Abstract
The mucosa-associated lymphoid tissue lymphoma-translocation protein 1 (MALT1) is a key protein in the adaptive immune response system in humans. This protein is widely expressed in the human body and is related to nuclear factor-κB (NF-κB) signaling activation in response to T-cell receptors. Due to this, MALT1 is key in the regulation of inflammatory events in a variety of tissues, where its dysregulation is associated with several types of cancer, especially hematological cancers. In this sense, its relevance makes MALT1 a valuable target to treat many diseases, drawing the attention of many researchers with the aim of proposing new MALT1 inhibitors. However, there is a lack of literature describing its complex dynamical behavior and allosteric inhibition, which considerably hampers the computational design of new MALT1 allosteric inhibitors. In that regard, the present work investigated the complex conformational behavior of MALT1 protein during its allosteric inhibition. For this, biased molecular dynamics simulations, sophisticated machine learning techniques such as neural networks, and docking calculations were used. From the performed investigation, it was observed that through allosteric inhibition, Loop 1 and 3 movements were crucial to reduce the catalytic site cavity volume, keeping cysteine unavailable for substrate mimic binding. In addition, statistical information over the explored ensemble showed that the great majority of the inhibited conformations presented an unavailable catalytic cysteine for substrate binding. Hence, the presented results can be used as an objective criterion for the computational proposal of new MALT1 allosteric inhibitors. However, despite mouse MALT1 and human MALT1 presenting 93% homology, the generalization of the findings to human MALT1 protein should be taken with care, and the obtained results apply specifically to the mouse MALT1 construct.