Roles and Regulation of Sphingosine Kinase 2 in Cancer

Abstract

The two mammalian sphingosine kinases, SK1 and SK2, produce the bioactive signalling lipid sphingosine-1-phosphate (S1P), which generally promotes cell survival, proliferation and migration. In line with this, SK1 is often found to be upregulated in human cancers, and overexpression of SK1 leads to neoplastic transformation of cells and tumour growth. However, despite generating the same product, most evidence to date suggests that SK2 acts in an opposing manner to promote cell death. In contrast, knockout mouse models indicate that there is at least some functional redundancy between the two SK isoforms, and targeting SK2 via genetic or pharmacological approaches in cancer models results in reduced tumour growth. Clearly, the roles of SK2 are poorly understood, but it is apparent that these unique and complex functions of SK2 are largely dictated by its differential subcellular localisation. SK1 is generally a cytoplasmic protein, but it can also be translocated to the plasma membrane where it mediates cell survival, proliferation and oncogenic signalling through the production of S1P. SK2, however, has been reported to localise to the nucleus, endoplasmic reticulum and mitochondria, and at these locations it appears to possess anti-proliferative and pro-apoptotic functions. SK2 has also been reported to localise to the plasma membrane, but its specific roles here have not been well characterised. Therefore, the main aims of this study were to explore the roles of SK2 in cancer, and to characterise novel mechanisms that regulate SK2 subcellular localisation, such as interacting proteins and post translational modifications, in order to gain a better understanding of this complex enzyme and the potential benefits of targeting SK2 in cancer. To explore the roles of SK2 in cancer, we examined the expression of SK2 in various human tumour samples using publically available datasets, and found that SK2 shows statistically significant upregulation in many cancers, but only to modest levels up to 2.5- fold over normal tissues. As high-level SK2 overexpression has been previously shown to cause cell death, we explored the effects of low, close to physiological levels of SK2 overexpression. By engineering a series of human and mouse cell lines overexpressing graded levels of SK2, we found that low-level SK2 overexpression increased cell survival and proliferation, and activated oncogenic signalling pathways. Notably, low-level SK2 overexpression (5- to 10-fold over endogenous levels) was sufficient to induce neoplastic transformation of mouse fibroblasts, resulting in efficient tumour formation in vivo. These findings coincided with decreased nuclear localisation and increased plasma membrane localisation of SK2, as well as increases in extracellular S1P formation. Hence, we have shown for the first time that SK2 can have a direct role in promoting oncogenesis. Furthermore, the Pitson laboratory previously identified a novel SK2-interacting protein, cytoplasmic dynein 1 intermediate chain 2 (IC-2), through a yeast two-hybrid screen, and characterising this interaction formed another part of these studies. We confirmed that SK2 interacts physiologically with the dynein complex in cells via the IC subunit, and being a retrograde-directed transport motor complex, we found that dynein mediates the translocation of SK2 away from the plasma membrane. Interestingly, although IC-2 was identified in the yeast two-hybrid screen, SK2 interacts more robustly with the highly-related IC-1 isoform, which is abundantly expressed in the brain. Strikingly, we found that IC-1 is downregulated 17-fold in glioblastoma multiforme (GBM) patient samples, which correlated with poorer survival of patients with this form of brain tumour. In line with a role for dynein in transporting SK2, low IC-1 expression in GBM cells coincided with more SK2 localised to the plasma membrane, where we had found it to accumulate in an oncogenic setting. Re-expression of IC-1 in these cells reduced plasma membrane localised-SK2 and extracellular S1P formation, and notably, decreased tumour growth and tumour-associated angiogenesis in vivo. Thus, these findings demonstrate a novel tumour-suppressive function of dynein IC-1, and uncover new mechanistic insights into SK2 regulation. Through previous mass spectrometric analyses performed by the Pitson laboratory, it is evident that SK2 contains multiple uncharacterised phosphorylation sites that are not shared with SK1. We explored the function of one such site, Ser363, and found it to potentially regulate nuclear localisation of SK2. Furthermore, we identified SK2 as a bona fide substrate of glycogen synthase kinase 3 (GSK3) in vitro and in cells, involving residues Ser437 and Ser441, and we found that other phosphorylation events may act to regulate SK2 catalytic activity. Overall, the studies outlined here have revealed a previously unreported role for SK2 in driving oncogenesis, and have described the characterisation of novel mechanisms that regulate the subcellular localisation of SK2. Therefore, these findings support the use of SK2 inhibitors as promising anti-cancer therapeutic agents. Furthermore, as the opposing functions of SK2 are largely dictated by it subcellular localisation, these findings may also assist in the development of new strategies to target oncogenic SK2 in cancer.