Cancer stem cells (CSCs) are an aggressive subpopulation of cancer cells, endowed with the ability to initiate primary and metastatic tumours, including in triple-negative breast cancer (TNBC). CSCs can exist across intermediary cell states, exhibiting phenotypic plasticity that facilitates this aggressive phenotype. Novel machine learning techniques developed in collaboration with A/Prof Krishnaswamy (Yale University) to specifically elucidate cell states residing along a phenotypic spectrum were applied to single-cell sequencing data derived from TNBC models. Resulting data from this preliminary analysis defined five archetypal cell states or ‘archetypes’ that describe the plasticity landscape of CSCs.
Archetypes were biologically characterised by hallmark, pathway and spatial transcriptomic analysis. These analyses revealed a range of biological differences between the archetypes, including a striking different between the metabolic phenotypes of two archetypes. These differences also correlated with differential expression of various glucose transporters (GLUT1, GLUT3, GLUT6 and GLUT8). Bioinformatic analysis revealed GLUT3 in particular to be associated with the cancer stem cell state (defined by high levels of CD44). GLUT3 is a high-affinity glucose transporter typically expressed in neurons, however upregulation has also been reported in some cancers.
We performed immunofluorescence and western blot techniques to validate the relationship between increased expression of GLUT3 and the cancer stem cell state in vitro. Knockdown of GLUT3 by siRNA and subsequent in vitro assays revealed cell changes relevant to cancer stem cell function. We have demonstrated that differential expression of glucose transporters is linked to distinct cell states and that GLUT3 over-expression correlates with an aggressive cancer stem cell state. Future work will focus on defining how increased levels of GLUT3 maintain cancer stem cell function and understanding the signalling networks that regulate GLUT3 in triple-negative breast cancer.