Background
Gastric cancer (GC) is the second leading cause of cancer-associated mortality worldwide and incidence rates are highest in Eastern Asia, Latin America, Central and Eastern Europe [
1,
2]. In China, gastric cancer is also a main malignant tumour and a chief reason of cancer deaths. The majority of GC patients are diagnosed at an advanced stage, 5-year survival rate of 11–42%. The prime determinant of survival following gastric carcinoma appears to be the development of liver metastasis [
3,
4]. Despite surgical resection and chemoradiotherapy can control most cancer cells [
5], a surgical resection has been rarely indicated for liver metastasis from gastric cancer [
6]. So far, the complex molecular mechanism of liver metastasis has still remained essentially unknown. Therefore, we need to explore novel molecules to better understand the mechanism of hematogenous metastasis.
Metastatic spreading and the formation of secondary neoplasms from primary site are not random, exhibiting organ selectivity [
7]. Recently the roles of intrinsic cancer cell properties have been investigated, such as selectin. In experimental metastasis studies, researchers demonstrate that liver sinusoidal endothelial cell lectin (LSECtin) mediated colon cancer cells metastasis to liver displays enhanced abilities to the specific organ [
8]. Also, serum of soluble E-selectin (sE-selectin) concentration in gastric cancer patients are detected by ELISA, but increasing only in gastric cancer patients with peritoneal metastasis [
9]. Similarly, hepatic sinusoidal endothelial E-selectin expression is up regulated by highly metastatic cells entering the liver [
10]. Moreover, using an E-selectin-specific monoclonal antibody reduces liver metastasis, showing that E-selectin is involved in metastatic formation in this organ [
11]. For further study, blocking colorectal carcinoma-induced hepatic endothelial E-selectin expression inhibits liver metastasis [
12]. These events suggest that selectin play a key role in tumour metastasis to the target organs. DC-SIGNR (DC-SIGN-related protein, also known as L-SIGN, CD299) as a member of C-type lectin belonging to selectin is found high serum concentration in colon cancer patients [
13]. Here, we ask whether DC-SIGNR contributes to hematogenous metastasis from gastric carcinoma.
Long ncRNAs (lncRNAs) which lengths are more than 200 nt are abundant in the human genome [
14]. Recently, several long ncRNAs have been reported to have a role in gastric cancer metastasis. The lncRNA HULC is higher expression in GC tissues than pair-matched adjacent normal tissues and is significantly associated with distal metastasis and lymphatic metastasis [
15]. While, FENDRR, as a tumour suppressor lncRNA, is downregulated in GC tissues and cell lines. Overexpression FENDRR exhibits the inhibiting capacity for cell migration and invasion in vitro and effectively reduces the number of metastatic nodules in vivo [
16]. According to situ hybridization analysis and microarray data, the lncRNA GAPLINC is associated with GC proliferation, migration and angiogenesis. These functions are reduced by CD44 repression [
17]. Another two extensively studied lncRNAs are HOTAIR and H19. They are correlated with GC development and poor prognosis. Gain and loss function analysis have showed that HOTAIR and H19 can drive gastric cell lines proliferation, migration and invasion [
18,
19].
Signal transducer and activator of transcription (STAT) molecules are ubiquitously expressed in many tumour, such as breast cancer, lung cancer and head and neck cancer [
20]. Interestingly, STAT family members can migrate to the nucleus to regulate gene expression at transcription level. Zhu et al. addressed that STAT5 controls some genes expression by binding to promoter sequences [
21]. CXCR4 is ubiquitously expressed in many human cancer cell lines [
22]. It binds to unique ligand CXCL12. So CXCR4 positive tumour cells move towards the high concentration CXCL12 organs to realize tumour metastasis.
In this study, DC-SIGNR was significantly increased in serum of gastric cancer patients and increased in middle-late patients. Then we characterized that DC-SIGNR promoted the biological function proliferation, migration and invasion on GC cell lines in vitro and promoted GC cells liver metastasis in vivo. Next, we employed LncPath chip to investigate the mechanism of DC-SIGNR mediated gastric cancer liver metastasis by the lncRNA HNRNPKP2. HNRNPKP2 was upregulated after DC-SIGNR knockdown. Furthermore, STAT5A promoted HNRNPKP2 expression after knockdown DC-SIGNR and CXCR4 was obviously decreased after knockdown HNRNPKP2. Collectively, we firstly found that DC-SIGNR was increased in gastric cancer patients serum and DC-SIGNR facilitated gastric cancer liver metastasis. A novel lncRNA HNRNPKP2 regulated by STAT5A was influenced by DC-SIGNR, and then DC-SIGNR promoted the expression of CXCR4.
Methods
Serum collection
Serum of gastric cancer patients and healthy samples were obtained from the First Hospital affiliated to Dalian Medical University and the Second Hospital affiliated to Dalian Medical University between 2013 and 2014. All gastric cancer cases were reviewed by pathologists and histologically confirmed as gastric cancer (stageI,II, III, IV; 7th Edition AJCC) based on histopathological evaluation. Clinical pathology information was available for all samples (Additional file
1: Table S1). No local or systemic treatment obtained from all patients. Recombinant human DC-SIGNR Fc chimera was used as standard substance (R&D Systems). The protocols and procedures were approved by the Dalian Medical University research ethics committee. Written informed consent was obtained from all patients. The standard curve of sDC-SIGNR was listed in Additional file
2: Figure S1.
Cell lines and culture conditions
Four gastric cancer cell lines (BGC823, SGC7901, MGC803 and HGC27), three colon cancer cell lines (SW620, LoVo and HCT116), three breast cancer cell lines (MCF-7, MDA-MB-231, MDA-MB-435), two liver cancer cell lines (PLC/PRF/5 and HepG2) and one human normal liver cell line (LO2) were purchased from the Institute of Biochemistry and Cell Biology of the Chinese Academy of Sciences (Beijing, China). Cells were cultured in RPMI 1640 or DMEM (Hyclone) medium supplemented with 10% fetal bovine serum (tBD Bioscience), 100 U/ml penicillin, and 100 mg/ml streptomycin (Beyotime) in humidified air at 37 °C with 5% CO2.
RNA extraction and qRT-PCR analyses
Total RNA was extracted from formalin fixed and paraffin embedded gastric cancer tissues using RNAprep pure FFPE kit (TIANGEN). Total RNA was extracted from cultured cells using TRIZOL reagent (TaKaRa Bio, Otsu, Japan). For qRT-PCR, RNA was reverse transcribed to cDNA by using random primer (TIANGEN Biotech). QRT-PCR analyses were performed with Power SYBRGreen (TaKaRa). Results were normalized to the expression of GAPDH. The primers were listed in Additional file
3: Table S3. QRT-PCR and data collection were performed on Thermal Cycler Dice Real Time System (TaKaRa Bio, Otsu, Japan)
PCR reactions were performed at 94 °C for 2 min, followed by 41 cycles of 94 °C for 15 s and 58 °C for 20 s. ΔCt was calculated by subtracting the Ct of internal control RNA from the Ct of lncRNA or the mRNA of interest, respectively. ΔΔCt was then calculated by subtracting the ΔCt of the internal control from the ΔCt of the target group. Fold change of DC-SIGNR or HNRNPKP2 was calculated by the equation 2-ΔΔCt.
Ectopic expression and knockdown of DC-SIGNR
Based on the expression of DC-SIGNR in GC cell lines, we selected SGC7901 cells for the enhanced expression study and BGC823 cells for the knockdown study. For ectopic expression, the full-length DC-SIGNR cDNA was subcloned into the LV5 lentiviruses (GenePharma, Suzhou) and infected into SGC7901 cells to generate SGC7901/DC-SIGNR.
For knockdown of DC-SIGNR expression, two complementary oligonucleotides of small hairpin RNA sequences were chemically synthesized, subcloned into the pGLV3 lentiviruses (GenePharma, Suzhou) and infected into BGC823 cells to generate BGC823/DC-SIGNR shRNA.
Briefly, a total of 3 × 104 cells were plated in 24-well plates for 24 h and then infected with the lentiviral vectors described above by means of polybrene (GenePharma, Suzhou) for 48 h. Concentration of puromycin (SIGMA) on SGC7901 and BGC823 is 0.6 μg/ml to get stably transfected cell lines. The negative control (GenePharma, Suzhou) was infected in parallel. The cells were then subjected to RNA extraction and further functional assays.
Knockdown HNRNPKP2, knockdown STAT5A and ectopic expression of STAT5A. Knockdown HNRNPKP2 was developed by transfecting siRNAs (GenePharma, Suzhou) against HNRNPKP2 into BGC823/DC-SIGNR shRNA to generate BGC823/DC-SIGNR shRNA/siHNRNPKP2. Negative siRNAs (GenePharma, Suzhou) were transfected as negative control and termed BGC823/DC-SIGNR shRNA/NC.
Knockdown STAT5A in SGC7901/DC-SIGNR cells was produced by transfecting siSTAT5A (GenePharma, Suzhou) and named SGC7901/DC-SIGNR shRNA/siSTAT5A. Negative siRNAs (GenePharma, Suzhou) were transfected as negative control and termed SGC7901/DC-SIGNR shRNA/NC.
For ectopic expression of STAT5A, pCMV3-Flag-STAT5A (CWBIO, Beijing) and pCMV3-empty were transfected into BGC823/DC-SIGNR shRNA.
Briefly, a total of 3 × 10
6 cells were plated in 6-well plates for 24 h and then transfected with the siRNA or plasmid described above by means of Lipofectamine 2000 (Invitrogen). The cells were then subjected to RNA extraction after 24 h and to protein extraction after 48 h for further functional assays. Target sequences for DC-SIGNR knockdown or overexpression, for HNRNPKP2 siRNA and for STAT5A siRNA were listed in Additional file
4: Table S4.
Cell proliferation assay
Cell proliferation was monitored by ELISA with MTT kit (Biosharp) according to the manufacturer’s instruction. Briefly, 100ul of cell suspension from each subgroup (2,000 cells/well) was placed in a 96-well plates and pre-incubated for 12 h. Then, 20 μl of MTT solution was added to each well and incubated for 4 h until purple precipitate was visible. After adding detergent agent for 2 h, the number of cells was counted every 24 h for 5 days by measuring the absorbance at 490 nm using Microplate (BIO-RAD). For colony formation assays, cells were placed into 6-well plates and maintained in media containing 10% FBS for 2 weeks. Colonies were fixed with methanol and stained with 0.1% crystal violet. Visible colonies were manually counted.
Flow-cytometric analysis of BGC823 and SGC7901
BGC823 cells and SGC7901 cells were harvested by trypsinization. After staining with CD299-APC (eBioscience) and CXCR4-PE (BioLegend), cells were analyzed with a flow cytometry (FACScan®; BD Biosciences) equipped with a Summit 5.0 software.
Wound-healing assay
For the wound healing assay, a total of 6 × 105 cells were seeded in 6-well plates and cultured overnight until 90% confluent. A sterile yellow pipette tip was used to make a straight scratch. The suspension cells were washed off twice gently and images of the scratch were acquired as baseline. The medium was then replaced and images of the same location were obtained using a microscope for next days.
Cell migration assay and invasion assays
Cell migration was assessed using transwell chamber of 8-μm pore size (Corning) according to the manufacturer’s instruction. Briefly, 200 μl of serum-free medium containing 3 × 105 cells from each subgroup were added to the upper chamber. For the invasion assays, 4 × 105 cells in 200 μl serum-free medium were added to the upper chamber coated with Matrigel (BD MatrigelTM). A volume of 0.6 ml of 10% FBS-containing medium was then added to the lower chamber as a chemoattractant. Cells were incubated for another 24 h at 37 °C in 5% CO2. After the incubation, cells on the upper surface of the membrane were scraped off with cotton swabs. Cells migrated to the bottom of the membrane were fixed and stained with 0.1% Crystal Violet Staining Solution. The cells on the bottom of the membrane were counted from five different microscopic fields and the average number was calculated.
4 ~ 6-week-old athymic BALB/C nude mice were purchased from Yangzhou University. For xenograft models, BGC823 cells infected negative control were collected at growing stage when confluence was about ~80%. 2 × 106, 4 × 106, 6 × 106, 8 × 106, 1 × 107 BGC823 cells were injected subcutaneously into the right flanks of the five BALB/C nude mice. The state of the mice were observed every three days, including tumour volumes. Fifteen days after implantation, mice were sacrificed and tumour were removed out. The formula V (mm3) = A × B2/2 was applied to calculate the volume (A means the largest diameter, B means the perpendicular diameter).
For metastasis assays, 5 × 106 BGC823 cells with DC-SIGNR knockdown (n =4 per group) and negative control (n =4 per group) in 100 μl PBS were inoculated into spleen of mice. At the first sigh of suffering post-inoculation, tumour in situ and hepatic metastasis were removed for IVIS (Carestream Health, Inc.) and sliced for IHC. The protocol was approved by the Committee on Ethics of Animal Experiments of the Dalian medical University.
Immunohistochemistory (IHC)
Paraffin-embedded, formalin-fixed tissues of mice liver and spleen were immunostained for Ki67, CK7 and CK20 proteins. The signal was amplified and visualized with diaminobenzidine-chromogen, followed by counterstaining with hematoxylin. Anti-Ki67 (1:50), anti-CK7 (1:50) and anti-CK20 (1:50) were purchased from Proteintech (CK7 and CK20) and Bioworld (Ki-67).
Microarray analysis
Samples preparation and microarray hybridization were performed by Kangchen Bio-tech, Shanghai P.R. China. Briefly, RNA was extracted from cultured flask at a concentration of 1 × 107 cells using TRIZOL reagent (TaKaRa Bio, Otsu, Japan). Then, each sample was amplified and transcribed into fluorescent cRNA along the entire length of the transcripts without 3’ bias utilizing a random priming method. The labeled cRNAs were hybridized onto the LncRNATM Arrays (6 × 7 K, Arraystar). After having washed the slides, the arrays were scanned by the Axon GenePix 4000BScanner. GenePix Pro software (version 6.0) was utilized to analyze acquired array images. Quantile normalization and subsequent data processing were performed using the R limma software package.
Western blot assay
The protein was extracted from cultured cells using a total protein extraction kit (KeyGen, Nanjing, China) and the protein concentration of the lysates was measured using the BCA Protein Assay Kit (Beyotime Biotechnology, Shanghai, China).
Equivalent amounts of protein was electrophoresed in 8% SDS polyacrylamide gels for STAT5A protein and 12% SDS polyacrylamide gels for β-actin protein, transferred onto 0.22 μm nitrocellulose membranes (Pall corporation). The membranes were incubated with specific antibodies and corresponding secondary antibodies. Signals were detected with electrochemical luminescence reagent (Advansta, USA). β-actin was used as control, anti-STAT5A (1:500) was purchased from Elabscience.
Statistical analysis
Student’s t-test (two-tailed) and one-way ANOVA test were performed to analyze the in vitro and in vivo data using GraphPad Software. P values less than 0.05 were considered significantly.
Discussion
Metastasis accompanies advanced stages of gastric cancer progression. However, the role of selectin in gastric cancer liver metastasis remains poorly understood. We focus on DC-SIGNR which has been associated with tumour biological functions. Therefore, DC-SIGNR has an important role in cancer metastasis. In this research, we explore whether DC-SIGNR mediates gastric cancer liver metastasis.
Our previous research results show that LSECtin of C-type lectin family plays an important role in tumour metastasis [
8]. LSECtin has no expression in colon cancer cells but a lot of expression in sinusoidal endothelial cells. During metastasis, some ligands maybe mediate engagement. In this study, we find another C-type lectin family member DC-SIGNR acts as oncogene by promoting cell proliferation, malignant and invasion in gastric cancer metastasis. Unlike colon cancer metastasis, DC-SIGNR expresses in gastric cancer cells mediated liver metastasis. The mechanisms potential the role of DC-SIGNR in metastasis regulation were complex. And which molecule in downstream can respond to DC-SIGNR to contribute to metastasis and invasion in gastric cancer, we focus on long non-coding RNA. Strikingly, Heterogeneous nuclear ribonucleoprotein K pseudogene 2 (HNRNPKP2), as a pseudogene, is found to be upregulated in BGC823 cell line after DC-SIGNR depletion. So far, there has been no related research on HNRNPKP2. Heterogeneous nuclear ribonucleoprotein K (hnRNP K) and HNRNPKP2 are highly homologous. According to reports, pseudogenes act as a antisense regulator to effect on protein-coding mRNAs, such as PTEN [
28]. Therefore, we speculate that HNRNPKP2 transcript may have the same function to suppress or enhance hnRNP K. HnRNP K as multifunctional signaling protein has several functions in tumorigenesis. First, hnRNP K interacts with K17 to regulate some pro-inflammatory expression, such as C-X-C chemokine, promoting skin tumour keratinoscyes to grow and invade by CXCR3 signaling pathway [
29]. Second, hnRNP K regulates gene expression of proliferation to affect tumour progression, such as c-myc, c-Sre and p53 [
30‐
33]. Third, focal adhesion can integrate internal and external signal to control cells migration and focal adhesion compound precursor contains hnRNP K. The content of hnRNP K plays an important role in metastasis, limiting the speed of cell proliferation [
34,
35]. Our results show that most of genes involve in signaling events mediated by focal adhesion kinase. Additionally, the negative correlation between DC-SIGNR and HNRNPKP2 had been found not only in gastric cell lines but also in gastric cancer tissues. From sDC-SIGNR of gastric cancer patients, we speculate DC-SIGNR should be mediated gastric liver metastasis in middle-late stage. Therefore, 17 paired formalin fixed and paraffin embedded gastric cancer tissues and corresponding para-carcinoma tissues can be divided into different stages to analyze if DC-SIGNR effected HNRNPKP2 in middle-late stage. It is better to get more middle-late stage gastric cancer solid tumours to explore the association between DC-SIGNR and HNRNPKP2. Based on them, we speculate that DC-SIGNR mediated gastric cancer liver metastasis maybe causes by HNRNPKP2 to effect on hnRNP K for further studies.
So far, there are some researches on transcription factors to regulate lncRNA expression. STAT3 binds to lncRNA on dendritic cell to effect its differentiation [
36]. ANRIL (lncRNA) regulated mTOR and CDK6/E2F1 pathway and then CDK6/E2F1 promotes ANRIL expression in form of positive feedback. At the same time, ANRIL suppresses miR-49a/449a expression in trans by binding to PRC2 which contained EZH2, SUZ12 and EED [
37]. Activated by c-Fos at transcription level, MALAT1 (lncRNA) interacts with EZH2 and SUZ12 to increased RCC progression [
38]. The numerous CCAT1 (lncRNA) is increased by c-Myc which binding to the E-box on CCAT1 promoter region [
39]. So, we speculate that some transcription factors regulated HNRNPKP2 expression with different DC-SIGNR expression. In solid tumors, persistent STAT molecules are involved in tumorigenesis in many cancers by regulating the expression of critical mediators during cancer formation and metastatic progression. That is to say, some gene expression more depends on STAT activity. STAT molecules migrate from cytoplasm to nucleus and occupied upstream specific promoter region to regulate gene expression at transcription level [
40,
41]. Similarly, we take effort to find which transcription factor to regulate HRNNPKP2 expression. From some candidate transcription factors, STAT5A can emerge as transcription activator to enhance HNRNPKP2 expression by binding to enhancer. The exact DNA sequences need to be further verified.
Chemokines and their receptors are one of the most representative patterns for organ-selective examples [
42]. CXCR4 is a seven-span transmembrane G protein-coupled chemokine receptor and CXC chemokine ligand (CXCL) 12 is its unique ligand. CXCR4 is usually overexpressed in a variety of human cancers [
43,
44], but absent or low in a lot of normal tissues [
45]. Inflammatory chemokine CXCL12 of enriched organs forms a concentration gradient and attracts CXCR4 positive tumour cells to homing by directed movement. Tumour cells moved towards the highest concentration CXCL12 gradients, realizing the directional migration and positive feedback mechanisms in an autocrine manner [
46]. Hence, CXCL12/CXCR4 biological axis plays an important role in regulating specific organs metastasis. After together with its corresponding receptor, there is a transducting signal by intracellular calcium influx and leading to the activation of downstream pathways, such as MAPK1/MAPK3 activation.
Acknowledgements
The authors thank all members in our lab for the excellent technical help.