Pattern Formation inHydraHead Regeneration

Abstract

SYNOPSIS. I have reinvestigated the “classical” <it>Hydra</it> transplantation phenomena, using experiments of simple design and large sample sizes. All experiments involved transplanting small tissue fragments from a donor hydra to a host. Both of the effects which are considered to demonstrate “head inhibition” in hydra (removing the head of the host, and placing transplants at different distances from the host's head) can be explained by assuming a diffusible inhibitory substance produced in the head; the apparent half-life of the substance is 2 hr and the apparent diffusion constant 2 × 10−6cm2/sec. In contrast, the two effects which are considered to demonstrate the “head activation” probably have different mechanisms. The “head activation gradient” is controlled by a factor with a half-life of 36 hr or more; the “activation increase” of head regeneration has an apparent half-life of about 12 hr. Further new findings are (i) that the head inhibition is restored quite late in head regeneration, (ii) that some head inhibitor appears to be produced in hydra's body, and (iii) that injury produces a local increase in the head activation. All these results can be rationalized by a new version of the Gierer-Meinhardt model; this model also accounts quantitatively for the changes in transplantation properties during head regeneration and for the proportioning of hydra's body and head. In simulations the model has 9 independent parameters but predicts 17 seemingly independent results. The model is thus more than trivially successful, and seems suitable as a guiding hypothesis for further work.