Knockdown of NUSAP1 Inhibits Cell Proliferation and Invasion Through Down-Regulation of TOP2A in Human Glioblastoma

Background: Nucleolar and spindle associated protein 1 (NUSAP1) is an indispensable mitotic regulator, which has been reported to be involved in the development, progression, and metastasis of several types of cancer. Here, we investigated the expression and biological function of NUSAP1 in human glioblastoma multiforme (GBM). Methods: The expression of NUSAP1 on GBM tissues and cells were determined by database analysis, immunohistochemistry and Western blot. EdU assay, transwell assay and ow cytometric analysis were performed to evaluate the effect of NUSAP1 knockdown on GBM cell proliferation, cell invasion and cell apoptosis. RNA sequencing was used to screen for downstream molecules altered in GBM cells after NUSAP1 depletion. An intracranial mice model and bioluminescent imaging were used to assess the effect of NUSAP1 on tumor growth and survival time in vivo. Results: Analysis of the molecular data in CGGA, TCGA and Rembrandt datasets demonstrated that NUSAP1 was signicantly up-regulated in GBM relative to low grade gliomas and non-neoplastic brain tissue samples. Kaplan-Meier analysis indicated that patients with tumors showing high NUSAP1 expression exhibited signicantly poorer survival in both CGGA (P = 0.002) and Rembrandt cohorts (P = 0.017). Analysis of RNA sequencing data from P3-cells with stable knockdown of NUSAP1 revealed topoisomerase 2A (TOP2A) as a possible molecule down-regulated by the loss of NUSAP1. SiRNA knockdown of either NUSAP1 or TOP2A in U251, T98 and GBM derived patient P3 cells inhibited GBM cell proliferation and invasion, and induced cell apoptosis. Finally, stable knockdown of NUSAP1 with shRNA led to decreased tumor growth in an orthotopic xenograft model of GBM in mice. Conclusions: Taken together, NUSAP1 gene silencing induced apoptosis possibly through the down-regulation of the candidate downstream molecule TOP2A. Interference with the expression of NUSAP1

Results: Analysis of the molecular data in CGGA, TCGA and Rembrandt datasets demonstrated that NUSAP1 was signi cantly up-regulated in GBM relative to low grade gliomas and non-neoplastic brain tissue samples. Kaplan-Meier analysis indicated that patients with tumors showing high NUSAP1 expression exhibited signi cantly poorer survival in both CGGA (P = 0.002) and Rembrandt cohorts (P = 0.017). Analysis of RNA sequencing data from P3-cells with stable knockdown of NUSAP1 revealed topoisomerase 2A (TOP2A) as a possible molecule down-regulated by the loss of NUSAP1. SiRNA knockdown of either NUSAP1 or TOP2A in U251, T98 and GBM derived patient P3 cells inhibited GBM cell proliferation and invasion, and induced cell apoptosis. Finally, stable knockdown of NUSAP1 with shRNA led to decreased tumor growth in an orthotopic xenograft model of GBM in mice.
Conclusions: Taken together, NUSAP1 gene silencing induced apoptosis possibly through the downregulation of the candidate downstream molecule TOP2A. Interference with the expression of NUSAP1 might therefore inhibit malignant progression in GBM, and NUSAP1 might thus serve as a promising molecular target for GBM treatment.

Background
Glioblastoma multiforme (GBM) is the most malignant type of primary human brain tumor characterized by a high proliferation rate and an unfavorable prognosis [1][2][3] , where the 5-year survival rate remains dismally at less than 5% [4][5][6] . Over the past 40 years, numerous therapeutic modalities have been implemented to enhance survival. However, due to the highly invasive nature of GBM, tumor tissue cannot be completely removed with surgery, which leads to tumor recurrence. Rigorous molecular analysis of GBM is ongoing to understand the development of the disease and its property of invasion for the identi cation of novel biomarkers and e cacious therapeutic avenues. Nucleolar and spindle associated protein 1 (NUSAP1), a 55-kD vertebrate protein is localized to spindle microtubules during mitosis and an indispensable mitotic regulator 7,8 . In the normal cell cycle, the expression of NUSAP1 protein reaches its peak in the G2 phase and decreases after the cell enters the division phase. However, high expression of NUSAP1 has been reported in several types of cancer. NUSAP1, for example, has been found to be associated with metastasis of cervical carcinoma 9 and involved in tumor recurrence in prostate cancer 10 . In breast cancer, NUSAP1 has been shown to affect the DNA damage response by controlling BRCA1 protein levels. NUSAP1 has also been identi ed as potential marker for breast ductal carcinoma in situ and a cause of resistance to therapy 7 . Knockout of the NUSAP1 gene inhibited cell proliferation, migration and invasion in colorectal cancer 11 , indicating that upregulation of NUSAP1 promotes cancer progression. However, whether increased NUSAP1 also has a role in driving the development of GBM remains unclear.
In this study, we investigated the expression and function of NUSAP1 in human GBM. We rst analyzed molecular and clinical data in publicly available databases and found that NUSAP1 expression is signi cantly up-regulated in GBM tissue samples and GBM cell lines. We subsequently used knockdown strategies in models in vitro and in vivo to explore the function of NUSAP1 in GBM and to identify coregulated molecules, such as topoisomerase 2A (TOP2A), that might mediate the function of NUSAP1. Our results thus provide a potential strategy for a new targeted therapy for the treatment of GBM.

SiRNA transfections and lentiviral transduction
Gene-speci c siRNAs against NUSAP1 and TOP2A (GenePharma; Shanghai, China) were transfected into U251, T98 and P3 cells for 48 h using Lipofectamine 2000 (Invitrogen, 11668-027) according to the manufacturer's instructions. Nonspeci c random sequences were used as the non-speci c negative control (Si-NC). Lentiviral vectors expressing human shRNA targeting NUSAP1 (shNUSAP1, LV2017-18615, GenePharma) or the scrambled-control (shNC) were used to generate stable cell clones expressing shNUSAP1 or a nonspeci c shRNA as the control. Transfected clones were selected using 1 mg/mL of puromycin (Selleckchem). Western blot analysis was used to evaluate siRNA and shRNA knockdown

Cell proliferation and invasion assays
The EdU incorporation assay (C103103, Ribobio; Guangzhou, China) was used to determine cell proliferation according to the manufacturer's protocol. Brie y, cells were seeded into 24-well plates at a density of 5.0 × 10 4 cells per well and incubated overnight. After treatment, EdU was incorporated into proliferating cells and detected through a catalyzed reaction with a uorescently labeled azide. Five random views from images acquired with a uorescence microscope were used to count EdU positive cells.
For invasion assays, inserts in transwell migration plates (8 μm pore size, Corning; Sigma-Aldrich; St. Louis, MO, USA) were coated with matrigel (Becton-Dickinson; Bedford, MA, USA) and incubated for 4 h at 37°C. Cells (2 x 10 4 ) in 100 μL DMEM without FBS were seeded into the upper chamber, and 600 μL of medium containing 10% FBS was added to the lower chamber. After 24 h at 37°C, cells remaining on top of the Matrigel in the upper chamber insert were removed with a cotton swab while cells that had migrated through the Matrigel to the lower side of the insert were xed with 4% formalin for 15 min, rinsed twice with PBS, and stained with 0.1% crystal violet for 10 min. Five random views from images acquired under a light microscope were used to count migrated cells.

Immunohistochemistry and western blotting analysis
Immunohistochemical (IHC) staining of the target proteins and evaluations of the intensities of staining were performed as previously described 24 . Brie y, para n-embedded samples were sliced and mounted on microscopic slides. Heat-induced epitope retrieval was performed with a microwave in 10 mmol/L citric acid buffer at pH 7.2. The sections were blocked with goat serum, incubated with primary antibodies at 4°C overnight, rinsed with PBS and incubated with a horseradish peroxidase-linked goat anti-rabbit antibody. Results were visualized through the reaction with diaminobenzidine and sections were counter stained with Mayer's hematoxylin. Western blot analysis of the target proteins in glioma cells and human glioma samples was performed as previously described 25 . β-actin and GAPDH were used as the loading controls. Primary antibodies used were the following: anti-NUSAP1 (#12024-1-AP, Proteintech; Chicago, IL, USA); anti-TOP2A (#MA5-32096; Invitrogen; Carlsbad, CA, USA); anti-MKI67 (#ab15580, Abcam; Cambridge, UK); mouse anti-β-actin (#TA-09, monoclonal, ZSGB-BIO; Beijing, China); and anti-GAPDH (#BA2913, BOSTER; Wuhan, China).
Flow cytometric analysis of apoptosis GBM cells were seeded in 6-well plates and incubated for 72h. Floating cells were harvested, resuspended in 1´ binding buffer, and incubated with Annexin V-FITC (BD Biosciences; San Jose, CA, USA) and propidium iodide according to the manufacturer's instructions. Apoptotic cells were detected with ow cytometry (ACEA Biosciences; San Diego, CA, USA), and the software Flowjo (Tree Star; Ashland, OR, USA) was used to analyze the data.

Statistical analysis
Three independent experiments were performed, and all statistical analyses and experimental graphs were performed using GraphPad Prism 8 software (San Diego, CA, USA). X 2 -tests were used to compare categorical variables, and continuous variables were analyzed using ANOVA and the Student's t-test. Pvalues determined from different comparisons are indicated as follows: *P < 0.05; ** P < 0.01; and *** P < 0.001. P-values 0.05 were considered statistically signi cant.

Results
NUSAP1 expression is up-regulated in high-grade gliomas and inversely associated with glioblastoma patient prognosis To begin to understand the role of NUSAP1 in the development of human gliomas, we rst examined expression of the gene in primary tumor samples by analyzing molecular data in publicly available datasets, including the Chinese Glioma Genome Atlas (CGGA), the Cancer Genome Atlas (TCGA) and Rembrandt. We found NUSAP1 to be signi cantly up-regulated in GBMs (grade IV, high grade gliomas, HGG) compared with low grade gliomas ( grade II and grade III, LGG) and normal brain tissue (NBT) (Fig. 1A). Kaplan-Meier analysis of the expression data in the CGGA and Rembrandt datasets demonstrated that high NUSAP1 expression in tumors predicted shorter overall survival in patients (Fig. 1B).
To determine whether NUSAP1 protein levels correspondingly increased, we performed immunohistochemistry (IHC) with an antibody against NUSAP1 on para n-embedded glioma (grade II, n = 10); grade III, n = 10; and grade IV, n = 15) and normal brain tissue (NBT; n = 5) samples. IHC staining and quantitative analyses of mean NUSAP1 IHC staining scores showed that the NUSAP1 protein levels also increased with increasing disease grade ( Fig. 1C and 1D). Western blot analysis con rmed that expression of NUSAP1 protein was signi cantly up-regulated in GBMs (grade IV) compared with low grade gliomas (Fig. 1E).
We also performed western blot analysis to determine NUSAP1 protein levels in GBM (U87, A172, U251, and T98) and patient-derived GBM (P3) cell lines. U251, T98, and P3 cell lines displayed higher NUSAP1 protein levels than in U87 and A172 (~ 3-5x; Fig. 1F). All together, these data demonstrated that NUSAP1 was signi cantly up-regulated in glioma samples and enhanced expression was correlated with disease progression in clinical glioma patients.

Down-regulation of NUSAP1 inhibits cell proliferation and invasion and promotes apoptosis in glioma cells
Further analysis of the molecular data in the CGGA revealed a strong correlation between the expression of NUSAP1 and proliferation-associated molecules, such as MKI67 and PCNA, in primary gliomas samples and recurrent gliomas samples ( Fig. 2A). Therefore, we next examined the possible function of NUSAP1 in the development of human glioma in siRNA knockdown studies of NUSAP1 in U251, T98, and P3 GBM cell lines in vitro. Each of the three siRNAs targeting NUSAP1 (siN1-1, siN1-2, and siN1-3) exhibited signi cant knockdown of NUSAP1 as demonstrated on western blot (Fig. 2B). In EdU assays, the number of EdU positive cells decreased by ~ 30% relative to the control in all three cell lines transfected with siN1-2 and siN1-3 demonstrating that loss of NUSAP1 inhibited cell proliferation (Fig. 2C  and 2D). Silencing of NUSAP1 with siN1-2 and siN1-3 also signi cantly inhibited invasion capabilities of U251, T98, and P3 GBM cell lines in transwell assays (~ 3⋅; Fig. 2E and 2F).
To identify possible cellular mechanisms involved in the inhibition of cell proliferation with loss of NUSAP1, we used ow cytometry to determine whether NUSAP1 knockdown induced apoptosis.

Down-regulation of NUSAP1 inhibits mRNA and protein expression levels of TOP2A in GBM cells
To further elucidate the molecular pathways mediating the putative oncogenic activity of NUSAP1, we rst performed mRNA sequencing to identify gene expression changes following the silencing of NUSAP1 in P3 cells. The results demonstrated that loss of NUSAP1 led to signi cant down-regulation of 8 genes, including MT1F, S1PR1, IGFN1, UTP14C, TOP2A, KRT6A, ANO1, and UFSP1, and up-regulation of 5 genes, including SERF1B, ADCY10P1, DIRC3, INMT-MINDY4, and TMLHE-AS1, in P3 cells (Fig. 3A and 3B). We then used the bioinformatics analysis tool TargetScanHuman 5.2 to identify potential signaling pathways associated with these genes. TargetScanHuman is a bioinformatics tool developed to aid in the prediction of the targets involved in the activities of protein or molecules. This analysis yielded a potential relationship between NUSAP1 and TOP2A (Fig. 3C). Western blot analysis con rmed a possible relationship, demonstrating that knockdown of NUSAP1 led to decreased levels of TOP2A in U251 and P3 cell lines (~ 60% and 40%, respectively; Fig. 3D). These results demonstrated that loss of NUSAP1 directly altered expression of genes involved in cell proliferation, such as TOP2A.

Loss of TOP2A inhibits proliferation and invasion, and induces apoptosis in GBM cells
Analysis of molecular data from the CGGA showed a strong correlation between the expression of TOP2A and NUSAP1 in primary and recurrent gliomas samples (Fig. 4A). Kaplan-Meier analysis of clinical and molecular data from the CGGA revealed that high TOP2A expression in primary and recurrent gliomas samples predicted shorter overall survival in patients (P < 0.0001 and P = 0.015, primary and recurrent glioma samples, respectively; Fig. 4B). We therefore further investigated the role of TOP2A in the development of human glioma by knocking down TOP2A expression in U251 and P3 GBM cells with three small-interfering RNAs, siTOP2A-1, siTOP2A-2, and siTOP2A-3. All three siRNAs achieved signi cant knockdown of TOP2A as assessed on western blot and we therefore used siTOP2A-1 and siTOP2A-2 in functional experiments (Fig. 4C). The percentage of EdU positive cells decreased by ~ 30% in U251 and P3 cells transfected with siTOP2A-1 and siTOP2A-2 relative to the control siRNA ( Fig. 4D and 4E). The number of invading cells in transwell assays also decreased signi cantly in U251 and P3 cells with knockdown of TOP2A (~ 50%; Fig. 4F and 4G) of glioma cells. Finally, to determine possible mechanisms contributing to the inhibition of cell growth, we used ow cytometry to determine whether TOP2A knockdown induced apoptosis. SiTOP2A-1 and siTOP2A-2 knockdown in both U251 and P3 cell lines led to up to ~ 4x and ~ 2x increase in apoptosis (Fig. 4G). Taken together, these results indicated that loss of TOP2A in GBM cells inhibited GBM cell growth by triggering apoptosis.

Down-regulation of NUSAP1 inhibits the expression of TOP2A and glioma growth and invasion in vivo
To investigate the effect of NUSAP1 on GBM tumor growth in vivo, we examined the effect of the loss of NUSAP1 in an orthotopic tumor model. We transduced U251 and P3 cells with a retrovirus containing an shRNA targeting NUSAP1, and implanted U251-shNC, U251-shNUSAP1, P3-shNC and P3-shNUSAP1 cells intracranially into nod-scid mice. Tumor size at 28 days was decreased in both U251-and P3-shNUSAP1 cells compared to controls based on animals' bioluminescence values (Fig. 5A). Overall survival was also enhanced in animals bearing U251-or P3-shNUSAP1 tumors compared to shNC control groups (P < 0.01, Fig. 5B). H&E staining of tumor sections showed that tumor sizes and invasion were markedly decreased in both U251-and P3-shNUSAP1 xenografts compared to controls (Fig. 5C).

Discussion
In this study, we demonstrated that NUSAP1, a gene involved in mitotic regulation, might also promote the malignant progression of human GBM. Analysis of the clinical and molecular data from CGGA, TCGA and Rembrandt databases revealed high expression of NUSAP1 in high-grade glioma tissues relative nonneoplastic brain tissue samples. Immunohistochemistry performed on our own cohort of primary tumor samples con rmed that expression of NUSAP1 increased with increasing glioma grade. Moreover, NUSAP1 overexpression was signi cantly correlated with poor patient survival. In functional studies, knockdown of NUSAP1 inhibited GBM cell growth in vitro and in vivo. Finally, loss of NUSAP1 triggered apoptosis which might be due to the concomitant loss of TOP2A. These results indicate that NUSAP1 might have a role as an oncogene in the development of human GBM and may serve as a novel molecular target in therapeutic treatment of the disease.
NUSAP1 plays an important role in spindle formation, bundling microtubules, attachment to chromosomes and mitotic regulation 12,13 . It has been implicated in the development, progression, metastasis, and poor prognosis in a variety of cancer types, including non-small-cell lung cancer 14 , bladder cancer 15 , prostate cancer 10 , astrocytoma 16 , colorectal cancer 11 , and cervical carcinoma 9 , and is a biomarker in prostate cancer following surgery 10 . Knockout of NUSAP1 has been shown to inhibit cell proliferation, migration and invasion through loss of DNMT1 expression in human colorectal cancer 11 . Our study also demonstrated that the knockdown of NUSAP1 with siRNA, as in the colorectal cancer model, inhibited proliferation and invasion in U251, T98, and P3 GBM cell lines. We also found that the cellular process underlying suppression of cell growth might be the induction of apoptosis.
To reveal possible pathways mediating this response, we used RNA sequencing to identify genes coregulated with loss of NUSAP1 in the P3 GBM cell line. RNA sequencing showed that knockdown of NUSAP1 signi cantly down-regulated expression of 8 genes MT1F, S1PR1, IGFN1, UTP14C, TOP2A, KRT6A, ANO1, and UFSP1, and up-regulated 5 genes, including SERF1B, ADCY10P1, DIRC3, INMT-MINDY4, and TMLHE-AS1. Using TargetScanHuman 5.2, we identi ed potential signaling pathways associated with these genes and showed that NUSAP1 expression was correlated with KIF11, TOP2A, NCAPG, DLGAP5, CDC20, CDK1 MKI67, NDC80, and CCNA2. Based on the results of our analysis of public databases and our RNA-seq data, we further investigated the relationship between NUSAP1 and TOP2A. Western blot analysis revealed that loss of NUSAP1 led to decreased protein levels of TOP2A in U251 and P3 GBM cells.
Topoisomerase 2A (TOP2A), a member of the DNA topoisomerases, is an enzyme that controls DNA superhelicity and unlinks double-stranded DNA segments during processes such as replication and transcription 17,18 . Many studies have shown that TOP2A plays a crucial role in the tumorigenesis and progression of a variety of malignancies, such as adrenocortical carcinoma 19 , bladder cancer 20 , prostate cancer 21 , liver cancer 22 , and breast cancer 18,23 . Analysis of the molecular data in the CGGA database con rmed that there was a strong correlation between TOP2A and NUSAP1 expression in gliomas samples. High TOP2A expression in primary and recurrent gliomas predicted shorter overall survival in patients. In functional studies, knockdown of TOP2A in U251 and P3 GBM cells suppressed cell proliferation and invasion capabilities possibly through the induction of apoptosis. Taken together, our data indicated that TOP2A might act downstream of NUSAP1 in the development and progression of GBM.

Conclusions
In conclusion, we demonstrated that NUSAP1 inhibits proliferation and invasion of GBM cells in vitro and in vivo possibly through the triggering apoptosis. The induction of apoptosis might be due to the concomitant down-regulation of the expression of TOP2A. Thus, our study provides new insights into the possible oncogenic role of NUSAP1 in the development of human GBM and furthermore demonstrates the therapeutic relevance of silencing NUSAP1 in malignant human brain tumors.     Down-regulation of NUSAP1 inhibits the expression of TOP2A, and leads to reduced glioma growth and invasion in vivo. A. U251 cells expressing luciferase were orthotopically implanted into athymic nude mice. Tumor growth was monitored using the IVIS-200 imaging system for detection of bioluminescence.

List Of Abbreviations
Bioluminescent signals were measured at days 7, 14, 21, and 28 after implantation. B. Kaplan-Meier survival analysis of overall survival, and log rank analysis to assess the statistical signi cance of the