Shaping destiny

The idea of one group of cells telling another what type of tissue it should become — a phenomenon known as induction — is a familiar theme in embryonic development. One of the earliest instances of this process is the formation of mesoderm, which, along with ectoderm and endoderm, is one of the three main tissue layers in early embryos. The discovery that mesoderm is induced from ectoderm, under instructions from endoderm, dates back to a seminal publication in 1969. It would be another 20 years, however, before researchers would understand what form those instructions take.

The amphibian embryo is a popular model organism for developmental biology research. At early (blastula) stages, the embryo consists of two relatively easily distinguishable parts — the 'animal' and 'vegetal' hemispheres. When excised and allowed to develop alone, the animal part develops into ectodermal tissues, whereas the vegetal half forms endodermis. But how does mesoderm form?

P. D. Nieuwkoop tackled this problem through a series of meticulous dissection experiments. By excising different portions of Ambystoma mexicanum (axolotl) blastulae, and culturing them alone or in combination with other portions, he concluded that the mesoderm develops from the ectodermal (animal) part of the embryo, but requires contact with the endodermal (vegetal) part to do so.

The implication was that signals released from the vegetal half instruct the animal half. Over the years, some attempts were made to purify the factors responsible, from embryos and from animal tissues, but without much success. Then, in 1987, J. M. W. Slack and colleagues published the results of testing various different growth factors on ectoderm explants from Xenopus blastulae. They found that, at low concentrations, basic fibroblast growth factor (bFGF) could induce mesoderm, as judged by histological criteria. Crucially, they also showed that when animal and vegetal explants were cultured together, the addition of heparin could prevent mesoderm induction — suggesting that the natural mesoderm-inducing signal from the vegetal half is a molecule that can be inhibited by heparin. FGF is one such molecule.

Later that year, David Kimelman and Marc Kirschner published the results of similar experiments. These authors used a different indicator of mesoderm induction — the levels of mRNA encoding cardiac actin — but they, too, found that bFGF can induce mesoderm. They also discovered that transforming growth factor-β (TGF-β) is required to boost cardiac actin expression to the level seen normally in embryos. Moreover, they found that FGF mRNA is present in early embryos. Three years later, J. C. Smith et al. discovered that in mammals, a likely mesoderm inducer is activin A — a protein that was best known until then for its roles in adult organisms, butwhich was now shown to have a similar sequence and activity to a member of the Xenopus TGF-β family that can induce mesoderm.