Sphingosine Kinase 1 in Cancer

1. INTRODUCTION

While sphingolipids, along with cholesterol and other phospholipids, serve as critical structural components of cell membranes, their most interesting functions involve their roles as bioactive signaling molecules. Although there is still much to be elucidated about specific mechanisms involved in sphingolipid signal transduction, some themes have emerged. Ceramide is known to be involved in apoptosis, cell senescence, differentiation, and cell stress (Hannun and Luberto, 2000; Mathias et al., 1998; Perry and Hannun, 1998). Sphingosine has also been revealed as an antigrowth signaling molecule (Taha et al., 2006b). In contrast, however, sphingosine-1-phosphate (S1P) is known to promote proliferation, survival, and inhibition of apoptosis (Spiegel and Milstien, 2003) as well as having roles in cell migration, vascular development, and inflammation (Hla, 2004). Sphingosine kinase (SK) is a key enzyme in the sphingolipid pathway, as it regulates the levels of all three of the aforementioned lipids to influence cell fate. Furthermore, its activity is necessary for the clearance of sphingolipids, as once it produces S1P, the latter is hydrolyzed by S1P lyase in an irreversible reaction to produce hexadecenal and ethanolamine phosphate which marks the only exit point for the sphingolipid metabolic pathway.

1.1. S1P signaling

Unlike the antiproliferative effects of ceramide and sphingosine, S1P has been shown to play a significant role in proliferation, migration, survival, angiogenesis, inflammation, and lymphocyte egress (Spiegel and Milstien, 2003). S1P can exert its various effects by binding five distinct and differentially expressed cell surface S1P receptors; these G-protein-coupled receptors were formerly referred to as endothelial differentiation gene (Edg) receptors (Taha et al., 2004), further highlighting S1P’s important prosurvival role.

S1P is capable of increasing cell survival and inhibiting the apoptotic process in a number of different cell types, including T cells (Goetzl et al., 1999; Kwon et al., 2001), and it has been also shown to induce a proliferative response in endothelial cells (Kimura et al., 2000) and vascular smooth muscle cells (Tamama et al., 2001). S1P, like potent angiogenic peptide growth factors such as VEGF and FGF-2, is now also considered one of the key regulators in angiogenesis. S1P acts on vascular endothelial cells via S1PR1 and S1PR3 receptors to stimulate migration and capillary-like tube formation in vitro (Kimura et al., 2000; Lee et al., 1999; Wang et al., 1999). This stimulatory activity of S1P on angiogenesis has recently been demonstrated in vivo in ischemic hindlimbs of mice (Oyama et al., 2008). Furthermore, siRNA-induced S1P receptor knockdown (Theilmeier et al., 2006) or FTY720-induced receptor downregulation has been shown to suppress tumor angiogenesis and tumor growth (LaMontagne et al., 2006; Nagaoka et al., 2008; Permpongkosol et al., 2002) as well as inhibit lymphocyte trafficking (Graler and Goetzl, 2004; Morris et al., 2005).

As S1P has been shown to be involved in cell survival, proliferation, and angiogenesis, it is easy to see how these S1P-mediated activities may contribute to the etiology of cancer (Olivera and Spiegel, 1993; Zhang et al., 1991). Interestingly, parenteral administration of S1P-specific antibodies has been shown to markedly slow human cancer xenograft progression and angiogenesis (Visentin et al., 2006). It is also important to note that S1P signaling has been implicated in the development of the drug resistant phenotype in cancer cells (Akao et al., 2006). Taken together, these findings implicate S1P as a very important signaling molecule in cancer biology that requires further study, and its manipulation may play a role in future anticancer therapies.

1.2. Sphingosine kinase

SK1 and SK2 are the enzymes responsible for catalyzing the conversion of sphingosine to S1P using adenosine triphosphate. Despite differences in structure (Okada et al., 2005) and cellular localization (Johnson et al., 2002) in vivo studies suggest that SK1 and SK2 may have some overlapping physiologic functions. For instance, although SK1 knockout (KO) mice and SK2 KO mice appear to develop normally, SK1/SK2 double KO mice suffer embryonic lethality due to inadequate angiogenesis and neurogenesis coupled with neuronal apoptosis (Mizugishi et al., 2005). SK1 and SK2 may in fact have some overlapping functions; however, a review of the current literature describes SK1 as being activated by several extracellular agonists and effecting extracellular S1P levels, whereas SK2 is considered to play more of a housekeeping role, and its function may be localized to the nucleus. Interestingly, it was recently shown that S1P produced by SK2 in the nucleus may regulate histone acetylation of the p21 promoter (Hait et al., 2009).

Further insights into the relative roles of SK1 and SK2 in S1P production have been complicated by the development of SK KO mice. Both the SK1 KO and SK2 KO mice show no abnormal gross or histological phenotypes and no compensation in message levels of the other SK isoform. However, SK1 KO mice show decreased S1P levels in their serum, while this was not detected in the sera of SK2 KO mice. To further investigate this, SK1 +/− mice were crossed with SK2 +/− mice. Interestingly, offspring lacking one to three SK alleles in any combination were indistinguishable from the wild type, but no animals lacking all four alleles were born, indicating that the SK1 KO SK2 KO genotype was embryonic lethal. These embryos had no detectable S1P and suffered hemorrhage and exencephaly due to vascular and neural tube defects (Mizugishi et al., 2005). It was also shown that SK1 KO SK2 +/− female mice were infertile due to defective decidualization with decreased cell mitosis, increased cell death, and defective decidual blood vessel formation (Mizugishi et al., 2007).

Despite the unclear consequences of SK1 KO on sphingolipid homeostasis, these mice have been useful for studying the oncogenic role of SK1. KO of SK1 was able to decrease the size but not the incidence of adenomas in Apc Min/+ mice by attenuating epithelial cell proliferation in the polyps. SK1 KOled to an elevation of sphingosine content and decreased expression of the G1/S cell cycle regulator CDK4 and c-myc in these adenomas, whereas S1P levels were not altered. Furthermore, this protection was not observed when Apc Min/+ mice were crossed with S1P receptor KOs. This highlights that the intracellular function of SK1, independent of the extracellular signaling of its product, S1P, is critical for the growth of intestinal adenomas (Kohno et al., 2006). Other studies have also shown that KO of SK1 can offer protection from chemically induced colon carcinogenesis (Kawamori et al., 2009; Snider et al., 2009).

The cancer preventative affects of SK1 KO are not confined to colon carcinogenesis but also apply to other chemically and genetically induced cancers. While 4-nitroquinoline-1-oxide (NQO) induced head and neck squamous cell carcinoma (HNSCC) in wild-type mice, genetic loss of SK1 prevented 4-NQO-induced HNSCC carcinogenesis with decreased tumor incidence, multiplicity, and volume. The SK1 KO mice showed decreased cell proliferation, increased caspase 3 activation, and decreased activation of AKT when compared with wild-type controls (Shirai et al., 2011). Work from our lab has shown that SK1 KO can protect p53 KO mice from developing their characteristic thymic lymphomas and increase their survival ~30%. Importantly, while KO of SK1 also protects p53 heterozygote mice from the development of lymphomas, it also protected them from the development of other tumor types including osteosarcoma and decreased tumor incidence while prolonging their survival (Heffernan-Stroud et al., 2012). Thus, studies with the SK1 KO mouse have confirmed the oncogenic character of SK1 in vivo.

2. SK1 IS AN ONCOGENE

Given its prime position in sphingolipid metabolism, SK1 expression affects the balance between prodeath and prosurvival sphingolipids to determine cell fate, a prime regulatory position for a potential oncogene. Likewise, screening of normal/tumor patient-matched thyroid and non-small-cell-lung cancer tissue, including carcinoid, squamous, and adenocarcinoma, exhibited over-whelmingly positive immunostaining for SK1 as compared with patient-matched normal tissue (Guan et al., 2011a). Moreover, a twofold elevation of SK1 mRNA expression was observed in cancer versus normal tissue for several types of solid tumors including breast, uterus, ovary, colon, small intestine, and rectum in addition to lung (Johnson et al., 2005). Increased SK1 and S1P were also found in a mouse model of colon cancer, and remarkably SK1 KO mice were resistant to the azoxymethane induced colon cancer observed in mice expressing SK1 (Kawamori et al., 2009).

2.6. SK1 inhibitors challenge cancer cell viability to enhance chemotherapy

Various studies have employed pharmacological inhibitors to show that inhibition of SK decreases cancer cell viability in neuroblastoma cells (Tavarini et al., 2000), glioblastoma cells (Bektas et al., 2009), and several types of leukemia cells (Cuvillier and Levade, 2001; Paugh et al., 2008; Zhang et al., 2008) as well as in various solid tumor cell lines (French et al., 2003). SK inhibitors (outlined in Table 7.5). were found to be antiproliferative toward a panel of tumor cell lines, including lines with the multidrug resistance phenotype because of overexpression of either P-glycoprotein or multidrug resistance phenotype 1, and they could induce apoptosis (French et al., 2003). Furthermore, resistance to doxorubicin- or etoposide-induced apoptosis in acute myeloid leukemia cells was attributed to sustained SK1 activity and reduced ceramide accumulation; likewise, apoptosis could be restored in these cells by treatment with an SK inhibitor, F-12509a (Bonhoure et al., 2006). Later results using the same inhibitor suggest that SK1 regulates imatinib-induced apoptosis in primary cells from CML patients, as inhibition of SK1 could kill both imatinib-sensitive and -resistant cells. This work places SK1 downstream of the Bcr–Abl/Ras/ERK pathway inhibited by imatinib and upstream of Bcl-2 family members (Bonhoure et al., 2008).

Table 7.5

Effects of SK1 inhibitors in cancer

Inhibitor	Mechanism of action
SKI-I (French et al., 2003)	Induces apoptosis and suppresses tumor growth rate and vascularization of glioblastoma xenografts (Kapitonov et al., 2009)
	Sensitizes pancreatic cancer cells to gemcitabine induced death (Guillermet-Guibert et al., 2009)
	Decreases growth and survival while enhancing apoptosis and Bcl-2 cleavage in leukemia cells (Paugh et al., 2008)
	Suppresses lysophosphatidic acid mediated proliferation in gastric cancer cells (Ramachandran et al., 2010)
	Suppressed S1P levels, decreased hemangiogenesis and lymphangiogenesis, reduced metastases to lymph nodes and lungs, and decreased overall tumor burden in murine model of breast cancer (Nagahashi et al., 2012)
	Blocked cell proliferation and induced apoptotic cell death in glioblastoma multiforme neurosphere cell line (Estrada-Bernal et al., 2010)
	Increases oxaliplatin-induced cytotoxicity in colon cancer (Nemoto et al., 2009)
	Increases sensitivity of non-small-cell-lung cancer cells to doxorubicin- or docetaxel-induced apoptosis (Song et al., 2011)
	Rescues daunorubicin sensitivity in leukemia cells (Sobue et al., 2008)
	Enhances ability of imatinib mesylate to inhibit cell growth, survival, and clonogenic potential in leukemia cells (Ricci et al., 2009)
	Sensitizes prostate and breast cancer cells to docetaxel (Sauer et al., 2009) (as does B-5354c)
	Reduces proliferation and sensitizes ovarian cancer cells to cytotoxic effect of N-(4-hydroxyphenyl) retinamide (Illuzzi et al., 2010a, 2010b)
	Impedes DHT-induced cell proliferation, DNA synthesis, and PSA secretion in prostate cancer cells (Dayon et al., 2009)
SKI-II (French et al., 2003)	Blocks expression of urokinase plasminogen activator and its receptor to inhibit glioma cell invasion (Young et al., 2009)
	Prevents accumulation and transcriptional activity of HIF-1α in prostate, brain, breast, and lung cancer cells (Ader et al., 2008)
	Reduces cell colony formation and activates of caspase-3 in glioblastoma cells (Bektas et al., 2009)
	Antimigratory and toxic effects on human glioma and glioma stem cells (Mora et al., 2010)
	Blocks breast cancer viability, clonogenic survival, and proliferation and dose-dependently decreases estrogen-stimulated estrogen response element transcriptional activity (Antoon et al., 2011)
	Sensitizes HNSCC to radiation-induced cytotoxicity (Sinha et al., 2011)
	Induces proteasomal degradation of SK1 and apoptosis in androgen-sensitive prostate cancer cells Loveridge et al., 2010
	Triggers lysosomal degradation of SK1 rather than direct inhibition of SK1 in HEK293 cells and lung cancer cells (Ren et al., 2010)
	Increases Tau hyperphosphylation by glutamate in neuroblastoma cells (Lopez-Tobon et al., 2009)
SKI-I and SKI-II	Induce apoptosis of T-LGL leukemia PBMCs but not normal PBMCs (Zhang et al., 2008)
SKI-I, SKI-II, and SKI-V	Reduce rate of solid tumor growth when JC mammary adenocarcinoma cells injected into BALB/c mice (French et al., 2006)
SKI-I, SKI-II, SKI-III, SKI-IV, and SKI-V	Induce apoptosis of bladder cancer cells (French et al., 2003)
SKI-II and DMS	Decreases TNF-α-stimulated DNA synthesis in glioma cells (Radeff-Huang et al., 2007)
DMS	Sensitizes prostate cancer cells to gamma-irradiation (Nava et al., 2000)
	Enhances cytochrome c and Smac/DIABLO release in leukemia cells (Cuvillier and Levade, 2001)
	Inhibits proliferation of Gastric Carcinoma cells (Ren et al., 2002)
	Increases sensitivity to imatinib-induced apoptosis in resistant chronic myeloid leukemia cells and returns BCR–ABL to baseline levels (Marfe et al., 2011)
DMS and DHS	Enhance apoptosis of leiomyoma cells in response to serum starvation by antagonizing the effects of endothelin-1 (Raymond et al., 2006)
DHS	Protects rhabdomyosarcoma cells from from TRAIL-induced apoptosis (Petak et al., 2003)
Phytosphingosine derivatives	Methylphytosphingosine induces apoptosis, decreases phophorylation of ERK, and inhibits daunorubicin-induced ERK activation in leukemia cells (Pewzner-Jung et al., 2010)
	Dimethylphytosphingosine suppresses cell growth and induces apoptosis in human leukemia cells; decreases phophorylation of ERK and inhibits daunorubicin-induced ERK activation to enhance its cytotoxicity in leukemia cells (Park et al., 2010)
F-12509a	Enhances imatinib-induced apoptosis in leukemia cells (Bonhoure et al., 2008)
B-5354c	Induces apoptosis and sensitizes prostate cancer cells to docetaxel and camptothecin to reduce tumor size in vivo (Pchejetski et al., 2005, 2008)

Inhibition of SK was also shown to reduce cell colony formation and activate caspase-3 in temozolomide-resistant glioblastoma multiforme cells (Bektas et al., 2009). And inhibition of SK1 or the S1P receptors can inhibit cell growth in camptothecin-resistant prostate cancer cells (Akao et al., 2006). Interestingly, treatment with SK inhibitors SK1-I or SK1-II, can selectively induce apoptosis in T cell large granular lymphocyte leukemia cells but not in normal peripheral blood mononuclear cells (Zhang et al., 2008). Likewise, treatment with SK1-I potently induced apoptosis in leukemic blasts isolated from patients with acute myelogenous leukemia but again was relatively sparing of normal peripheral blood mononuclear leukocytes (Paugh et al., 2008). In vivo, treatment with SK1-I induced apoptosis and reduced tumor vascularization leading to a markedly decreased tumor growth rate of glioblastoma xenografts and enhancing the survival of mice harboring LN229 intracranial tumors (Kapitonov et al., 2009). Recently, production of an aspirinyl analog of SK1-I has been shown to increase its half-life for better in vivo delivery (Sharma et al., 2010).

Moreover, combining SK1-I and the proteasome inhibitor, bortezomib, synergistically increases apoptosis, decreases colony formation, and induces downregulation of BCR/ABL and Mcl-1 in human leukemia cells and was even effective in imatinib-resistant cells (Li et al., 2011). Whereas, SK1-II on its own has been shown to induce proteasomal degradation of SK1 in human pulmonary artery smooth muscle cells, androgen-sensitive LNCaP prostate cancer cells, MCF-7 and MCF-7 HER2 breast cancer cells (Loveridge et al., 2010). Although FTY720 has been reported to be more proficient at inhibiting SK2, some studies claim that FTY720 and its analogue (S)-FTY720 vinylphosphonate behave as typical SK1 inhibitors and induce proteasomal degradation of SK1 and apoptosis in breast and prostate cancer cells (Tonelli et al., 2010) as well as preventing S1P-stimulated rearrangement of actin in MCF-7 cells (Lim et al., 2011). Of note, it has also been reported that FTY720 inhibits SK1 to increase radiation sensitivity of prostate cancer tumor xenografts to reduce tumor growth and metastasis in mice (Pchejetski et al., 2010).

The first putative-SK inhibitor to enter phase 1 clinical trials for advanced solid tumor therapy was Safingol, 1-threo-dihydrosphingosine, in 2011, which was found to decrease plasma levels of S1P and prolong stable disease/inhibit cancer progression when used in combination with cisplatin in a small cohort of patients (Dickson et al., 2011). However, new SK1 inhibitors with increased specificity are being developed. As the upregulation of SK1 in HNSCC cells correlates with their radioresistance, a group developed gold nanorod-SK1 siRNA nanocomplexes as a method of radiosensitization of head and neck cancer to achieve over 50% greater tumor regression as compared to controls (Masood et al., 2012).

Recently developed amidine-based SK1 nanomolar inhibitors were shown to significantly reduce S1P levels in human leukemia U937 cells (Kennedy et al., 2011). Two sphingo-guanidine salts named LCL146 and LCL135 have been reported to be cytotoxic to cancer cells and decrease the migration rate of human prostate (DU145) cells through their inhibition of SK (Sharma, 2011). Interestingly, the anticancer agent resveratrol and its dimers, ampelopsin A and balanocarpol, have been shown to downregulate SK1 in MCF-7 human breast cancer cells (Lim et al., 2012). The AKT inhibitor, BI-69A11, was shown to reduce NF-κB signaling through its inhibition of SK1 to inhibit melanoma growth (Feng et al., 2011). All of these results indicate that an SK1 inhibitor could be a beneficial addition to the chemotherapeutic strategies employed for treating various types of tumors, leading to extensive review of the subject (Cuvillier, 2007, 2008; Edmonds et al., 2011; Gault and Obeid, 2011; Pallis, 2002; Pitman and Pitson, 2010; Pyne et al., 2011; Shida et al., 2008b; Uddin and Platanias, 2008).

2.1. Increased expression of SK1 in cancer

The potent oncogene, Bcr/Abl, was shown to upregulate SK1 through various pathways leading to myeloid cell leukemia-1 (Mcl-1) expression in chronic myelogenous leukemia (Li et al., 2007a). In multiple myeloma cells, prokineticin-1, an endocrine gland-derived vascular endothelial growth factor, upregulates SK1 and Mcl-1 in addition to its activation of the MAPK, PI3K-AKT, and Jak-STAT3 pathways in order to protect cells from starvation-induced apoptosis (Li et al., 2010). Transcriptional upregulation of SK1 has also been associated with the tumorigenic phenotype of transgenic proerythroblasts in a mouse model of erythroleukemia (Le Scolan et al., 2005). It has also been shown in MKN1 gastric cancer cells that lysophosphatidic acid signaling through ERK-1 activation upregulates SK1 to exert its proliferative effects (Ramachandran et al., 2010). Increased SK1 expression has been shown to correlate with carcinogenicity in several other cancers including acute leukemia (Sobue et al., 2008), endometrial cancer (Knapp et al., 2010), prostate cancer (Malavaud et al., 2010), HNSCC (Facchinetti et al., 2010; Shirai et al., 2011; Sinha et al., 2011), and thyroid cancer (Guan et al., 2011a).

As for the possible mechanisms for this transcriptional upregulation of SK1 in cancer, in human glioblastoma cells, upregulation of SK1 transcription is induced by IL-1 and is mediated by a novel AP-1 element located within the first intron of the SK1 gene, and this upregulation of SK1 can be blocked by inhibition of JNK (Paugh et al., 2009). Rapidly growing tumors often undergo hypoxia, and this results in HIF-1α and HIF-2α production which has been shown to stimulate two putative hypoxia-inducible factor-responsive-elements in the SK1 promoter region (Anelli et al., 2008; Schwalm et al., 2008). Another pathway of SK1 upregulation has been found in rat pheochromocytoma cells in which nerve growth factor is able to upregulate SK1 expression in a pathway involving transcription factor specificity protein 1 (Sobue et al., 2005). Therefore, it seems that various transcription factors can promote SK1 expression for it to enact its oncogenic effects (Table 7.1).

2.2. SK1 as a prognostic indicator

The elevation of SK1 in various cancers implies that it could be a molecule of prognostic value. In fact, SK1 expression is higher in estrogen receptor (ER) negative breast tumors and its high expression correlates with worse survival and a higher chance of metastasis in patients, with carcinoma cells being the major source of SK1 expression in the tumors (Ruckhaberle et al., 2008). Furthermore, expression of SK1, but not SK2, in glioblastoma multiforme tissue correlated with short patient survival. Patients with low SK1 survived a median of 357 days, whereas those with high levels of SK1 survived a median of 102 days (Van Brocklyn et al., 2005). Also in glioblastoma multiforme, it was found that SK is necessary for basal activity of the urokinase plasminogen activator system and glioma cell invasion (Young et al., 2009). High expression of SK1 in non-Hodgkin lymphomas (Bayerl et al., 2008), astrocytomas (Li et al., 2008), gastric cancer (Li et al., 2009), salivary gland carcinoma (Liu et al., 2010a), esophageal carcinoma (Pan et al., 2011), non-small-cell-lung cancer (Song et al., 2011), and head and neck squamous cell carcinoma (Facchinetti et al., 2010) also correlates with grade and shorter survival time. Thus not only is SK1 over-expressed in cancer, but it can also enhance tumor invasiveness and serve as a prognostic indicator.

Patients with ER-positive breast tumors that expressed high levels of both cytoplasmic SK1 and ERK-1/2 had significantly shorter recurrence times than those that expressed low levels of cytoplasmic SK1 and cytoplasmic ERK-1/2, with a difference in recurrence time of 10.5 years (Watson et al., 2010). In contrast, for patients with ER-positive breast cancer, a low HER1–3/SK1 expression ratio is correlated with improved prognosis compared to patients that have a high HER1–3/SK1 expression ratio. This is thought to occur since although the HER2 oncogene increases SK1 expression in these tumors, then SK1 limits HER2 expression in a negative-feedback manner, this phenomenon has been termed “oncogene tolerance” (Long et al., 2010).

Although treatment failure and Gleason score were found to correlate with tumor SK1 activity in prostate cancer patients (Malavaud et al., 2010), patients with prostate cancer were shown to have lower circulating levels of S1P and decreased SK1 activity in their erythrocytes when compared to healthy subjects. They found that decreased circulating S1P was an early marker of prostate cancer progression to hormonal unresponsiveness and correlated with prostate-specific antigen levels and lymph node metastasis as well as mortality (Nunes et al., 2012). Thus, implying that tumor versus serum SK1/S1P levels may have opposite prognostic indications in certain cancer patients.

2.3. Elevated SK1 confers resistance to chemotherapy

With increased SK1 correlated to poor prognosis in several cancers, it seems likely that overexpression of SK1 may confer resistance to current chemotherapeutics. In fact, imatinib-induced apoptosis in K562 human chronic myeloid leukemia (CML) cells involves an increase in C18-ceramide; however, imatinib-resistant cells have increased expression of SK1 which elevates their S1P to C18-ceramide ratio sixfold and prevents apoptosis (Baran et al., 2007). Furthermore, impaired SK1/S1P signaling enhanced the growth-inhibitory effects of nilotinib against 32D/T315I-Bcr–Abl1-derived mouse allografts, demonstrating SK1 inhibition as a potential chemosensitizer in CML (Salas et al., 2011).

Likewise, it has been found that camptothecin-resistant prostate cancer PC3 cells show higher expression of SK1 and S1PR1 and -3 receptors when compared to camptothecin-sensitive, LNCaP cells, and inhibition of SK1 or the S1P receptors can inhibit cell growth in PC3 cells (Akao et al., 2006). Increased SK1 also appears to be involved in cisplatin resistance in lung cancer cells (Min et al., 2005), daunorubicin resistance in leukemia cells (Sobue et al., 2008), N-(4-hydroxyphenyl) retinamide resistance in ovarian carcinoma cells (Illuzzi et al., 2010), oxaliplatin resistance in colon cancer cells (Nemoto et al., 2009), and doxorubicin, docetaxel, and tamoxifen resistance in breast cancer cells (Ling et al., 2009; Nava et al., 2002; Sukocheva et al., 2009). Accordingly, inhibition or knockdown of SK1 has been found to increase drug efficacy in acute myeloid leukemia cells (Bonhoure et al., 2006), chronic myelogenous leukemia cells (Li et al., 2007a; Salas et al., 2011), colon cancer cells (Nemoto et al., 2009), and ovarian carcinoma cells (Illuzzi et al., 2010).

2.4. SK1 functions to suppress apoptosis of cancer cells

In vivo cells undergo apoptosis, cell cycle arrest, or senescence when their survival and growth no longer benefit the organism. When aberrant signaling overrides this response, these cells can become tumorigenic. Bioactive sphingolipids have been shown to be important signaling molecules for such cell fate decisions and SK is at crucial point for contributing to this since its product, S1P, is a prosurvival signaling molecule. Here, we review cell studies that illustrate the prosurvival role of the SK1/S1P pathway, in order to show how their functions at the cellular level might contribute to tumor initiation at the organismal level. For instance, in an elegant study by Le Scolan et al., they showed that overexpression of SK1 in erythroleukemic cells increased proliferation, clonogenicity, and resistance to apoptosis in reduced serum by a mechanism involving ERK1/2 activation and the PI3K/Akt pathway leading to increased tumorigenicity when engrafted in vivo. However, expression of a dominant-negative mutant of SK1 or treatment with a pharmacological inhibitor was able to reduce both cell growth and apoptosis-resistance in the tumorigenic erythroleukemia cells (Le Scolan et al., 2005). Furthermore, a novel hepatic growth factor that protects hepatocarcinoma cells, hepatopoietin Cn, was recently found to work through SK1 signaling to upregulate Mcl-1 expression to induce its antiapoptotic effects (Chang et al., 2010).

Work from our lab has shown in breast cancer cells, knockdown of SK1 causes cell cycle arrest and induces apoptosis involving effector caspase activation, cytochrome c release, and Bax oligomerization in the mitochondrial membrane, thus placing SK1 knockdown upstream of the mitochondrial pathway of apoptosis (Taha et al., 2006a). Knockdown or downregulation of SK1 has also been shown to lead to ceramide accumulation, decreased S1P, and apoptosis in other cancer cell lines including prostate cancer (Pchejetski et al., 2005), glioblastoma (Van Brocklyn et al., 2005), chronic myelogenous leukemia cell lines (Li et al., 2007a), and neuroblastoma cells (Gomez-Brouchet et al., 2007). (Table 7.2) Likewise, when SK1 is overexpressed in breast cancer cells, it results in enhanced proliferation and resistance to tamoxifen-induced growth arrest and apoptosis (Sukocheva et al., 2009) and it confers resistance to other anticancer drugs, sphingosine, and TNF-α (Nava et al., 2002). Interestingly, other studies in breast cancer cells have suggested that overexpression of SK1 stimulates autophagy. And it was shown that nutrient starvation stimulates SK1 activity and that SK1/S1P-induced autophagy can protect cells from apoptosis under nutrient-deplete conditions (Lavieu et al., 2006). Overexpression of SK1 was also shown to upregulate the Mcl-1 protein in multiple myeloma cells, leading to increased cell proliferation and survival. Furthermore, activation of SK1 was shown to mediate the suppressive effects of IL-6 on multiple myeloma cell apoptosis (Li et al., 2007b). Overexpression or activation of SK1 has been shown to suppress the apoptotic response in a variety of other cell lines including rat pheochromocytoma cells (Edsall et al., 2001; Misasi et al., 2001) and Jurkat, U937, and HL-60 leukemia cells (Cuvillier and Levade, 2001) (Table 7.3).

SK1 status appears to be tightly linked with caspase-dependent apoptosis. KO of SK1, but not SK2, was shown to enhance caspase-3-mediated apoptosis induced by staurosporine (Hofmann et al., 2008). In several human leukemia cell lines (Jurkat, U937, and HL-60), S1P can inhibit caspase-3 activation by inhibiting translocation of cytochrome c and Smac/DIABLO from mitochondria to cytosol (Cuvillier and Levade, 2001). In addition to its prosurvival product S1P, SK1 also regulates the level of its proapoptotic substrate, sphingosine. Recent studies have shown that binding of sphingosine to 14-3-3 proteins renders them phosphorylatable at the dimer interface, an event that abolishes their prosurvival signaling. Thus by reducing the availability of sphingosine for interaction with 14-3-3, SK1 can inhibit cell death (Woodcock et al., 2010), underlining the important position of SK1 in balancing growth and death signaling sphingolipids.

2.5. SK1 affects apoptosis through controlling the cellular ceramide:S1P ratio

To illustrate this balancing act controlled by SK1, in noncancerous human endothelial cells, TNF activates sphingomyelinase and generates ceramide, but these cells are resistant to TNF-induced apoptosis since TNF also activates SK and increases S1P. However, in a transformed endothelial cell line (C11), TNF fails to activate SK and thus induces apoptosis, which treatment with S1P can prevent (Xia et al., 1999). Clearly, with its important role of regulating the balance between prodeath and prosurvival sphingolipids, modifications of SK1 could lead to aberrantly enhanced cell survival leading to tumorigenesis.

In fact, overexpression of SK1 in prostate cancer cells can increase resistance to chemotherapy by decreasing the ceramide:S1P ratio (Pchejetski et al., 2005). Furthermore, a significantly higher ceramide:S1P ratio was found in melanoma cells that are resistant to ceramide- and Fas-induced death than those sensitive to programmed cell death. Overexpression of SK1-reduced sensitivity of A-375 melanoma cells to Fas- and ceramide-mediated apoptosis while siRNA of SK1 decreased resistance of Mel-2a cells to apoptosis. Interestingly, overexpression of prosurvival protein Bcl-2 in A-375 cells stimulated SK1 expression and activity, and downregulation of Bcl-2-reduced SK1 expression (Bektas et al., 2005). Furthermore, it has been shown that Dasatinib (Gencer et al., 2011) and resveratrol (Kartal et al., 2011) induce apoptosis through downregulating SK1 and upregulating ceramide synthase genes to increase the ceramide:S1P ratio in K562 CML cells (Gencer et al., 2011; Kartal et al., 2011). This opposing regulation of ceramide synthases and SK1 has also been observed in breast cancer samples (Erez-Roman et al., 2009). Moreover, our work has shown that through KO of SK1, an increased ceramide:S1P ratio is protective from thymic lymphoma development and may initiate tumor cell senescence in vivo (Heffernan-Stroud et al., 2012) (Table 7.4).

2.6. SK1 inhibitors challenge cancer cell viability to enhance chemotherapy

Various studies have employed pharmacological inhibitors to show that inhibition of SK decreases cancer cell viability in neuroblastoma cells (Tavarini et al., 2000), glioblastoma cells (Bektas et al., 2009), and several types of leukemia cells (Cuvillier and Levade, 2001; Paugh et al., 2008; Zhang et al., 2008) as well as in various solid tumor cell lines (French et al., 2003). SK inhibitors (outlined in Table 7.5). were found to be antiproliferative toward a panel of tumor cell lines, including lines with the multidrug resistance phenotype because of overexpression of either P-glycoprotein or multidrug resistance phenotype 1, and they could induce apoptosis (French et al., 2003). Furthermore, resistance to doxorubicin- or etoposide-induced apoptosis in acute myeloid leukemia cells was attributed to sustained SK1 activity and reduced ceramide accumulation; likewise, apoptosis could be restored in these cells by treatment with an SK inhibitor, F-12509a (Bonhoure et al., 2006). Later results using the same inhibitor suggest that SK1 regulates imatinib-induced apoptosis in primary cells from CML patients, as inhibition of SK1 could kill both imatinib-sensitive and -resistant cells. This work places SK1 downstream of the Bcr–Abl/Ras/ERK pathway inhibited by imatinib and upstream of Bcl-2 family members (Bonhoure et al., 2008).

Table 7.5

Effects of SK1 inhibitors in cancer

Inhibitor	Mechanism of action
SKI-I (French et al., 2003)	Induces apoptosis and suppresses tumor growth rate and vascularization of glioblastoma xenografts (Kapitonov et al., 2009)
	Sensitizes pancreatic cancer cells to gemcitabine induced death (Guillermet-Guibert et al., 2009)
	Decreases growth and survival while enhancing apoptosis and Bcl-2 cleavage in leukemia cells (Paugh et al., 2008)
	Suppresses lysophosphatidic acid mediated proliferation in gastric cancer cells (Ramachandran et al., 2010)
	Suppressed S1P levels, decreased hemangiogenesis and lymphangiogenesis, reduced metastases to lymph nodes and lungs, and decreased overall tumor burden in murine model of breast cancer (Nagahashi et al., 2012)
	Blocked cell proliferation and induced apoptotic cell death in glioblastoma multiforme neurosphere cell line (Estrada-Bernal et al., 2010)
	Increases oxaliplatin-induced cytotoxicity in colon cancer (Nemoto et al., 2009)
	Increases sensitivity of non-small-cell-lung cancer cells to doxorubicin- or docetaxel-induced apoptosis (Song et al., 2011)
	Rescues daunorubicin sensitivity in leukemia cells (Sobue et al., 2008)
	Enhances ability of imatinib mesylate to inhibit cell growth, survival, and clonogenic potential in leukemia cells (Ricci et al., 2009)
	Sensitizes prostate and breast cancer cells to docetaxel (Sauer et al., 2009) (as does B-5354c)
	Reduces proliferation and sensitizes ovarian cancer cells to cytotoxic effect of N-(4-hydroxyphenyl) retinamide (Illuzzi et al., 2010a, 2010b)
	Impedes DHT-induced cell proliferation, DNA synthesis, and PSA secretion in prostate cancer cells (Dayon et al., 2009)
SKI-II (French et al., 2003)	Blocks expression of urokinase plasminogen activator and its receptor to inhibit glioma cell invasion (Young et al., 2009)
	Prevents accumulation and transcriptional activity of HIF-1α in prostate, brain, breast, and lung cancer cells (Ader et al., 2008)
	Reduces cell colony formation and activates of caspase-3 in glioblastoma cells (Bektas et al., 2009)
	Antimigratory and toxic effects on human glioma and glioma stem cells (Mora et al., 2010)
	Blocks breast cancer viability, clonogenic survival, and proliferation and dose-dependently decreases estrogen-stimulated estrogen response element transcriptional activity (Antoon et al., 2011)
	Sensitizes HNSCC to radiation-induced cytotoxicity (Sinha et al., 2011)
	Induces proteasomal degradation of SK1 and apoptosis in androgen-sensitive prostate cancer cells Loveridge et al., 2010
	Triggers lysosomal degradation of SK1 rather than direct inhibition of SK1 in HEK293 cells and lung cancer cells (Ren et al., 2010)
	Increases Tau hyperphosphylation by glutamate in neuroblastoma cells (Lopez-Tobon et al., 2009)
SKI-I and SKI-II	Induce apoptosis of T-LGL leukemia PBMCs but not normal PBMCs (Zhang et al., 2008)
SKI-I, SKI-II, and SKI-V	Reduce rate of solid tumor growth when JC mammary adenocarcinoma cells injected into BALB/c mice (French et al., 2006)
SKI-I, SKI-II, SKI-III, SKI-IV, and SKI-V	Induce apoptosis of bladder cancer cells (French et al., 2003)
SKI-II and DMS	Decreases TNF-α-stimulated DNA synthesis in glioma cells (Radeff-Huang et al., 2007)
DMS	Sensitizes prostate cancer cells to gamma-irradiation (Nava et al., 2000)
	Enhances cytochrome c and Smac/DIABLO release in leukemia cells (Cuvillier and Levade, 2001)
	Inhibits proliferation of Gastric Carcinoma cells (Ren et al., 2002)
	Increases sensitivity to imatinib-induced apoptosis in resistant chronic myeloid leukemia cells and returns BCR–ABL to baseline levels (Marfe et al., 2011)
DMS and DHS	Enhance apoptosis of leiomyoma cells in response to serum starvation by antagonizing the effects of endothelin-1 (Raymond et al., 2006)
DHS	Protects rhabdomyosarcoma cells from from TRAIL-induced apoptosis (Petak et al., 2003)
Phytosphingosine derivatives	Methylphytosphingosine induces apoptosis, decreases phophorylation of ERK, and inhibits daunorubicin-induced ERK activation in leukemia cells (Pewzner-Jung et al., 2010)
	Dimethylphytosphingosine suppresses cell growth and induces apoptosis in human leukemia cells; decreases phophorylation of ERK and inhibits daunorubicin-induced ERK activation to enhance its cytotoxicity in leukemia cells (Park et al., 2010)
F-12509a	Enhances imatinib-induced apoptosis in leukemia cells (Bonhoure et al., 2008)
B-5354c	Induces apoptosis and sensitizes prostate cancer cells to docetaxel and camptothecin to reduce tumor size in vivo (Pchejetski et al., 2005, 2008)

Inhibition of SK was also shown to reduce cell colony formation and activate caspase-3 in temozolomide-resistant glioblastoma multiforme cells (Bektas et al., 2009). And inhibition of SK1 or the S1P receptors can inhibit cell growth in camptothecin-resistant prostate cancer cells (Akao et al., 2006). Interestingly, treatment with SK inhibitors SK1-I or SK1-II, can selectively induce apoptosis in T cell large granular lymphocyte leukemia cells but not in normal peripheral blood mononuclear cells (Zhang et al., 2008). Likewise, treatment with SK1-I potently induced apoptosis in leukemic blasts isolated from patients with acute myelogenous leukemia but again was relatively sparing of normal peripheral blood mononuclear leukocytes (Paugh et al., 2008). In vivo, treatment with SK1-I induced apoptosis and reduced tumor vascularization leading to a markedly decreased tumor growth rate of glioblastoma xenografts and enhancing the survival of mice harboring LN229 intracranial tumors (Kapitonov et al., 2009). Recently, production of an aspirinyl analog of SK1-I has been shown to increase its half-life for better in vivo delivery (Sharma et al., 2010).

Moreover, combining SK1-I and the proteasome inhibitor, bortezomib, synergistically increases apoptosis, decreases colony formation, and induces downregulation of BCR/ABL and Mcl-1 in human leukemia cells and was even effective in imatinib-resistant cells (Li et al., 2011). Whereas, SK1-II on its own has been shown to induce proteasomal degradation of SK1 in human pulmonary artery smooth muscle cells, androgen-sensitive LNCaP prostate cancer cells, MCF-7 and MCF-7 HER2 breast cancer cells (Loveridge et al., 2010). Although FTY720 has been reported to be more proficient at inhibiting SK2, some studies claim that FTY720 and its analogue (S)-FTY720 vinylphosphonate behave as typical SK1 inhibitors and induce proteasomal degradation of SK1 and apoptosis in breast and prostate cancer cells (Tonelli et al., 2010) as well as preventing S1P-stimulated rearrangement of actin in MCF-7 cells (Lim et al., 2011). Of note, it has also been reported that FTY720 inhibits SK1 to increase radiation sensitivity of prostate cancer tumor xenografts to reduce tumor growth and metastasis in mice (Pchejetski et al., 2010).

The first putative-SK inhibitor to enter phase 1 clinical trials for advanced solid tumor therapy was Safingol, 1-threo-dihydrosphingosine, in 2011, which was found to decrease plasma levels of S1P and prolong stable disease/inhibit cancer progression when used in combination with cisplatin in a small cohort of patients (Dickson et al., 2011). However, new SK1 inhibitors with increased specificity are being developed. As the upregulation of SK1 in HNSCC cells correlates with their radioresistance, a group developed gold nanorod-SK1 siRNA nanocomplexes as a method of radiosensitization of head and neck cancer to achieve over 50% greater tumor regression as compared to controls (Masood et al., 2012).

Recently developed amidine-based SK1 nanomolar inhibitors were shown to significantly reduce S1P levels in human leukemia U937 cells (Kennedy et al., 2011). Two sphingo-guanidine salts named LCL146 and LCL135 have been reported to be cytotoxic to cancer cells and decrease the migration rate of human prostate (DU145) cells through their inhibition of SK (Sharma, 2011). Interestingly, the anticancer agent resveratrol and its dimers, ampelopsin A and balanocarpol, have been shown to downregulate SK1 in MCF-7 human breast cancer cells (Lim et al., 2012). The AKT inhibitor, BI-69A11, was shown to reduce NF-κB signaling through its inhibition of SK1 to inhibit melanoma growth (Feng et al., 2011). All of these results indicate that an SK1 inhibitor could be a beneficial addition to the chemotherapeutic strategies employed for treating various types of tumors, leading to extensive review of the subject (Cuvillier, 2007, 2008; Edmonds et al., 2011; Gault and Obeid, 2011; Pallis, 2002; Pitman and Pitson, 2010; Pyne et al., 2011; Shida et al., 2008b; Uddin and Platanias, 2008).

3. SK1/S1P SIGNALS FOR TUMOR GROWTH AND SPREAD

Not only does SK1 protect cell survival, it also plays an important role in signaling for cell proliferation. It appears that SK1-dependent Akt activation plays a significant role in cell proliferation induced by TNF-α in 1321N1 glioblastoma cells (Radeff-Huang et al., 2007) and by 11,12-epoxyeicosatrienoic acid in human umbilical vein endothelial cells (Yan et al., 2008). SK1 has also been found to act as a proproliferative oncogenic kinase in thyroid cancer where it appears to activate the Akt/glycogen synthase kinase-3β/β-catenin pathway (Guan et al., 2011a). Moreover, various growth factors have been shown to function in part by upregulating SK1 expression and/or activity, these include PDGF (Francy et al., 2007), NGF (Edsall et al., 2001), IGF-BP3 (Martin et al., 2009), VEGF (Shu et al., 2002), and EGF (Doll et al., 2005). VEGF induces DNA synthesis in a pathway which sequentially involves protein kinase C, SK, Ras, Raf, and ERK1/2 (Shu et al., 2002). And epidermal growth factor both acutely and transcriptionally upregulates SK1 in a MAPK-, PKC-, and PI3K-dependent manner. Accordingly, when cells are depleted of SK1, but not SK2, by siRNA, EGF-induced proliferation and migration are drastically reduced (Doll et al., 2005). Also through the ERK1/2 and PI3K signaling pathways, glial cell line-derived neurotrophic factor induces transcription of SK1 which leads to neurite formation, GAP43 expression, and cell growth in TGW human neuroblastoma cells and growth and anchorage-independent colony formation in a model of multiple endocrine neoplasia type 2 tumor (Murakami et al., 2007).

3.1. Hormonal regulation of SK1

Hormones have also been shown to activate SK1. Progesterone induced SK1 expression (30-fold) in the rat uterus during pregnancy in the glandular epithelium, vasculature, and myometrium. Overexpression of SK1 in a rat myometrial leiomyoma cell line (ELT3) resulted in increased levels of the cell cycle regulator cyclin D1 and increased myosin light-chain phosphorylation and increased proliferation rates (Jeng et al., 2007). Prolactin and 17-β estradiol were also shown to activate SK1, but not SK2, in MCF-7 breast adenocarcinoma cells both within minutes and within hours after STAT5-, PKC-, and MAPK-dependent promoter activation. Either inhibition of SK1 activation with glucocorticoids or direct knockdown of SK1 with siRNA could abolish the hormone-induced cell proliferation and migration observed in the breast cancer cells (Doll et al., 2007).

Interestingly, the SK1-II inhibitor binds the ER directly in the antagonist ligand-binding domain. SK1-II was shown to dose-dependently decrease estrogen-stimulated estrogen response element transcriptional activity and diminish mRNA levels of the ER regulated genes progesterone receptor and steroid derived factor-1 to block breast cancer viability, clonogenic survival, and proliferation (Antoon et al., 2011). In addition to hormonal regulation of SK1 in breast cancer, dihydrotestosterone was shown to signal for proliferation through an androgen receptor/PI3K/Akt-dependent stimulation of SK1 in hormone-sensitive prostate cancer cells (Dayon et al., 2009). This hormonal regulation of SK1 has even been shown to be important in neuroblastoma cell lines in which 17β-estradiol upregulates SK1 to protect from Tau hyperphosphorylation and glutamate toxicity (Lopez-Tobon et al., 2009).

3.2. SK1 functions to enhance proliferation of cancer cells

SK1 can signal to a cell to proliferate through its product S1P. For instance, in the presence of retinoic acid receptor alpha, retinoic acid upregulates neutral sphingomyelinase and downregulates SK1 to inhibit cell growth, but in the absence of that receptor, retinoic acid activates SK1 and promotes cell growth through S1P (Somenzi et al., 2007). Addition of S1P increases proliferation in neuroblastoma cells (Tavarini et al., 2000) and extracellular S1P is required for basic fibroblast growth factor-induced growth stimulation of astrocytes (Bassi et al., 2006). In a glioma cell line, S1P induces expression of early growth response-1, an essential transcription factor for fibroblast growth factor 2 (Sato et al., 1999). This was the first report of S1P acting as a transcription factor, and its role as a molecule capable of epigenetic regulation was more recently elucidated when it was shown to regulate histone acetylation (Hait et al., 2009). Clearly, SK1 and its product S1P could employ their potent proliferative signaling to enhance tumor growth.

3.3. SK1 can stimulate angiogenesis to increase blood supply to tumors

In order to maintain rapid growth, tumors require a sufficient vascular supply. As tumor cells rapidly proliferate, they undergo hypoxia which results in their signaling for new blood vessels to be formed. Hypoxia increases SK1 mRNA, protein, and activity in human glioma U87MG cells, followed by intracellular S1P production and S1P release and this could all be inhibited with knockdown of HIF-2α by siRNA (Anelli et al., 2008). Similar results were observed in endothelial cells themselves, where their production of HIF-1α and HIF-2α and their binding to two putative hypoxia-inducible factor and -responsive elements in the SK1 promoter led to increases in SK1 (Schwalm et al., 2008). Furthermore, our lab showed that conditioned medium from hypoxia-treated tumor cells results in neoangiogenesis in human umbilical vein endothelial cells in an S1P receptor-dependent manner (Anelli et al., 2008). Interestingly, inhibition of SK1 activity can prevent the accumulation of HIF-1α and its transcriptional activity in several human cancer cell lineages including prostate, brain, breast, kidney, and lung (Ader et al., 2008). Thus, SK1 signaling appears to play an important role in creating new vasculature in order for oxygen and nutrients to be delivered to oxygen-starved tumors.

In addition to angiogenesis, SK1 has recently been shown to play a role in lymphangiogenesis. SK1 overexpression in HEK cells or its downregulation in glioma or breast cancer cells modulated extracellular S1P levels accordingly, which in turn increased or decreased both migration and tube formation in cocultured vascular or lymphatic endothelial cells. Furthermore, S1P initiated endothelial cell sprouting in three-dimensional collagen matrices (Anelli et al., 2010). Activation of SK1 has been shown to be necessary for the angiogenic effects of 11,12-epoxyeicosatrienoic acid, a product of cytochrome p450 epoxygenases (Yan et al., 2008), TNF-α (Xia et al., 1998), and VEGF (Kolmakova et al., 2009). Likewise, inhibition of SK has been shown to be responsible for the antiangiogenic properties of phenoxodiol (2H-1-benzopyran-7-0, 1,3-[4-hydroxyphenyl]), an anticancer drug undergoing clinical trials for its apoptotic properties (Gamble et al., 2006), and alphastatin, the potent antiangiogenic molecule that has been shown to inhibit gastric cancer angiogenesis (Chen et al., 2006; Li and Chen, 2009). In NIH3T3 cells with dominant negative CBF1, SK1 is required for FGF1 export and acceleration of cell growth to produce highly angiogenic tumors in nude mice (Kacer et al., 2011). Therefore, the angiogenic signaling properties of SK1 appear to be important for vascularization, and hence expansion, of tumors.

3.4. SK1 may enhance tumor metastasis

In addition to its angiogenic and lymphangiogenic properties, other aspects of SK1 signaling that could potentially contribute to tumor metastasis have been elucidated. For instance, upregulation of SK1 led to a migratory response in endothelial cells (Schwalm et al., 2008) and promotes migration of thyroid follicular carcinoma cells (Bergelin et al., 2009). Whereas, depletion of SK1 can inhibit EGF-induced (Doll et al., 2005; Sarkar et al., 2005), as well as hormone-induced migration of breast cancer cells (Doll et al., 2007), and it can abrogate migration toward EGF in HEK293 cells (Hait et al., 2005). SK1 was also shown to be required for nucleotide-induced migration of renal mesangial cells (Meyer zu Heringdorf and Jakobs, 2007). Moreover, the antimetastasis molecule KAI1 has been shown to suppress migration of pancreatic carcinoma cells through downregulation of SK activity (Guo et al., 2006), and it has been shown to downregulate hepatocyte growth factor induced activation of SK1 in order to impede migration of hepatoma cells (Mu et al., 2008). These results indicate that the regulation of SK1 is a crucial regulator of cell migration.

The ability to migrate as well as invasiveness and the aggressiveness of tumor cells is what distinguishes a benign mass from malignant cancer. Importantly, NIH3T3 fibroblasts overexpressing SK acquire a transformed phenotype, as determined by focus formation, colony growth in soft agar, and the ability to form tumors in NOD/SCID mice, and SK activity appears to be involved in transformation mediated by the potent oncogene, H-Ras (Xia et al., 2000). Furthermore, TGF-β’s activation of SK and formation of S1P contribute to non-Smad signaling that could be important for progression of esophageal cancer (Miller et al., 2008), and S1P is able to induce invasion of glioblastoma cells (Bryan et al., 2008). Studies show that hepatocellular carcinoma cell migration and invasion may also be promoted by SK1/S1P signaling through S1PR1 (EDG1) (Bao et al., 2011). Moreover, S1P has been shown to signal through S1PR2 for apoptotic cell extrusion, which because live cells can also be extruded, it is postulated that this may serve as a mechanism for the invasion of cancer cells (Gu et al., 2011). Other studies have suggested that SK1 is a convergence point of multiple cell surface receptors for LPA, EGF, and S1P, which have all been implicated in the regulation of motility and invasiveness of cancer cells (Shida et al., 2008a). Thus, it appears that SK1 signaling can contribute to the malignant transformation of tumors and their metastasis.

4. CONCLUSION

Through both regulating ceramide and sphingosine accumulation within the cell and producing S1P, increased expression of SK1 results in an oncogenic phenotype. Decreased ceramide and sphingosine levels can confer chemotherapeutic resistance to cancer cells and increased S1P can stimulate the proliferation, vascularization, and metastasis of tumors. Therefore, SK1 represents an important target for chemotherapy.

Acknowledgments