Syntheses of Protein Degraders and Compounds for Targeted Drug Release

Abstract

The thesis consists of two main projects: a) the development of a novel strategy to gain genetic control with spatiotemporal precision, and b) the optimisation of inhibitors of the protein kinase p38α, which has a major role in the homeostasis of tumours and in the regulation of metastasis, invasion and proliferation, amongst other processes in cancer.

a) Guaymoxifen is a molecule that is inactive with respect to the Cre method for genetic modification, but can be cleaved by a xeno-enzyme (LigF, artificially expressed in the metastatic cells of our mouse model, but not in the innate cells of the model) to liberate the active compound hydroxytamoxifen (4OHT). When guaymoxifen is administered to a CreER-GFP mouse (which has the potential to express green fluorescent protein upon exposure to 4OHT) that possesses LigF metastatic tumours, 4OHT will only be released around these metastatic tumour cells. Therefore, genetic modification (expression of GFP in this case) will only occur in the cells surrounding the tumour sites (stroma), allowing visualisation and isolation of key stromal cells during tumour development. A drawback was that the liver of the mouse was also capable of cleaving guaymoxifen and releasing 4OHT, which caused genetic recombination to occur away from the stroma and thus generated background noise, which limited the use of this technique in vivo. In order to address this problem, we developed a family of compounds based on guaymoxifen that would be more resistant to oxidative metabolism in the liver, by substitution of the beta- ether bond. Initial biological testing of these compounds identified 2 analogues that were more resistant to liver cleavage, yet could still be effectively cleaved by LigF in vitro. Ongoing work will complete the data needed for proof of concept of this strategy. 4OHT, (the active part), possesses a double bond with four distinct substituents, which leads to the existence of 2 stereoisomers. Only the Z stereoisomer is active against CreER, and furthermore, the proZ isomer of guaymoxifen is preferentially cleaved by LigF over the proE isomer. We therefore synthesised a modified analogue of 4OHT: 2Me-4OHT which favours the formation of the active isomer due to steric repulsion from the additional methyl group. This analogue was demonstrated to have comparable activity to 4OHT in CreER assays and in vivo. We therefore prepared the corresponding compound including the guaiacol moiety and observed that it could be cleaved by LigF to liberate active 2Me-4OHT in vitro. The liberated compound possessed a Z:E isomer ratio that was approximately 8x higher than that of 4OHT. In vivo examination is ongoing to determine the performance of these compounds in the complete mouse models.

b) Our objective was to design and synthesise a new family of compounds based on the existing p38α inhibitor PH-797804 (PH), developed by Pfizer. We incorporated a variety of functional groups at different lengths distant from a crucial amide moiety of the molecule, which could be modified without adversely affecting the affinity for the target protein. The added functionalities were amenable to conjugation to moieties

for the accumulation in tumour cells, such as peptide sequences or gold nanoparticles. The selective accumulation of these drug conjugates in the tumour cells would avoid the side effects associated with p38α inhibitors, such as skin rash and liver toxicity, identified in a number of phase II clinical trials. Our free analogues, before incorporation of the directing groups, were tested in a number of biological assays and in vivo. They maintained the activity of PH, and in the case of 4-13a even displayed superior activity. A new series of conjugates bound to RGD (an integrin recognition fragment) and analogues of somatostatin (a ligand for somatostatin receptor) were synthesised and shown to possess activity in vitro and preliminary in vivo tests. The conjugates were predicted to have an affinity for the corresponding receptors, which are overexpressed on the surface of many types of tumour cells and would thereby cause the accumulation of the conjugates in tumours. Endosomic assisted internalisation and cleavage would then release the active analogues within the tumour cells. Ongoing testing aims to establish the therapeutic benefit of these conjugates. We also designed a PROTAC (“Proteolysis Targeting Chimera”) using PH as the warhead. After screening of appropriate ligands for E3 ubiquitin ligase and optimisation of the linker, we obtained a family of compounds that induced the ubiquitination and degradation of p38α and p38 in cellular assays at nanomolar concentrations. The degradation took less than 8 hours and was maintained for over 48 hours, in a range of cell lines. These PROTACs were also effective in vivo, whereby p38α was degraded in lung tissue after 16 h when administered intratracheally, but not in nearby tissues such as heart or lung. Ongoing optimisation to improve the solubility and bioavailability of this PROTAC hopes to achieve a novel therapy for use against cancer.

La tesis consiste en dos proyectos principales: a) desarrollo de una estrategia novedosa para la modificación genética con control espaciotemporal, y b) mejora de inhibidores de la proteína p38α, que tiene un papel muy importante en la homeostasis de tumores y en la regulación de la metástasis, invasión y proliferación, etc. a) Una molécula (que denominamos Guaymoxifen) inactiva respeto al método “Cre” de modificar genes, que mediante la acción de una xeno-enzima (LigF, sólo expresada en bacterias, no en las células humanas), se parte y libera hydroxytamoxifen (4OHT, la parte activa). Cuando se administra guaymoxifen a un ratón con tumores que expresan LigF, libera 4OHT sólo alrededor de las células tumorales, y entonces solo se modifica la genética de dichas células. El problema es que en el hígado también se libera 4OHT y causa ruido de fondo en los datos. Para afrontar ésta problema del metabolismo en el hígado, desarrollamos una familia de nuevos análogos de guaymoxifen que fueron más resistentes. Las pruebas biológicas identificaron dos análogos más resistentes, pero 4OHTam todavía fue liberado por la enzima LigF. También sintetizamos un nuevo análogo de 4OHT (la parte activa): 2Me-4OHT, que favorece la formación del estereoisómero activo debido a la repulsión estérica del grupo metilo añadido. Esté análogo tenía una actividad en los ensayos Cre-ER parecido a 4OHT. Preparamos el compuesto correspondiente incluyendo la parte “guay” y vimos en los ensayos de escisión que LigF rompe la molécula, liberando 2Me-4OHT. Así pues, cuando tenga lugar la isomerización se generará una mezcla enriquecida en el isómero activo.

b) Nuestro objetivo es diseñar y sintetizar una nueva familia de compuestos basada en un inhibidor ya existente, PH-797804 (PH), desarrollado por Pfizer. Hemos incorporado una variedad de grupos funcionales en estos análogos para que pudiéramos unir grupos directores, como por ejemplo nanopartículas de oro, o secuencias peptídicas, para direccionar el inhibidor al tumor con más especificidad. Así ganaríamos más eficacia y superaríamos los efectos secundarios asociados con los inhibidores de p38α ya descritos. Nuestros análogos, antes de incorporar los grupos directores, fueron probados en ensayos celulares y también in vivo en tumores implantados en ratones y mantuvieron la actividad de la inhibición de p38α, o incluso más alta que la del PH. Una nueva serie de análogos fue sintetizada con nuestros análogos enlazado covalentemente a RGD, y a análogos de somatostatina. El mismo análogo de la serie también fue modificado para incluir un iniciador de “E3 ubiquitin ligase”, para obtener un PROTAC (“Proteolysis Targeting Chimera”) con capacidad de degradar p38α. Después de un programa de optimización de los iniciadores de E3 ubiquitin ligase y optimización del “linker,” conseguimos una familia de compuestos “PROTAC” que degradaban eficazmente la proteína. Estas son las primeras moléculas pequeñas con capacidad para degradar únicamente p38α y p38beta descritas hasta el momento. La degradación fue probada en una variedad de líneas células y también in vivo.