Understanding the conformational dynamics of PI3Kα due to helical domain mutations: insights from Markov state model analysis.

Phosphoinositide 3-kinases (PI3Ks) phosphorylate phosphoinositides on the membrane, which act as secondary signals for various cellular processes. PI3Kα, a heterodimer of the p110α catalytic subunit and the p85α regulatory subunit, is activated by growth factor receptors or mutations. Among these mutations, E545K present in the helical domain is strongly associated with cancer, and is known to disrupt interactions between the regulatory and catalytic subunits, leading to its constitutive activation. However, while the mutation's role in disrupting autoinhibition is well documented, the molecular mechanisms linking this mutation in the helical domain to the structural changes in the kinase domain remain poorly understood. This study aims to understand the conformational events triggered by the E545K mutation, elucidate how these changes propagate from the helical domain to the kinase domain, and identify crucial residues involved in the activation process. Molecular dynamics (MD) simulations combined with Markov state modeling (MSM) were employed to explore the conformational landscapes of both the wild-type and mutant systems. Structural and energetic analyses, including Molecular Mechanics Poisson-Boltzmann Surface Area (MM-PBSA) calculations, revealed that the E545K mutation significantly reduces the binding affinity between the regulatory and catalytic subunits. The mutation was found to induce a sliding motion of the regulatory subunit along the catalytic subunit, leading to the disruption of key salt-bridges between these domains. This disruption releases the inhibitory effect of the regulatory subunit, resulting in increased domain motion, particularly in the adaptor-binding domain (ABD). Enhanced flexibility in the ABD, helical, and C2 domains facilitates the rearrangement of the two lobes of kinase domain, thereby promoting activation. Additionally, the mutation appears to enhance PI3Kα's membrane affinity via the Ras-binding domain (RBD). Network analysis helped to identify key residues that may involve in allosteric signaling pathways, providing insights into the communication between domains. Druggable pockets in the metastable states were predicted followed by its docking with a PI3K inhibitor library. Docking studies revealed the crucial residues that may be participating in inhibitor binding. The identification of residues and regions involved in activation mechanisms using MSM helped to reveal the conformational events and the knowledge on probable allosteric pockets, which may be helpful in designing better therapeutics.