The anti-osteosarcoma cell activity by the sphingosine kinase 1 inhibitor SKI-V

Sphingosine kinase 1 (SphK1) expression and activity are elevated in human osteosarcoma (OS) and is a promising target of therapy. SKI-V is a non-competitive and highly-efficient non-lipid SphK1 inhibitor. The potential anti-OS cell activity by the SphK1 inhibitor was studied here. In primary OS cells and immortalized cell lines, SKI-V robustly suppressed cell survival, growth and proliferation as well as cell mobility, and inducing profound OS cell death and apoptosis. The SphK1 inhibitor was however non-cytotoxic nor pro-apoptotic in human osteoblasts. SKI-V robustly inhibited SphK1 activation and induced accumulation of ceramides, without affecting SphK1 expression in primary OS cells. The SphK1 activator K6PC-5 or sphingosine-1-phosphate partially inhibited SKI-V-induced OS cell death. We showed that SKI-V concurrently blocked Akt-mTOR activation in primary OS cells. A constitutively-active Akt1 (ca-Akt1, S473D) construct restored Akt-mTOR activation and mitigated SKI-V-mediated cytotoxicity in primary OS cells. In vivo, daily injection of SKI-V potently suppressed OS xenograft tumor growth in nude mice. In SKI-V-administrated OS xenograft tissues, SphK1 inhibition, ceramide increase and Akt-mTOR inhibition were detected. Together, SKI-V exerts significant anti-OS activity by inhibiting SphK1 and Akt-mTOR cascades in OS cells.


INTRODUCTION
Cancer statistical studies have estimated that osteosarcoma (OS) accounts for about one-fifth of all primary bone malignancies [1,2]. The current treatment options for OS include a three-drug (cisplatin, doxorubicin, and methotrexate) chemotherapy regimen together with surgical OS resection [3][4][5][6][7]. In the past decades, the five-year overall survival (70-80%) of OS have reached plateau [3][4][5][6][7]. A large number of OS patients can bediagnosed at advanced stages or with recurrent tumors. The prognosis of these patients is often poor, possibly due to tumor metastasis [3][4][5][6][7]. It is therefore important to explore the key molecular targets of OS development and progression, and to explore novel and efficient therapeutic agents [3,7,8].
SphK1 expression is significantly elevated in clinical OS specimens [15]. Recent studies have supported that targeting SphK1 could produce significant anti-OS activity. Wei et al. showed that furowanin A inhibited SphK1 to exert significant anti-proliferative and pro-apoptotic activities in OS cells [16]. Phenoxodiol and doxorubicin co-treatment synergistically inhibited SphK1, suppressing OS cell growth [17]. Yao et al. showed that microRNA-3677 (miR-3677) silenced its target SphK1 to arrest OS cell growth [15]. Zhou et al. also found that microRNA-124, by silencing SphK1, suppressed the proliferation and invasion of OS cells [18].
SKI-V is a non-competitive, potent, and non-lipid SphK1 small molecule inhibitor with an IC 50 of 2 μM [19,20]. It inhibited SphK1 activity and S1P contents, causing apoptosis in T24 bladder cancer cells [20]. Intraperitoneal injection of SKI-V significantly inhibited mammary adenocarcinoma xenograft growth in immunocompetent BALB/c mice [20]. It however failed to induce immediate or delayed toxicity at doses up to 75 mg/kg in Swiss-Webster mice and BALB/c nude mice [20]. We here tested its potential anticancer effect in OS cells.
decreased the CCK-8 OD (Fig. 1A). At 1 μM SKI-V was ineffective (Fig. 1A). In addition, SKI-V required 48 h to exert a significant activity in C1 OS cells, showing a time-dependent response (Fig. 1A). Moreover, SKI-V-induced viability reduction in C1 primary cells lasted for 96 h (Fig. 1A). By employing Trypan blue staining assays, we showed that SKI-V dose-dependently induced C1 cell death (Fig. 1B). Furthermore, treatment with the SphK1 inhibitor (5-50 μM) largely inhibit the number of viable C1 cell colonies (results quantified in Fig. 1C), further supporting the cytotoxic activity of SKI-V against primary OS cells.
We also tested whether SKI-V could exert similar activity in other OS cells. C2 OS cells (derived from another primary OS patient) and the immortalized OS cell lines (U2OS and MG-63) were cultivated in complete medium and treated with SKI-V at 25 μM. The SphK1 inhibitor potently decreased viability (CCK-8 OD) in the immortalized and primary OS cells (Fig. 1G). Increased Trypan blue staining, indicating cell death, was observed in the SKI-V-treated primary and immortalized OS cells (Fig. 1H). Moreover, SKI-V robustly suppressed proliferation (tested by reduction of the EdUstained nuclei ratio, Fig. 1I) and migration (quantification from the "Transwell" assays, Fig. 1J) in the OS cells. Thus SKI-V exerted robust anti-OS cell activity.

human osteoblastic cells and primary osteoblasts
The potential effect of SKI-V in non-cancerous osteoblasts was examined next. As described, hFOB1.19 osteoblastic cells and the primary human osteoblasts ("osteoblasts") were cultured and treated with SKI-V (25 μM). As demonstrated, SKI-V did not significantly decrease the viability (CCK-8 OD) in hFOB1.19 osteoblastic cells and primary osteoblasts (Fig. 3A). Moreover, the ratio of EdU-stained nuclei was not significantly altered in SKI-V-treated hFOB1.19 osteoblastic cells and primary osteoblasts (Fig. 3B). In addition, results from Trypan blue staining assay ( Fig. 3C) and nuclear TUNEL staining assay (Fig. 3D) showed that SKI-V (25 μM) did not induce significant cell death and apoptosis in hFOB1.19 cells and primary osteoblasts.

SphK1 inhibition in SKI-V-treated OS cells
Since SKI-V is a SphK1 inhibitor. As shown, in primary OS cells, C1 and C2, treatment with SKI-V (25 μM) led to 80-90% inhibition of SphK1 activity (Fig. 4A). Conversely, the cellular ceramide contents increased over 4-5 folds (Fig. 4B). Notably, the SphK1 mRNA/ protein (Fig. 4C, D) expression was not significantly altered by the SKI-V in the primary OS cells.

Akt-mTOR inhibition in SKI-V-treated OS cells
A CRISPR/Cas9-SphK1-KO construct [15] was applied to stably knockout (KO) SphK1 in C1 primary OS cells ("SphK1-KO" cells) (Fig. 5A, B). As shown in C1 primary cells SphK1 KO induced significant ceramide production (Fig. 5C), cell death (Fig. 5D), and apoptosis (evidenced by the TUNEL-stained nuclei ratio increasing, Fig. 5E). Importantly, although SKI-V treatment did not affect SphK1 expression (Fig. 5A, B) and ceramide production in SphK1- D and I) were tested; Cell apoptosis was tested by nuclear TUNEL staining (E and J) and Annexin V-PI FACS (F and K) assays, with results quantified. The C1 primary OS cells were pretreated with the pan caspase inhibitor z-VAD-fmk (50 μM) or 0.25% DMSO for 30 min, followed by SKI-V (at 25 μM) treatment for 72 h, cell viability and death were examined by CCK-8 and Trypan blue staining assays, respectively (G). "Veh" stands for the vehicle control. Data were presented as mean ± standard deviation (SD, n = 5). *P < 0.05 vs. "Veh" group. # P < 0.05 vs. DMSO group (G). Experiments were repeated five times with similar results obtained. Scale bar = 100 μm (D and E).
KO cells (Fig. 5C), it did induce further cytotoxicity by enhancing cell death (Fig. 5D) and apoptosis (by measuring the ratio of the TUNEL-stained nuclei, Fig. 5E). Thus, SKI-V was still cytotoxic in SphK1-KO OS cells, further supporting the existence of SphK1independent mechanisms for SKI-V-induced anti-OS cell activity.
The anti-OS cell activity of SKI-V in vivo At last the C1 primary OS cells were s.c. injected to the flanks of nude mice. After 20 days, OS xenografts (100 mm 3 per tumor) were established (as "Day-0"). The xenograft-bearing nude mice were then assigned into two different groups. The treatment group received intraperitoneal (i.p.) injection of SKI-V (at 30 mg/kg body weight, daily administration for 18 consecutive days). The other group received vehicle control administration ("Veh"). The tumor growth curve, recording tumor volumes every six days, showed that daily SKI-V injection robustly suppressed OS xenograft growth in nude mice (Fig. 6A). The volumes of OS xenografts with SKI-V injection were dramatically lower than those of vehicle administration (Fig. 6A). The formula, (Tumor volume at Day-36 subtracting tumor volume at Day-0)/36, was utilized to calculate the estimated daily tumor growth was calculated under. Results showed that the growth of OS xenografts in the nude mice was largely inhibited following SKI-V injection. At Day-36, all mice were anaesthetized and decapitated, OS xenografts were isolated carefully and weighted. OS xenografts with SKI-V injection were significantly lighter than the xenografts with the vehicle control treatment (Fig.  6C). The mice body weights were however not significantly different among the two mice groups (Fig. 6D). No significant animal toxicities, including fever, vomiting, hair loss, and neurological symptoms, were detected in the SKI-V-treated nude mice.
At the experimental Day-5 and the experimental Day-10, one tumor of each group was isolated. Tissue lysates of the four isolated OS xenografts were analyzed. As shown, the SphK1 activity was significantly decreased in SKI-V-treated OS xenograft tissues (Fig. 6E). Conversely, increased ceramide contents were detected (Fig. 6F). SphK1 mRNA/protein expression (Fig. 6G, H) in tumor tissues was unchanged with SKI-V administration. Levels of phosphorylated-Akt (Ser-473) and phosphorylated-S6K1 were significantly inhibited in SKI-V-treated OS xenograft tissues Fig. 3 The effect of SKI-V in hFOB1.19 human osteoblastic cells and primary osteoblasts. The hFOB1.19 human osteoblastic cells or the primary human osteoblasts ("osteoblasts") were cultured and treated with SKI-V (25 μM) for applied time periods; cell viability, proliferation, death, and apoptosis were tested by CCK-8 (A), nuclear EdU staining (B), trypan blue staining (C) and nuclear TUNEL staining (D) assays, respectively. "Veh" stands for the vehicle control. Data were presented as mean ± standard deviation (SD, n = 5). Expression of the SphK1 mRNA (E) and protein (F) in the stable hFOB1.19 cells with the SphK1-expressing lentiviral construct ("oe-SphK1") or the empty vector ("Vec") was shown. The oe-SphK1 hFOB1.19 cells were treated with SKI-V (25 μM) or vehicle control for 72 h; cell viability and death were tested by CCK-8 (G) and trypan blue staining (H) assays, respectively. "N. S." stands for non-statistical difference (P > 0.05). *P < 0.05 vs. "Vec"/"Veh" group. Experiments were repeated five times with similar results obtained. Scale bar = 100 μm (B and D).

DISCUSSION
Three conventional OS subtypes are recognized, including osteoblastic, chondroblastic and fibroblastic OS [34,35]. Recent molecular profiling studies together with exploring the tissue banks have led to an increased understanding for the biology and pathological mechanisms of OS progression [3,7,8]. Yet, the clinical outcomes for the advanced OS is still far from satisfactory, with a poor median survival time particularly in patients with metastatic or recurrent OS [3,7,8]. It is therefore urgent to explore new therapeutic agents that effectively target OS [3,7,8].
Pharmacological or genetic means were applied to silence or inhibit SphK1, inhibiting OS cell growth and inducing cell apoptosis [15][16][17][18]. SKI-V is a non-competitive and highly-efficient SphK1 inhibitor. SKI-V intraperitoneal injection largely suppressed mammary adenocarcinoma xenograft growth in mice [20]. Gong et al. found that SKI-V facilitated bortezomib-induced ceramide production and apoptosis in pancreatic cancer cells [36].
We showed that SKI-V exerted significant anti-tumor activity in OS cells. In the primary and immortalized OS cells, treatment with the SphK1 inhibitor inhibited cell survival, growth, proliferation and cell mobility, and inducing profound cell death and apoptosis. It however failed to significant cytotoxicity in human osteoblasts. SKI-V inhibited SphK1 activation and induced ceramide accumulation, without affecting SphK1 expression in primary human OS cells. The SphK1 activator K6PC-5 or S1P partially inhibited SKI-Vinduced OS cell death and apoptosis. In vivo, daily injection of SKI-V robustly suppressed OS xenograft tumor growth. The experimental mice were well-tolerated to the treatment regimen, as no significant toxicities were reported. In SKI-V-treated OS xenograft tissues, SphK1 inhibition, ceramide increase, and apoptosis activation marker were detected.
We found that SKI-V inhibited Akt-mTOR activation in primary human OS cells, an effect that was parallel to SphK1 inhibition. Akt-mTOR inactivation was observed as well in OS xenograft tumor tissues after SKI-V injection. The caAkt1 restored Akt-mTOR activation and ameliorated SKI-V-induced OS cell death. Therefore, concurrent inhibition of Akt-mTOR cascade by the SphK1 inhibitor could explain its superior anti-OS activity. Indeed, SKI-V-induced cytotoxicity against primary human OS cells was significantly more potent than two other established SphK1 inhibitors (PF-543 and SKI-II).

Fig. 4 SphK1 inhibition in SKI-V-treated OS cells.
The primary OS cells ("C1"/"C2") were cultured in FBS-containing complete medium and treated with SKI-V (at 25 μM) for applied time periods, the relative SphK1 activity (A) and ceramide contents (B) were shown; expression of SphK1 mRNA and protein was tested by qRT-PCR (C) and western blotting (D) assays. "C1" and "C2" primary cells were cultured in FBScontaining compete medium and treated with 25 μM of SKI-V, SKI-II or PF-543 for applied time periods, cell viability, death, and apoptosis were tested by CCK-8 (E), trypan blue staining (F) and nuclear TUNEL staining (G) assays, respectively. The primary human OS cells ("C1"/"C2") were pretreated with K6PC-5 (20 μM), S1P (20 μM) or vehicle control (0.1% DMSO), followed by SKI-V (at 25 μM) treatment for 60 h/72 h, cell viability, death, and apoptosis were examined by CCK-8 (H) and trypan blue staining (I) and nuclear TUNEL staining (J) assays, respectively. "Veh" stands for the vehicle control. Data were presented as mean ± standard deviation (SD, n = 5). *P < 0.05 vs. "Veh". # P < 0.05 vs. SKI-V group (E-G). # P < 0.05 vs.  Cell culture U2OS and MG63, as well as the primary human OS cells (from two writteninformed consent primary patients, "C1/C2"), were from Dr. Cao [46,47]. Fig. 5 Akt-mTOR inhibition in SKI-V-treated OS cells. The primary OS cells ("C1"), stably expressing a CRISPR/Cas9-SphK1 KO construct ("SphK1-KO" cells), were treated with or without 25 μM of SKI-V, and cultured for applied time periods; control cells were stably expressing the control CRISPR/Cas9 vector ("Vec"); expression of listed proteins and mRNA were tested by western blotting (A and F) and qRT-PCR (B) assays. Cellular ceramide contents were examined (C); cell death and apoptosis were tested by trypan blue staining (D) and nuclear TUNEL staining assays (E), respectively, with results quantified. The primary OS cells ("C1"), stably expressing the constitutively-active Akt1 (ca-Akt1, S473D) or empty vector ("Vec") were treated with 25 μM of SKI-V for applied time periods; expression of listed proteins was tested by Western blotting assays (G). Cell death (Trypan blue staining assays, H) and apoptosis (by measuring TUNEL-stained nuclei ratio, I) were tested. "Veh" stands for the vehicle control. Data were presented as mean ± standard deviation (SD, n = 5). *P < 0.05 vs. "Cas9-C"/"Veh" cells. # P < 0.05 (D, E, H and I). "N. S." stands for non-statistical difference (P > 0.05) (B and C). Experiments in this Figure were repeated five times with similar results obtained. Fig. 6 The anti-OS cell activity of SKI-V in vivo. The "C1" OS xenograft-bearing nude mice were subject to intraperitoneal (i.p.) injection of SKI-V (30 mg/kg body weight, daily for 18 days) or vehicle control ("Veh"), with ten mice per group. Tumor volumes (A) and mice body weights (D) were recorded every six days (from Day-0 to Day-36). The estimated daily tumor growth, in mm 3 per day, was calculated (B). At Day-36, all mice were anaesthetized and decapitated, OS xenografts were weighted (C). At experimental Day-5 and Day-10, one tumor of each group was isolated (total for tumors), the relative SphK1 activity (E), ceramide contents (F), and expression of SphK1 mRNA (G) and listed proteins (H) were shown. Data were presented as mean ± standard deviation (SD). *P < 0.05 vs. "Veh" group.
The primary human osteoblasts and hFOB1.19 osteoblastic cells were from Dr. Ji [48], and cells cultivated under the described protocols [49,50]. The protocols using human cells were approved the Ethic Committee of Taizhou People's Hospital, according with the principles expressed in the Declaration of Helsinki.

Other cellular function studies
Cells were seeded into ploy-L-lysine-coated tissue-culturing plates at optimal seeding density, and the detailed protocols of the cellular functional assays, including CCK-8 viability, Trypan blue staining, colony formation, EdU (5-ethynyl-20-deoxyuridine) staining, and the in vitro cell migration and migration ("Transwell" assays) as well as Annexin V FACS, TUNEL staining, mitochondrial depolarization JC-1 staining and single strand DNA (ssDNA) ELISA were described in the previous studies [51,52].

Quantitative real time-PCR (qRT-PCR) and Western blotting assays
The detailed protocols for qRT-PCR and Western blotting assays were described in detail elsewhere [46,47].

SphK1 activity assay
After the designated treatments, cells and tissues were homogenized and centrifuged at 12,000 rpm to obtain the supernatant. SphK1 activity was determined by a SphK1 activity assay kit (Abnova). The attached SphK1 substrate was added to the supernatant (25 μL per treatment) for 5 min at 30°C, and the SphK1 activity was detected by a microplate reader.

Ceramide assay
The detailed protocols of analyzing cellular ceramides were described previously [36]. Ceramides were expressed as fmol in nmol of phospholipid and were always normalized to that of control treatment.

CRISPR/Cas9-induced knockout of SphK1
OS cells were first stably transduced with the Cas9-expressing construct (Genechem, Shanghai, China). Cells were further transfected with a lentiCRISPR SphK1-KO plasmid (from Dr. Yao [15]), and thereafter the single stable cells were established after puromycin selection and PCRmediated screening of SphK1 KO.

Constitutively-active mutant Akt1
The constitutively-active Akt1 ("caAkt1", S473D) adenoviral construct was from Dr. Liu [21]. The adenovirus was directly added to the cultured OS cells. The caAkt1 expression was always verified by Western blotting in puromycin-selected stable cells.

SphK1 overexpression
A GV369 SphK1-expressing lentiviral construct [15] was transduced to the hFOB1.19 cells, and the stable cells were established with puromycin selection, with SphK1 overexpression verified regularly.

OS xenograft studies
The animal procedures were conformed to the Ethics Committee and Animal Care Committee of Taizhou People's Hospital (Taizhou, China). The five-six week-old BALB/c nude mice (half male half female) were provided by Shanghai Slake Laboratory Animal Co. (Shanghai, China). The nude mice were maintained in controlled standard environmental conditions. The C1 primary human OS cells (at six million cells per mouse, in 200 µL serum-free DMEM/Matrigel) were subcutaneously (s.c.) injected into the flanks of nude mice. After 20 days of cell injection, OS xenografts were established. The xenograft-bearing nude mice were subject to intraperitoneal (i.p.) injection of SKI-V (30 mg/kg body weight, daily for 18 consecutive days) or the vehicle control. The volume of each xenograft tumor was calculated using the following formula: π/6 × L (long diameter) × W (short diameter) 2 .

Statistical analyses
Data with normal distributions were presented as mean ± standard deviation (SD). The significance between groups were determined by the two-tailed Student's t test (Excel 2007, for two groups) or ANOVA analysis and Student-Newman-Keuls post hoc test (SPSS 23.0, for multiple groups). P values < 0.05 were considered as statistically significant.

DATA AVAILABILITY
All data are available upon request.