Overview
The questions of how an embryo decides where to place its organs (“positional information”) and how these organs are correctly organized into functional three dimensional structures (“morphogenesis”) are of fundamental importance. We study these processes in the frog, Xenopus, and in the zebrafish, Danio. We have two major areas of interest: the nervous system, including very early patterning events as well as later events that build the three dimensional structure of the brain, and the extreme anterior of the embryo that forms the primary mouth, and is an evolutionarily conserved and important region. Frog and fish embryos are ideal for these studies, since the events we analyze take place very early in development, when mammalian embryos are tiny and inaccessible. Genes that are important for frog and fish embryogenesis are conserved in mammals, and our research is therefore relevant for understanding normal and abnormal human development.

Research Summary
Mechanism by which inhibitors of BMP signaling activate neural determination. We have shown that in fish and frogs, the embryo decides to make a nervous system by the onset of gastrulation. This is a very early decision, corresponding to a two and one half week old human embryo. Expression of the transcription factor zic1 at the onset of gastrulation is one of the earliest molecular indicators of neural fate determination in Xenopus. Inhibition of bone morphogenetic protein (BMP) signaling is critical for activation of zic1 expression and fundamental for establishing neural identity in both vertebrates and invertebrates. The mechanism by which interruption of BMP signaling activates neural-specific gene expression is not understood. We have identified of a 215 bp genomic module that is both necessary and sufficient to activate Xenopus zic1 transcription upon interruptionof BMP signaling. Transgenic analyses demonstrate that this BMP inhibitory response module (BIRM) is required for appropriate spatial and temporal expression in the whole embryo. Multiple consensus binding sites for specific transcription factor families within the BIRM are required for its activity and some of these regionsare phylogenetically conserved between orthologous vertebrate zic1 genes. These data suggest that interruption of BMP signaling facilitates neural determination via a complex mechanism, involving multiple regulatory factors that cooperate to control zic1 expression.

Hindbrain patterning. The embryonic hindbrain gives rise to the cerebellum and medulla. In the embryonic zebrafish hindbrain, we have shown that cells in the posterior hindbrain acquire unique anterior/posterior (A/P) identity when they both express the homeodomain transcription factor vhnf1 and are exposed to fibroblast growth factor (FGF) signals. Each of these factors alone is required for formation of other tissues, but together they regulate a distinct set of genes that specify rhombomeres 5 + 6 (r5+6), including valentino (val), krox20 and hoxB3. In order to understand better the genetic program that is initiated by the combination of vhnf1 and FGF signals, we have tried to recapitulate the determination of rhombomeres 5+6 in explants of zebrafish embryonic ectoderm. We have obtained the exciting result that expression of vhnf1 + FGF is sufficient to activate val and krox20 expression in this simple explant system. We have performed expression microarray experiments in which we compare genes regulated by the two factors together to the targets of each factor alone. From these assays, we have identified two new genes that are targets of vhnf1 expression and may play a pivotal role in regulation of posterior hindbrain determination.

Brain ventricle morphogenesis. A unique and conserved feature of the vertebrate nervous system is that it is tubular. Why is this so? One answer may be that a tubular structure allows exposure of the nervous system to different “inside” and “outside” environments. The inside of the neural tube forms a system of cavities, the brain ventricles. Brain ventricles are required for normal brain development and function, and ventricular abnormalities are apparent in several neurodevelopmental disorders, including autism, indicating that this system is essential for normal mental health. Several functions have been assigned to the adult brain ventricular system and the cerebrospinal fluid it contains, however, embryonic development and function of this system are poorly studied. We have begun to analyze brain ventricle development, using the zebrafish as a model, and taking a forward genetic approach. The zebrafish is an excellent system for this study as imaging the brain in live embryos is feasible, and as we have collected more than thirty brain ventricle mutants, identified in several mutagenesis screens, but not generally studied further.We have shown that the neural tube expands into primary forebrain, midbrain, and hindbrain ventricles rapidly, over a four-hour period during mid-somitogenesis. Two mutants that do not develop brain ventricles are nagie oko and snakehead. Mutants in nagie oko, which encodes a MAGUK family protein, fail to undergo ventricle morphogenesis. This correlates with an abnormal brain neuroepithelium, with no clear midline and disrupted junctional protein expression. In contrast, the snakehead neural tube undergoes normal ventricle morphogenesis, however the ventricles do not inflate, likely due to impaired ion transport and disrupted osmotic gradient formation. We have shown that snakehead is allelic to small heart, which has a mutation in the Na+K+ ATPase gene atp1a1a.1. We are using transplantation techniques to generate mosaic snakehead embryos to examine the tissue requirement for Atp1a1a.1 function during brain ventricle formation. This study defines three steps required for brain ventricle development and that occur independently of circulation: 1) morphogenesis of the neural tube, requiring nok function; 2) lumen inflation requiring atp1a1a.1 function, and 3) localized cell proliferation. We suggest that mechanisms of brain ventricle development are conserved throughout the vertebrates.

The extreme anterior is a conserved region. In most animals, the three first cell types to form, the ectoderm, mesoderm and endoderm, are present in all parts of the embryo, with ectoderm on the outside, endoderm on the inside and mesoderm sandwiched between them. However, at the extreme front (anterior) of deuterostome embryos, such as sea urchins and all vertebrates including humans, there is a mesoderm-free region, where the ectoderm and endoderm are directly juxtaposed. It is our hypothesis that this region is an evolutionary relic of the diploblasts, animals with only ectodermal and endodermal germ layers. This anterior region forms the first opening between the gut and the outside, the primary mouth, which may be the most evolutionarily conserved part of the head. In amphibians, this region also forms the mucus-secreting cement gland.

Development of the primary mouth. The extreme anterior of the deuterostome embryo is unique in that ectoderm and endoderm are directly juxtaposed, without intervening mesoderm. From this region the initial opening between the gut and outside of the embryo breaks through. We have termed this the primary mouth, which is essential for eating and therefore life. In vertebrates, the neural crest grows around the primary mouth to form the face and a secondary mouth opening must therefore form. The primary mouth then becomes the pharyngeal opening. In order to establish a molecular understanding of this important process we have examined primary mouth formation during Xenopus development. We find that multiple steps are involved. An early step involves dissolution of the basal lamina separating ectoderm and endoderm. A subsequent step requires cell death in the ectoderm. Later, the ectodermal and endodermal layers overcome their normally separate identities and intercalate to generate a single cell layer. The final step is perforation, where the primary mouth breaks through. From fate mapping, we have defined the ectodermal and endodermal regions that will form the primary mouth. Extirpations and transplants indicate that, surprisingly, ectoderm from any region of the embryo can be used to generate the primary mouth. In contrast, endoderm specifically in the presumptive primary mouth region provides essential signals. We have performed expression microarray analysis to define genes expressed during primary mouth formation, and have defined at least two that are required for this process.

The cement gland as an anterior paradigm. The amphibian cement gland is a mucus-secreting epithelium that forms from extreme anterior ectoderm, and that we have used to analyze positional cues. We have proposed that the overlap of anterodorsal identity (AD), ventrolateral identity (VL), and ectodermal outer layer identity (EO) results in determination of the cement gland primordium, formulated in the equation A+VL+EO=CG. We are constructing the hierarchy of transcription factors involved in cement gland determination. Anterior identity is conferred by the transcription factor Otx2, which is necessary and sufficient to activate the cement gland program. Using hormone-inducible fusion proteins, we have shown that the homeodomain genes pitx1 and pitx2C lie immediately downstream of Otx2. Neither Otx2, not Pitx proteins are able to directly activate cement gland-specific gene expression, but require expression of intervening or additional genes. In order to identify these, we have examined the cis-acting elements by which transcription of two related genes is activated specifically in the cement gland. These are Xag1 and gob4, which are members of the conserved agr gene family we have defined. Ets and ATF/CREB sites are required for expression of Xag1 in the cement gland. We have also defined a 140bp region of the gob4 promoter that is necessary and sufficient for robust cement gland specific expression. Linker scan and point mutagenesis implicate AP1 sites as crucial for gob4 regulation. Binding sites for AP1 and ATF/CREB factors are likely to bind related bZip family members. Thus, comparison of these promoters suggests that the AP1 superfamily is pivotal in directing gene expression specifically to the cement gland.

Selected Publications
Lowery, L.A. and Sive, H. Initial formation of zebrafish brain ventricles occurs independently of circulation and requires the nagie oko and snakehead/atp1a1a.1 gene products. Development, 132:2057-2067 (2005).

Sagerstrom, C.G., Gammill, L.S., Veale, R. and Sive, H. Specification of the enveloping layer and lack of autoneuralization in zebrafish embryonic explants. Dev. Dynamics, 232:85-97 (2005).

Wiellette, E., Grinblat, Y., Austen, M., Hirsinger, E., Amsterdam, A., Walker, C., Westerfield, M. and Sive, H. Combined haploid and insertional mutation screen in the zebrafish. Genesis, 40:231-240 (2004).

Lowery, L.A. and Sive, H. Strategies of vertebrate neurulation and a re-evaluation of teleost neural tube formation. Mech. Dev., 121:1189-1197 (2004).

Wiellette, E.L. and Sive, H. vhnf1 and FGF signals synergize to specify rhombomere identity in the zebrafish hindbrain. Development, 130:3821-3829 (2003).

Tropepe, V. and Sive, H. Can zebrafish be used as a model to study the neurodevelopmental causes of autism? Genes, Brain and Behavior, 2:268-281 (2003).

Wardle, F. and Sive, H. What's your position: The Xenopus cement gland as a paradigm of regional specification. BioEssays, 25:717-726 (2003).

Wardle, F., Wainstock, D. and Sive, H. Cement gland-specific activation of the Xag1 promoter is regulated by co-operation of putative Ets and ATF/CREB transcription factors. Development, 129:4387-4397 (2002).

Gamse, J. and Sive, H. Early anteroposterior division of the presumptive neurectoderm in Xenopus. Mech. Dev., 104:21-36 (2001).

Search PubMed for Sive Lab publications.