Netrin-1 promotes gastric cancer cell proliferation and invasion via the receptor neogenin through PI3K/AKT signaling pathway

Netrin-1 is a laminin-related protein found to promote proliferation and invasion in multiple types of cancers. Recent studies have identified the function role of netrin-1 in several cancers; however, the influence of netrin-1 in human gastric cancer(GC) remains largely unknown. In this study, we found netrin-1 was upregulated in human GC tissues, where its expression correlated inversely with cancer stage and lymph node metastasis. We detected netrin-1 and its receptor knockdown significantly suppressed GC cells proliferation and invasion, while overexpression netrin-1 reversed these effects. Xenografted analyses using GC cells displayed significantly inhibition of tumor growth and metastasis by netrin-1 depletion. Furthermore, we identified that netrin-1 as a regulator of PI3K/AKT pathway to modulate GC cells proliferation and invasion abilities via its receptor neogenin. Taken together, our findings argued that netrin-1 and its receptor neogenin might act synergistically in promoting GC cells proliferation and invasion through the PI3K/AKT signaling pathway. It is conceivable that netrin-1 could be new therapeutic target to GC therapy.


INTRODUCTION
Gastric cancer (GC) remains a major public health issue as the fourth most commonly diagnosed cancer. It is the third leading cause of cancer-associated mortality in male and the fifth in female worldwide [1]. Most of GC patients are diagnosed at an advanced stage and with lymph nodes or distal metastasis [2]. It is important to improve the understanding of the mechanisms underlying growth and metastasis in GC.

Research Paper
Oncotarget 51178 www.impactjournals.com/oncotarget Although it is clear that netrin-1 and their receptors play important role in cancer progression, the detailed molecular mechanisms involved are not well understood. The PI3K/AKT signaling pathway was considered as an important player in cancer cell proliferation and invasion [18]. Recently, it was reported that netrin-1 induced cancer cell migration and invasion through PI3K/AKT pathway in hepatic cancer [5].In this study, we investigated the role of netrin-1 in GC development. We have demonstrated that blockage of endogenous netrin-1 resulted in the suppression of GC cells proliferation and invasion in vitro and in vivo. Our studies also identified netrin-1 functions as a novel regulator for the PI3K/AKT pathway via the receptor neogenin to mediate GC cells proliferation and invasion. We concluded that the netrin-1/neogenin pathway held promise as a novel therapeutic target to inhibit GC growth and metastasis.

Netrin-1 was overexpressed in GC tissues, and its expression correlated with lymph node metastasis and cancer stage
The specimens into groups including GC tissues and adjacent normal tissues were collected from 86 GC patients. We found netrin-1 mRNA expression level of GC tissues was upregulated compared with the matched non-cancerous tissues ( Figure 1A). GC tissues with more than two-fold enrichment in netrin-1 mRNA expression level were defined as the high expression group, whereas those with a less than two-fold increase in netrin-1 mRNA expression level were defined as the low expression group. As shown in Table 1, netrin-1 expression level  was correlated with lymph node metastasis and cancer stage. Next, we determined netrin-1 protein expression in human GC tissues by immunohistochemistry(IHC) and confirmed that netrin-1 protein expression was almost absent in adjacent normal gastric tissue. Additionally, netrin-1 expression was upregulated in GC tissues, and was further increased in those with lymph node metastasis (LNM; Figure 1B-1D). Moreover, we assessed the expression levels of netrin-1 mRNA and protein in GC cell lines by qRT-PCR and western blotting. As observed in Figure 1E-1F, HGC27 and AGS cell lines expressed netrin-1 at high levels, SGC7901 and MGC803 cell lines expressed netrin-1 at moderate levels, BGC823 cell line expressed netrin-1 at low level, while no expression was observed in MKN45 cell line.

Netrin-1 silencing inhibited GC cells proliferation, migration, and invasion in vitro
To address the efficacy of netrin-1 on GC cells, we knocked down netrin-1 in HGC27 and AGS cells by using Oncotarget 51179 www.impactjournals.com/oncotarget two different shRNA. As shown in Figure 2A-2B, each GC cell line transfected with netrin-1 shRNA showed efficient silencing of netrin-1 expression, as determined by qRT-PCR and western blotting. We next observed the effect of netrin-1 silencing on the proliferation ability of the GC cells. With CCK-8 assay, we found that netrin-1 inhibition significantly decreased the proliferation ability of GC cells compared with the control group ( Figure 2C). We also examined the effect of shNTN1 by colony formation assay and found that netrin-1 knockdown notably suppressed GC cells proliferation ( Figure 2D, 2E). Taken together, our results demonstrated that netrin-1 suppression inhibited GC cells proliferation.
We next investigated whether netrin-1 knockdown could regulate GC cells migration and invasion. We conducted Transwell assay to further illustrate the impact of netrin-1 on migration and invasion abilities of GC cells. We discovered that netrin-1 knockdown markedly reduced the number of migrated HGC27 and AGS cells ( Figure 2F, 2G). In addition, the number of invasive HGC27 and AGS shNTN1 cells were obviously decreased compared with negative control cells ( Figure 2H, 2I). Thus, our date suggested that netrin-1 knockdown inhibited GC cells migration and invasion abilities in vitro.

Netrin-1 overexpression promoted GC cells proliferation, migration, and invasion in vitro
To further assess the role of netrin-1 in the proliferation, migration, and invasion abilities of GC cells, we overexpressed netrin-1 in BGC823 and MKN45 cell lines whose netrin-1 expression level was lower than other GC cell lines. The expression level of netrin-1 in BGC823 and MKN45 cell lines which were transfected into netrin-1 lentivirus were significantly higher than that in negative control group( Figure 3A, 3B). CCK-8 and colony formation assays revealed that the overexpression of netrin-1 obviously enhanced cell proliferation ability in both BGC823 and MKN45 cells ( Figure 3C-3E). These data supported netrin-1 overexpression promoted GC cells proliferation, confirming the role of netrin-1 in regulation of cell proliferation ability.
To prove the role of netrin-1 in GC cells migration and invasion abilities, we determined the role of netrin-1 overexpression in BGC823 and MKN45 cells motility by using Transwell assay. Transwell assay also discovered that netrin-1 overexpression increased the number of migrated and invaded GC cells ( Figure 3F-3I).

Netrin-1 increased GC cells proliferation and invasion through receptor neogenin
Netrin-1 exerted its effects by binding to its receptor on cell membrane. We found neogenin and UNC5B expression levels were higher than other receptors in GC cell lines ( Figure 4A, 4B). To further address the role of UNC5B and neogenin in the proliferation and invasion abilities of GC cells, we knocked down both neogenin (named siNeo) and UNC5B (named siUNC5B) in HGC27 cells. Western blotting showed that UNC5B and neogenin siRNA efficiently reduced protein expression in HGC27 cells, respectively ( Figure 4C). The CCK-8 and colony formation assays indicated that siNeo significantly decreased the proliferation ability of HGC27 cells, while siUNC5B did not block cells proliferation ( Figure 4D and Supplementary Figure 1A). There was no additional effect on GC cells proliferation using a combination of UNC5B and neogenin siRNA. In addition, silencing of neogenin also decreased HGC27 cells invasion, while siUNC5B has no effect ( Figure 4E, 4F). Because the expression level of netrin-1 was highest in HGC27 cells, we next knocked down both netrin-1 and neogenin ( Figure 4G). Our results showed that combination of netrin-1 and neogenin siRNA strongly suppressed GC cells proliferation ability by using CCK-8 and colony formation assays ( Figure 4H and Supplementary Figure 1B). Meanwhile, Transwell assay showed that GC cells invasion ability was suppressed significantly when netrin-1 and neogenin were both silencing ( Figure 4I, 4J).These results suggested that the netrin-1/neogenin loop could be a target to repress the proliferation and invasion abilities of GC cells.

In vivo analysis on the role of netrin-1 in regulating GC cells growth and metastasis
HGC27 cells with netrin-1 inhibition were injected into flanks of nude mice to form xenograft tumors ( Figure 5A). As shown in Figure 5B-5C, the tumor size and weight of netrin-1 inhibition group were markedly smaller than the control group. In converse, the tumor size and weight of netrin-1 overexpression group were larger than the control group ( Supplementary Figure 2A-2C). We next employed IHC assay to analyze GC cells proliferation in vivo. As shown in Figure 5D-5E, the positive nuclei rate of Ki-67 was remarkably lower in netrin-1 inhibition group. On the contrary, the positive nuclei rate of Ki-67 was higher in netrin-1 overexpression group (Supplementary Figure 2D-2E). Thus, our date indicated that GC cells growth was impaired after netrin-1 silencing.
To investigate whether netrin-1 could affect metastasis of GC cells in vivo, HGC27 cells with netrin-1 knockdown and negative control cells were injected into tail vein of BALB/c nude mice. After about 6 weeks, we found that the nude mice which were injected into HGC27 cells with netrin-1 knockdown suppressed metastasis of GC cells in vivo ( Figure 5F). Thus, our results showed that knockdown netrin-1 significantly decreased cancer metastasis to the liver. As shown in Figure 5G-5H, HGC27 control cells formed large liver metastases, while the netrin-1 inhibition group has no metastasis. Taken together, these results suggested that netrin-1 knockdown could suppress metastasis of GC cells in vivo, which was consistent with our experiments results in vitro.

Effects of netrin-1 and neogenin on PI3K/AKT signaling pathway
In order to determine the molecular mechanism underlying the netrin-1 induced expression changes of proliferation and invasion-related biomarkers, we analyzed the potential downstream signaling pathway affected by netrin-1 knockdown and overexpression. FAK was an important downstream effector of netrin-1 [19]. We found FAK phosphorylation was declined in HGC27 and AGS cells with netrin-1 knockdown, while total FAK has no change. Conversely, we detected increased FAK phosphorylation in netrin-1 overexpression group( Figure 6A). Next, the activities of main downstream of FAK-the PI3K-AKT signaling pathway-was examined following netrin-1 treatment. We found that AKT phosphorylation was reduced in HGC27 and AGS cells with netrin-1 knockdown compared to negative control group, while no change was detected in their total protein levels. In contrast, the phosphorylation of AKT was increased in BGC823 and MKN45 cells with netrin-1 overexpression, and its total protein levels has no change ( Figure 6A). In order to further understand how netrin-1 induced GC cells invasion, we examined the expression of matrix metalloproteinase 9 (MMP-9), a invasion-related biomarkers. As shown in Figure 6A, when blocking the netrin-1 expression, the expression level of MMP-9 was significantly downregulated. Conversely, overexpression of netrin-1 increased MMP-9 expression level ( Figure 6A).
We also detected that neogenin knockdown decreased the phosphorylation of FAK, AKT, and MMP-9 in BGC823 cells ( Figure 6B).We found neogenin knockdown significantly decreased the proliferation ability of BGC823 cells by using CCK-8 assay ( Figure 6C). In addition, silencing of neogenin also decreased BGC823 cells invasion ability by using Transwell assay ( Figure 6E, 6F). Moreover, we demonstrated that LY294002 (AKT inhibitor) suppressed BGC823 cells proliferation and invasion abilities ( Figure 6D, 6G, 6H). In a word, our results declared that the activation of PI3K/ AKT axis was essential for netrin-1 to promote GC cells proliferation and invasion.

DISCUSSION
Our primary goal in this study was to identify molecular factors that induce proliferation and invasion Oncotarget 51183 www.impactjournals.com/oncotarget Oncotarget 51184 www.impactjournals.com/oncotarget of GC. In tumors, netrin-1 acted as an oncogene that promoted cancer cells proliferation, adhesion and invasion [2,11,16]. In this study, we demonstrated that increased netrin-1 expression in GC tissues was associated with clinical cancer stage and lymph node metastasis. The results from in vitro and in vivo studies showed that silencing netrin-1 expression significantly decreased GC cells proliferation, migration, invasion, growth, and metastasis, suggesting that netrin-1 functions as a novel tumor inducer in GC. Moreover, we found the oncogenic effects of netrin-1 on GC cells proliferation and invasion via its receptor neogenin through PI3K/AKT signaling pathway.
The netrin family consists of three secreted netrins which include netrin-1, netrin-3 and netrin-4. Netrin-1 was known to promote cancer cell proliferation and invasion during tumor development [5,6,12,20,21]. Netrin-3 and its receptor neogenin were expressed in glioblastoma cells and associated with cell migration [22]. Previous work in GC cell lines revealed that netrin-4 administration resulted in receptor-mediated regulation of cell invasiveness, angiogenesis and cancer progression [23]. In addition, netrin-4 was found to promote glioblastoma cell proliferation via UNC5B [24].Like netrin-4, our results also indicated that netrin-1 was able to stimulate GC cells proliferation and invasion, implying the oncogenic effect of netrin-1 in GC progression.
Netrin-1 is a secreted protein, which binds to its receptor to exert its effects. We found netrin-1 and its receptor were both expressed in several GC cells, suggesting the possibility of an autocrine functional loop. However, the receptors and downstream signaling Oncotarget 51185 www.impactjournals.com/oncotarget pathways of netrin-1 mediating these functions are not clear. Recent studies have reported that netrin-1 could bind to neogenin and UNC5B to promote medulloblastoma cell invasiveness [10]. Of all netrin-1 receptors, UNC5B and neogenin seem to be the most predominant in all six GC cell lines in our study. Using siRNAs inhibited neogenin or UNC5B protein expression respectively, we found GC cells proliferation and invasion abilities were markedly suppressed by interacting specifically with neogenin. These findings consistent with previous studies reporting that neogenin promoted the growth and invasive potentials of GC cells [25]. Interestingly, it was reported that netrin-1 suppressed growth of pancreatic cancer through receptor UNC5B [26].This disparity may be because of the presence of bifunctional receptors and interaction with them (UNC5A-D, DCC, neogenin and DSCAM) in different tumors.
Netrin-1 exerts its functions by activation of intracellular signaling pathways. FAK was one of the main downstream effectors of netrin-1 through the receptors Oncotarget 51186 www.impactjournals.com/oncotarget DCC and neogenin [27][28][29][30]. Our studies suggested that FAK phosphorylation was decreased in GC cells with netirn-1 inhibition, while netrin-1 overexpression reversed this effect, suggesting that netrin-1 could regulate FAK activity. In addition, PI3K/AKT signaling pathway was activated after netrin-1 treatment, which has been confirmed a downstream of netrin-1 in other studies [5,31]. Moreover, netrin-1 stimulated schwann cells migration by activating PI3K/AKT signaling pathway [32]. Reportedly, netrin-4 induced GC cells proliferation and invasion by stimulating phosphorylation of AKT [23]. Consistently, our results indicated that netrin-1 knockdown decreased the phosphorylation of AKT, while netrin-1 overexpresssion increased its phosphorylation. Our previous studies suggested that MMP-9 was a key molecule for cell invasion and associated with netrin-1 in pancreatic cancer [33]. In this study, our data provided the evidence that MMP-9 was required for the netrin-1 signaling pathway in promoting GC cells invasion. Intimate crosstalk between the neogenin and FAK was observed [34].Our results also found neogenin knockdown decreased the expression levels of p-FAK, p-AKT and MMP-9. In a word, it is possible that netrin-1 promotes the proliferation and invasion of GC cells in a neogenindependent manner through further activation of PI3K/ AKT signaling pathway.
Taken together, our in vitro and in vivo studies indicated that netrin-1 functions as a novel tumor inducer in GC development and progression. In addition, we revealed a critical mechanism for netrin-1 in regulation of GC cells proliferation and invasion via its receptor neogenin to regulate the PI3K/AKT pathway. This may highlight a new entry point for treating GC by targeting the netrin-1/neogenin/AKT signaling axis.

Samples and patients
GC tissue samples were collected from 86 patients who underwent radical resection at the First Affiliated Hospital of Nanjing Medical University. No chemotherapy or radiation therapy was administered before surgery. Written informed consent was obtained from all patients or from their relatives. The use of all tissue blocks for this study was approved by the Ethics Committees of Nanjing Medical University. The Ethics Committee permission number is 2015-SRFA-027.

Quantitative real-time PCR
Total cellular RNA was isolated using Trizol (Invitrogen Life Technology, CA, USA) and cDNA was synthesized using PrimeScript RT Reagent (Takara, Dalian, China) according to the manufacturer's protocol.

Immunohistochemical analysis
The GC tissues and xenograft tumor samples were fixed with 4% formaldehyde solution. After being embedded within paraffin, the 4μm of slices were incubated with diluted primary antibody. The following primary antibodies were used: Ki-67 and netrin-1 (Abcam, Cambridge, UK; 1:200 dilution). Ki-67 and netrin-1 expression levels were evaluated according to the staining intensity (0 for absent, 1 for weak, 2 for moderate and 3 for strong staining). The proportion of positive cells was scored as 0(negative), 1 (< 10%), 2 (10-50%), 3 (50-75%) and 4 (> 75%). The two scores were multiplied, and the median value was calculated to be 6. Values of ≥ 6 were defined as high expression, and values <6 were regarded as low expression. All immunostanined sections were determined respectively by two pathologists who www.impactjournals.com/oncotarget did not know the research contents. The sum of the intensity and proportion scores were used to analyze the immunoreactivity levels.

Cell proliferation and clonogenic assay
For cell proliferation detection, cells were transfected with lentivirus vectors and selected with puromycin at a concentration of 1 μg/ml after transfection for 48 h. Then, the cells were seeded into 96-well plates 1000 cells per well in triplicate and incubated with RPMI 1640 (10% FBS) for 5 days for CCK-8 colorimetric assay (Dijindo, Japan) according to the manufacture's specifications. For the clonogenic assay, 500 cells were seeded into six-well plates and cultured for 14 days. The colonies on the plates were fixed with 4% paraformaldehyde for 30 minutes, then stained with 0.1% crystal violet and counted colony numbers.

Transwell assay
For invasion assay, Matrigel solution (BD Biosciences, USA) was prepared in serum-free cell culture medium at a dilution of 1:9, coated with the 24-well transwell chambers (Corning Costar, USA) overnight at 37°C before cell seeding. Complete culture medium containing 10% FBS was used as the chemoattractant in the lower chamber, 3 × 10 4 cells were seeded in the upper chamber in serum-free medium and incubated in 5 % CO2 atmosphere at 37°C for 24 h. The noninvaded cells were gently wiped from the upper surface of the membrane with cotton-tipped swabs. The invaded cells on the lower surface of the membrane were stained with 0.1% crystal violet for 20 min and counted. The experiments were performed in triplicate. The same experimental method was used for the migration assay, but the membranes were not pre-coated with Matrigel.

Western blotting assay
GC cells were processed for protein extraction according to standard procedures. The following primary antibodies were used: Netrin-1 (Abcam, Cambridge, UK; 1:200 dilution), GAPDH, p-FAK, FAK, p-AKT, AKT and MMP-9 (Cell Signaling Technology, Danvers, USA; 1:1000 dilution), UNC5B,neogenin (Sigma-Aldrich, St Louis, USA; 1:1000 dilution). GAPDH was used as an internal control. Protein was separated on the 10%SDS polyacrylamide gel and transferred to PVDF membranes (Bio-Rad Laboratories) which was then blocked in TBST containing 5%milk. The membranes were incubated with specific first antibodies in dilution buffer at 4 °C overnight. After the membranes were washed with TBST, the blotted membranes were incubated with HRR-conjugated antimouse or anti-rabbit IgG (1:1000) at room temperature for 2 h. Targeting protein expression levels were detected using an enhanced chemiluminescence (Millipore, Billerica, MA, USA) detection system.

Tumor xenograft of human GC cells in nude mice
Four-week-old male nude mice(BALB/c nude mice) were purchased from the Department of Laboratory Animal Centre of Nanjing Medical University. HGC27 and BGC823 cells were implanted by subcutaneous injection of 2 × 10 6 cells in 100μl of PBS into flanks of mice to make tumor. Four weeks later, mice were sacrificed and tumors were harvested. The xenograft diameters were measured using a slide caliper every other day until day28. The xenograft tumor volume was calculated using the following formula: v = 0.5 ab 2 (a = the long diameter of the tumor, b=the short diameter of the tumor, and v= volume) [35]. Negative control HGC27 cells and netrin-1 knockdown cells (5 × 10 6 cells in 100 μl PBS) were injected into tail vein of nude mice respectively. After 6 weeks, IVIS Imaging system (Caliper life Sciences, Hopkinton, MA) was used to observe the occurrence of distant metastasis. Liver tissues were collected for metastatic foci evaluation and standard histopathologic study. Care of experimental animals were in accordance with Nanjing Medical University Institutional Animal Care and Use Committee.

Statistical analysis
The statistical analyses were performed using SPSS version 22.0 (SPSS, Inc., Chicago, USA). The data were presented as the mean ± standard error of the mean (SEM) unless indicated otherwise. The statistical significance