Toxigenics: strategic cell death in the embryo

Abstract

Transgenic mice have offered a unique opportunity for molecular biologists to study DNA sequences involved in controlling tissue specific gene expression in the intact animal. The knowledge gained from such studies is now being used to the benefit of mammalian developmental biologists. Or at least, this is the claim in two recent papers using transgenics to annihilate particular cell populations l'z. Both demonstrate that it is possible to ablate, or anyway parfi'ally destroy, certain tissues which express a common marker gene, and this is seen as a prelude to being able to effect specific cell lineage ablation. The basic strategy behind both these studies is the fusion of part of the diphtheria toxin gene to a tissue spec~Jic promoter or enhancer. The diphtheria toxin gene, from C o ~ e bacterium diphtheriae, codes for a polypeptide which is cleaved into two subunits. Subunit A comprises an ADP-ribosyltransferase which catalyses the transfer of ADP-ribose from NAD to elongation factor 2, thus inhibiting protein synthesis and eventually killing the celP. So effective is this enzyme that it is estimated that a single molecule in the cytoplasm can prove fatal to a cell 4. Subunit B, on the other hand, binds to the cell surface and is instrumental in ensuring that the toxin is taken into cells by endocytosis. Without subunit B the ADP-ribosyltransferase cannot get into cells and is essentially harmless. Fusing the sequence encoding subunit A (DT-A) to a tissue specific promoter should, therefore, produce cell autonomous death in all cells in which that promoter is active. There is undoubtedly something very attractive about such nicely directed proscription. The two tissues which have been used to demonstrate that the strategy does indeed work are the pancreas and the lens. The rat e!~stase gene codes for a serine protease which is expressed almost exclusively in the exocrine acinar cell population of the pancreas. The enhancer/promoter of this gene lies in a 133 bp region 5' to the gene and has been shown to be capable of directing the expression of human growth hormone (hGH) exclusively to the exocrine pancreas s. SimiI~rly, the mouse yz-crystallin promoter is active only in lens fibre cells and the /ac Z gene fused to this promoter is expressed only in these cells in transgenic mice e. Thus, fusion of DT-A to the elastase enhancer should delete, at the very least, the acinar cell population of the pancreas, while Yzcrystallin-DT-A constructs should destroy the lens fibre cells. Some transgenlcs carrying the DT-A sequence (which, for reasons that are not clear, was only expressed when incorporated into the original rat elastase-hGH construct) did bck a normal pancreas. However, the frequency of this abnormality, presumably coincident with expression of the construct, was less than half that expected from the expression of hGH in transgenics without the DT-A insert (and this did not seem to be due to specific selection of wild-type pancreatic cells in otherwise mosaic founder animals). In affected transgenics, pancreatic development seemed to be arrested either at the embryonic rudiment stage, or during the early stages of differentiation when it is composed only of duct-like structures or, in one animal, had formed a few acinar-like cells. Lmmunocytochemistry demonstrated that amylase, a major exocrine product, was greatly reduced in the most advanced of these pancreatic structures and that there was a deficiency in endocrine products, such as insulin, compared with controls. All in all, the genetic ablation seems to have achieved a 10--100-fold reduction in islet and duct cells and almost complete depletion of exocrine cells I. In the absence of mosaic founder transgenics it is not possible to test the inheritance of this trait and to examine its pheno-