Synthesis of TGF-Beta inhibitors and compounds for spatiotemporal drug release

Abstract

During this doctoral thesis, we have synthesised hundreds of grams of the TGF-β inhibitor LY2157299 (Galunisertib). This product was used by the group of Eduard Batlle in their investigations into the roles that TGF-β plays in colorectal cancer. During the initial synthesis and subsequent scale up, we overcame hurdles including optimisation of the penultimate step of the reaction - giving reproducible results - as well as optimising conditions of the final solid form to provide a product which is suitable for formulation for the in vivo experiments. After discovering some of the drawback of LY2157299, we designed and synthesised a novel TGF-β inhibitor which we named HOLY (3-10). HOLY proved to be nearly 10x more potent than the original compound and shows a very significant reduction in the negative physical effects that were observed when administering LY2157299 at a therapeutic dose in vivo. HOLY was designed to be a compound which could be transformed into a prodrug, and therefore modulation of various properties could be achieved with careful modification of the structure. This was demonstrated by modification of the phenol of HOLY to give a substituted carbamate inspired by the structure of Irinotecan and was subsequently named IrinOLY. It was hypothesised that this carbamate would be enzymatically cleaved in vivo, releasing the active TGF-β inhibitor. When tested in vivo, this IrinOLY was shown to be a potent TGF-β inhibitor as well as a compound much more comfortable to formulate when compared to LY2157299. Other potential advantages of this novel compound, such as infusion via an osmotic pump, are yet to be explored and will be done so in the near future.

We then moved on to a separate project where we designed a molecule which was capable of being cleaved at a specific point by an exogenous enzyme (LigF), releasing the active compound 4-hydroxy tamoxifen. After the cleavage, 4-OHT would then be able to activate the Cre toolkit and therefore performing specific genetic modifications to cells in the vicinity from where the 4-OHT was released. In order for the 4-hydroxytamoxifen to be released, a beta ether bond needed to be cleaved by the enzyme LigF. The fragment containing the beta ether bond had various points in which modifications could be possible to modulate properties such as solubility and bioavailability, however specificity for the LigF enzyme needed to be maintained. In order to synthesise the optimum compound(s), first a structure activity relationship study was performed on the recognition fragment of the part of the desired compound that LigF would cleave. A small library of compounds bearing on one side various strategic modification of the original ligand to LigF, and on the other side the reporter 4methylumberliferone. This enabled us to perform modifications on the recognition fragment, test the new modifications in front of the LigF enzyme, and then perform further modifications based upon these results. A synthesis the optimised recognition fragment covalently bound to 4-hydroxy tamoxifen was then designed, synthesised, and tested both in vitro. The tests showed that 4-OHT was indeed cleaved in all three of the aforementioned settings, and so we sought to synthesise the target compound in its pure isomeric form. Following many attempts at complex coupling reactions, we eventually reached our target compound isomerically-pure following preparative HPLC of a key intermediate of the initial synthesis. The corresponding isomers were then transformed successfully into both the E and the Z isomers of Guaymoxifen.

Durante esta tesis doctoral, se han sintetizado cientos de gramos del inhibidor de TGF-Beta LY2157299 (Galunisertib). Este producto ha sido utilizado por el grupo de Dr. Eduard Batlle en sus investigaciones de cáncer colorrectal. Antes de llevar a cabo la síntesis a grande escala, se ha optimizado cada paso a pequeña escala– mejorando la síntesis para que sea más reproducible. Para superar algunas dificultades durante la formulación de LY2157299, se ha decidido diseñar y a continuación sintetizar un nuevo inhibidor de TGF-Beta para mejorar dichas desventajas. Después de realizar los experimentos in vivo, se demostró que el nuevo inhibidor es 10 veces más potente que LY2157299. Más adelante, este nuevo inhibidor de TGF-Beta ha sido transformado en un profármaco para mejorar sus propiedades como por ejemplo la solubilidad en agua. El segundo proyecto de esta tesis doctoral ha sido el diseño y la síntesis de una molécula para liberar 4hidroxitamoxifeno únicamente cuando esté en presencia de una enzima exógena especifica - LigF. Una vez liberado, el 4-hidroxitamoxifeno puede activar el sistema/la enzima Cre para llevar a cabo ciertas modificaciones genéticas como activar o silenciar un gen en específico. El enlace que rompe LigF para liberar el 4-OH es de tipo beta éter. Se sintetizó una librería de moléculas con modificaciones en sus estructuras para optimizar el fragmento que LigF reconozca. Mas adelante se llevó a cabo la síntesis de que Guaymoxifen – una molecula basada en 4-hidroxitamoxifeno y el fragmento que LigF reconozca. Se realizó ensayos en células para confirmar la hipótesis de activar Cre en células en el alrededor de células que contiene LigF.