Quantitative structure‐selectivity relationship (QSSR)‐based molecular insight into the cross‐reactivity and specificity of chemotherapeutic inhibitors between PI3Kα and PI3Kβ

Abstract

Selective inhibition of phosphoinositide 3-kinase (PI3K) isoforms α and β with small-molecule inhibitors can result in distinct biological effects on anticancer chemotherapy. However, many existing PI3K inhibitors have moderate or high promiscuity and cross-reactivity between the 2 kinase isoforms. Here, a quantitative structure-selectivity relationship-based statistical modeling scheme was used to characterize the relative contribution of independent kinase residues to inhibitor selectivity and to predict the selectivity and specificity for existing PI3K inhibitors. It is found that the residue type and distribution of kinase's active site play an important role in inhibitor selectivity, while rest of the kinase may contribute to the selectivity through long-range chemical interactions and indirect allosteric effect. The selectivity is also determined by the configuration difference between PI3Kα and PI3Kβ kinase domains. Larger inhibitor compounds have lower binding potency to PI3Kβ than PI3Kα and thus possess higher selectivity for PI3Kα over PI3Kβ.KEYWORDS chemotherapeutic agent, inhibitor selectivity, phosphoinositide 3-kinase, quantitative structure-selectivity relationship, statistical modeling

Statistical analysis and heuristic identification of unexpected interactions from the neurokinase‐inhibitor interactome in trigeminal neuralgia pharmacological intervention

“…At a first glance of the heat map, it is observed that most kinase‐inhibitor bindings are favorable and can be spontaneous ( Score < 0, the score is related to binding free energy); this is not unexpected if considering that most protein kinases are conserved in their active site and thus kinase inhibitors generally exhibit promiscuity and broad specificity over diverse kinases, leading to UKIIs. In fact, the UKIIs are a common phenomenon that was also observed in, for example, inhibitor targeting glioblastoma‐related mTOR and cancer‐related PI3K signaling pathways . However, there are few kinase‐inhibitor pairs that were predicted as unfavorable interactions ( Score > 0), although the predicted unbinding potential for these unfavorable interactions is quite modest or moderate ( Score < 1 kcal/mol).…”