Micellar gangliosides mediate the lipid insertion of cholera toxin protomer A

The topology of the interaction of cholera toxin with ganglioside and detergent micelles was studied with the technique of hydrophobic photolabelling. Cholera toxin α and γ polypeptide chains appear to penetrate into the hydrophobic core of ganglioside micelles. Micelles of SDS cause the labelling also of the β polypeptide chains, while Triton X-100 micelles have little ability to mediate the labelling of the toxin. The specific reduction of the α-γ disulfide bond allows the penetration of the α polypeptide chain into Triton X-100 micelles, but does not affect the interaction of cholera toxin with either ganglioside or SDS micelles. Thus, ganglioside micelles appear to cause a conformational change of the native toxin, such as to induce the penetration of the α chain into the micelle hydrophobic core.     

Lipid phase separations induced by the association of cholera toxin to phospholipid membranes containing ganglioside GM1

The interactions of cholera toxin and their isolated binding and active subunits with phospholipid bilayers containing the toxin receptor ganglioside GM1 have been studied by using high-sensitivity differential scanning calorimetry and steady-state and time-resolved fluorescence and phosphorescence spectroscopy. The results of this investigation indicate that cholera toxin associates with phospholipid bilayers containing ganglioside GM1, independent of the physical state of the membrane. In the absence of Ca2+, calorimetric scans of intact cholera toxin bound to dipalmitoylphosphatidylcholine (DPPC) large unilamellar vesicles containing ganglioside GM1 result in a broadening of the lipid phase transition peak and a slight decrease (less than 5%) in the transition enthalpy. In the presence of Ca2+ concentrations sufficient to cause ganglioside phase separation, the association of the intact toxin to the membrane results in a significant decrease of enthalpy change for the lipid transition, indicating that under these conditions the toxin molecule perturbs the hydrophobic core of the bilayer. Calorimetric scans using isolated binding subunits lacking the hydrophobic toxic subunit did not exhibit a decrease in the phospholipid transition enthalpy even in the presence of Ca2+, indicating that the binding subunits per se do not perturb the hydrophobic core of the bilayer. On the other hand, the hydrophobic A1 subunit by itself was able to reduce the phospholipid transition enthalpy when reconstituted into DPPC vesicles. These calorimetric observations were confirmed by fluorescence experiments using pyrene phospholipids.(ABSTRACT TRUNCATED AT 250 WORDS)     

Comparison of the interaction of mono- and oligovalent ligands with cholera toxin. Demonstration of aggregate formation at low ligand concentrations.

The stimulation by cholera toxin of adenylate cyclase in Chinese hamster ovarian cells could be inhibited by various ligands. The latter have been shown to contain the structural oligosaccharide entities required for binding to cholera toxin, established as Galbeta1 leads to 3GalNAcbeta1 leads to 4Gal3 comes from 2alphaNeuAc. The different inhibitory potency of the ligands thereby correlates with the size of the aggregates formed with the toxin, which in turn depends on the valency of the ligands. The conclusion is drawn from a comparison of the interaction of cholera toxin and its B-protomer with ganglioside II3NeuAc-GgOse4-Cer, the newly synthesized bis-(monosialo-gangliotetraityl)amine and monosialogangliotetraose. In a double diffusion test cholera toxin B-protomer precipitated with the ganglioside II3 NeuAcGgOSE4-Cer and the divalent ligand bis(monosialo-gangliotetraityl)amine, suggesting the formation of high molecular weight aggregates, whereas no precipitation was observed with the monovalent monosialo-gangliotetraose. By ultracentrifugation analysis, aggregate formation of the cholera toxin B-protomer could be demonstrated with the ganglioside II3 NeuAc-GgOse4-Cer and bis(monosialo-gangliotetraityl)amine at a concentration at which the ganglioside was assumed to be monodisperse. Ganglioside/cholera toxin B-protomer complexes sediment faster than those of the toxin and bis(monosialo-gangliotetraityl)amine, suggesting higher aggregation of cholera toxin B-protomer with the former. On the other hand, no sedimentation with monosialo-gangliotetraose was observed. By equilibrium displacement dialysis, however, a comparable high affinity of binding to cholera toxin B-protomer of both the mono- and divalent oligosaccharides was demonstrated. Furthermore, values for the maximal concentration of the bound ligand from these binding experiments with cholera toxin B-protomer established molar ratios of ligand to protein of 4 to 1 and 2 to 1 for monosialo-gangliotetraose and bis(monosialo-gangliotetraityl)amine, respectively. From the results it is concluded that the lipophilic moiety of the ganglioside is not directly involved in the binding process to the toxin protein but leads to an oligovalency of this ligand, due to formation of micellar or submicellar structures.     

Pertussis toxin B-subunit-induced Ca2(+)-fluxes in Jurkat human lymphoma cells: the action of long-term pre-treatment with cholera and pertussis holotoxins

The exotoxins of Bordetella pertussis and Vibrio cholera have been used to investigate signal transduction in the human T-cell lymphoma Jurkat. Stimulation of the cells, leading to an increase in cytoplasmic free calcium, could be achieved by the anti-T-cell receptor complex antibody OKT3 and by pertussis holotoxin (PTHT), or its B-subunit (PTB), but not by cholera holotoxin (CTHT) or its B-subunit (CTB). Both holotoxins ADP-ribosylated specifically G-proteins in the plasma membrane of intact cells, while their B-subunits had no ADP-ribosyltransferase activity. Incubation of the cells with CTHT led to a state of unresponsiveness to all stimulants. CTB was without any effect, indicating that the ADP-ribosyltransferase activity of cholera toxin (located in the A-subunit of the holotoxin) was necessary for the inhibition of cellular signalling. The inhibitory effect of cholera toxin on the pertussis toxin action was not due to a blockade of pertussis toxin interaction with the cell surface, because pertussis toxin was still able to ADP-ribosylate membrane proteins in cholera toxin treated intact cells. In addition, the cholera toxin mediated inhibition was not due to elevated levels of cyclic-AMP, as forskolin (a direct activator of the adenylate cyclase) and no inhibitory effect. The stimulating effect of PTHT was independent of its ADP-ribosyltransferase activity, because it could also be obtained by the B-subunit alone. In addition, the increase of cytoplasmic free calcium after stimulation by PTHT clearly preceded the ADP-ribosylation. Pre-treatment with PTHT, PTB or OKT3, led to a long lasting increase in the level of intracellular Ca2+ in Jurkat cells, which could not, therefore, be stimulated further. Inhibition by cholera holotoxin of the stimulation by OKT3 and pertussis toxin (PTHT and PTB) imply that the mitogenic effect of pertussis toxin is perhaps mediated via the T-cell antigen receptor signalling cascade. The presented data do not support the idea that a pertussis toxin-sensitive G-protein is involved in coupling the T-cell antigen receptor to the phospholipase C.     

Cholera toxin and adenylate cyclase: properties of the activated enzyme in liver plasma membranes.

Adenylate cyclase (EC 4.6.1.1) activity in mouse liver plasma membranes is increased fivefold when animals are pretreated with cholera toxin. The increase in activity is detectable within 20 min of an intravenous injection of the toxin. The response of the control and cholera-toxin-activated adenylate cyclase to hormones, GTP, and NaF is complex. GTP causes the same fold stimulation of control and toxin-activated cyclase, but glucagon and NaF remain the most potent activators of liver adenylate cyclase irrespective of whether the enzyme is activated by cholera toxin. Determination of kinetic parameters of adenylate cyclase indicates that cholera toxin, hormones, and NaF do not change the affinity of the enzyme for ATP-Mg nor do they alter the Ka for free Mg2+. High concentrations of Mg2+ inhibit adenylate cyclase that is stimulated by either cholera toxin, glucagon, or NaF. These same Mg2+ concentrations have no effect on the basal activity of the enzyme or its activity in the presence of GTP.     

Accessory cholera enterotoxin, Ace, from Vibrio cholerae: structure, unfolding, and virstatin binding

Vibrio cholerae accessory cholera enterotoxin (Ace) is the third toxin, along with cholera toxin (CT) and zonula occludens toxin (Zot), that causes the endemic disease cholera. Structural characterization of Ace has been restricted because of the limited production of this toxic protein by V. cholerae. We have cloned, overexpressed, and purified Ace from V. cholerae strain O395 in Escherichia coli to homogeneity and determined its biological activity. The unfolding of the purified protein was investigated using circular dichroism and intrinsic tryptophan fluorescence. Because Ace is predominantly a hydrophobic protein, the degree of exposure of hydrophobic regions was identified from the spectral changes of the environment-sensitive fluorescent probe 4,4'-dianilino-1,1'-binaphthyl-5,5'-disulfonic acid (bis-ANS) that quenches the fluorescence of tryptophan residues of Ace in a concentration-dependent manner. Results showed that bis-ANS binds one monomeric unit of Ace with a 1:1 stoichiometry and a K' of 0.72 μM. Ace exists as a dimer, with higher oligomeric forms appearing upon glutaraldehyde cross-linking. This study also reports the binding of virstatin, a small molecule that inhibits virulence regulation in V. cholerae, to Ace. The binding constant (K=9×10(4) M(-1)) and the standard free energy change (ΔG°=-12 kcal mol(-1)) of Ace-virstatin interaction have been evaluated by the fluorescence quenching method. The binding does not affect the oligomeric status of Ace. A cell viability assay of the antibacterial activity of Ace has been performed using various microbial strains. A homology model of Ace, consistent with the experimental results, has been constructed.     

Structure-function studies of cholera toxin and its A and B protomers. Modification of tryptophan residues

The tryptophan residues on cholera toxin and its A and B protomers have been modified by reaction with 2-nitrophenylsulfenyl chloride and 2,4-dinitrophenylsulfenyl chloride. Modification of the tryptophan residues of cholera toxin results in complete loss of toxicity measured in a skin permeability assay. Modification of cholera toxin and its B protomer results in the complete loss of binding activity toward membrane receptors, the ganglioside galactosyl-N-acetylgalactosaminyl-[N-acetylneuraminyl]-galactosylceramide (GM1), and the oligosaccharide moiety of the ganglioside GM1. Modification of cholera toxin and its A protomer results in a complete loss of the ADP-ribosylation activity exhibited by their native counterparts. Modification of the A protomer results in no apparent change in its physical properties by sedimentation velocity in the ultracentrifuge or by gel filtration chromatography. Modification of the B protomer, either directly or when it remains a component part of the holo toxin structure, results in a change in its sedimentation value and its elution from gel filtration columns. The changes are compatible with a conversion of the B protomer from a pentameric moiety in aqueous solvents to its existence as a monomer unit, i.e. to the individual polypeptide chains comprising the native B pentamer. Thiolysis of the 2,4-dinitrophenylsulfenyl chloride derivative of the B protomer reaggregates the individual-polypeptide chains but does not return its ability to interact with GM1.     

Analysis of transmembrane dynamics of cholera toxin using photoreactive probes

Using sodium dodecyl sulfate--polyacrylamide gel electrophoresis and autoradiography, we have shown that ¹²⁵I-labeled cholera toxin binds to Newcastle disease virus. Pretreatment of Newcastle disease virus with “cold” cholera toxin (at 37°C for 30 minutes) inhibits the binding of ¹²⁵I-labeled toxin in a subsequent incubation (at 37°C for 30 minutes). These results suggest that cholera toxin binds to Newcastle disease virus in a specific manner. The precise receptor for toxin is unknown in Newcastle disease virus but it is presumed to be the ganglioside GM₁. We have previously shown that the photoreactive probe 12-(4-azido-2-nitrophenoxy)stearoylgucosamine[1-¹⁴C] labels the membrane proteins of Newcastle disease virus. Since the reactive group of the probe, ie, N₃, resides within the membrane bilayer, studies were initiated to determine which, if any, of the subunits of cholera toxin cross the membrane of Newcastle disease virus and become radioactively labeled upon photoactivation of the probe at 360 nm. After a 15-minute incubation of cholera toxin with Newcastle disease virus containing the photoreactive probe, irradiation effected the ¹⁴C-labeling of the active A₁ subunit of cholera toxin. Irradiation of cholera toxin in solution with an equivalent amount of probe but without virus resulted in no labeling of toxin subunits.     

Temperature-dependent alteration of cellular morphology by cholera toxin in rat liver epithelial cells which are ts for maintenance of transformed properties

Cholera toxin via its ability to increase intracellular cyclic AMP levels can induce drastic changes in cell morphology. This report describes a temperature sensitive mutant of chemically transformed rat liver epithelial cells which only display cell shape alterations in response to cholera toxin at the permissive temperature. Shift up-shift down experiments indicate that the change in the response occurs fairly rapidly, i.e., within 2 hours at the new temperature. The behavior of the temperature sensitive cells at the nonpermissive temperature mimics that of the untransformed rat liver epithelial cells (i.e., no morphological change in response to cholera toxin) while at the permissive temperature the positive cell shape change is identical to that exhibited by chemically transformed rat liver epithelial cells. The temperature sensitive response to cholera toxin is not a function of cyclic AMP production, since the amount of cyclic AMP found as a function of either time or concentration of cholera toxin is quite similar in cells treated at either temperature.     

Biphasic regulation of adenylate cyclase by cholera toxin in neuroblastoma x glioma hybrid cells is due to the activation and subsequent loss of the alpha subunit of the stimulatory GTP binding protein (GS)

Exposure of neuroblastoma x glioma hybrid (NG108-15) cells to low concentrations of cholera toxin produced a stimulation of both basal and forskolin-amplified adenylate cyclase activity in membranes prepared from these cells. Higher concentrations of cholera-toxin reversed this effect. Mn2+ activation of adenylate cyclase indicated that this effect was not due to a modification of the intrinsic activity of this enzyme. Cholera toxin was demonstrated to produce a concentration and time-dependent loss of GS alpha from membranes of these cells. Loss of GS alpha from membranes of these cells was preceded by its ADP-ribosylation. The effects of cholera toxin were specific for GS alpha, as no alterations in levels of the pertussis toxin-sensitive G-proteins Gi2, Gi3 and Go, were noted in parallel. Equally, no alteration in levels of G-protein beta-subunit were produced by the cholera toxin treatment. These experiments demonstrate that cholera toxin-catalysed ADP-ribosylation does not simply maintain an activated population of GS at the plasma membrane and that alterations in levels of GS at the plasma membrane can modify adenylate cyclase activity.     

cAMP-independent effects of cholera toxin on B cell activation. I. A possible role for cell surface ganglioside GM1 in B cell activation

Cholera toxin has been used as a tool to study the effects of cAMP on the activation of B cells but may have effects independent of its ability to elevate cAMP. We found five lines of evidence which suggested that cholera toxin suppressed mitogen-stimulated B cell activation through a cAMP-independent pathway. 1) Cholera toxin (1 microgram/ml) was consistently more suppressive than forskolin (100 microM) despite the induction of higher intracellular cAMP levels by forskolin. 2) Cholera toxin was more suppressive at 1 microgram/ml than at 0.1 microgram/ml despite equivalent elevations of cAMP. 3) Washing B cells following their incubation with cholera toxin reversed much of the inhibition without altering intracellular cAMP levels. 4) The A subunit of cholera toxin, which at high concentrations (10 micrograms/ml) induced levels of cAMP comparable to those induced by cholera toxin (1 and 0.1 microgram/ml), did not inhibit B cell activation. 5) cAMP derivatives at high concentrations were much less effective than was cholera toxin in suppressing B cell activation. Although the elevation of cAMP may cause a mild inhibition of B cell proliferation, we found that even a marked elevation of cAMP did not suppress B cell proliferation, unless the elevation was persistent. We did, however, observe that the degree of toxin inhibition more closely paralleled binding of the toxin to B cells than toxin stimulation of cAMP. This result raised the possibility that binding of cholera toxin to its ganglioside GM1 receptor mediated an inhibitory signal which suppressed B cell proliferation.     

Modulation of thyroid hormone nuclear receptors by cholera toxin in cultured GH1 cells

The cellular actions of the thyroid hormones L-thyroxine and L-triiodothyronine are mediated by the association of hormone with a chromatin-associated receptor. In cultured GH1 cells, a hormone-responsive rat pituitary cell line, thyroid hormone decreases the concentration of its receptor at early incubation times by reducing the accumulation of newly synthesized receptor. In this study, we demonstrate that cholera toxin also reduces the amount of nuclear receptor in GH1 cells in a time- and dose-dependent fashion, without altering the affinity of the receptor for hormone. The reduction of receptor mediated by cholera toxin is not secondary to a generalized inhibition of cell protein synthesis or cell replication rates and this effect can be abolished by pretreatment of the cholera toxin with soluble ganglioside II3-alpha-N- acetylneuraminosylgangliotetraosylceramide . This effect requires an intact cholera toxin molecule and does not occur at similar concentrations of the membrane-binding B subunit of cholera toxin. In order to study the influence of cholera toxin on thyroid hormone receptor turnover, we have used a dense amino acid-labeling technique. The results indicate that cholera toxin does not change the half-life of receptor, but decreases the rate of appearance of newly synthesized receptor. This decreased rate completely accounts for the lowered steady state receptor levels. The extent of cAMP stimulation by cholera toxin does not correlate with the extent of receptor reduction and forskolin, which stimulates cAMP 25- to 500-fold, does not decrease thyroid hormone receptor abundance. These studies suggest that cholera toxin modulates receptor levels by a mechanism(s) that is not mediated by cAMP in GH1 cells.     

Acetylcholine and cholecystokinin receptors functionally couple by different G-proteins to phospholipase C in pancreatic acinar cells

We have studied the involvement of GTP-binding proteins in the stimulation of phospholipase C from rat pancreatic acinar cells. Pretreatment of permeabilized cells with activated cholera toxin inhibited both cholecystokinin-octapeptide (CCK-OP) and GTPγS but not carbachol (CCh)-induced production of inositol trisphosphate. Pertussis toxin had no effect. Neither vasoactive intestinal polypeptide, a stimulator of adenylyl cyclase, nor the cAMP-analogue, 8-bromo cAMP, mimicked the inhibitory effect of cholera toxin on agonist-induced phospholipase C activation. This indicates that inhibition by cholera toxin could not be attributed to a direct interaction of cholera toxin activated G_s with phospholipase C or to an elevation of cAMP. In isolated rat pancreatic plasma membranes cholera toxin ADP-ribosylated a 40 kDa protein, which was inhibited by CCK-OP but not by CCh. We conclude from these data that both CCK- and muscarinic acetylcholine receptors functionally couple to phospholipase C by two different GTP-binding proteins.     

NMR study of the complexes between a synthetic peptide derived from the B subunit of cholera toxin and three monoclonal antibodies against it

The contact interactions between a synthetic peptide and three different anti-peptide monoclonal antibodies have been studied by nuclear magnetic resonance (NMR). The synthetic peptide is CTP3 (residues 50-64 of the B subunit of cholera toxin) suggested as a possible epitope for synthetic vaccine against cholera. The hybridoma cell lines TE33 and TE32 derived after immunization with CTP3 produce antibodies cross-reactive with the native toxin. The cell line TE34 produces anti-CTP3 antibodies that do not bind the toxin. Selective deuteriation of the antibodies has been used to simplify the proton NMR spectra and to assign resonances to specific types of amino acids. The difference spectra between the proton NMR spectrum of the peptide-Fab complex and that of Fab indicate that the combining site structures of TE32 and TE33 are very similar but differ considerably from the combining site structure of TE34. By magnetization transfer experiments with selectively deuteriated Fab fragment of the antibody, we have found that in TE32 and TE33 the histidine residue of the peptide is buried in a hydrophobic pocket of the antibody combining site, formed by a tryptophan and two tyrosine residues. The hydrophobic nature of the pocket is further demonstrated by the lack of any pH titration effect on the chemical shift of the C4H of the bound peptide histidine. In contrast, for TE34 we have found only one tyrosine residue in contact with the histidine of the peptide.(ABSTRACT TRUNCATED AT 250 WORDS)     

Potentiation of cholera toxin-stimulated cyclic AMP production in cultured cells by inhibitors of RNA and protein synthesis

Cyclic AMP increased 8- to 10-fold after a 3-h treatment with 6 nM cholera toxin in rat C6-2B astrocytoma cells. In the presence of cycloheximide, cholera toxin increased intracellular cyclic AMP about 50-fold. Qualitatively similar potentiation of cholera toxin action by cycloheximide was observed in isolated swine aortic vascular smooth muscle cells. Cycloheximide, by itself, had no effect upon cyclic AMP levels and did not alter the apparent Ka for cyclic AMP generation by cholera toxin in the cells. Also, cycloheximide did not appear to augment cholera toxin action via inhibition of cyclic nucleotide phosphodiesterase. Puromycin and actinomycin D also augmented cholera toxin action in C6-2B cells. Potentiation of cholera toxin-increased cyclic AMP formation by cycloheximide was correlated with the inhibition of [14C]leucine incorporation into protein. These results indicate that the ability of cholera toxin to stimulate cyclic AMP production in C6-2B astrocytoma and swine vascular smooth muscle cells is enhanced by inhibition of de novo protein synthesis.     

Structural organization of cholera toxin gene and its expression in an environmental non-pathogenic strain ofVibrio cholerae

Non-pathogenic, environmental strain ofVibrio cholerae, ELTOR Ogawa EW6 carries a copy of the cholera toxin gene in its chromosome. Restriction enzyme digestion followed by Southern blot analysis revealed that the structure of the cholera toxin gene in this organism is different from that found in the virulent strains. The xbaI site which has been found to be conserved in the cholera toxin of the virulent strains examined so far, is absent here. Results of the RNA dot blot analysis indicated that the cholera toxin gene in EW6 is transcribed much less efficiently compared to the cholera toxin gene present in the virulent strainVibrio cholerae classical Inaba 569B.     

The effect of cholera toxin on the inhibition of vasopressin-stimulated inositol phospholipid hydrolysis is a cyclic AMP-mediated event at the level of receptor binding.

Incubation of L6 skeletal myoblasts for 16 h with cholera toxin but not with pertussis toxin, led to the inhibition of inositol phosphate generation induced by subsequent exposure to vasopressin. The effects of the toxin on inositol lipid metabolism were accompanied by the total ADP-ribosylation of the available cholera-toxin substrates within the cells. Immunological analysis demonstrated that the two polypeptides modified in vivo by cholera toxin were different forms of Gs alpha (alpha subunit of Gs). No novel cholera-toxin substrate(s) were detected. The cholera-toxin-mediated inhibition of vasopressin-stimulated inositol phosphate generation could be mimicked by both forskolin and dibutyryl cyclic AMP, but not by the separated subunits of the toxin. Receptor-binding studies demonstrated that the inhibition of agonist-stimulated inositol phosphate generation was accompanied by a decrease in cell-surface vasopressin-binding sites, with no effect on the affinity of these for the hormone. We suggest that the effect of cholera toxin and agents which increase intracellular cyclic AMP on vasopressin-stimulated inositol lipid hydrolysis is an effect on receptor number, and that there is no requirement to postulate a role for a novel G-protein, which is a substrate for cholera toxin, in the regulation of inositol phospholipid metabolism.     

Indomethacin inhibits cholera toxin-induced cyclic AMP accumulation in Chinese hamster ovary cells

Abstract Indomethacin was examined for its capacity to inhibit increases in adenosine-3',5'-monophosphate (cAMP) concentrations in Chinese hamster ovary (CHO) cells treated with cholera toxin. When added to the culture medium 1 h prior to cholera toxin (100 ng/ml), indomethacin (500 μg/ml) exhibited maximum protection against the typical increase in cAMP. Application of indomethacin at the same time as cholera toxin or up to 3 h after the toxin progressively decreased the drug's capacity to block further increases in cAMP. The drug appeared to block adenylate cyclase activity because addition of forskolin to drug-treated cells did not elicit a cAMP response. Binding of ¹²⁵I-labeled cholera toxin to indomethacin-treated cells was also reduced by at least 50%. These data indicate that indomethacin's inhibitory effect on cAMP formation in cholera toxin-treated cells could be explained by its capacity to alter adenylate cyclase activity and cholera toxin binding.     

Crystallization of isoelectrically homogeneous cholera toxin

Past difficulty in growing good crystals of cholera toxin has prevented the study of the crystal structure of this important protein. We have determined that failure of cholera toxin to crystallize well has been due to its heterogeneity. We have now succeeded in overcoming the problem by isolating a single isoelectric variant of this oligomeric protein (one A subunit and five B subunits). Cholera toxin purified by our procedure readily forms large single crystals. The crystal form (space group P2(1), a = 73.0 A, b = 92.2 A, c = 60.6 A, beta = 106.4 degrees, one molecule in the asymmetric unit) has been described previously [Sigler et al. (1977) Science (Washington, D.C.) 197, 1277-1278]. We have recorded data from native crystals of cholera toxin to 3.0-A resolution with our electronic area detectors. With these data, we have found the orientation of a 5-fold symmetry axis within these crystals, perpendicular to the screw dyad of the crystal. We are now determining the crystal structure of cholera toxin by a combination of multiple heavy-atom isomorphous replacement and density modification techniques, making use of rotational 5-fold averaging of the B subunits.     

Action of cholera toxin on dispersed acini from rat pancreas. Post-receptor modulation involving cyclic AMP and calcium

In dispersed acini from rat pancreas, cholera toxin caused a significant increase in cellular cyclic AMP but little or no change in amylase secretion. The presence of a secretagogue that causes mobilization of cellular calcium (e.g., cholecystokinin, carbamylcholine, bombesin or ionophore A23187) caused a substantial increase in the effect of cholera toxin on enzyme secretion. Cholera toxin did not alter calcium transport or the changes in calcium transport caused by other secretagogues, and secretagogues that mobilize cellular calcium did not alter cellular cyclic AMP or the increase in cyclic AMP caused by cholera toxin. These results indicate that in dispersed acini from rat pancreas there is post-receptor modulation of the action of cholera toxin by secretagogues that mobilize cellular calcium and that this modulation is a major determinant of the effect of the toxin on enzyme secretion.