The Bordetella pertussis RTX (repeat in toxin family protein) adenylate cyclase toxin-hemolysin (ACT) acquires biological activity upon a single amide-linked palmitoylation of the ⑀-amino group of lysine 983 (Lys 983 ) by the accessory fatty-acyltransferase CyaC. However, an additional conserved RTX acylation site can be identified in ACT at lysine 860 (Lys 860 ), and this residue becomes palmitoylated when recombinant ACT (r-Ec-ACT) is produced together with CyaC in Escherichia coli K12. We have eliminated this additional acylation site by replacing Lys 860 of ACT with arginine, leucine, and cysteine residues. Two-dimensional gel electrophoresis and microcapillary high performance liquid chromatography/tandem mass spectrometric analyses of mutant proteins confirmed that the two sites are acylated independently in vivo and that mutations of The adenylate cyclase toxin-hemolysin (ACT, 1 AC-Hly, or CyaA) is a key virulence factor of the whooping cough agent Bordetella pertussis and a promising protective antigen candidate for acellular pertussis vaccines (1-5). ACT belongs to the RTX (repeats in toxin) protein family (6) and has the capacity to form small cation-selective membrane channels, which account for its weak hemolytic activity (8 -11). The major cytotoxic activity of the 1706-residue-long protein, however, consists in its capacity to invade a variety of eukaryotic cells directly across their cytoplasmic membrane (12-14) and to deliver into cells a catalytic adenylate cyclase (AC) domain. This intoxicates cells by unregulated conversion of ATP to cAMP (15-18) and causes impairment of microbicidal functions of immune effector cells and apoptosis of lung macrophages (19).The capacity of ACT to penetrate into target cell membranes and to intoxicate cells depends on a posttranslational activation by the accessory protein, CyaC (20, 21). It was first established for the Escherichia coli ␣-hemolysin (HlyA) that the activation of RTX toxins consists in amide linked fatty-acylation (22), and Hackett et al. (23) have demonstrated by mass spectrometric analysis that native ACT produced by Bordetella (Bp-ACT) is mono-acylated by a palmitoyl residue at the ⑀-amino group of lysine 983. In contrast, the HlyA from E. coli was found to be acylated at two lysines, both in vitro and in vivo (24,25). Moreover, two highly conserved RTX acylation sites (25), corresponding to lysine 983 (Lys 983 ) and lysine 860 (Lys 860 ) are also found in ACT, and for an unknown reason, the CyaCactivated recombinant ACT produced in E. coli (r-Ec-ACT) is palmitoylated also at Lys 860 , in addition to acylation of Lys 983 (26). When compared with the native mono-acylated Bp-ACT, the doubly acylated r-Ec-ACT exhibits about four times lower specific hemolytic activity on sheep erythrocytes (20) and about 10 times lower specific channel-forming activity in artificial planar lipid bilayers (10). At the same time, however, the characteristics of the channels formed by both proteins are identical, and both proteins have identical capacity to insert ...

Bordetella pertussis adenylate cyclase (AC) toxin-hemolysin (ACT-Hly) can penetrate a variety of eukaryotic cells. Recombinant AC toxoids have therefore been recently used for delivery of CD8(+) T-cell epitopes into antigen-presenting cells in vivo and for induction of protective antiviral, as well as therapeutic antitumor cytotoxic T-cell responses. We have explored the carrier potential of the ACT molecule by insertional mutagenesis scanning for new permissive sites, at which integration of two- to nine-residue-long peptides does not interfere with membrane interaction and translocation of ACT. A model CD8(+) T-cell epitope of ovalbumin was incorporated at 10 of these permissive sites along the toxin molecule, and the capacity of ACT constructs to penetrate into cell cytosol and deliver the epitope into the major histocompatibility complex (MHC) class I antigen processing and presentation pathway was examined. While all six constructs bearing the epitope within the Hly portion of ACT failed to deliver the epitope to the MHC class I molecules, all four toxoids with inserts within different permissive sites in the AC domain efficiently delivered the epitope into this cytosolic pathway, giving rise to stimulation of a specific CD8(+) T-cell hybridoma. The results suggest that, in contrast to the AC domain, the hemolysin moiety of ACT does not reach the cytosolic entry of the MHC class I pathway.

The capacity of adenylate cyclase toxin (ACT) to penetrate into target cells depends on post-translational fatty-acylation by the acyltransferase CyaC, which can palmitoylate the conserved lysines 983 and 860 of ACT. Here, the in vivo acylating capacity of a set of mutated CyaC acyltransferases was characterized by two-dimensional gel electrophoresis and mass spectrometric analyses of the ACT product. Substitutions of the potentially catalytic serine 20 and histidine 33 residues ablated acylating activity of CyaC. Conservative replacements of alanine 140 by glycine (A140G) and valine (A140V) residues, however, affected selectivity of CyaC for the two acylation sites on ACT. Activation by the A140G variant of CyaC generated a mixture of bi-and monoacylated ACT molecules, modified either at both Lys-860 and Lys-983, or only at Lys-860, respectively. In contrast, the A140V CyaC produced a nearly 1:1 mixture of nonacylated pro-ACT with ACT monoacylated almost exclusively at Lys-983. The respective proportion of toxin molecules acylated at Lys-983 correlated well with the cell-invasive activity of both ACT mixtures, which was about half of that of ACT fully acylated on Lys-983 by intact CyaC. These results show that acylation of Lys-860 alone does not confer cell-invasive activity on ACT, whereas acylation of Lys-983 is necessary and sufficient.The whooping cough agent, Bordetella pertussis, secretes a 1706-residue-long RTX 1 adenylate cyclase toxin-hemolysin (ACT, AC-Hly, or CyaA), which can invade a variety of eukaryotic cells (1, 2). ACT delivers into cells a catalytic adenylate cyclase domain (AC) that is activated by intracellular calmodulin and catalyzes unregulated conversion of ATP to cAMP (3-6). This impairs microbicidal functions of immune effector cells and induces apoptosis of lung macrophages (7,8). In addition, ACT has the capacity to form small cation-selective membrane channels that account for its weak hemolytic activity (7-13).The capacity of ACT to form hemolytic channels and to penetrate target cell membranes and deliver the AC domain (cell-invasive activity) depends on a covalent post-translational fatty-acyl modification (14 -16). This is catalyzed by a dedicated protein acyltransferase, CyaC, which can acylate the ⑀-amino groups of two internal lysine residues of ACT, Lys-983 and Lys-860, located within conserved RTX acylation sites (14 -17). The mechanism of this novel type of protein acylation was recently analyzed in substantial detail for the prototype RTX toxin-activating and acyl-ACP-dependent protein acyltransferase HlyC, which acylates the homologous lysines 564 and 690 of the Escherichia coli ␣-hemolysin HlyA (18 -21). Several residues, including Ser-20 and His-23, were identified as being potentially involved in acyl transfer catalysis by . A model of the reaction mechanism was proposed, and formation of an intermediary acyl-ACP⅐HlyC complex was demonstrated (22-25). Various acyl-ACP-carrying fatty acids, including the most common in E. coli, the palmitoyl (C16:0) and pamitoleil (C16:...

No abstract