JAG1 is associated with the prognosis and metastasis in breast cancer

Jagged canonical Notch ligand 1 (JAG1) regulates the progression of many cancers by the Notch signaling pathway, but its role in breast cancer (BC) remains unclear. In this research, JAG1 protein expression in BC tissues was detected by immunohistochemistry. The association between JAG1 and clinical significance was analyzed. The effect of JAG1 on malignant behaviors of BC cells was demonstrated by in vitro experiments. JAG1 expression in BC tissues was higher than that in para-carcinoma tissues. High JAG1 expression was significantly linked to advanced lymph node metastasis, distant metastasis, and the TNM stage. JAG1 was an independent prognostic factor for BC patients. JAG1 knockdown inhibited the proliferation, motility, migration, and invasion of BC cells, and weakened adhesion and penetration abilities to the blood–brain barrier, whereas JAG1 overexpression had the opposite effects. JAG1 has the potential to be a prognostic marker and therapeutic target for BC patients.

High JAG1 expression is linked to a poor prognosis. Univariate analysis demonstrated that risk factors for overall survival (OS) of BC patients included JAG1 expression (P < 0.001), the TNBC type (P = 0.031), lymph node metastasis (P = 0.001), distant metastasis (P < 0.001), and TNM stage (P < 0.001). Multivariate analysis confirmed that poor OS was significantly associated with high JAG1 expression (HR, 7.097; 95% CI:  Table 2). Kaplan-Meier survival analysis demonstrated that BC patients with high JAG1 expression showed poorer survival than other patients (Fig. 2).

Knockdown of JAG1 inhibits the malignant behaviors of MDA-MB-231 cells.
To further ascertain the function of JAG1, we downregulated JAG1 expression in the MDA-MB-231 (Fig. 3A, B). CCK-8 assays demonstrated that JAG1 knockdown significantly inhibited BC cell proliferation (Fig. 3C). A wound-healing assay confirmed that downregulation impaired the motility of MDA-MB-231 cells (Fig. 3D). Transwell assays revealed the inhibitory influence of JAG1 downregulation on migration (Fig. 3E, upper panels) and invasion (Fig. 3E, lower panels) of BC cells. Additionally, a BBB model revealed that JAG1 knockdown significantly weakened the tumor cell adhesion and penetration into the BBB (Fig. 3F, G).

Overexpression of JAG1 promotes the malignant behaviors of MDA-MB-231 cells.
Subsequently, we upregulated JAG1 in the MDA-MB-231 (Fig. 4A, B). CCK-8 assay demonstrated that JAG1 overexpression significantly promoted BC cell proliferation (Fig. 4C). Wound healing assays confirmed that JAG1 upregulation facilitated the motility of MDA-MB-231 cells (Fig. 4D). Transwell assays confirmed that high JAG1 expression contributed to enhanced cell migration (Fig. 4E, upper panels) and invasion (Fig. 4E, lower panels). Furthermore, the BBB model showed that JAG1 upregulation significantly increased tumor cell adhesion and penetration into the BBB (Fig. 4F, G).   19 demonstrated that JAG1 was overexpressed in hepatocellular carcinoma, and showed a significant connection to age and albumin level. In gastric cancer, JAG1 expression in tissues was associated with Borrmann type, and patients with high JAG leve had poorer survival than others 20 . Qiu et al. 21 confirmed that JAG1 in glioblastoma tissues was significantly increased than non-neoplastic tissues. In addition, high JAG1 expression was connected to advanced clinical features such as the Karnofsky performance scale and symptom duration, as well as poor survival 21 . Zohny et al. 18 detected JAG1 protein in 90 BC tissues and 42 benign lesions and found the positive rate of JAG1 in BC was higher than that in benign lesions (73.33% vs 26.19%). This research detected JAG1 expression in 200 BC tissues and 47 para-cancer breast tissues and proved that JAG1 protein was highly expressed in BC tissues compared to adjacent non-cancerous tissue (67.5% vs 23.4%), which was consistent with the literature. Correlation analysis confirmed that high JAG1 expression was significantly correlated to advanced lymph node metastasis, distant metastasis, and the TNM stage. In addition, BC patients with high JAG1 expression had a poorer OS than other patients, and JAG1 expression was an independent biomarker for BC patients.  25 found that cyclin D1 is a downstream transcription target of the JAG1-Notch pathway, and knockdown of the JAG1 gene decreases cyclin D1 level, arresting the cell cycle in the G1 phase. In basal-like subtypes of BC, high JAG1 expression induced by the NF-KB pathway could activate the Notch pathway, and then accelerate selfrenewal and replication of tumor stem cells 26 . Epithelial-mesenchymal transition (EMT) is essential for BC cells to break through the basement membrane 27 . JAG1-Notch pathway activation facilitates EMT of BC cells by upregulation of the transcriptional suppressor Slug and inhibition of E-cadherin 28 . JAG1 promotes urokinasetype plasminogen activator and enhances the invasive ability of BC cells, resulting in disease progression 29 . Our study showed that JAG1 knockdown in MDA-MB-231 cells inhibited their proliferation, migration, and www.nature.com/scientificreports/ invasion, whereas JAG1 overexpression had the opposite effects. The above phenomena indicate that JAG1 is a major driving factor promoting the malignant behavior of BC cells. The BBB is a tightly connected structure formed by cerebrovascular endothelial cells, basement membrane, and astrocytes 30 . It protects the brain from foreign macromolecules and microorganisms, and prevents most chemotherapeutic drugs and antibodies from entering the brain 31 . However, it cannot prevent the invasion of circulating metastatic cells, resulting in unsatisfactory therapeutic effects on brain metastasis. Therefore, it is urgent and necessary to further explore the molecular mechanism of BC cells crossing the BBB. Our results confirmed that JAG1 downregulation decreased the adhesion and penetration of BC cells into the BBB, and high JAG1 resulted in the opposite trend. This finding suggests the vital role of JAG1 in brain metastasis of BC.
In conclusion, we verified that JAG1 is overexpressed in BC tissues, and JAG1 is an independent prognosis biomarker for BC patients. JAG1 promotes BC progression and brain metastasis. This research affords a strong experimental basis for JAG1 to become a therapeutic target for advanced BC patients.

Methods
All experimental protocols were approved by Inner Mongolia Medical University. All methods were carried out in accordance with relevant guidelines and regulations.
Tissues and clinical data. Tissue samples analyzed in the research were collected from patients who accepted BC resection in Bayannur Hospital from January 2010 to December 2013. All patients did not receive any cancer treatments such as chemotherapy, radiotherapy, endocrine therapy, targeted therapy, or immune therapy. A total of 200 BC tissues and 47 para-cancer breast tissues were collected. Comprehensive clinicopathological information and follow-up data of patients were also collected. This research gained approval from the Human Research Ethics Committee of Bayannur Hospital. www.nature.com/scientificreports/ Immunohistochemistry (IHC). IHC was performed according to Envision two-step method. The microarray was dewaxed by xylene and hydrated by graded alcohol. Tris/EDTA solution (PH = 9.0) was selected as the antigen repair buffer, and hot antigen repair was carried out in a pressure cooker using the high-pressure steam method. The tissue microarray was soaked with 3% H 2 O 2 solution for 10 min and then incubated with 10% goat serum for 1 h. The tissue chips and rabbit anti-human JAG1 antibody (1:200, # ab85763, Abcam, USA) were incubated overnight at 4 °C. The tissue chip was incubated with the goat anti-rabbit antibody (1:500, #ab6721, Abcam, USA) at room temperature for 30 min. DAB dye was used for visualization processing, and the tissue chip was counterstained with hematoxylin. The expression of JAG1 protein was assessed semi-quantitatively through the combination of staining positive intensity and the proportion of staining positive cells 32 . The IHC cut-off value was 90 via the X-tile software 33 . A score of less than 90 was low or no expression, and the score higher or equal to 90 was a high expression.

Quantitative real-time polymerase chain reaction (qRT-PCR).
TRIzol reagent (Invitrogen, Thermo, USA) was utilized to leach the total RNA of each cell group, and the RNA samples were reversely transcribed into cDNA using the BeyoRT II cDNA synthesis kit (#D7170S, Beyotime, China). Samples were amplified using Bey-oFast Probe qPCR Mix (#D7273, Beyotime, China) on an ABI 7500 real-time fluorescent quantitative PCR system. The reference gene was GAPDH, and the relative expression of JAG1 was calculated by the 2 −△△Ct method.  Wound healing assay. The cells of each group were sown on 6-well plates (5 × 10 5 cells/well). When the cells converged 60-70% of the well, the original medium was replaced by serum-free DMEM medium. A sterile tip was used to scratch the cell surface. The cell wounds were photographed at 0 h, 12 h, and 24 h.

Statistical analysis.
All data analysis of this study was carried out by SPSS 22.0 software. The measurement data were expressed as mean ± standard deviation. T test was used for the difference between the two groups in accordance with normal distribution, or the M-W-U test was used. The χ 2 test was used to compare the rates between groups. Kaplan-Meier method and Log-rank method were used to calculate and compare the survival rate of patients in different groups. Univariate and multivariate Cox regression models were used to analyze the risk factors affecting the prognosis and survival of patients with BC, and P < 0.05 was considered statistically significant.
Ethics approval and consent to participate. This study acquired the approval of the Human Research Ethics Committee of Bayannur Hospital.
Informed consent. Informed consent was obtained from all the patients participating in the study.

Data availability
The data in this study are available from the corresponding author.