Notch2 as a promising prognostic biomarker for oesophageal squamous cell carcinoma

We aimed to examine Notch2 expression in oesophageal squamous cell carcinoma (ESCC) patients and to evaluate its prognostic potential. Immunohistochemical (IHC) staining, quantitative real-time polymerase chain reaction (qRT-PCR) and western blot analysis were utilized to investigate the Notch2 expression status and prognostic value. Furtherly, CCK8 and clonogenic assays were conducted to determine if Notch2 inhibition by shRNA could lead to a decrease in the proliferation and survival of ESCC cells. A notably higher Notch2 expression level was found in ESCC tissues at the mRNA (P < 0.0001) and protein levels (IHC: P = 0.004; western blot: P = 0.021). Log-rank analysis demonstrated that Notch2 overexpression was significantly associated with worse overall survival (OS) (29.1% vs. 49.1%; P = 0.013) and progression-free survival (PFS) (15.3% vs. 34.4%; P = 0.006) rates in ESCC patients. The multivariate analysis revealed Notch2 as an independent prognostic factor for OS and PFS (P = 0.002 and 0.006, resp.). Besides, in vitro assays showed that OD450 values and colony formations were significantly reduced in Notch2-shRNA group (all P < 0.0001). In conclusion, these results show that Notch2 is up-regulated in ESCC tissues and could serve as a promising biomarker for identifying individuals with poor prognostic potential.

Song Y et al. 22 conducted whole-genome sequencing and whole-exome sequencing in ESCC patients and identified significantly mutated genes, including genes involved in Notch signalling. Similarly, Notch2 was reported to be more frequently altered in ESCC compared with oesophageal adenocarcinoma via a comprehensive genomic profiling method [23][24][25] and immunohistochemical (IHC) staining 26 . At present, the precise association between Notch2 and ESCC prognosis has yet to be elucidated. The present study used IHC, qRT-PCR and western blot to examine the expression of Notch2 in human ESCC tissues and paracancerous tissues. In addition, the association between Notch2 and various clinicopathological characteristics was investigated, along with the predictive potential of Notch2 in ESCC. Further in vitro assays were used to verify its effects in proliferation and survival of ESCC cells.

Results
Notch2 expression in frozen ESCC tissues. We investigated the Notch2 mRNA and protein expression levels in 30 paired cancerous tissues and matched paracancerous tissues using qRT-PCR, IHC and western blot analysis. As shown in Fig. 1, Notch2 expression was primarily localized to the cytoplasm of cancer cells. IHC staining revealed that Notch2 was overexpressed in 63.3% (19/30) of cancerous tissues and 26.7% (8/30) of paracancerous tissues, and the difference between these levels was statistically significant (P = 0.004, Table 1). The qRT-PCR analysis revealed that Notch2 was up-regulated in cancerous tissues compared with matched paracancerous tissues in 19 (63.3%) cases (P < 0.0001). The mean fold increase in Notch2 mRNA in cancerous tissues was 4.71 ± 1.19 vs 1.04 ± 0.11 ( Fig. 2A). Furthermore, we randomly selected 8 pairs of cancerous and paracancerous tissues to identify the Notch2 protein level using western blot analysis (Fig. 2B). The results showed higher Notch2 protein expression in cancerous tissues than in adjacent tissues (Notch2/β -actin: 0.70 ± 0.20 vs. 0.51 ± 0.14, P = 0.021, Fig. 2C,D).
Notch2 inhibition with shRNA decreased EC-9706 cell proliferation and survival ability. ESCC cell lines (Eca 109 and EC 9706) were firstly transfected with shRNA-Notch2. qRT-PCR and western blot were used to test the efficacy of transfection. As shown in Fig. 4A,B, Notch2 mRNA level in test group were significantly induced compared with the control group (Notch2/β -actin: 0.26 ± 0.04 vs 1.01 ± 0.10, P < 0.0001 and 0.30 ± 0.04 vs 1.03 ± 0.08, P < 0.0001, resp.). Fig 4C,D showed that the protein level of Notch2 was also decreased in transfection group (Notch2/β -actin: 0.24 ± 0.01 vs 0.57 ± 0.02, P < 0.0001, and 0.20 ± 0.02 vs 0.50 ± 0.04, P < 0.0001, resp.). To determine if knock-down of Notch2 expression by shRNA could lead to a decrease in the proliferation and survival of ESCC cells, CCK8 and clonogenic assay were conducted. The OD450 values of the Eca 109 and EC 9706 cells transfected with shRNA-Notch2 showed significant decrease at 24, 48, and 72 h (all P < 0.0001), compared with those cells in the control groups (Fig. 5A,B). The colony formations of transfected Eca 109 and EC 9706 were also significantly reduced compared with the control groups (all P < 0.0001, Fig. 5C,D).

Discussion
Notch signalling has several demonstrated essential roles in the regulation of tumour growth, invasion, metastasis and angiogenesis 27 . Furthermore, the overexpression and oncogenic role of Notch2 have been observed in numerous human cancer types, such as lung adenocarcinoma 8 , glioma 9 , cervical cancer 10,11 , hepatoblastoma 12 , gastric cancer 13,14 and salivary adenoid cystic carcinoma 15 . By contrast, there is evidence supporting a suppressive role for Notch2 in bladder cancer 16 , lung cancer 17 , breast cancer 19 and colorectal cancer 20 . In the present study, we investigated the Notch2 expression level in ESCC tissues and analysed its prognostic value. For the first time, via qRT-PCR, IHC and western blot analyses, we found that Notch2 was up-regulated in frozen cancerous tissues compared with paracancerous tissues at the mRNA (P < 0.0001) and protein (IHC: P = 0.004; western blot: P = 0.021) levels. Similarly, Notch2 was overexpressed in 51.3% of FFPE tissues. Additionally, IHC staining revealed that Notch2 was primarily localized to the cytoplasm of cancer cells. These findings suggest an oncogene role for Notch2 in ESCC. However, Notch2 expression showed no significant relationship with clinicopathological features, such as the age, gender, smoking, drinking, T stage, N stage and differentiation, which may be due to the small quantity of sampling.
Notch2 has the potential to serve as a predictive biomarker in a variety of cancers. Accordingly, Notch2 mutations have been associated with a poor prognosis in splenic marginal zone lymphoma 6 , and the loss of Notch2 positively predicts survival in subgroups of patients with glial brain tumours 28 . A similar prognostic value was also demonstrated for Notch2 in liver cancer 29 . By contrast, a synergistic effect of positive Notch1 and negative Notch2 coexpression in predicting poor overall survival has been demonstrated 30 , and high Notch2 expression was shown to predict good survival for breast cancer patients 19 . In the present study, log-rank and multivariate analyses demonstrated that Notch2 expression in cancer tissues served as an independent prognostic factor for OS and PFS in ESCC patients. In particular, high Notch2 protein expression significantly predicted decreased 5-year OS and PFS; the T and N stages were also identified as prognostic indicators. The ROC-AUC was provided for the Cox regression models. It is a value that indicates the concordance level between observed and expected ordering of the data, and represents the percentage of concordance of all pairs of data with different outcome values, with concordance defined as occurring within a pair when the observation with the higher outcome value also has the higher predicted probability of the outcome. The AUC value for Notch2 was significant according to the OS and PFS predictions. Furthermore, CCK8 and clonogenic assay were conducted to determine if Notch2 inhibition could lead to a decrease in the proliferation and survival of ESCC cells. As expected, the results indicated that OD values and colony formations were significantly reduced in transfected group. Thus, we can conclude that Notch2 can be used as a biomarker for predicting ESCC survival in patients who underwent surgery. Besides, Notch1 has been widely reported in esophageal cancer 22,23 . It is also a novel potential prognostic biomarker for ESCC patients 31,32 . Further studies are necessary to investigate associations of Notch pathway signal molecules and synergistic effect of Notch1 and Notch2 coexpression in predicting survival of ESCC.
The detailed mechanisms of Notch2 have been studied and reported in several tumours. The silencing of Notch2 inhibits glioma cell proliferation by inducing cell cycle arrest and apoptosis in vitro and in vivo 9,33 . Constitutive Notch2 signalling in neural stem cells has been reported to promote tumorigenic features and  astroglial lineage entry 34 . As a target gene of miR-107 35 , miR-204-5p 10 and miR-23b 13 , Notch2 may also regulate cell migration and tumour invasion. In particular, Notch2 may negatively regulate cell invasion by inhibiting the PI3K-Akt signalling pathway in gastric cancer 36 . In salivary adenoid cystic carcinoma tissues, Notch2 may target HEY2 and CCND1 to regulate cell proliferation, invasion, and migration. Furthermore, the loss of the Notch pathway was shown to promote the epithelial-mesenchymal transition in bladder cancer cells, which was partially mediated by the loss of HES1 16 . Moreover, Notch2 activation by ZER inhibits its proapoptotic and anti-migratory response in breast cancer cells 37 . However, the mechanisms governing the role of Notch2 in ESCC have not been reported, and further studies are necessary to elucidate these cellular processes.
In conclusion, Notch2 is up-regulated in ESCC tissue compared to matched paracancerous tissue, and its overexpression could serve as a promising biomarker to identify individuals with poor prognostic potential.

Methods
Patient recruitment and data collection. Thirty pairs of ESCC and paracancerous tissue from patients were collected from October 2014 to March 2015 from Qilu Hospital of Shandong University. In addition, 140 FFPE cancer tissue samples from patients who underwent subtotal esophagectomy and esophagogastric anastomosis plus regional lymph node dissection in Qilu Hospital during 2009 were collected. Because 25 patients were lost to follow-up, we included 115 cases in the prognostic analysis. All cases were pathologically confirmed as ESCC. Patients did not receive chemotherapy or radiotherapy before surgery. We obtained the relevant data, including age, sex, smoking and drinking habits, histologic grade, invasion depth (T stage), lymph node metastasis (N stage), distant metastasis (M stage), differentiation degree and number of dissected lymph nodes, from clinical or pathologic records. The tumour, node, metastasis (TNM) classification was performed according to the American Joint Committee on Cancer staging manual (7th edition, 2010). The study protocol was approved by the ethics boards of Qilu Hospital, and tissue specimen acquisition was performed in accordance with the institutional guidelines. The written informed consent was obtained from all subjects.    primary rabbit anti-Notch2 polyclonal antibody (1:100, Abcam, Cambridge, MA, USA) overnight at 4 °C in a high humidity chamber, followed by incubation for 30 min at 37 °C with biotinylated secondary antibodies and streptavidin-peroxidase complex. Finally, a 3,3′-diaminobenzidine solution was added, and the slides were counterstained with haematoxylin and mounted with neutral balsam. For negative controls, sections were incubated with PBS instead of the primary antibodies. The sections were observed under a light microscope and independently scored by three investigators. Conflicting scores were resolved by selecting the value that was consistent between two observers or the average    CCK8 assay. The Eca 109 and EC 9706 cells proliferation was detected using 2-(4-indophenyl)-3-(4-nitrophenyl)-5-(2,4-disulphophenyl)-2 Htetrazolium monosodium salt (cell counting kit-8 (CCK8)). After transfection, the logarithmically growing cells were seeded in 96-well plates at a cell density of 5 × 10 4 /well and incubated for 0, 24, 48, and 72 h. At different time intervals, the cells were incubated with CCK8 reagent for 1 h at 37 °C. The absorbance of each well was measured at 450 nm using Thermo Scientific Varioskan Flash (Thermo Scientific, Finland). Percentage of viable cells = (OD450 of treated sample − OD450 of blank sample)/(OD450 of control sample − OD450 of blank sample) × 100%. The results shown were mean values of three independent experiments.
Clonogenic assay. Clonogenic assay was also used to evaluate the function of Notch2. Transfected cells were trypsinized to generate a single cell suspension and seeded in 6-well plates at 500 cells per well. 14 days after seeding, colonies were stained with crystal violet, and the number of colonies containing at least 50 cells was counted. The colony survival fraction was calculated for each treatment.
Statistical analysis. The difference in the Notch2 mRNA and protein level between cancerous and paired paracancerous tissues, as well as difference in Notch2 level, OD450 value and colony formation between shRNA-Notch2 and control group was compared using a paired Student's t-test. The chi-square test was used to test the correlation between Notch2 expression and clinicopathological factors. The Kaplan-Meier method was used to calculate the survival curves, and the log-rank test was used to compare the survival difference between patient subgroups. Multivariate Cox regression analysis was used to identify significant independent prognostic factors. The ROC-AUC was also provided for the Cox regression models. AUC values are calculated for each of the adjusted models to illustrate the predictive ability of the independent variables. Differences between groups were considered significant for P values < 0.05. All statistical analyses were performed with SPSS 17.0 statistical software (SPSS Inc., Chicago, IL, USA).