Cholera toxin (CT), produced by Vibrio cholerae and the causative agent of the disease cholera, is a typical A-B toxin. The B subunit binds to ganglioside GM1 on the surface of the intestinal mucosal cell whereas the A subunit after being reduced to generate the A1 peptide (CT-A1), activates adenylyl cyclase. CT-A1 is an ADP-ribosyltransferase that modifies the a subunit of the stimulatory G protein (Gs) of adenylyl cyclase, thus blocking its intrinsic GTPase activity and keeping the cyclase persistently activated. We have been investigating the detailed mechanism of cellular processing and activation of CT, using as a model human intestinal CaCo-2 cells, that behave in culture as differentiated enterocytes, the natural target for CT. We are particularly interested in events during the lag period between toxin binding and cyclase activation. We previously showed that the holotoxin binds to the cell surface with the A subunit facing away from the membrane and is internalized during the lag period. Then small amounts of CT-A1 are generated by the cells, and the cyclase becomes activated. We identified the cellular activity that reduces the toxin to CT-A1 as protein disulfide isomerase (PDI), an endoplasmic reticulum (ER)- resident protein. Furthermore, we found that CT reduction and action is blocked in cells treated with brefeldin A (BFA), which disrupts the Golgi apparatus and inhibits membrane trafficking. It has been proposed that CT is internalized through caveolae, noncoated invaginations on the plasma membrane that are enriched in cholesterol, sphingomyelin and glycolipids. We recently demonstrated that exposing CaCo-2 cells to the cholesterol-binding drug filipin, that perturbs caveolae and their function, blocks the internalization of CT, its reduction to CT-A1, and its ability to activate adenylyl cyclase. By contrast, drugs that disrupt endocytosis via clathrin-coated pits do not have these effects on CT. Taken together, our results support a model in which CT enters the cell through caveolae, and undergoes reduction by PDI in an intracellular compartment (most likely the ER) to generate CT-A1. In order to further identify components of the intracellular pathway that CT follows, we are using fluorescence microscopy in combination with BFA. The cellular effects of BFA are primarily directed towards the inhibition of antegrade transport from the ER to the Golgi apparatus. As a consequence, BFA induces a morphological and functional disassembly of the Golgi that results in the retrograde movement of Golgi-resident proteins to the ER. As BFA does not directly inhibit the reduction of CT by PDI, we are considering several other possible mechanisms to explain the effects of BFA on CT activation. These include: i) CT must be transported through a functional Golgi to reach the site where reduction takes place; ii) the location of an essential component (such as an intracellular receptor) required for the intracellular transport of CT is altered and no longer accessible to the internalized toxin or its subunits; or, iii) the colocalization of CT and PDI is prevented by alterations in Golgi-ER transport. With the use of CT subunit-specific antisera and organelle-specific antibodies, we have begun to monitor the trafficking of the toxin within the cell with time. CT initially appeared to remain intact as both A and B subunits gave similar fluorescence patterns. Beyond 30 min, however, a distinct separation of toxin subunits became evident. Whereas CT-B accumulated and remained in the Golgi, CT-A exhibited an increasingly reticular fluorescence pattern with significant colocalization with PDI indicative of the ER. In contrast, when the cells were treated with BFA, CT-B accumulated in a structure juxtaposed to the nucleus and no longer colocalized with Golgi markers. Although BFA treatment did not affect the morphology of the ER (as PDI continued to display a fine reticular pattern throughout the cell), CT-A exhibited a fluorescence pattern similar to that displayed for the B subunit. Whereas this structure is clearly not of Golgi origin, its identity remains unknown. While the observed pattern was similar to that described for markers of the intermediate compartment, colocalization studies with antisera to these antigens have not yet been performed. Likewise, it is not known whether these patterns represent intact holotoxin and to what extent the subunits are occupying the same site. Further studies using confocal microscopy are needed to confirm true colocalization of toxin subunits at specific intracellular sites.
