A bacterial two-hybrid system based on a reconstituted signal transduction pathway has been developed to identify protein-protein interactions in *Escherichia coli*. This system utilizes the catalytic domain of *Bordetella pertussis* adenylate cyclase, which is split into two complementary fragments, T25 and T18. When these fragments are genetically fused to interacting proteins, they functionally complement each other to produce cAMP. This cAMP then activates catabolic operons, such as lactose or maltose, leading to a detectable phenotype. The system allows for the spatial separation of protein-protein interactions from transcriptional activation, enabling versatile screening procedures. It can be used to identify ligands that bind to a given "bait" or to detect molecules/mutations that block specific protein interactions. The system is based on the modular structure of the adenylate cyclase catalytic domain, which requires both T25 and T18 fragments to form an active enzyme in the presence of calmodulin. When these fragments are fused to interacting proteins, they reassociate and produce cAMP, which activates catabolic genes. The system was tested using various chimeric proteins, including the yeast splicing factors Prp11 and Prp21, and was able to detect their interaction. It was also used to screen for interacting proteins in *E. coli*, where the presence of cAMP synthesis was used as a readout. The system can be used for both screening and selection, allowing the identification of interacting proteins among a large number of irrelevant clones. The system is particularly useful for analyzing protein interactions that occur in the cytosol or at the inner-membrane level. It offers the possibility of both positive and negative selections, making it a versatile tool for studying protein interactions. The system is based on the fundamental principle of signal transduction, specifically signal amplification, and can be made highly sensitive by using the full catalytic activity of the adenylate cyclase. The system is efficient for screening complex libraries and analyzing protein interaction networks. It is also easy to adapt for various applications, including the use of antibiotic resistance or toxic products as reporter genes.A bacterial two-hybrid system based on a reconstituted signal transduction pathway has been developed to identify protein-protein interactions in *Escherichia coli*. This system utilizes the catalytic domain of *Bordetella pertussis* adenylate cyclase, which is split into two complementary fragments, T25 and T18. When these fragments are genetically fused to interacting proteins, they functionally complement each other to produce cAMP. This cAMP then activates catabolic operons, such as lactose or maltose, leading to a detectable phenotype. The system allows for the spatial separation of protein-protein interactions from transcriptional activation, enabling versatile screening procedures. It can be used to identify ligands that bind to a given "bait" or to detect molecules/mutations that block specific protein interactions. The system is based on the modular structure of the adenylate cyclase catalytic domain, which requires both T25 and T18 fragments to form an active enzyme in the presence of calmodulin. When these fragments are fused to interacting proteins, they reassociate and produce cAMP, which activates catabolic genes. The system was tested using various chimeric proteins, including the yeast splicing factors Prp11 and Prp21, and was able to detect their interaction. It was also used to screen for interacting proteins in *E. coli*, where the presence of cAMP synthesis was used as a readout. The system can be used for both screening and selection, allowing the identification of interacting proteins among a large number of irrelevant clones. The system is particularly useful for analyzing protein interactions that occur in the cytosol or at the inner-membrane level. It offers the possibility of both positive and negative selections, making it a versatile tool for studying protein interactions. The system is based on the fundamental principle of signal transduction, specifically signal amplification, and can be made highly sensitive by using the full catalytic activity of the adenylate cyclase. The system is efficient for screening complex libraries and analyzing protein interaction networks. It is also easy to adapt for various applications, including the use of antibiotic resistance or toxic products as reporter genes.