The six clones pTB112, pTB324, pTBs12, pCd122, pCd14 and pCd13 cover the tox locus of Clostridium difficile VPI 10463. This region of 19 kb of chromosomal DNA contains four open reading frames including the complete toxB and toxA genes. The two toxins show 63% amino acid (aa) homology, a relatedness that had been predicted by the cross-reactivity of some monoclonal antibodies (mAb) but that is in contrast to the toxin specificity of polyclonal antisera. A special feature of ToxA and ToxB is their repetitive C-termini. We define herein 19 individual CROPs (combined repetitive oligopeptides of 20-50 aa length) in the ToxB C-terminus, which are separable into five homologous groups. Comparison of the aa sequences of the N-terminal two-thirds of ToxA and ToxB revealed three marked structures, a cluster of 172 hydrophobic, highly conserved aa in the centre of both toxins, a sequence of 120 residues with an accumulation of highly conserved arginine, cysteine, histidine, methionine, and tryptophan residues, and a stretch of 248 less conserved aa. The probable function of these domains is discussed. Structural and functional homologies of ToxA and ToxB indicate that both genes have a common ancestor and may have evolved by gene duplication, with subsequent recombination and mutation, as has been reported for streptococcal glucosyltransferases (Gtf).

Receptors coupled to heterotrimeric guanine nucleotide-binding proteins (G proteins) activate phosphatidylinositol 4,5-bisphosphate (PtdIns(4,5)P2)-hydrolyzing phospholipase C (PLC) enzymes by activated alpha of free beta gamma subunits of the relevant G proteins. To study whether low molecular weight G proteins of the Rho family are involved in receptor signaling to PLC, we examined the effect of Clostridium difficile toxin B, which glucosylates and thereby inactivates Rho proteins, on the regulation of PLC activity in human embryonic kidney (HEK) cells stably expressing the m3 muscarinic acetylcholine receptor (mAChR) subtype. Toxin B treatment of HEK cells did not affect basal PLC activity, but potently and efficiently inhibited mAChR-stimulated inositol phosphate formation. PLC activation by the endogenously expressed thrombin receptor and by the direct G protein activators, A1F-4 and guanosine 5'-[gamma-thio]triphosphate (GTP gamma S), studied in intact and permeabilized cells, respectively, were also inhibited by toxin B treatment. C3 exoenzyme, which ADP-ribosylates Rho proteins, mimicked the inhibitory effect of toxin B on GTP gamma S-stimulated PLC activity. Finally both toxin B and C3 exoenzyme significantly reduced, by 40 to 50%, the total level of PtdIns(4,5)P2 in HEK cells, without affecting the levels of phosphatidylinositol and phosphatidylinositol 4-phosphate. Accordingly, When PLC activity was measured with exogenous PtdIns(4,5)P2 as enzyme substrate, Ca(2+)- as well as GTP gamma S- or A1F-4-stimulated PLC activities were not altered by prior toxin B treatment. In conclusion, evidence is provided that toxin B and C3 exoenzyme, apparently by inactivating Rho proteins, inhibit G protein-coupled receptor signalling to PLC, most likely by reducing the cellular substrate supply.