LncRNA-SNHG1 promotes paclitaxel resistance of gastric cancer cells through modulating the miR-216b-5p-hexokianse 2 axis

Abstract Gastric cancer (GC) is one of the most malignant tumors with high incidence and poor prognosis. Currently, the combination of surgery with chemo- or radiotherapy is widely applied therapeutic strategy against GC. However, development of drug resistance severely limited the clinical application of chemotherapy. Small nucleolar RNA host gene 1 (SNHG1) has been reported to be frequently overexpressed in diverse human tumors. Yet, the biological roles and mechanisms of SNHG1 in chemoresistant GC remain unclear. Expressions of lncRNA and miRNA were detected by qRT-PCR. Responses of GC cells to Taxol treatments were evaluated by cell viability assay and apoptosis assay. Glucose metabolism rate was examined by glucose uptake and extracellular acidification rate (ECAR). The lncRNA–miRNA interaction was validated by RNA pull-down assay and luciferase assays. This study reports that expressions of SNHG1 were significantly elevated in patients with GC and gastric cancer cell lines. Silencing SNHG1 effectively suppressed GC cells migration and increased the Taxol sensitivity of GC cells. Moreover, we detected remarkedly upregulated SNHG1 expression and increased glucose metabolism in Taxol resistant cell line, MKN-45 TXR. Low glucose supply rendered Taxol resistant cells more susceptible to Taxol treatment compared with that from MKN-45 parental cells. Bioinformatical analysis, RNA pull-down and luciferase assays verified that SNHG1 functioned as a ceRNA of miR-216b-5p in GC cells. Consistently, we detected miR-216b-5p was significantly downregulated in GC tumor specimens and Taxol resistant GC cells. The hexokinase 2 (HK2), a glucose metabolism key enzyme, was predicted and validated as a direct target of miR-216b-5p in GC cells. Finally, restoration of miR-216b-5p in SNHG1-overexpressing MKN-45 TXR cells successfully overrode the SNHG1-promoted Taxol resistance through targeting the HK2-glycolysis axis. This study uncovered new biological roles and molecular mechanisms of the lncRNA-SNHG1-mediated Taxol resistance of gastric cancer, suggesting targeting the SNHG1-miR-216b-5p-HK2 axis could be a potentially therapeutic approach against chemoresistant gastric cancer.


Introduction
Gastric cancer (GC) is one of the most human malignancies worldwide associating with poor prognosis and high mortality rate [1].Currently, surgery is the primary and effective approach for treating early-stage gastric cancer [2].In addition, the combined strategy with chemo-or radio-therapy is applied to treat GC in the late, metastatic or recurrent stage [2,3].Helicobacter pylori infection, age and high salt intake are the most frequent risk factors of GC [3].Gastric cancer is diagnosed histologically using endoscopic biopsy and the routine approaches for staging GC include CT, PET, endoscopic ultrasound and laparoscopy [1].The programmed cell death ligand 1 (PD-L1) and human epidermal growth factor receptor 2 (HER2) are most representative biomarkers of GC [1].Currently, targeting PD-L1 by pembrolizumab and HER2 by trastuzumab has been widely applied for anti-GC treatment [1].
Paclitaxel (Taxol) is an antitumor drug that binds to and stabilizes microtubules to prevent depolymerization of microtubules of cancer cells to induce dysfunction of tubulin, leading to cancer cell death [4].Although applications of new therapeutic approaches improved survival rates of patients with GC, a large fraction of them developed Taxol resistance, which severely limited abroad application of Taxol-based chemotherapy, resulting in tumor recurrence and high mortality incidence [5].Currently, mechanisms of Taxol resistance in cancer includes: (a) Alterations in microtubule dynamics that prevents the binding of Taxol on microtubules; (b) Altered expression/posttranslational modifications of tubulin and (c) Alterations in oncogenic/tumor suppressive signaling pathways [6].Yet, the precise molecular mechanisms of Taxol resistance are still under investigation.Therefore, understanding the underlying cellular processes and molecular mechanisms of the Taxol resistant gastric cancer cells is urgent to develop effectively therapeutic approaches to prolong the overall survival rates of patients with GC.
LncRNAs are groups of long-chain, non-coding RNAs that were known to play essential roles in cancers through modulating their target miRNAs expressions [7,8].Accumulating studies revealed that lncRNAs play critical roles in various processes of cancer cells such as proliferation, differentiation, metastasis, migration, epithelial to mesenchymal transition and drug resistance [9][10][11][12][13][14]. LncRNAs were known to locate in cytoplasm or nucleus of cells [15].Evidence uncovered that lncRNAs located in nucleus are involved in gene modulation at epigenetic and transcription levels [15].Moreover, lncRNAs that locate in cytoplasm are majorly participated in transcriptional and post-transcriptional modulating gene expression through interaction with miRNAs to form a competitive endogenous RNA (ceRNA) network [15].Furthermore, recent studies revealed that the exosomes that originate from endosomal processing contain functional lncRNAs [16].Particularly, lncRNAs such as NEAT1, HOTAIR and UCA1 have been shown to participate in gastric cancer progressions [17].LncRNA-SNHG1 (small nucleolar RNA host gene 1) has been reported to be frequently overexpressed in human cancers [18][19][20].In addition, SNHG1 is correlated with worse prognosis and low survival rates of patients with cancer [18][19][20], suggesting SNHG1 might be an effectively therapeutic target for anti-cancer treatments.
MicroRNAs (miRNAs) are endogenous, small ($22nt) noncoding RNAs that are dysregulated during cancer progressions and play critical roles in regulating mRNAs expressions through binding on 3 0 UTR regions of them [21].MiR-216b-5p has been reported to be negatively associated with diverse cancers [22][23][24].However, the precise functions and molecular targets of SNHG1 as well as the regulatory relationship between SNHG1 and miR-216b-p in Taxol resistant gastric cancer remain unknown.
As a consequence of metabolic reprogramming, cancer cells displayed elevated dependence on anerobic glycolysis, a phenomenon called 'Warburg effect' [25].Accumulating studies revealed that blocking the dysregulated cellular glycolysis rate of cancer cells enhanced anti-cancer effects of traditional chemo-or radiotherapy [26].Based on these background, we hypothesized that lncRNA SNHG1 promotes Taxol resistance in gastric cancer cells via positive regulating the glycolysis pathway.Therefore, a series of experiments was performed to investigate the downstream targets of SNHG1 as well as the precise molecular mechanisms.This study suggests SNHG1 could be a promising biomarker for GC diagnosis and a potentially anti-cancer agent against chemoresistant GC.
Transfection of siRNA, miRNA and plasmid DNA Gastric cancer cells were transfected with siRNA, miRNA or plasmid by using the Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's protocols.siSNHG1, miRNA-216b-5p and their negative control were synthesized from Genepharma (Shanghai, China) and transfected at 25 nM for 48 h.HK2 overexpression plasmid was purchased from Origene.com.Plasmid transfection was performed in 6-cm dish with 70% cell confluence.Plasmid was transfected at 1 mg/mL.

Bioinformatics analysis
The predictions of lncRNA-miRNA and miRNA-mRNA interactions were performed from the starBase of ENCORI http://starbase.sysu.edu.cn/.Expressions of SNHG1, miR-126b-5p and HK2 in human normal gastric tissues and cancer tissues were analyzed from the http://ualcan.path.uab.edu.Survival rates of patients with GC with low or high miR-216b-5p were analyzed by Kaplan-Meier plotter from http://kmplot.com.

RNA pull-down assay
Scramble, sense and antisense SNHG1 probes were biotin-labeled by RiboBio Co. Ltd (Guangzhou, China).RNA probes were mixed with GC cell extracts for 2 h, followed by adding streptavidin agarose beads (50 mL) (Thermo Fisher Scientific, Shanghai, China) into the mixture.After washing, the samples and inputs were digested by Proteinase K.The enrichment of miR-216b-5p was then pull-down from the RNA-RNA complex and detected by qRT-PCR.

Luciferase assay
The luciferase assay was performed by measurements of the chemiluminescence of luciferase activity of luciferase vectors using a dual-luciferase reporter assay (Promega, USA) according to the manufacturer's protocol.The WT-or Mut-SNHG1 or 3 0 UTR of HK2 was cloned into pmirGLO Dual-Luciferase miRNA Target Expression Vector (Promega, USA), followed by co-transfection of control miRNA or miR-216b-5p precursor into the above constructed vectors by lipofectamine 2000 in gastric cancer cells.The relative firefly luciferase activity was normalized to renilla luciferase activity.Experiments were performed in triplicate and repeated three times.

Wound healing assay
The GC cells (2 Â 10 5 per well) were plated in 12-well plates for 24 h to reach 100% confluence.A sterile 200 mL pipette tip were used to scratch the wound uniformly.Cell motility was assessed by measuring the movement of cells into a scraped wound after 16 h by measuring the distance of the wound from 0 h.Each experiment was repeated three times.

Cell viability assay
Responses to drug treatments of gastric cancer cells were examined by cell viability assay and apoptosis assay.Cell viability was determined by 5-diphenyltetrazolium bromide (MTT) assay.GC cells (5 Â 10 3 / well) were seeded on 96-well plates and incubated at 37 C for 24 h.Cells were then incubated with 100 lL of 0.5 mg/mL MTT for 4 h at 37 C followed by adding of 150 lL dimethylsulfoxide (DMSO) per well.The optical density value of each well was detected at 570 nm using a microplate reader (Thermo Fisher Scientific, Shanghai, China).Experiments were performed in triplicate and repeated three times.

Clone formation assay
GC cells (1000 cells/well) were plated into 6-well plates with control or Taxol treatments for 48 h, then were placed in a 5% CO 2 incubator at 37 C for 2 weeks.Cell culture medium was refreshed every three days.Cells were stained by 0.1% crystal violet for 5 min, at room temperature.After washing by PBS, the survival colonies were examined and recorded using a bright-field microscopy.

Apoptosis assay
The cell death rates of GC cells in response to Taxol treatments were examined by Annexin V/FITC apoptosis assay using the FITC Annexin V Apoptosis Detection Kit I (BD Biosciences, San Jose, CA, USA) according to the manufacturer's instruction.Cells were collected and resuspended in 300 mL of binding buffer followed by adding Annexin V FITC (3 mL) and 5 mL of propidium iodide (PI) staining solution.The mixture was incubated for 30 min in the dark at room temperature.The fluorescence intensity was analyzed using a FACSCalibur Flow Cytometer (BD Biosciences).

Measurements of glucose metabolism
The glucose metabolism rate was evaluated by glucose uptake (Kit No. BC2490, Solarbio Science & Technology Co., Ltd.Beijing China) and extracellular acidification rate (ECAR) assay using the Seahorse XF Glycolysis Stress Test Kit (Agilent, Santa Clara, CA, USA) according to the manufacturer's instructions.The data were normalized to cell number of each group.Experiments were repeated three times.

Western blot
Whole-cell lysates of GC cells were isolated using the RIPA buffer (Sigma, Shanghai, China) with 1Â protease inhibitor cocktail (Sigma, Shanghai, China).Proteins were separated by 10% SDS polyacrylamide gel followed by transferring onto PVDF membranes (Thermo Fisher Scientific, CA, USA).The PVDF membranes were blocked with 5% BSA in TBST at room temperature for 1 h.Membranes were incubated with primary antibodies (1:1000) for overnight at 4 C.After complete washing by PBST, the PVDF membrane was incubated with secondary anti-rabbit IgG antibody (Abcam; ab150077) (1:5000) for 1 h at room temperature.Protein expressions were detected using ECL reagent (Santa Cruz Biotechnology).b-actin was a loading control.Three independent experiments were performed.

Statistical analysis
Statistical analysis was performed using the Prism 7.0 software (GraphPad Software, La Jolla, CA, USA).All values were expressed as the mean ± standard deviation (SD) and three independent experiments were performed.The comparison between two groups was analyzed by Student's t test.The comparisons among multiple groups were analyzed by one-way ANOVA (analysis of variance) followed by Tukey's test.P < 0.05 was considered statistically significant.

SHHG1 is upregulated in gastric cancer cells and promotes Taxol resistance
Previous studies revealed potentially oncogenic roles of SNHG1 in diverse cancers [18][19][20].We then started to evaluate the clinical relevance of SNHG1 in GC.From TCGA cancer database, we observed that SNHG1 was significantly upregulated in multiple cancers including gastric tumors (Supplementary Figure S1).Specifically, bioinformatics analysis showed that SNHG1 expression was remarkedly elevated in gastric cancers (Figure 1(A)), indicating SNHG1 is positively associated with GC progressions.Consistently, qRT-PCR results demonstrated that SNHG1 was significantly upregulated in human GC cell lines, AGS, HCG27, MKN-45, BGC-823 and SGC-7901 compared with normal gastric epithelial cells GES-1 (Figure 1(B)).To evaluate the biological roles of SNHG1 in gastric cancer, SNHG1 was knocked down by siRNA in GC cell lines, MKN-45 and AGS (Figure 1(C)).Expectedly, GC cells with lower SNHG1 expressions exhibited significantly suppressed cell migration capacity (Figure 1(D)) and increased chemosensitivity (Figure 1(E,F)).The IC50s of control MKN-45 and AGS cells in response to Taxol were 5.72 nM and 13.75 nM, respectively.Silencing SNHG1 effectively decreased the IC50s of MKN-45 and AGS cells to 1.62 nM and 4.31 nM (Figure 1(E,F)), suggesting SNHG1 is a potential target for anti-gastric cancer treatments.
We then established a Taxol resistant GC cell line (MKN-45-TXR) originating from MKN-45 cells by treating parental cells with gradually increased concentrations of Taxol.The Taxol resistance of MKN-45-TXR cells was validated by cell viability assay which demonstrated that TXR cells could withstand higher concentrations of Taxol treatments (Figure 1(G)).The IC50 of MKN-45-TXR cells jumped to 37.22 nM, which was sevenfold higher than that from MKN-45 parental cells (Figure 1(G)).Furthermore, SNHG1 was detected to be significantly upregulated in Taxol resistant cells compared with MKN-45 parental cells (Figure 1(H)).To test whether inhibition of SNHG1 could re-sensitize TXR cells to Taxol, MKN-45-TXR cells without or with SNHG1 knockdown (Supplementary Figure S2) were treated with Taxol.Consistently, silencing SNHG1 effectively decreased the resistance of TXR cells (Figure 1(I)).Taken together, these results suggest SNHG1 plays oncogenic roles and is a potentially therapeutic target for antichemoresistant GC treatment.

SNHG1 inhibits miR-216b-5p expression by sponging it in GC cells
Given the above-described functions of SNHG1 in Taxol resistant GC cells, we then explored the molecular mechanisms of the SNHG1-promoted Taxol resistance.Accumulating studies revealed that lncRNAs inhibited their target miRNAs expression by sponging them to form a ceRNA network [8].The target miRNAs of SNHG1 were subsequently predicted from online non-coding RNA service, starBase.Bioinformatics analysis indicated that miR-216b-5p, which has been reported to be negatively associated with multiple cancer progressions [22][23][24], contains binding sites of SNHG1 (Figure 2(A)).Expectedly, a significantly negative correlation between SNHG1 and miR-216b-5p expression was observed from gastric cancer TCGA database (Figure 2(B)).Silencing SNHG1 effectively upregulated miR-216b-5p expressions in GC cells (Figure 2(C)).To further validate the binding of miR-216b-5p on SNHG1, RNA pulldown and luciferase assays were performed.Results from RNA pull-down assay showed miR-216b-5p was only detected by the antisense SNHG1-precipitated RNA complex compared with control and sense SNHG1 probe-precipitated RNA complexes (Figure 2(D)).Furthermore, co-transfection of miR-216b-5p with luciferase vector containing wild type (WT) SNHG1 significantly blocked the luciferase activity.While co-transfection of miR-216b-5p with luciferase vector containing the binding site mutant (Mut) SNHG1 could not inhibit the luciferase activity (Figure 2(E,F)).Taken together, these results consistently validated that SNHG1 sponged miR-216b-5p by direct sponging it.

microRNA-216b-5p is downregulated in GC cells and negatively correlated with Taxol resistance
Since we showed that SNHG1 promoted Taxol resistance and inhibited miR-216b-5p in GC cells, we thus hypothesized that miR-216b-5p functioned as a tumor suppressor that decreased the Taxol resistance of gastric cancer cells.The Kaplan-Meier plotter analysis showed higher miR-216b-5p expression was associated with relatively better survival rates of patients with gastric cancer (Figure 3(A)).Consistently, qRT-PCR results demonstrated that miR-216b-5p was remarkedly downregulated in GC cell lines compared with normal gastric epithelial cells (Figure 3(B)).The effects of miR-216b-5p on the Taxol sensitivity of gastric cancer cells were assessed.As we expected, overexpression of miR-216b-5p significantly attenuated cell survival rate of MKN-45-TXR cells under Taxol treatments (Figure 3(C,D)), indicating miR-216b-5p plays an anti-chemoresistant role in gastric cancer cells.

Elevated glucose metabolism is observed in Taxol resistant gastric cancer cells
A bulk of studies uncovered that cancer cells exhibited dysregulated glucose metabolism that was abnormally elevated in cancer cells compared with their untransformed counterparts [25].To investigate which cellular process was regulated by the SNHG1-miR-216b-5p ceRNA network, we examined the glucose metabolism of Taxol resistant gastric cancer.Consequently, the glucose uptake and the extracellular acidification rate (ECAR), two glucose metabolism rate readouts were assessed in MKN-45 parental and TXR cells.As we expected, the glucose uptake rate and ECAR were significantly increased in MKN-45-TXR cells (Figure 4(A,B)).In addition, MKN-45 TXR cells were more dependent on glucose supply than MKN-45 parental cells under Taxol treatment that Taxol resistant cells displayed higher Taxol-induced cell death rate by glucose starvation compared with that under normal condition (Figure 4(C)).While MKN-45 cells were only slightly affected under low glucose condition (Figure 4(C)), suggesting blocking the dysregulated glucose metabolism could be an effective approach to treat chemoresistant gastric cancer.Consistently, blocking glucose metabolism by glycolysis inhibitor, 2-DG effectively synergized with Taxol to induce apoptosis of Taxol resistant cells (Figure 4(D)).In summary, these results demonstrated a unique glucose metabolism characteristic of Taxol resistant GC cells.

SNHG1 and miR-216b-5p reversely regulate glucose metabolism of GC cells
To assess the roles of SNHG1 and miR-216b-5p in regulating glucose metabolism, the effects of SNHG1 silencing and miR-216b-5p overexpression on glucose Expression of miR-216b-5p was determined by qRT-PCR.D: RNA pull-down assay was performed in GC cells as described in method section.E and F: Luciferase assay was performed in GC cells as described in method section.ÃÃ P < 0.01; ÃÃÃ P < 0.001.uptake and ECAR were evaluated in GC cells.Silencing SNHG1 effectively suppressed the glucose uptake and ECAR rates of GC cells (Figure 5(A,B)).On the other way, GC cells with overexpression of miR-216b-5p displayed similar metabolic phenotypes that the glucose uptake and ECAR were significantly inhibited by miR-216b-5p (Figure 5(C,D)).

Overexpression of miR-216b-5p sensitizes gastric cancer cells to taxol by direct targeting HK2
Accumulation studies revealed microRNAs bond to the 3 0 UTR of mRNAs to regulate target mRNAs expressions [21].Since miR-216b-5p has been reported to be negatively associated with cancers, we explored the putative targets of miR-216b-5p in gastric cancer cells.Bioinformatics analysis was performed through the TargetScan databases.Among potential targets, we primarily focused on molecules that are important glycolysis key enzymes or positive regulators.Interestingly, we found 3 0 UTR of hexokinase 2 (HK2), which catalyzed the conversion of glucose to glucose-6-phosphate [26], contained a putative miR-216b-5p binding region (Figure 6(A)).Expectedly, HK2 was significantly upregulated in various cancers including GC analyzed from the TCGA human cancer database (Supplementary Figure S3 and Figure 6(B)).In addition, qRT-PCR results demonstrated the expression of HK2 was remarkedly increased in human GC cell lines, AGS, HCG27, MKN-45, BGC-823 and SGC-7901 compared with normal gastric epithelial cells GES-1 (Figure 6(C)).Notably, miR-216b-5p and HK2 were inversely correlated in human gastric tumors (Figure 6(D)), suggesting miR-216b-5p could inhibit HK2 expression in gastric cancer.We then evaluated the roles of HK2 in Taxol resistance in gastric cancer cell.Expectedly, HK2 was significantly upregulated in MKN-45-TXR cells (Figure 6(E)), indicating HK2 could be a potential target for anti-Taxol resistance treatment.To verify this, HK2 was knocked down by siRNA in gastric cancer cells.Expected results demonstrated that silencing HK2 successfully increased the Taxol sensitivity of MKN-45 (Figure 6(F)) and AGS (Figure 6(G)).To test whether miR-216b-5p could downregulate HK2 in GC cells, we transfected control miRNAs or miR-216b-5p precursor into AGS and MKN-45 cells.Western blot results demonstrated overexpression of miR-216b-5p significantly blocked protein expression of HK2 (Figure 6(H,I)).To validate   whether miR-216b-5p could directly bind to the 3 0 UTR of HK2 mRNA, luciferase reporter assays were performed by co-transfection of control miRNA, miR-216b-5p with WT-or binding site mutant-3 0 UTR of HK2 into GC cells.Expectedly, exogenous overexpression of miR-216b-5p significantly blocked the luciferase activity of vector containing WT-HK2 3 0 UTR compared with that from the binding site mutant HK2 3 0 UTR (Figure 6(J,K)).These results consistently validated that HK2 is directly targeted by miR-216b-5p in gastric cancer.
To examine whether the miR-216b-5p-promoted Taxol sensitization was through targeting HK2, rescue experiments were performed.MKN45-TXR cells were co-transfected with control miRNA, miR-216b-5p alone or with HK2 overexpression plasmid.Western blot results from Figure 7(A) demonstrated co-transfection of miR-216b-5p with HK2 successfully recovered the HK2 protein expressions.Consequently, the glucose uptake and ECAR of the HK2-recovered cells were rescued compared with those from control cells (Figure 7(B,C)).Moreover, cell viability assay demonstrated that restoration of HK2 in miR-216b-5p-overexpressing cells recovered the Taxol resistance of MKN-45-TXR cells (Figure 7 Silencing SNHG1 synergizes with Taxol to inhibit GC by targeting the miR-216b-5p-HK2 axis Finally, we evaluated the SNHG1-miR-216b-5p-HK2 in Taxol resistance of gastric cancer cells.Bioinformatics analysis indicated that SNHG1 was positively associated with HK2 in human gastric cancer (Supplementary Figure S4).Thus, mechanism rescue experiments were performed in MKN-45 TXR cells by transfection of control plasmid, SNHG1 alone or plus miR-216b-5p.qPCR and Western blot results showed successful rescues of miR-216b-5p and HK2 in the SNHG1 plus miR-216b-5p-cotransfected cells (Figure 8(A,B)).Expectedly, the glucose uptake and ECAR of the miR-216b-5p-recovered Taxol resistant cells were rescued compared with those from control cells (Figure 8(C,D)).Consistently, cell viability assay and apoptosis assay demonstrated that recovery of miR-216b-5p in SNHG1-overexpressing cells successfully rescued the Taxol sensitivity of MKN-45-TXR cells (Figure 8(E,F)).Taken together, these results validated that the SNHG1-promoted Taxol resistance   of GC cells was through modulating the miR-216b-5p-HK2 axis (Figure 9).

Discussion
GC is a human malignant tumor with poor prognosis and high mortality rates [1].Currently, surgery is considered the primary and effective approach for early-stage GC [2].In addition, traditional chemotherapy, radiotherapy and the implementation of neoadjuvant therapy are strategies against GC in the late, metastatic or recurrent stage [3].However, development of drug resistance impaired wide application of chemotherapy [4,5].This study uncovered a non-coding RNA-mediated Taxol resistance in GC cells through modulating the glucose metabolism.Previous studies have reported that lncRNA SNHG1 was involved in the progressions of multiple cancers [18][19][20].Here, we showed lncRNA SNHG1 was significantly upregulated in GC tumors and cancer cell lines, consistent with previous reports that SNHG1 displayed potentially oncogenic roles in diverse cancers [18][19][20].Moreover, SNHG1 is positively associated with Taxol resistance by sponging miR-216b-5p, indicating SNHG1 is a diagnosis marker and therapeutic target against Taxol resistant GC.
Cancer cells display reprogramed glucose metabolism characteristics that they utilize glucose toward aerobic glycolysis for supporting various biochemical reactions instead of mitochondrial oxidative phosphorylation [25].This phenomenon is called 'Warburg effect'.Consequently, the metabolic reprogramming of cancer cells renders them an environmental advantage that contributes to avoiding apoptosis or development of drug resistance [26].Accumulating studies have revealed a tight relationship between drug resistance and the cellular glucose metabolism of gastric cancer cells [26], suggesting that the dysregulated glucose metabolism plays essential roles in chemoresistant GC.However, the underlying molecular linkage between Taxol resistance and glucose metabolism in gastric cancer remains largely unknown.From the established Taxol resistant gastric cancer cell line, our study validated that the glucose metabolism rate in Taxol resistant GC cells was significantly elevated.Moreover, under low glucose supply, Taxol resistant GC cells exhibited increased sensitization to Taxol compared with that from parental cells.Importantly, bioinformatics analysis and luciferase assay validated that HK2, which catalyzes the speed-limiting reaction of glycolysis, was directly targeted by miR-216b-5p in GC cells.Rescue experiments verified the miR-216b-5p-mediated Taxol sensitivity was through targeting HK2.Subsequently, blocking glucose metabolism effectively sensitized Taxol resistant GC cells to Taxol, suggesting targeting the non-coding RNA-mediated glucose metabolism is a novel strategy for anti-chemoresistant GC treatment.
A number of studies revealed that lncRNAs function through sponging target miRNAs to de-repress downstream mRNA targets of miRNAs [8].Here, bioinformatics analysis indicated SNHG1 contained putative binding sites of miR-216b-5p, suggesting an SNHG1-miR-216b-5p ceRNA network which plays important roles in Taxol resistance.Consequently, the RNA-RNA complex was verified by RNA pull-down assay and luciferase assay.Finally, mechanism rescue experiments demonstrated the SNHG1-mediated Taxol resistance of GC cells was through modulating the miR-216b-5p-HK2 axis.Although the miR-216b-5p-HK2 axis has been reported by Liu et al. recently [27], our results integrated the SNHG1-promoted Taxol resistance with miR-216b-5p-HK2-glycolysis pathway for the first time.
In summary, this study unveiled a new SNHG1-miR-216b-5p-HK2-glycolysis axis, leading to Taxol resistance in GC.Our conclusions will contribute to developing effectively therapeutic strategy against chemoresistant gastric cancer.

Figure 1 .
Figure 1.SNHG1 is positively associated with GC and Taxol resistance.A: Expression of SNHG1 in normal gastric tissues and GC tissues from TCGA database analyzed from http://ualcan.path.uab.edu.B: Expression of SNHG1 in normal gastric cell line and five GC cell lines by qRT-PCR.C: SNHG1 expressions in AGS and MKN-45 cells with control or SNHG1 siRNA transfection.D: Wound healing assay from AGS and MKN-45 cells without or with SNHG1 silencing.E and F: GC cells without or with SNHG1 silencing were treated with Taxol at the indicated concentrations.Cell viability was determined by MTT assay.G: The cell viability of MKN-45 parental and TXR cells under Taxol treatments at the indicated concentrations.H: Expression of SNHG1 in MKN-45 parental and TXR cells by qRT-PCR.I: MKN-45 TXR cells were transfected with control siRNA or SNHG1 siRNA, followed by treating with Taxol.Cell apoptosis was determined by Annexin V apoptosis assay.Ã P < 0.05; ÃÃ P < 0.01; ÃÃÃ P < 0.001.

Figure 2 .
Figure2.SNHG1 sponges miR-216b-5p.A: Prediction of the SNHG1-miR-216b-5p interaction from starBase.com.B: Correlation analysis of SNHG1 and miR-216b-5p from starBase.com.C: GC cells were transfected with control or SNHG1 siRNA.Expression of miR-216b-5p was determined by qRT-PCR.D: RNA pull-down assay was performed in GC cells as described in method section.E and F: Luciferase assay was performed in GC cells as described in method section.ÃÃ P < 0.01; ÃÃÃ P < 0.001.

Figure 4 .
Figure 4. Glucose metabolism is elevated in Taxol resistant GC cells.A: The glucose uptake and (B) ECAR from MKN-45 parental and TXR cells were measured.C: MKN-45 parental and TXR cells were cultured under normal or low glucose medium followed by Taxol treatment.Responses of cells were determined by clonogenic assay.D: MKN-45 TXR cells were treated with control or Taxol plus 2-DG at the indicated concentrations.Cell viability was determined by MTT assay.Ã P < 0.05; ÃÃ P < 0.01; ÃÃÃ P < 0.001.

Figure 5 .
Figure 5. SHNG1 and miR-216b-5p reversely regulate glucose metabolism of GC cells.A and B: The glucose uptake and lactate product from GC cells without or with SNHG1 silencing were measured.C and D: The glucose uptake and lactate product from GC cells without or with miR-216b-5p overexpression were measured.Ã P < 0.05; ÃÃ P < 0.01.

Figure 6 .
Figure 6.miR-216b-5p targets HK2 in GC cells.A: Prediction of the miR-216b-5p and HK2 3 0 UTR association from starBase.com.B: Expression of HK2 in normal gastric tissues and GC tissues from TCGA database analyzed from http://ualcan.path.uab.edu.C: Expression of HK2 in normal gastric cell line and five GC cell lines by qRT-PCR.D: Negative correlation between miR-216b-5p and HK2 in GC analyzed from starBase.com.E: Expression of HK2 in MKN-45 parental and TXR cells by qRT-PCR.F and G: GC cells without or with HK2 silencing were treated with Taxol at the indicated concentrations.Cell viability was determined by MTT assay.H and I: GC cells were transfected with control miRNA or miR-216b-5p.Expressions of HK2 were examined by Western blot.J and K:Luciferase assay was performed in GC cells as described in method section.Ã P < 0.05; ÃÃ P < 0.01.

Figure 7 .
Figure 7. Rescue of HK2 recovers the miR-216b-5p-promoted Taxol sensitization.A: MKN-45 TXR cells were transfected with control, miR-216b-5p alone or plus HK2.Protein expression of HK2 was determined by Western blot.B: The glucose uptake and (C) ECAR of the above transfected cells were measured.D: The above cells were treated with Taxol, cell viability was determined by MTT assay.Ã P < 0.05; ÃÃ P < 0.01; ÃÃÃ P < 0.001.

Figure 8 .
Figure 8.The SNHG1-miR-216b-5p-HK2 axis in Taxol resistant GC cells.A: MKN-45 TXR cells were transfected with control, SNHG1 alone or plus miR-216b-5p.Expressions of miR-216b-5p and (B) HK2 were determined by qRT-PCR and Western blot, respectively.C: The glucose uptake and (D) ECAR of the above transfected cells were measured.E: The above cells were treated with Taxol, cell viability was determined by MTT assay and (F) Annexin V apoptosis assay.Ã P < 0.05; ÃÃ P < 0.01.