Biochemical and computational studies towards selective inhibition of the immunoproteasome

The proteasome pathway degrades >90% of cytosolic proteins deemed redundant, misfolded or toxic, thereby influencing key regulatory pathways including: cell cycle control, DNA repair and apoptosis. As such, proteasome inhibitors (PI) have exhibited broad therapeutic applications, particularly for multiple myeloma and mantle cell lymphoma with 3 inhibitors gaining FDA approval. However, covalent binding and lack of targeted action cause severe toxicity. Upon stimulation by inflammatory cytokines, constitutive proteasome (CP) active sites β1c, β2c and β5c are replaced with corresponding β1i, β2i and β5i subunits; forming the immunoproteasome (IP). The abundant CP is required for regular cell function, however due to upregulation in diseased states selective IP inhibition is associated with an increased therapeutic index. Recent identification of structural differences between CP and IP specificity pockets (S1-4) allows structure-based drug design.
The cyclic peptides argyrin A and F exhibit potent, reversible CP inhibition with mechanisms distinct to existing therapeutics. In this project, argyrin B inhibition and binding interactions between the CP and IP are investigated, using purified enzyme assays alongside computational molecular modelling. Kinetic assays revealed argyrin B IC_50 values of 146.5 μM and 8.76 μM at β1c and β1i, respectively; a 16-fold difference with statistical significance. Whilst argyrin B also showed slight preference towards β5i over β5c, with low micromolar IC_50 values. The same trends were supported by Ki values and molecular docking estimated binding energies. AutoDock and FRED simulations suggest increased β1i S1 pocket hydrophobicity, T21S and G97H substitutions from β1c to β1i as key towards favourable β1i binding. At β5c, small, hydrophobic characteristics of S2 become polar in β5i that enhances argyrin B interactions. These findings facilitate design of further IP selective inhibitors, whilst the identification of the first known β1i selective and non-covalent PI shows great therapeutic potential with reduced toxicity proposed in comparison to existing therapeutics.