Hazel L. Sive

Scholar: 1992

Awarded Institution
Massachusetts Institute of Technology
Whitehead Institute


Research Interests

Developmentof the Embryonic Axes in Vertebrates

We study formation of the anteroposterior (A/P) axis in the frog, Xenopus laevis, and in the zebrafish, Danio rerio. Our focus is on the earliest events involved in patterning the dorsal ectoderm, that gives rise to all neural tissue (including the brain, spinal cord and sense organs) and to the anterior cement gland (the chin primordium). Frog and fish embryos are ideal for these studies, since the decision to form a nervous system takes place very early in development, when mammalian embryos are tiny and inaccessible. Molecules that are important for frog and fish embryogenesis are also important in mammals, and our research is therefore relevant for understanding normal and abnormal human development.

Cell interactions during ectodermal patterning. The zebrafish, Danio, is an excellent subject for genetic analyses of development. However, a knowledge of basic zebrafish embryology is lacking. We have developed an assay for asking what cell interactions ("inductions") are involved in patterning the zebrafish nervous system. In these "explant" assays, small pieces of tissue from one region of the embryo are cultured alone or together with tissue from a different region. After culture, the cell types that differentiate are examined and provide information about cell interactions that control tissue fate. Using these assays we have, for the first time, demonstrated neural induction in zebrafish, and have shown that anterior and posterior neural determination are separable events. In addition to providing key information about normal development, our assays will help analyses of mutant embryos, showing whether a mutation maps to an inducing or to a responding tissue.

The timing of anteroposterior patterning. A major unanswered question in vertebrate development is when does the A/P ectodermal pattern form? This has been difficult to answer because inital induction and patterning (or "determination") of the neurectoderm occur many hours before the differentiated nervous system is visible. Classical experiments in Amphibians showed that induction was underway by the middle of "gastrulation" (the cell movements that form the embryonic "germ layers"), however, these experiments were rather crude. We decided to answer this question by isolating genes that are very early markers of A/P patterning. Using the powerful technique of subtractive cloning, we isolated from Xenopus more than forty genes that are expressed during gastrulation in induced dorsal ectoderm but not in uninduced ectoderm. Using these genes as markers, we have shown that neurectodermal induction has occurred by early gastrula, many hours earlier than was previously thought, while, by mid-gastrula a highly complex A/P pattern is in place. Currently, we are constructing molecular maps of the ectoderm at different stages of gastrulation to unequivocally identify steps in A/P patterning. A parallel study is underway in zebrafish, since we do not know whether patterning occurs at a similar time in fish as it does in frogs.

The molecular basis for determination. As the A/P pattern forms, a large spectrum of regulatory genes is induced in the dorsal ectoderm to control subsequent region-specific differentiation. From our subtractive screen, we have identified many new proteins that are excellent candidates for regulators of ectodermal patterning. One of these is a zinc finger protein that we have called opl. opl is the earliest neural-specific marker yet isolated in any vertebrate. opl can activate neural differentiation and may be a primary regulator of neural fate. Further, opl profoundly alters morphogenesis of the nervous system and may be a key regulator of neural tube formation.

Inducers of ectodermal patterning. The A/P ectodermal pattern is induced by signals originating from the dorsal mesoderm. One ectodermal inducer is retinoic acid. Exogenous retinoids induce posterior (but not anterior) neural tissue, while a dominant negative retinoid receptor mutant prevents endogenous retinoid function and ablates expression of the gene HoxD1 that is a posterior neural marker. Multiple other inducers exist, and in an ongoing screen we have identified many potential ectodermal inducers.

The cement gland as an anterior paradigm. The cement gland is a simple, mucus-secreting organ that marks the chin primordium and is an excellent paradigm for determining why an organ forms at a particular position in the embryo. We have found that three positive inducing signals and two negative (inhibitory) signals position the cement gland. We are currently searching for inducer molecules that can modulate cement gland formation. We are also defining genes expressed within the ectoderm that regulate cement gland formation. One such gene, is the homeodomain gene otx2 that activates formation of ectopic cement gland when misexpressed in embryos. In order to regulate the timing of otx2 expression, we used a system of hormone-inducible fusion proteins that we have shown is very effective in embryos. Our data shows that the ectoderm is able ("competent") to respond to otx2 and form a cement gland only during a limited period of development and within a limited region of the embryo, helping to explain the precise positioning of the cement gland.