Glioblastoma display vast cellular heterogeneity, with glioblastoma stem cells (GSCs) at the apex. The critical role of GSCs in tumour growth and resistance to therapy highlights the need to delineate mechanisms that control stemness and differentiation potential of GSC. Dual-specificity tyrosine phosphorylation-regulated kinase 1A (DYRK1A) regulates neural progenitor cell differentiation, but its role in cancer stem cell differentiation is largely unknown. Herein, we demonstrate that DYRK1A kinase is crucial for the differentiation commitment of glioblastoma stem cells. DYRK1A inhibition insulates the self-renewing population of GSCs from potent differentiation-inducing signals. Mechanistically, we show that DYRK1A promotes differentiation and limits stemness acquisition via deactivation of CDK5, an unconventional kinase recently described as an oncogene. DYRK1A-dependent inactivation of CDK5 results in decreased expression of the stemness gene SOX2 and promotes the commitment of GSC to differentiate. Our investigations of the novel DYRK1A-CDK5-SOX2 pathway provide further insights into the mechanisms underlying glioblastoma stem cell maintenance.

The unique structural and physicochemical properties of adamantane and its derivatives have attracted considerable attention in the field of medicinal chemistry. Substituting phenyl rings for adamantane or its derivatives has provided a promising strategy to introduce lipophilicity and escape the ‘flat land’ of modern drug discovery. Additionally, the unique three-dimensional structure of adamantane facilitates the precise positioning of substituents allowing for a more effective exploration of drug targets. Evidently, we have seen an increased use of adamantane in pharmaceutically relevant molecules. The following Account highlights our group’s research in five drug discovery programs over the past 15 years showcasing the use of adamantane and its analogues in these studies.

Human β-galactoside α-2,6-sialyltransferase I (ST6Gal I) catalyses the synthesis of sialylated glycoconjugates. Overexpression of ST6Gal I is observed in many cancers, where it promotes metastasis through altered cell surface sialylation. A wide range of sialyltransferase inhibitors have been developed, with analogues structurally similar to the transition state exhibiting the highest inhibitory activity. To improve synthetic accessibility and pharmacokinetics of previously reported inhibitors, the replacement of the charged phosphodiester linker with a potential neutral isostere such as a carbamate or a 1,2,3-triazole has been investigated. Extensive molecular dynamics simulations have demonstrated that compounds with the alternate linkers could maintain key interactions with the human ST6Gal I active site, demonstrating the potential of a carbamate or a 1,2,3-triazole as a phosphodiester isostere. Free energy perturbation calculations provided energetic evidence suggesting that the carbamate and 1,2,3-triazole were slightly more favourable than the phosphodiester. Further exploration with free energy component, quasi-harmonic and cluster analysis suggested that there is an enthalpy-entropy compensation accounting for the replacement of the flexible charged phosphodiester with a neutral and rigid isostere. Overall, these simulations provide a strong rationale for the use of a carbamate or 1,2,3-triazole as a phosphodiester isostere in the development of novel inhibitors of human ST6Gal I.