Brain-Targeting Prodrug Design for Thyrotropin Releasing Hormone




Nguyen, Vien
Del La Cruz, Daniel
Prokai, Laszlo
Prokai-Tatrai, Katalin


Journal Title

Journal ISSN

Volume Title



Purpose: Thyrotropin releasing hormone (TRH) has many neuromodulatory effects throughout the brain, however, treatment using this peptide induces unwanted peripheral side-effects. Based on our novel prodrug design that synergistically employs lipoamino acid residues (LAAs) and a brain-enzyme sensitive linker for prolyl oligopeptidase (POP), we have developed a set of lead compounds in silico. Computationally assessing lipophilicity and POP-binding affinity of our virtual prodrugs led to the selection of a representative, termed Prodrug (1), for membrane affinity studies to predict brain access from circulation. Method: Prodrugs with different LAAs and POP-sensitive linkers were designed in silico for docking with POP’s binding site using SCIGRESS and AutoDock Vina software. The prodrugs’ calculated logP (clogP) and POP-binding affinity were determined with built-in SCIGRESS applications. AutoDock results were represented as Gibbs free energy of binding (ΔG). The molecules with the most negative ΔG and adequate clogP were selected for synthesis. Prodrug (1) was compared to TRH in membrane affinity studies via immobilized artificial membrane chromatography (IAMC), an established method to predict membrane affinity (i.e., BBB permeability) with a chromatographic column comprised of immobilized synthetic lipids that mimic biological membranes. A high IAMC retention time correlates to a greater membrane affinity, with IAM Chromatographic Hydrophobicity Index (CHIIAM) values being the quantitative measure of retention. A range of reference compounds with known CHIIAM values were selected; plotting these values against their experimentally determined gradient retention times through the IAM column provided the linear relationship equation used to convert Prodrug (1)’s gradient retention time into its CHIIAM value. Results: A virtual library of prodrugs having various LAAs and POP-sensitive linkers were designed and docked to POP’s binding site. The binding of Prodrug (1) was comparable to the calculated ΔG of POP’s published, co-crystalized ligand. This prodrug also showed favorable clogP for transport into the brain and a significantly increased CHIIAM compared to TRH. Conclusion: Based on the prediction of POP-binding, using in silico docking and the favorable membrane affinity of Prodrug (1), we expect this prodrug to efficiently deliver TRH into the brain and serve as a template for fine-tuning future prodrug constructs for the efficacious brain-delivery of TRH.