The therapeutic value of the SphK1-targeting microRNA-3677 in human osteosarcoma cells

Sphingosine kinase 1 (SphK1) is a potential therapeutic target for human osteosarcoma (OS). SphK1-targeting microRNAs (miRNAs) could have important therapeutic value for OS. We discovered that micorRNA-3677 (miR-3677) is a SphK1-targeting miRNA, inhibiting OS cell progression. The results of RNA-Pull down assay confirmed direct binding between biotinylated-miR-3677 and SphK1 mRNA in primary human OS cells. In established and primary human OS cells forced overexpression of miR-3677, by a lentiviral construct, decreased SphK1 3’-UTR (untranslated region) activity and downregulated SphK1 expression. Both were however enhanced with miR-3677 inhibition in OS cells. Function studies demonstrated that OS cell growth, proliferation and migration were inhibited with miR-3677 overexpression, but augmented with miR-3677 inhibition. MiR-3677 overexpression-induced anti-OS cell activity was reversed with re-expression of the 3’-UTR-depleted SphK1. Additionally, in SphK1 knockout OS cells (by CRISPR/Cas9 strategy), altering miR-3677 expression failed to further alter cell functions. Finally, we show that miR-3677 expression was significantly downregulated in primary human OS tissues, correlating with SphK1 mRNA upregulation. We conclude that targeting SphK1 by miR-3677 inhibits human OS cell progression.


INTRODUCTION
Osteosarcoma (OS) is a common malignant bone tumor [1,2]. Each year it is estimated that over three million new cases of OS will be diagnosed, mostly in children and adolescents [1,2]. OS survival has been significantly improved over the past decades, owing to progress in the early disease diagnosis techniques and latest developments in molecularly-targeted and/or combination therapies [1,2]. For the recurrent and metastatic OS, the current clinical therapies are limited [1,2]. Therefore, it is important to explore novel and reliable molecular targets for OS efficient therapy [1,2]. It is also the research focus of our group [3][4][5].
Sphingosine kinase (SphK) family proteins, including SphK1 and SphK2, catalyze sphingosine phosphorylation to form sphingosine-1-phosphate (S1P) AGING [6], the latter is a key lipid mediator with intracellular and extracellular functions [7]. SphK1 regulates the balance between lipid mediators, including ceramide, sphingosine, and S1P. SphK1 inhibition, silencing or loss-of-function mutation will lead to S1P depletion and ceramide accumulation, causing significant cell death and apoptosis [7]. However, in many types of human cancers, SphK1 overexpression and/or over-activation would promote cancer cell growth and proliferation [7]. Our previous studies have shown that SphK1 is overexpressed in human OS, representing as an important therapeutic target [5].
The miR-3677's activity in other OS cells was studied next. In U2OS/MG63 cells and primary human OS cells (OS-2 and OS-3, derived from two other patients), infection of lv-pre-miR-3677 for 48h led to upregulation of mature miR-3677 ( Figure 1L), leading to SphK1 mRNA reduction ( Figure 1M). These results show that miR-3677 targets and silences SphK1 in human OS cells.

MiR-3677 is downregulated in human OS tissues
At last we tested the expression of miR-3677 in human tissues. In human OS tissues ("T", from eight independent primary OS patients [18]), miR-3677 expression was 40% lower than that in the surrounding normal bone tissues ("N") ( Figure 4G). Importantly, miR-3677 reduction in OS tissues was correlated with SphK1 mRNA elevation ( Figure 4H). The latter was over 6-fold higher in OS tissues than that in normal bone tissues ( Figure 4H). Therefore, in human OS tissues miR-3677 expression is downregulated, correlating with SphK1 mRNA upregulation.

DISCUSSION
Few studies have focused on the function of miR-3677 in human cells. Studies revealed miR-3677 is negatively correlated with overall survival of hepatocellular carcinoma (HCC) patients [19,20]. Zorniak et al., have shown that the mucosal miR-3677 expression is elevated in patients with cirrhotic gastric antral vascular ectasia (GAVE) [21]. Its potential targets and its expression in human OS are, however, largely unknown.

Figure 2. Ectopic overexpression inhibits OS cell progression in vitro.
Sable OS-1 cells with the pre-miR-3677-expressing lentivirus ("lv-pre-miR-3677", s-L1/s-L2, two lines) or with non-sense control miRNA ("lvmiC"), as well as the parental control OS-1 cells ("Ctr"), were cultured, with cell growth curve shown in (A); Cell colony formation (B), proliferation (EdU incorporation, C) and migration ("Transwell" assay, D) were tested by mentioned assays, with cell apoptosis examined by TUNEL staining (E) and Annexin V FACS (F) assays. U2OS cells and MG63 cells as well as primary human OS cells (OS-2 and OS-3) were infected with lv-pre-miR-3677 or lvmiC for indicated time periods, cell proliferation and apoptosis were tested by EdU incorporation (G) and TUNEL staining (H), respectively. For in vitro cell functional assays, the exact same number of viable cells with different genetic modifications were initially plated into each well/dish (at 0h/Day-0, same for all figures). Data were presented as mean ± SD (n=5), and results were normalized. ***P< 0.001 vs. "lvmiC"/"miC" cells. Experiments in this figure were repeated five times with similar results obtained. Bar=100 μm (C-E).

Figure 3. Forced inhibition increases SphK1 expression, promoting OS cell progression in vitro. Expression of listed genes in
parental control OS-1 cells ("Ctr"), OS-1 cells with pre-miR-3677 anti-sense lentivirus ("lv-antagomiR-3677", L1/ L2, two lines) or non-sense anti-sense construct ("lv-an-C"), was tested by qPCR (A and C) and Western blotting (D) assays, with the relative SphK1 3'-UTR activity examined (B); Cell proliferation and migration were tested by EdU incorporation (E) and "Transwell" (F) assays, respectively. The listed OS cells were infected with lv-antagomiR-3677 or lv-an-C for 48h, expression of mature miR-3677 (-3p, G) and SphK1 mRNA (H) was tested, with cell proliferation examined by EdU incorporation assays (I). Expression of listed proteins was quantified, normalized to the loading control Tubulin (G). Data were presented as mean ± SD (n=5), and results were normalized. ***P< 0.001 vs. "lv-an-C" cells. Experiments in this figure were repeated five times with similar results obtained.

AGING
The results of this study suggest that miR-3677 is a SphK1-targeting miRNA, which inhibits human OS cell progression. RNA pull down assay results demonstrated that biotinylated-miR-3677 directly associated with SphK1 mRNA in OS-1 cells. Forced overexpression of miR-3677 in OS cells inhibited SphK1 3'-UTR activity, causing downregulation of SphK1 mRNA and protein.
Functional studies revealed that OS cell growth, proliferation and migration were largely inhibited with ectopic miR-3677 overexpression, but augmented with forced miR-3677 inhibition. Furthermore, miR-3677 overexpression induced apoptotic death in OS cells. In human OS tissues, miR-3677 expression was downregulated, correlating with SphK1 mRNA upregulation. These results implied that miR-3677 targets and silences SphK1, efficiently inhibiting human OS cell progression.
Our results further imply that SphK1 silencing should be the primary reason of miR-3677-induced anti-OS cell activity. First, mimicking lv-pre-miR-3677induced actions, SphK1 KO (using CRISPR/ Cas9 strategy) inhibited OS cell proliferation and migration. Second, in SphK1 KO OS-1 cells exogenously altering miR-3677 expression, by lv-pre-miR-3677 or lv-antagomiR-3677, failed to change cell functions. Third, lv-pre-miR-3677induced anti-OS-1 cell activity was reversed by forced expression of an UTR-depleted SphK1. Therefore, targeting SphK1 by miR-3677 induced significant anti-OS cell activity.
OS is a neoplastic growth in bone tissue, with malignant proliferation and metastasis of OS cells [22]. The current anti-OS therapies are very limited. Therefore, exploring novel therapeutic targets and developing alternative treatment strategies are urgently required [22]. The current study demonstrated that targeting SphK1 by miR-3677 inhibited human OS cell progression. MiR-3677, and possible other SphK1targeting miRNAs, could be novel therapeutic advance for OS treatment. lentivirus ("lv-pre-miR-3677", s-L1) were further transfected with lentiviral SphK1-expresing construct ("+wt-SphK1"), control cells were transduced with the lentiviral construct with control miRNA ("lvmiC"), expression of listed proteins was shown (A); Cells were further cultured, cell proliferation and migration were tested by EdU incorporation (B) and "Transwell" assay (C), respectively; and results were quantified (B and C). Expression of listed proteins in stable OS-1 cells with the CRISPR-Cas9-SphK1-KO-GFP construct ("koSphK1-s-L1/koSphK1-s-L2", two lines) or control construct ("Cas-9-C") was shown (D). The koSphK1 cells were further infected with lv-antagomiR-3677 or lv-pre-miR-3677 for 48h, with cell proliferation (E) and migration (F) tested; and results were quantified. Expression of miR-3677 (G) and SphK1 mRNA (H) in eight (n=8) different human OS tissues ("T") and surrounding normal bone tissues ("N") was tested. Data were presented as mean ± SD, and results were normalized. ***P< 0.001 (A-C). ***P< 0.001 vs. "Cas-9-C" cells (D). ***P< 0.001 vs. "N" tissues (G and H). Experiments in this figure were repeated five times with similar results obtained.

Chemicals and reagents
Puromycin and polybrene were provided by Sigma-Aldrich Chemicals (St. Louis, Mo). All the antibodies were obtained from Abcam (Cambridge, MA). RNA reagents and Lipofectamine 2000 were obtained from Thermo-Fisher Invitrogen (Shanghai, China). All sequences and plasmids were provided by Shanghai Genechem Co. (Shanghai, China), unless otherwise mentioned.

Human tissues
The protocols of this study were approved by the Ethic Committee of Nanjing Medical University, according to Declaration of Helsinki. The human OS tissues and the surrounding normal bone tissues from eight (8) writteninformed OS patients, were provided by Dr. Liang at Zhejiang University [18]. Tissues were incubated with the described lysis buffer [18], stored in liquid nitrogen before further biochemical analyses.

OS cells
U2OS and MG-63 established human OS cells were provided by Dr. Liang at Zhejiang University [18]. The OS cells were subjected to mycoplasma/microbial contamination examination every three months. Short Tandem Repeat (STR) profiling, population doubling time (PDT), and cell morphology were routinely (every 3-4 months) checked to verify their genotypes. Three independent patient-derived primary human OS cells [23], OS-1, OS-2 and OS-3, were provided by Dr. Ji at Nanjing Medical University [23], cultured under described protocols [23,24]. The primary OS cells at passage 3-10 were utilized.

qPCR
Total cellular RNA, extracted using the TRIzol reagents, was revere transcripted [25]. Under the ABI Prism 7900 system the quantitative real time-PCR (qPCR) was performed (using described protocols [26,27]). The melt curve analyses were applied to calculate product melting temperature. Glyceraldehyde-3-phosphatedehydrogenase (GAPDH) was tested as the reference gene and the internal control, with quantification through the 2 −∆∆Ct method. MiR-3677 expression was normalized to U6. The primers utilized in this study were listed in Table 1.

Transfection of miR mimic
OS cells were seeded into the six-well plates (at 50% confluence), transfected with 500 nM of the applied miR mimic through using a Lipofectamine 2000 protocol [29].

SphK1 3'-UTR activity assay
Briefly, the human SphK1 3'-UTR with the putative binding sites of miR-3677 (position 235-242) was amplified by PCR, then inserted into the firefly luciferase reporter vector, pGL4.13 (luc2/SV40) (Promega) at the XbaI site and downstream from the stop codon of the luciferase gene. The plasmid, along with the Renilla luciferase reporter vector and pRL-SV40 (Promega), were co-transfected to human OS cells by Lipofectamine 2000. Cells were then subjected to applied genetic modifications, with the SphK1 3'-UTR luciferase activity tested through a Promega kit [29].

RNA-pull down assay
The detailed protocols of RNA-Pull down assay, using the Pierce Magnetic RNA Pull-Down Kit, were described early [30,31]. OS-1 cells were transfected with biotinylated miR-3677 mimic or control mimic (100 nmol/L) for 36h, and cells were harvested [31]. The lysates were incubated with streptavidin-coated magnetic beads to pull-down biotin-captured RNA complex [30], and the latter was purified by the RNeasy Mini Kit (QIAGEN), with expression of SphK1 mRNA tested by qPCR. Its level was always normalized to input controls.

Western blotting
OS cells, with the applied genetic modifications, were harvested using the described lysis buffer [32]. Twenty µg lysate proteins per sample were separated by 10- GCTGGCAGCTTCCTTGAACCAT GTGTGCAGAGACAGCAGGTTCA GAPDH GTCTCCTCTGACTTCAACAGCG ACCACCCTGTTGCTGTAGCCAA 15% SDS gels, and transferred onto PVDF blots (Millipore, Shanghai, China). The blots were blocked, probed with applied primary and second antibodies [32]. The enhanced chemiluminescence (ECL) detection system was applied to visualize the targeted protein bands (based on the molecular weights), using x-ray films. For all the Western blotting assays, each lane was loaded with exact same amount of quantified protein lysates, then the same set of lysate samples were run in parallel ("sister") gels. The ImageJ software was utilized for data quantification [33,34].

Colony formation
OS-1 cells (5 ×10 3 per well) with the applied genetic modifications were trypsinized and re-suspended in complete medium with agarose (0.25%). Cells were then plated on the top of 10-cm dishes. Medium was renewed every two days. At day-10, the viable cell colonies were counted manually.

In vitro cell migration
OS cells with applied genetic modifications (0.2 × 10 5 cells of each treatment in 250 μL medium) were plated in the upper chambers (12 μm pore, Corning, New York, NY) [35], with the lower chambers filled with completed medium (with 12% FBS). After 24h, OS cells invading into the lower chambers were fixed, stained and counted. To exclude cell proliferation mitomycin (3.0 μg/mL, Sigma) was added [35].

Annexin V FACS
OS cells with the applied genetic modifications were stained with Annexin V and propidium iodide (PI), analyzed by fluorescent-activated cell sorting (FACS) on a FACSCalibur machine (BD Biosciences) [35].

TUNEL staining
OS cells with the applied genetic modifications were tested by a TUNEL (Terminal deoxynucleotidyl transferase dUTP nick end labeling) In Situ Cell Death Detection Kit (Roche, Shanghai, China). At least 600 cell nuclei per preparation in six random views were counted to determine TUNEL ratio (TUNEL/DAPI×100%).

Ceramide assay
Using the previously-described protocol [36], the cellular ceramide contents in OS cells were analyzed, with the values expressed as fmol by nmol of phospholipid.

Ectopic overexpression of SphK1
The lentiviral SphK1 (with no 3'-UTR region) expression GV369 construct was designed, synthesized and sequence-verified by Shanghai Genechem, then transduced to OS-1 cells with lv-pre-miR-3677. Cells were selected by puromycin for two passages, with SphK1 expression confirmed by qPCR and Western blotting.

CRISPR/Cas9-induced knockout of SphK1
The small guide RNA (sgRNA) against human SphK1 (Target DNA Sequence: ACCGATAAGGAGCTG AAGGC, PAM sequence AGG) was selected from Dr. Zhang's lab at MIT, inserted into the lentiCRISPR-GFP plasmid (from Dr. Zhang at Soochow University [37]) containing a puromycin selection gene (Addgene) [37]. OS-1 cells were plated into the sixwell plates (1×10 5 cells per well), and transfected with lentiCRISPR SphK1-KO plasmid. Cells were subjected to FACS-mediated GFP sorting. The single stable cells were screened for SphK1 KO by Western blotting/qPCR. Two stable SphK1 KO cell lines were established. Control cells were transfected with the empty vector.

Statistical analysis
Date were expressed as means ± standard deviation (SD). The statistical differences were analyzed through the one-way analysis of variance (ANOVA) by the Tukey's post hoc multiple comparison tests (SPSS 21.0, SPSS co. Chicago, CA). Comparisons between two specific groups were performed by the two-tailed Student t tests (Excel 2007, Microsoft). P < 0.05 was considered statistically significant.