Integrating chemical and biological approaches to target NAMPT with ADC technology in cancer

Abstract

Cancer is a large group of diseases and has become a major societal problem; it is the second leading cause of mortality in EU countries after cardiovascular diseases and a leading cause of death worldwide. Current chemo- and radiotherapeutic approaches lack selectivity and efficacy, leading to low quality of life and poor outcomes in patients. In recent years, targeted therapy using antibody drug conjugates (ADCs) has become an increasingly important treatment option. ADCs are hybrid molecules that use an antibody to deliver a cytotoxic drug specifically to tumour sites to reduce the systemic toxicity associated with traditional chemotherapy. Currently, 11 ADCs have been approved by the FDA, more than 50% of which use potent microtubule metabolism-disrupting agents, such as auristatins or maytansins. They have shown outstanding results in clinical trials over a variety of malignancies from liquid to solid tumours. However, their use is limited due to the mode of action of the cytotoxic payload, since only actively dividing tumour cells can be targeted, leaving dormant or non-dividing tumour cells untouched. Therefore, this thesis focuses on NAMPT inhibitors as promising payload candidates due to their high potency in affecting cellular energetics by depleting intracellular NAD+ levels, a biological feature essential for both dividing and non-dividing cells. First, the in vitro efficacy of published NAMPT inhibitors on CD19+ and HER2+ human cancer cell lines, both as free payloads and ADCs, was explored. Optimisation of the free payload and linker-payloads for ADC delivery was initially chemically guided based on the pharmacophores identified on the inhibitor molecule. Despite the promising results of the free inhibitors, ADCs loaded with these NAMPT inhibitor linker-payloads showed poor in vitro efficacy. Accordingly, the effects of NAMPT inhibitors on cellular endocytosis processes were investigated in more detail. The aim was to determine the feasibility of ADC-target complex internalisation when using CD19 or HER2 surface receptors as targets for ADCs loaded with NAMPT inhibitors as payloads. Initial findings suggested that ADC-surface receptor complex internalisation is inefficient in CD19+ cell lines when targeted with NAMPT inhibitors, prompting the exploration of alternative targets. CD30 was identified as an optimal tumour-associated antigen for NAMPT inhibitor-ADCs due to the efficacy of ADCs in depleting cellular NAD+ levels, which was supported by promising in vitro results. However, the ADCs based on such NAMPT inhibitors lacked full-blown cytotoxicity, even in CD30+ cell lines. Biochemical exploration subsequently revealed that the hydrophobic nature of these NAMPT inhibitors is counterproductive for their use as ADC payloads due to excessive cytoplasmic clearance through passive diffusion, resulting in cellular escape. To increase efficacy, a computationally aided lead optimisation campaign for NAMPT inhibitors was performed, leading to a payload showing significantly higher hydrophilicity. This resulted in full-blown cytotoxicity of the corresponding ADC with EC50 values in the picomolar range on a CD30+ cell line in vitro, as well as promising 80-day survival in mouse xenograft models in vivo. However, lysosomal in vitro stability assays showed instability of the new NAMPT inhibitors, as demonstrated by more than 50% lysosomal hydrolysis of the delivered NAMPT inhibitor. The resulting lower intracellular payload concentration limited its applicability to L540, a CD30+ cell line particularly sensitive to NAMPT inhibition. The cause of this sensitivity was demonstrated through q-RT-PCR experiments, which showed significantly lower expression of the NAMPT enzyme in L540 cells compared to other cell lines tested in vitro. In the latter, free NAMPT inhibitors had a cytotoxic effect, whereas ADCs loaded with NAMPT inhibitors as payloads were ineffective due to lysosomal hydrolysis. Finally, a new family of NAMPT inhibitors resistant to lysosomal conditions was designed through computationally assisted virtual screening of a 3000-compound database generated by reaction-based compound enumeration methods and using the first lead candidate as the initial molecule. This set the basis for new NAMPT inhibitors optimised for ADC delivery targeting different tumour-associated antigens, such as CD30, CD19, and HER2. The investigation of these new and highly efficacious ADC payloads holds promise to enable the development of ADCs with a new mode of action targeting non-dividing cancer cells for improved treatment of various types of malignancies.