Activated interferon signaling by down-regulation of CENP-N contributing to inhibited tumor growth in breast cancer

Background Centromere proteins (CENPs) are primary components for chromosomal segregation in the mitotic stage. CENP-N is a member of CENPs, and is a key factor for recruitment of other CENPs and formation of a link between the centromere and micro-tubules, which facilitate cell division. Methods In order to clarify the role of CENP-N in breast cancer, RNA sequences data were downloaded from TCGA online database and the CENP-N expression was knocked down in breast cancer cells. Results The results show that the expression of CENP-N was higher in breast cancer comparing with the paracancerous tissues. In breast cancer, patients with high expression of CENP-N have a short-term overall survival compared with low expression of CENP-N. Both in vitro and in vivo, the growth of breast cancer cells was inhibited by down-regulation of CENP-N. In the gene-chip analysis, it reveals that down-regulation of CENP-N is primarily associated with functions of immune response and anti-tumor ef-fects. Of these changed canonical pathways, the activated interferon signaling was the most significant in CENP-N down-regulated breast cancer cells. In the western blot as-say, up-regulated expressions of molecules involved in interferon signaling were also confirmed. Conclusions Our results suggest that CENP-N can be a potential therapeutic target in the treatment of breast cancer, and the involved interferon signaling needs to be mainly fo-cused on. of CENP-N using gene-chip analysis. 5’-TGACTTCAACAGCGACACCCA-3’; downstream, 5’- CACCCTGTTGCTGTAGCCAAA. of determine of to determine to examine Down-regulation CENPN decreased cell proliferation, cell viability in MDA-MB-231 (g, i-j) MCF7 cells (h, k-l). Down-regulation of CENPN increased apoptosis and caspase3/7 activity percentage in MDA-MB-231 (m, n) and MCF7 cells (o,

formation of CCAN complex and assembly of kinetochore [8,9]. Additionally, CENP-A nucleosome retention at centromeres also requires being fastened by CENP-N [10]. Deficiency of CENP-N was demonstrated to lead to disrupted chromosome alignment and mitotic arrest [8,11].
As evidence accumulated, centromere proteins were linked with many cancer types, and consider as the potential targets for cancer treatment. High expressions of centromere proteins were found in many kinds of cancers, such as hepatocellular carcinoma [12,13], ovarian cancer [14], Lung adenocarcinoma [15], cervical cancer [16], and breast cancer [17][18][19]. CENP-A was indicated as a prognostic indicator for relapse in estrogen receptor (ER) positive breast cancer [18], and an independent prognostic factor in lung adenocarcinoma [15]. CENP-H was found to promote proliferation of gastric cancer cells, and highly expressed in hepatocellular carcinoma [12]. CENP-K was identified to be specifically up-regulated in ovarian cancer cells, and its overexpression was associated with poor patient survival [14]. These centromere proteins are all proved involving in assembly of kinetochore, in which process CENP-N is required [8].
However, the relationship between CENP-N and tumor is less understood. In a multivariable analysis, CENP-N, CETN1, CYP1A1, IRF2, LECT2, and NCOA1 were reported to be important predictors for both breast carcinoma recurrence and mortality among smokers [20]. For all the above supporting, we postulate CENP-N may associate with breast cancer malignant phenotype and progression, and can be a potential therapeutic target. In our study, we will focus on the expression of CENP-N in breast cancer tissues and the effect of reducing CENP-N expression on proliferation and -tumor formation of breast cancer cells in vitro and in vivo, and find the underling mechanisms by comparing the changed expression of genes after down-regulation of CENP-N using gene-chip analysis.

Ethics approval and consent to participate
All animal procedures were supervised and approved by the Jinzhou Medical University Laboratory Animal Welfare and Animal Experimental Ethical Committee.

Study design
Downloaded clinical data of CENP-N mRNA expression on breast cancer and paracancerous tissues were analyzed to compare the differences. In order to confirm the potential role of CENP-N on breast cancer, we examined the malignant phenotype of breast cancer cells with or without deficient expression of CENP-N in vitro or in vivo. Finally, microarray analysis for genome-wide effects of silencing CENP-N in breast cancer MDA-MB-231 was performed to find the changed and associated signaling pathways, and diseases and functions.

Online data and analysis of breast invasive carcinoma (BRCA)
For the TCGA set, clinical data and CENP-N mRNA expression (RNA-seq and RNA-seq V2) were downloaded form the TCGA data portal (http://tcga-data.nci.nih.gov/tcga/). The data included 1094 samples with the CENP-N mRNA expression, of the 106 samples have their paired CENP-N mRNA data in the paracancerous tissues. The differential expression of CENPN between breast cancer and adjacent normal tissues were evaluated. Overall survival analyses between breast cancer patients separated by median of CENPN mRNA expression were performed using the downloaded data and online database (http//:kmplot.com). Cells were maintained in DMEM (Coring, New Jersey, USA) medium containing 10% fetal bovine serum (Gibco, thermo Fisher Scientific, Inc., USA), and 1% penicillin and streptomycin in a 5% CO2 humidified atmosphere. Trypsin-EDTA was used to detach cells. In the next five days, the number of cells was counted using Celigo Image Cytometer each time a day.

MTT assay
Lentivirus transfected MDA-MB-231 (2×10 4 cells/ml) and MCF7 (3×10 4 cells/ml) cells suspension were prepared and seeded into 96 well plates (100 µl per well). After attachment, cells were incubated in 5% CO2 at 37˚C for 1 to 5 days. Next, cells were added with 10 µl MTT reagent (5mg/ml in PBS) and incubated with cells for 4h. In order to dissolve the formazan product, 100 µl of dimethyl sulfoxide (DMSO) was added. At 490 nm, the absorbance was measured using a microplate reader, and continuous 5-day changes of the optical density (OD) value were analyzed.

Apoptosis assay
Cells of MDA-MB-231 and MCF7 were trypsinized, counted and planted in 6-well plates. In the second day, cells were transfected with shCENPN and shCtrl lentivirus. 72 h later, according instruction of Apoptosis detection kit (88-8807, eBioscience, USA), transfected cells reach ~80% confluence, were trypsinzed, washed, and 10 6 cells were resuspended in 200µl 1× biding buffer. Afterward, 10µl Annexin V-APC staining solution was added and incubated with cell suspension at room temperature avoiding from light for 10-15 min. Finally, 800 µl 1× biding buffer was added and cell apoptotic percentage was examined within 1 h.

Caspase3/7 assay
Cells were transfected with lentivirus as described in apoptosis assay. Afterward, cells were trypsinzed, counted and seeded into each 96 well plate at a density of 10000 cells, and were cultured at 37˚C in a humidified incubator containing 5% CO2 overnight. Caspase3/7 activity was examined by Caspase-Glo ® 3/7 assay kit (G8901, Promega, USA). According to manufacturer's instruction, Caspase-Glo solution was prepared. After, each well was added 100 µl Caspase-Glo solution and shaking at a speed of 300-500 rpm using a 96-well plate shaker for 30 min. For another 2 h incubation at room temperature, micoplate reader was used to determine intensity of fluorescence signals. Caspase3/7 activity was evaluated by analyzing the examined data.

Nude mice and in vivo Tumor growth assay
4-week old female BALB/c nude mice averagely weighted 21 gram were obtained from Shanghai Lingchang Biotechnology Co., Ltd. The mice were bred in specific pathogen free housing, and 3-5 female mice were kept in a cage, which was away form light. The bedding material and food were all A total of 20 mice were randomized and allocated to two groups using statistical software (JMP software). MDA-MB-231 cells were transfected with shCENPN and shCtrl lentivirus. Afterwards, 10 7 transfected MDA-MB-231 cells were inoculated subcutaneously into right axillae of nude mice (10 miles for each group). When tumor could be observed, tumor length and width were measured and tumor volume was calculated as volume =length×width 2 /2. The measurement was conducted each three days until tumor volume up to 1cm 3 . At this time, nude mice were killed by cervical dislocation, and tumor were removed and weighted up. Photos of nude mice and tumor were taken and preserved.

Microarray analysis
Using a GeneChip® PrimeView™Human Gene Expression array (901838, Affymetrix, USA), which can measure gene expression of more than 36,000 transcripts and variants per sample, the genome-wide effects of CENP-N knockdown were evaluated. Total RNA of transfected MDA-MB-231 cells were extracted, and examined for quality control using Thremo Nanodrop 2000 spectrophotometer and Alient Bioanalyzed 2100. Amplified RNA (aRNA) was then prepared using GeneChip® 3' IVT labeling kit (Affymetrix) according to manufactuters' instructions. Next, GeneChip Hybridization Wash and Stain Kit (Affymetrix) were used for target hybridization and fluidics setup, according to manufacturers' instructions. Finally, array scanning was performed using CeneChip Scanner 3000 (Affymetrix). Data were collected and normalized to find the differentially expressed gene that the absolute of gene expression fold change is more than 1.5.

Ingenuity Pathway Analysis
Datasets of differentially expressed genes from microarray analysis were imported into the Ingenuity Pathway Analysis (IPA) tools (Ingenuity® Systems, USA). According to manufactures' instructions, IPA is an integrative online analysis software containing more than 5 millions of biological information from a number of literatures and more than 30 public databases. Different functional modules were set to help understand interaction network among different molecules and diseases and functions.
Classical pathway analysis was performed to find the possible activated or inhibited signaling pathways involved in CENP-N knockdown. In the present study, CENP-N associated diseases and functions were also analyzed. During the analytic process, two statistical indicators were used, Zscore was to evaluate which process is activated (Z-score>2) or suppressed (Z-score<-2), as for P value was to assess whether the difference is significant.

Statistical analyses
JMP software (version 11, SAS Institute Co. Ltd, China) was used for statistical analysis. Student's t test was used to analyze difference between two groups. One-way ANOVA followed by Turkey-Kramer HSD method was used to determine differences among multiple groups. Analysis of variance of repeated measures was conducted to determine differences of tumor growth in continuous 5 days.
Kaplan-Meier method was performed to examine differences of overall survival between patients in low-expression CENPN group and high-expression CENPN group. P<0.05 was identified as a statistical significance.

Up-regulation of CENP-N in breast cancer tissues and correlated with decreased overall survival
To determine the role of CENP-N on progression of breast cancer, we downloaded RNAseq and clinical characteristic data of 106 paired breast cancer and adjacent noncancerous breast tissues (Table1), as well as of 1094 breast cancer tissues from TCGA database ( Table 2). Differential expression of CENP-N between breast cancer and normal tissues was significant (P<0.0001) (Fig.1a), and fold change (FC) calculated as Log2 (cancer/Normal) was plotting in Fig.1b, with an average value of 2.177. Kaplan-Metier analysis showed higher expression of CENP-N was associated with a short-term overall survival for patients with breast cancer (P=0.042) (Fig.1c). Consistently, the same result was also found in the online database (http://kmplot.com) (Fig.1d-g). Moreover, CENP-N expression on tumor with negative expression of estrogen receptor (ER) (Z=-16.58, P<0.0001) or progesterone receptor (PR) (Z=-15.35, P<0.0001) or positive expression of human epidermal growth factor receptor-2 (HER2) (Z=3.71, P =0.0006) was higher (Table 2). Altogether, these results identified a significantly positive association between progression of breast cancer and CENP-N expression.

Inhibited proliferation and expansion of breast cancer cells by silencing of CENP-N in vitro and in vivo
To investigate effect of CENP-N on tumorigenesis of breast cancer, CENP-N expression was assayed in human breast cancer cell lines. It was found that CENP-N expression was great, and then lentivirus vector virus (shCtrl), then CENP-N mRNA and protein expression were examined to confirm efficiency of CENP-N silencing (Fig. 2a-f). Findings of MTT, colony formation and proliferation assays were extraordinarily similar, these results all indicated that interference of CENPN-N expression lead to inhibited cell proliferation, viability and colony formation (Fig. 2g-l). Moreover, apoptosis assay revealed that apoptotic percent of MDA-MB-231 and MCF7 cells were increased by down-regulation of CENP-N, and caspase3/7 assay conferred to the same result ( Fig.2m-p). Consistently, in vivo experiment, the result showed inoculated tumor growth was dramatically slowed in CENP-N knockingdown group (shCENP-N) compared with control group (shCtrl) in nude mice (Fig.3). Taken together, we could conclude that CENP-N may be a key regulator during breast cancer proliferation and expansion.

Microarray analysis for genome-wide effects of silencing CENP-N in MDA-MB-231 cells
To disclose mechanisms underlying influences causing by CENP-N knockdown, differentially expressed gene profiles were examined by GeneChip® PrimeView™ Human Gene Expression assay. Compared with control group, 199 up-regulated and 278 down-regulated genes were found in MDA-MB-231 cells with CENP-N knockdown. To further illustrate the potential biological interactions among these differently regulated genes, the Ingenuity Pathway Analysis (IPA) were conducted. The results showed these differentially expressed genes were largely enriched in cancer, cell death and survival, and infectious diseases with the absolute value of Z-score more than 2 which indicated theses functions and diseases were significantly activated (Z-score>2) or suppressed (Z-score<-2) (Fig. 4a). In the canonical pathway analysis, activated Interferon signaling and inhibited PI3K/AKT signaling were highlighted (Fig. 4b). Moreover, up-regulated interferon signaling was more significant than depressed PI3K/AKT signaling in CENP-N silenced MDA-MB-231 cells (Fig. 4b). These results suggested that activated Interferon signaling casing by down-regulated CENP-N might be the main factor contributing to inhibited cell growth of breast cancer cells.

Activation of interferon signaling after down-regulation of CENP-N
In the canonical signaling pathway analysis, we found that genes involved in interferon signaling, such as STAT2, IRF1, IFIT1, CCND1, CASP1, ISG15 were all increased in the microarray analysis (Fig.   4c). Western blot assay was performed to determine the protein expressions of these factors. In accordance with canonical pathway analysis, compared with shCtrl group, molecules of STAT2, CASP1, and ISG15 were all upregulated in shCENP-N group, which is accompanied by decreased expression of CCND1 (Fig. 4d). These results indicate that inhibited breast cancer cell growth by knocking down CENP-N might be partially attributed to activated interferon associated signaling.

Discussion
Tumor development is partially dictated by rapid proliferation that associated with uncontrolled cell cycle process [21]. Several centromere proteins (CENPs), which are necessary for cell division, are highly expressed in malignant tumors, and targeting CENPs have been demonstrated to inhibit tumor growth or enhance effect of chemotherapy in cancer cells [22,23]. As the role of CENP-N that is required for recruitment of many other CENPs and formation of the platform for chromosomal separation [8], we investigate the effect of deficiency of CENP-N in breast cancer cells. Our findings provided evidence supporting the hypothesis that targeting CENP-N leaded to inhibited cell proliferation and viability coupled with increased apoptosis by activating interferon associated signaling in breast cancer cells.
We showed that the RNA expression of CENP-N was overexpressed in breast cancer tissues and associated with unfavorable overall survival. Grouped by status of ER, PR, or HER-2, higher expression of CENP-N was found in the groups of negative ER, negative PR or positive HER-2. For patients with ER and PR negative or HER-2 positive breast cancer usually predict poor prognosis [24,25], our results suggest that CENP-N may be an indicator of aggressive phenotype of breast cancer cells. As expected, reducing the expression of CENP-N in breast cancer cells lead to inhibited cell proliferation and increased apoptosis, which was in accordance with the finding that tumor growth was slowed in vivo.
As considerable role of CENPs has been reported, CENPs were identified as therapeutic targets in cancer treatment. High expression of CENP-A was reported in ER negative breast cancer compared with ER positive subjects, and positively correlated with Ki67 expression [18]. By targeting CENP-A, the aggressive phenotype of lung adenocarcinoma cells could be attenuated [23]. Using a lung cancer mouse model with down-regulation of CENP-E, researchers found that low rates of chromosome missegregation can promote tumorigenesis, however, missegregation of high numbers of chromosomes results in tumor inhibition [26]. Knockdown the expression of CENP-K decreased cell viability in triple negative breast cancer cells [27]. On account of the core status of CENP-N during CCAN assembly, kinetochore formation and chromosome segregation [8,9,11], CENP-N may be an effective therapeutic target in the treatment of breast cancer, especially in the aggressive subjects.
Gene profiling was evaluated to clarify the changed molecules involving in down-regulating expression of CENP-N in breast cancer cells. Analyses of diseases or functions annotation using differential expressions of molecules after CENP-N knocking-down showed that many tumor development associated diseases or functions were significantly suppressed, which were accompanied by activated functions of immune response, cell apoptosis and cell death. Within these diseases and functions, the mostly affected is the immune response. Recently, enhanced tumor immunity has been considered to be associated with the anti-cancer effect [28]. Patients with more tumor-infiltrated lymphocytes (TILs) usually have a long-term overall survival, especially in estrogen and progesterone receptor negative breast cancer [29,30]. Moreover, more infiltration of immunologic effector cells after chemotherapy indicates a good response to the treatment [31]. Thus, the enhanced immune response is closely associated with inhibited tumor progression and increased therapeutic effect. Therefore, significantly activated immune response after down-regulation of CENP-N may be potential stimuli to enhance anti-tumor effect. Moreover, increased apoptosis and cell death effect by down-regulation of CENP-N may coordinately inhibit breast tumor growth. These findings support the notion that targeting CENP-N can block breast cancer development by promoting apoptosis and enhancing immune response for killing cancer cells.
In our study, further analysis reveal that significantly activated Interferon signaling by reducing the expression of CENP-N may contribute to blocked tumor growth. Interferon was found to be against tumor primarily owing to their anti-proliferation and immune-stimulating activities, which can function through activation of specific signal transducers and activators of transcription (STATs) and interferon stimulated genes (ISGs) [32][33][34]. In our study, by knocking down of CENP-N, STAT2 and a number of ISGs were identified, such as interferon-induced transmembrane protein 1 (IFITM1), interferon induced protein with tetratricopeptide repeats 3 (IFIT3), interferon induced protein with tetratricopeptide repeats 1 (IFIT1), Interferon alpha-inducible protein 6 (IFI6) and interferon regulatory factor 1 (IRF1), which indicated the activation of interferon signaling. Consistent the anti-tumor effect by knocking down of CENP-N, IRF1 was reported to be associated with the growth inhibitory activity of IFN-γ in human breast carcinoma cells [35][36][37]. Additionally, IRF1 is widely considered as a tumor suppressor to promote apoptosis, inhibit tumor growth and progression in breast cancer. Furthermore, high IRF1 expression is associated with prolonged survival in patients with breast cancer. In addition, down-regulated cyclin D1 (CCND1), up-regulated caspase 1 (CASP1) and interferon-stimulated gene together, activated interferon signaling through down-regulation of CENP-N, which is characterized by up-regulated IRF1, may attribute to inhibited tumor growth in breast cancer.
In addition to anti-tumor effect on tumor growth, the interferon signaling also participates in immune responses. Free ISG15 is found to show anti-tumor activity by increasing NK cell infiltration, and therefore suppress tumor growth. Thus, our study provide the concept that targeting CENP-N may be used in the treatment of breast cancer, which is dependent on its role in both directly inhibiting tumor growth and indirectly promoting immune responses. However, overexpression of ISG15 is also linked to tumorigenesis and unfavorable prognosis in breast cancer. Though the targeting CENP-N therapy could be a promising therapeutic regimen, the mechanisms still need to be further examined and elucidated.

Conclusion
Centromere proteins are essential for mitosis and cell proliferation, and associated with tumor initiation and progression. In the present study, we found that targeting CENP-N could promote apoptosis and inhibit tumor growth, which could be partially attributed to the activated interferon signaling that is characterized by increased expression of IRF1. Our study provides a new strategy for the treatment of breast cancer, and targeting CENP-N therapy may not only have a function on breast cancer, but also on immune responses.

Abbreviations
List of Abbreviations