Investigation of emerging cancer cell molecular vulnerabilities linked to the NAD(H) metabolome and opportunities for novel therapeutics

Introduction: Traditional anti-cancer drugs target proliferating cells and are generally potent but have poor selectivity towards cancer cells leading to dose limiting toxicities. Targeted therapies offer cancer selectivity but may be effective only against cancers with particular lesions and the development of resistance can be a problem. The focus of this PhD research has been on the essential metabolite NAD+ and NAD+ dependent cellular processes. An emerging hallmark of cancers is deregulated cellular metabolism and it was hypothesised that differences between cancer and non-cancer cells in NAD+ metabolism and function may offer potential novel therapeutic opportunities that are both cancer selective and potent.

Aims: The aim was to evaluate specific putative cancer cell molecular vulnerabilities linked to the NAD+ metabolome. Four key hypotheses were investigated, a) that inhibition of NAD+ salvage enzyme NAMPT could reduce activity of NAD+-dependent PARP DNA repair enzymes selectively in cancer cells, b) that inhibition of NAD+-dependent deacetylase SIRT1 is preferentially cytotoxic towards cancer cells and that it can modulate non-cancer and cancer cell differentiation, c) that inhibition of glycolytic enzyme LDH-A affects cancer cell epigenetics and d) that small molecule KHS101 has selective activity against cancer cells through metabolic effects and perturbation of NAD+/NADH balance.

Methods: Small molecule inhibitors of the indicated cellular targets, or their knock-down by RNAi, were utilised to analyse effects on a panel of cancer cell lines and non-cancer cell models in vitro. Cytotoxicity was analysed by chemosensitivity and image cytometry-based assays. Metabolic effects were analysed using a Seahorse XFp analyser and NAD(H) assays. mRNA expression was analysed by qPCR and protein expression by immunoblotting and IF, with colorimetric/fluorometric-based enzyme assays used to analyse effects on PARP1, SIRT1 and GPD2 activity.

Results: KHS101 and NAMPT inhibitor FK866 both showed promising in vitro activity towards the cancer cell line panel and, for the most part, were more active towards the cancer cells than the non-cancer cell models tested. Short term NAMPT inhibition by FK866 at non-toxic doses preferentially depleted NAD(H) levels in cancer cells which was independent of p53 status and was sufficient to induce a cancer selective decrease in PARP activity. Consistent with this, FK866 was able to potentiate several DNA damaging agents, in particular temozolomide, selectively in cancer cells. Preferential activity of SIRT1 inhibitor EX-527 towards
cancer cell lines than non-cancer cells over 96 hours was modest but depended on the cell line. Phenotypically, EX-527 induced a neuronal-like phenotype in two non-cancer cell models suggesting the cells undergo neuronal transdifferentiation. This was accompanied by changes in mRNA levels of stem cell factors and neuronal markers and was also observed in a glioblastoma stem cell-like model. LDH-A suppression led to alterations in the epigenetic state of cancer cells with global levels of H3K9me3 and H3K9Ac changing and mRNA levels of the epigenetically silenced TSG E-cadherin increasing. KHS101 mechanism of action studies revealed depletion of mitochondrial respiratory function and cancer cell ATP levels. Autophagy was induced by KHS101 in both cancer and non-cancer cells. Activity of mitochondrial G3P shuttle enzyme GPD2 was decreased and NAD+/NADH balance perturbed in cancer cells.

Discussion and Conclusions: In vitro proof of principle is provided that through NAMPT suppression, NAD+-dependent PARP activity can be reduced preferentially in cancer cells and certain DNA damaging chemotherapeutic agents can be potentiated selectively in cancer cells. This raises the possibility of widening the therapeutic window of clinically used drugs if such effects can be translated into in vivo. SIRT1 can affect cell fate or differentiation status, its inhibition promoting neuronal (trans)differentiation of non-cancer cells and of GBM stem-like cells, indicating potential to target cancer stem cells and as a differentiation therapy. LDH-A suppression affected global levels of histone H3K9 acetylation and methylation. Whilst effects at specific gene promoters have yet to be investigated, this identifies LDH-A as a novel target for selective modification of the cancer cell epigenome. Future work will investigate whether it can promote re-expression of epigenetically silenced tumour suppressors without adverse epigenetic effects on non-cancer cells. Cancer cells were unable to sufficiently metabolically compensate in response to small molecule KHS101 with in vitro results suggesting selective activity against multiple cancer cell types. Overall, these results suggest a number of potential therapeutic opportunities linked to the NAD+ metabolome that warrant further preclinical investigation.