The Organizer: Spemann and Mangold

Move a small piece of tissue from one newt embryo into the side of another, and the host can build a second neural tube, mostly out of its own cells that were on their way to becoming skin.

That result came from grafts Hilde Mangold made in Hans Spemann’s laboratory in Freiburg in 1921 and 1922, published with Spemann in 1924. The summary line for this unit puts it as a rule: “Neural Induction: A cell fate is determined in part by the signals to which it is exposed.” This lesson builds that rule from the experiments and ends on the question the graft left open: what is the signal?

Before the graft: half embryos, whole larvae and a loop of hair

The question began as an argument about the first cell divisions. In 1888 Wilhelm Roux reported experiments on the two-cell embryo of a frog. He killed one of the two cells with a hot needle and left the dead cell in place. The living cell could go on to make something like half an embryo, with structures on one side and not the other. Roux read this as a mosaic: each early cell carries its own share of the future body.

A few years later Hans Driesch shook two-cell sea-urchin embryos until the cells came apart. A separated cell, instead of making half a larva, could make a smaller whole one. The two results are less opposed than they look. The experiments differed in species, and in one detail that mattered: Roux’s dead cell stayed attached to its living sister. In later experiments that separated frog cells completely, the survivor often formed a whole embryo.

Spemann wanted a gentler tool. By 1901 he was tying a loop of very fine hair around newt eggs and tightening it slowly, dividing the embryo partly or completely without an open wound. Partial constriction gave embryos with doubled front ends. Complete separation sometimes gave two whole embryos. In one case from 1901, one half made a normal embryo while its partner stayed alive, kept dividing and formed layers of cells, but made no neural plate and no notochord. Cells could live and divide without building an axis.

The reading that entered the textbooks, which Spemann developed over the following years, is that a half makes a whole embryo only if it contains material from the future dorsal side, where gastrulation will begin. Something on that side was needed, and the loop could not say what.

Gastrulation: “Different embryonic layers give rise to distinct organs”

After fertilization the egg divides without growing, a process called cleavage, into a ball of smaller cells. In amphibians the ball becomes a blastula, a sphere of cells around a fluid-filled cavity. Then comes gastrulation, the stage at which cells move from the surface to the inside and the embryo sorts itself into three germ layers. The outer layer, the ectoderm, makes skin and the nervous system. The middle layer, the mesoderm, makes muscle, bone, blood, kidney and the notochord, a stiff rod of tissue along the back. The inner layer, the endoderm, lines the gut and forms organs such as the liver.

In an amphibian, cells move inside through a slit in the surface called the blastopore. It first appears as a short groove on one side of the embryo, and the edge of that groove is the dorsal lip, on what will become the back. Cells that roll inward over the dorsal lip become the mesoderm under the back, including the notochord.

Now the poll: “What layer of the gastrula gives rise to the nervous system?” The answer is ectoderm. The brain and spinal cord come from a strip of the outer layer along the back, the neural plate, which later rolls up into the neural tube. The puzzle is why that strip becomes neural while the ectoderm on the belly becomes skin, when both start as the same kind of surface tissue. The graft answered with a place: the tissue that moves in underneath it.

The graft: pigment tells host from donor

Earlier grafts within a single species had shown that a dorsal lip moved to a new place could give extra axial structures. But in a same-species graft the donor’s cells look like the host’s. A second axis might simply be the graft growing according to its origin.

Mangold’s design solved this with color. She used two newt species, then called Triton cristatus and Triton taeniatus. Cells of cristatus carry little pigment. Cells of taeniatus carry pigment granules that stay visible after the embryo is fixed and cut into thin sections. Natural color became a lineage marker.

She cut a small piece of the dorsal lip region from a pale cristatus gastrula and placed it in the flank or belly of a pigmented taeniatus gastrula of the same stage, in tissue that would normally become belly skin. Most operated embryos died. By one historical reconstruction she performed a few hundred operations over two breeding seasons, and the paper described six embryos in detail. In the survivors, a second neural plate rose beside the host’s own, folded, and closed into a second neural tube.

The sections settled the question. In one embryo, most of the second neural tube was pigmented, and so made of host cells, with a narrow pale strip from the graft along its floor. In another, the second neural tube was entirely host tissue, while the pale graft lay beneath it and formed a second notochord and much of the nearby mesoderm. It had changed the fate of host ectoderm that would otherwise have made skin. Spemann and Mangold called the dorsal lip region the organizer.

Mangold did this work for her doctoral thesis. The paper appeared in 1924 with Spemann’s name first, and Mangold died in 1924, the year it appeared. In 1935 Spemann alone received the Nobel Prize in Physiology or Medicine, for the organizer effect, and in his Nobel lecture he said the experiment was first carried out at his instigation by her.

Primary induction: competence, the organizer and the axis

The lecture name for this event is “Primary Induction”: mesoderm that has moved inside signals to the ectoderm above it, and that ectoderm becomes neural. Three words carry the idea.

Induction is a change in the fate of one tissue caused by a signal from another. Competence is the ability of a tissue to respond to that signal. It belongs to the responder, and it does not last. In amphibians, ectoderm taken from older and older embryos responds less and less to an organizer, and by the end of gastrulation it largely stops forming neural tissue in such tests. An organizer is a region whose signals do more than switch on one fate. The second axis in the grafts could have a neural tube, notochord and somites, aligned head to tail and back to belly. The organizer’s own cells become notochord and other dorsal mesoderm, it recruits neighbouring mesoderm into dorsal fates such as somites, and it induces the ectoderm.

Why is the dorsal lip special? The lecture heading is “Dorsalizing Factors - Wnt”. In the frog Xenopus, where this has been worked out best, events just after fertilization stabilize β-catenin, the Wnt pathway’s transcriptional co-activator, on the future dorsal side, where it enters nuclei (see that lesson). Dorsal β-catenin, together with signals from the yolky cells below, switches on organizer genes in the tissue that will form the dorsal lip. An early example was goosecoid, a homeobox gene expressed in the organizer, cloned by Eddy De Robertis’s laboratory in 1991.

Design problem

You have a dorsal lip from a pale newt species and host gastrulae from a pigmented species. Design experiments to show (a) whether the induced neural tube is made by host or donor cells, (b) whether the signal needs the two tissues to touch, and (c) when the responding ectoderm stops being able to respond.

Dead tissue that still induces

If the organizer acts through a signal, the signal should be a substance. In 1932 a note from Spemann’s circle reported that organizer activity could survive destructive treatment, and from 1933 Johannes Holtfreter tested this at length. He froze, dried and heated pieces of newt embryos, in some series to boiling, then placed them against ectoderm from early gastrulae. On its own such ectoderm became skin. Next to dead organizer it often became neural tissue. A living organizer was not needed for the neural response.

Then the result grew stranger. Many other killed tissues induced too. In 1933 C. H. Waddington, Joseph Needham and Dorothy Needham began physico-chemical tests of the organizer, and by 1934 the group reported that ether extracts of amphibian tissue could also induce, and they suspected a fat-soluble substance; the Cambridge group went on to test sterols. Through the 1930s more and more unrelated materials, from adult tissues to synthetic chemicals, were found to “induce” in some cases. A test that answers yes to almost everything cannot pick out one molecule.

Waddington described the trap. If competent ectoderm already holds a hidden tendency to become neural, an implanted substance might be the natural signal, or it might merely injure the tissue and release that tendency. Holtfreter’s later finding, in the 1940s, that newt ectoderm could become neural after brief damage that did not kill it, is one explanation for why the chemical hunt stalled for decades.

The confusion also pointed toward the answer. The next lesson shows that frog ectoderm whose cells are separated from one another becomes neural tissue with no inducer at all, and that the organizer’s products, Noggin, Chordin and Follistatin, act by blocking a signal of the TGF-beta family (see that lesson). If neural is what ectoderm becomes when a brake is released, then many injuries and chemicals that weaken the brake would be expected to “induce” as well.