Inversion of the dorso-ventral axis in amphibian embryos by unilateral restriction of oxygen supply.

ACCORDING to the hypothesis advanced by L0vtrup (1958) the supply of oxygen is one of the factors responsible for the determination of bilateral symmetry in amphibian embryos. The protein coat covering the outside of the egg is known to have a very low permeability (Holtfreter, 1943), and it was suggested in the hypothesis that the formation of the grey crescent consists in a stretching of this coat by which the permeability is increased (cf. the work of Dalcq & Dollander (1948) and of Dollander & Melnotte (1952) on permeability of Nile blue), in this way the radial symmetry of the egg is changed to a bilateral symmetry from a metabolic point of view. As a consequence of the increase in permeability those oxidative, energy-supplying processes which are associated with gastrulation are enabled to proceed at a higher rate at one side of the egg. This hypothesis can of course only be valid for eggs which require oxygen for gastrulation. In some amphibian species, particularly in toads, gastrulation may occur under anaerobic conditions (Waddington, 1956). It seems necessary to assume that the bilateral symmetry is somehow built into the structure of these eggs. In order to test the hypothesis mentioned one might cut off the supply of oxygen from the dorsal side of the egg, where the grey crescent is located, but not from the ventral side. This may be accomplished by placing the eggs in closed tubes with the grey crescent towards the closed end of the tube. Failure to develop would, of course, only permit the already-known conclusion that oxygen is necessary for gastrulation. Support for the suggested hypothesis might be found, however, if inversions of the dorso-ventral axis occurred, i.e. if gastrulation could be forced to take place at the ventral, aerobic side. In the axolotl egg with which we have carried out our experiments it may be difficult to observe the grey crescent, and a different approach was therefore necessary. The first division plane does not invariably coincide with the future plane of bilateral symmetry (cf. the discussion by Ancel & Vintemberger, 1948).