatally, many NIC-infected cells became periventricular astrocytes, cells previously shown to be neural stem cells in the adult. These results suggest that Notch1 promotes radial glial identity during embryogenesis, and that radial glia may be lineally related to stem cells in the adult nervous system.

A method for rapid gain-of-function studies in the mouse embryonic nervous system.

We used ultrasound image-guided injections of high-titer retroviral vectors to obtain widespread introduction of genes into the mouse nervous system in utero as early as embryonic day 8.5 (E8.5). The vectors used included internal promoters that substantially improved proviral gene expression in the ventricular zone of the brain. To demonstrate the utility of this system, we extended our previous work in vitro by infecting the telencephalon in vivo as early as E8. 5 with a virus expressing Sonic Hedgehog. Infected embryos showed gross morphological brain defects, as well as ectopic expression of ventral telencephalic markers characteristic of either the medial or lateral ganglionic eminences.

Transplantation as a tool to study progenitors within the vertebrate nervous system.

In recent years, many studies have focused on the fate and potential of neural progenitors in vertebrates. While much progress has been made, many questions remain about the mechanisms which lead to neural diversity, in terms of both the regionalization of the nervous system and specification of cell fates within those regions. Studies aimed at addressing these questions have fallen into three main categories: in vivo lineage tracings, in vitro differentiation analyses, and in vivo cell transplantation studies. This body of work has pointed to the existence of both pluripotent and unipotent neural progenitors, and has suggested that both cell intrinsic and extrinsic cues play a role in the determination of neural cell fate. In addition, the existence of neural "stem cells" maintained into adulthood has been suggested. This review will focus on transplantation studies in mammals, and will emphasize how this method has been useful as a means of determining the changing potential of neural precursors and their environments within the developing nervous system.

Insertional mutagenesis in zebrafish identifies two novel genes, pescadillo and dead eye, essential for embryonic development.

Recently our laboratory described an efficient method for generating retroviral provirus insertions in the zebrafish germ line, and we showed that provirus insertions induce embryonic mutations at a frequency of roughly one mutant per 70 insertions. To date we have isolated four insertional mutants and, using the proviruses as a molecular tag, have cloned the genes disrupted in three of them. The proviruses in all three mutants lie within or just 5' of the first coding exon, point in the opposite transcriptional orientation from the gene, and disrupt transcription. Here we present a molecular characterization of two genes identified by this method and describe the associated mutant phenotypes. The pescadillo (pes) gene is predicted to encode a protein of 582 amino acids with no recognizable functional motifs, which is highly conserved from yeast to humans. pes mRNA is expressed widely and dynamically during the first 3 days of embryogenesis. Prominent sites of expression are the eyes and optic tectum on day 1, the fin buds, liver primordium, and gut on day 2, and the branchial arches on day 3. Beginning at day 3 of embryogenesis, pes mutant embryos exhibit small eyes, a reduced brain and visceral skeleton, shortened fins, and a lack of expansion of the liver and gut, and then die on the sixth day of development. The dead eye (dye) gene encodes a protein of 820 amino acids that is homologous to genes of unknown function in human, mouse, and Xenopus, and that has weak homology with the yeast NIC96 (nucleoporin-interacting component) gene. dye mutants can be recognized on day 2 of embryogenesis by the presence of necrotic cells in the tectum and eyes. dye mutants die on day 5 of development. These results demonstrate the power of insertional mutagenesis in zebrafish for rapidly finding and characterizing novel genes essential for embryonic development. Using our current methodology, we estimate that our laboratory could screen approximately 25,000 insertions in 2-3 years, identifying perhaps 250-350 embryonic lethal genes. Assuming that all genes are accessible to proviral insertion, the wider application of this approach could lead to the rapid identification of the majority of genes that are required for embryonic development of this vertebrate.

Insertional mutagenesis and rapid cloning of essential genes in zebrafish.

Large-scale chemical mutagenesis screens in zebrafish have led to the isolation of thousands of lethal mutations in genes that are essential for embryonic development. However, the cloning of these mutated genes is difficult at present as it requires positional cloning methods. In Drosophila, chemical mutagenesis screens were complemented with P-element insertional mutagenesis which facilitated the cloning of many genes that had been identified by chemical lesions. To facilitate the cloning of vertebrate genes that are important during embryogenesis, we have developed an insertional mutagenesis strategy in zebrafish using a retroviral vector. Here, in a pilot screen of 217 proviral insertions, we obtained three insertional mutants with embryonic lethal phenotypes, and identified two of the disrupted genes. One of these, no arches, is essential for normal pharyngeal arch development, and is homologous to the recently characterized Drosophila zinc-finger gene, clipper, which encodes a novel type of ribonuclease. As it is easy to generate tens to hundreds of thousands of proviral transgenes in zebrafish, it should now be possible to use this screening method to mutate and then rapidly clone a large number of genes affecting vertebrate developmental and cellular processes.

Highly efficient germ-line transmission of proviral insertions in zebrafish.

An important technology in model organisms is the ability to make transgenic animals. In the past, transgenic technology in zebrafish has been limited by the relatively low efficiency with which transgenes could be generated using either DNA microinjection or retroviral infection. Previous efforts to generate transgenic zebrafish with retroviral vectors used a pseudotyped virus with a genome based on the Moloney murine leukemia virus and the envelope protein of the vesicular stomatitis virus. This virus was injected into blastula-stage zebrafish, and 16% of the injected embryos transmitted proviral insertions to their offspring, with most founders transmitting a single insertion to approximately 2% of their progeny. In an effort to improve this transgenic frequency, we have generated pseudotyped viral stocks of two new Moloney-based genomes. These viral stocks have titers up to two orders of magnitude higher than that used previously. Injection of these viruses resulted in a dramatic increase in transgenic efficiency; over three different experiments, 83% (110/133) of the injected embryos transmitted proviral insertions to 24% of their offspring. Furthermore, founders made with one of the viruses transmitted an average of 11 different insertions through their germ line. These results represent a 50- to 100-fold improvement in the efficiency of generating transgenic zebrafish, making it now feasible for a single lab to rapidly generate tens to hundreds of thousands of transgenes. Consequently, large-scale insertional mutagenesis strategies, previously limited to invertebrates, may now be possible in a vertebrate.

Integration and germ-line transmission of a pseudotyped retroviral vector in zebrafish.

The zebrafish is rapidly becoming a popular model system for the study of vertebrate development because it is ideal for both embryological studies and genetic analysis. To determine if a retroviral vector pseudotyped with the envelope glycoprotein of the vesicular stomatitis virus could infect zebrafish embryos, and in particular, the cells destined to become the germ line, a pseudotyped virus was injected into blastula-stage zebrafish embryos. Fifty-one embryos were allowed to develop and eight transmitted proviral DNA to their progeny. Founders were mosaic, but as expected, transgenic F1's transmitted proviral DNA in a Mendelian fashion to the F2 progeny. Transgenic F1 fish inherited a single integrated provirus, and a single founder could transmit more than one viral integration to its progeny. These results demonstrate that this pantropic pseudotyped vector, originally developed for human gene therapy, will make the use of retroviral vectors in zebrafish possible.