Solid cancers like melanoma have the capability to exhibit spatially different behaviours as they are mostly comprised of irregular zones of both actively proliferating and quiescent cells. This microenvironment-driven intra-tumoral dynamic heterogeneity is a leading cause of melanomas acquiring drug resistance, thus promoting poor patient outcome. Studies have shown that compared to rapid proliferating sub-populations, slow cycling sub-populations within a tumour are more resistant to current therapies. NDRG1 has been reported in other cancers to be a key regulator of tumour progression and metastasis. However, the molecular mechanisms have not been fully understood.
Utilising Hoechst dye diffusion and subsequent FACS, we separated the proliferating peripheral from the dormant central sub-populations of 3D melanoma spheroids generated from fluorescent ubiquitination-based cell cycle indicator (FUCCI)-transduced melanoma cells. Our previous data revealed that the proliferating sub-population expresses high levels of MITF, in contrast to the dormant sub-population. Proteomics and transcriptomics analysis confirmed upregulation of MITF in the proliferating and identified NDRG1 as upregulated in the dormant sub-population. To observe the molecular mechanism behind this phenomenon, we knocked down NDRG1 in melanoma cells. Confocal microscopy of 3D spheroids displayed an increase in FUCCI-green cells within the G1-arrested core of spheroids generated from NDRG1-knockdown compared to non-targeting control cells. This observation mimics the behaviour of MITF-overexpressing spheroids. Immunofluorescence for both NDRG1 and MITF, confirmed the hypothesised high NDRG1 expression towards the centre of spheroids contrasting with high MITF expression towards the periphery. Based on these results, it is evident that NDRG1 does play a role in regulating the dormant cell subpopulation in melanoma. Pathway enrichment analysis based on our proteomics and transcriptomics data, we speculate that NDRG1 inhibits proliferation and MITF expression through inhibition of PI3K/AKT and canonical WNT-signalling. Overexpression of NDRG1 and specific modulation of downstream effectors will elucidate this hypothesis.