Photorhabdus luminescens is a nematode-symbiotic bacterium that secretes a plethora of effectors in order to kill insects and use them as a source of food. One of the effectors perforates the host membrane by forming channels that translocate toxic enzymes into the host. These so-called Tc toxin complexes are made of three components (TcA, TcB, and TcC). TcA forms a pentameric translocation device and TcB and TcC form a cocoon that shields the actual toxic enzyme in its lumen. The binding of TcB-TcC to TcA is the first step of Tc toxin activation, which we describe here in detail.
We determined a near-atomic resolution structure of an ABC holotoxin complex (TcA-TcB-TcC) which revealed something unexpected. What we noticed was a different conformation at the bottom of the cocoon where it binds to TcA. The distorted six-bladed β-propeller, which closes the cocoon like a gate in free TcB-TcC, is ordered in the holotoxin and the toxic enzyme moves into the translocation channel of TcA. Our results indicate that the toxic enzyme enters with its C-terminus first and has to pass through a 10 Å constriction site.
A direct transition from the closed to the open β-propeller state would result in clashes and the formation of a protein knot as the intermediate. Since this initially puzzled us, we decided to perform molecular dynamics simulations to understand the conformational changes involved. Without destabilizing the β-propeller, no transition took place. After removing this initial obstacle, however, the transition began: two of the six b-propeller blades switched successively from the closed conformation to an unfolded form prior to refolding in reverse order into the open conformation. This mechanism clearly avoids knot formation and results in the opening of the gate.
What exactly triggers the start of the conformational transition? Interestingly, a routine experiment in our lab provided us with hints to answer this question. When we recorded electron micrographs of an empty cocoon, that is TcB-TcC without toxic enzyme in its lumen, we noticed that there was almost no holotoxin formed! Quantification of the affinity of empty TcB-TcC to TcA showed that it was decreased by three orders of magnitude compared to the wild type. More perplexing was that the crystal structures of the empty cocoon and the toxin-filled wild type were practically identical and there was no change at the β-propeller region.
What solved the mystery was the next experiment: hydrogen-deuterium exchange mass spectrometry. This in-solution protein dynamics sensor provided us with the last piece of the puzzle to complete the whole story. Despite the fact that it is not more surface-exposed, the gatekeeper hairpin which sits in the center of the β-propeller, is more dynamic in the filled cocoon than in the empty cocoon. The presence of the toxic enzyme inside, therefore, destabilizes the gatekeeper. The clash between other loops of the β-propeller and TcA during complex formation further destabilizes the domain and triggers its opening. This results in the activated ABC holotoxin.
We present here a novel cargo-driven activation mechanism of Tc toxins that could only be clarified by combining various methods: structural biology, computational biology, biochemistry and mass spectrometry. Our study lays a strong foundation for further investigation and manipulation of Tc toxins of insect, plant and human pathogens.
The associated manuscript in Nature is here: