Monastrol Targeted KIF11 Showed Potential Treatment Effective of Small Cell Lung Cancer

Objective: This study is to identify Small Cell Lung Cancer (SCLC) driver genes, annotate enrichment functions and key pathways, and also verify Monastrol therapeutic effect. Methods: The gene expression profiles of GSE40275 and GSE43346 was analyzed to identify the DEGs (Differentially Expressed Genes) between SCLC and the normal tissue. GO (Gene Ontology), KEGG (Kyoto Encyclopedia of Genes and Genomes) analysis and PPI (Protein-protein interaction) network analysis were conducted to find out the enrichment functions, pathways and hub genes. Moreover, in vitro, MTT assay, colony-forming assay, and the scratch assay were performed to verify the effect of Monastrol. Results: There were common 129 up-regulated and 176 down-regulated DEGs between SCLC samples and normal lung samples. KIF11, NDC80 and PBK were identified as hub genes after PPI network analysis. The q-PCR results showed that genes KIF11, NDC80 and PBK consistently expressed higher in cancer cells than normal cell lines. And in vitro assay showed that Monastrol inhibited SCLC cellular viability, proliferation and migration (P < 0.01). Conclusion: KIF11, NDC80 and PBK were aberrantly expressed and could be potentially applied as diagnostic biomarkers, therapeutic targets and prognostic biomarkers. Monastrol was a promising drug in treatment of SCLC patients.

about 35,000 new cases occur annually [2] . In underdeveloped countries, the percentage of SCLC cases is higher. More than 130,000 new diagnoses of SCLC and 100,000 deaths from this disease were estimated to have occurred in China in 2013 [3] .
Small cell cancer is highly aggressive. This kind of tumor cells is characterized by rapid doubling time and not suitable for traditional surgical therapy but sensitive to both chemotherapy and radiation. Unfortunately, these conventional treatment methods are both short in duration and not curative in most cases with an average 5-year survival rate below 7%. No major treatment advances have occurred over the past 30 years [4] . In patients with extensive-stage disease, chemotherapy alone can palliate symptoms and prolong survival in most patients; however, long-term survival is rare [5] . The management of SCLC is still very challenging cause of disease outcome has remains stubbornly poor due mainly to limited options for effective treatment. Although according to some preclinical studies about the understanding of SCLC, the c-kit inhibitor and other agents targeting angiogenesis could show an inhibition on it, the results of the actual use were somewhat disappointing and unexpected. Meanwhile, there is no clear expatiation about the neoplastic processes of SCLC [6] . In this way, conducting such a study which comprehensively depicts molecular pathogenesis of SCLC development as well as identifies novel therapeutic agents is necessary and crucial.
The hallmarks of cancer are the most crucial to the development of SCLC, including genomic instability and mutations, evading growth suppressors, resisting cell deaths, sustaining proliferative signaling and enabling replicative immorality. The genomic instability of SCLC is higher than most cancers. SCLC has a mutation rate with 5.6 to 7.4 mutations per Mb, and about 175 mutations per tumor [7] . However, few specific driver genes have been proved related to SCLC pathogenesis. To further clarify the molecular pathogenesis of this tumor, our study also used bioinformatics methods combining with GO, KEGG analysis and PPI network analysis, employing mRNA microarray datasets to screen out the hub genes and key pathways associated with SCLC.

Microarray data
The gene expression profiles of GSE40275 and GSE43346 were collected from GEO database (The Gene Expression Omibus, http://www.ncbi.nlm.nih.gov/geo). These profiles contained 40 SCLC samples and 44 normal samples totally.

Identification of DEGs
The analysis was conducted based on raw data using software GeneSpring (version 11.5, Aglient, USA). The category of the mRNA expression data was realized with hierarchical clustering analysis. The probe quality control in GeneSpring was limited by virtue of principal component analysis (PCA), and probes with intensity values below 20th percentile were filtered out using the "filter probesets by expression" option. Then, the DEGs were identified using classical t test with P value cutoff of <0.05 and a change >2 fold, which were applied for statistically significant definition. At last, the Venn plot analysis regarding DEGs was conducted (http://bioinformatics.psb.ugent.be/webtools/Venn/).

Gene ontology and pathway enrichment analysis of DEGs
The DAVID database (Database for Annotation, Visualization and Intergrated Discovery, http://david.abcc.ncifcrf.gov/), provided a comprehensive set of functional annotation tools to understand biological meaning underlying plenty of genes. GO (Gene Ontology) was a useful method for analyzing biological process, molecular function and cell component of genes. And KEGG (Kyoto encyclopedia of Genes and Genomes) was a base for gene function analysis and genomic information link. In this study, GO and KEGG pathway enrichment analysis were performed using DAVID for DEGs functions analysis.

Colony-forming assay
The cancer cells (H446, A549) were cultured in Petri dishes with 50 cells/cm 2 . After 24 hours, those cells were treated with different doses of Monastrol respectively. After ten days in vitro growth, colonies were counted. Then, colonies were rinsed with PBS, fixed in 4% paraformaldehyde, stained with 5% crystal violet for half hours, and rinsed twice with water.
In vitro scratch assay The H446 and A549 cells were cultured on 24-well Permanox TM plates. A 1ml pipette tip across each well was used to creat a consistent cell-free area. The loose cells were washed out gently using DMEM. Then, the cells were exposed to different doses of Monastrol. After the scratch and at 0, 12, 24 hours, the images of the scraped area were captured with phase contrast microscopy. The remaining wounded area and the scratch width at six different points per image were measured.

Statistical analysis
All statistic data were entered into SPSS 20.0 (SPSS Inc., Chicago, Illinois, USA) for analysis. Independent-samples t test was conducted to analyze quantitive data. P values < 0.05 were set as significance level.

Indentification of DEGs and gene function analysis
The gene expression level of GSE40275 and GSE43346 were showen in the volcano plots ( Figure 1A). Altogether 305 DEGs were picked up between normal and SCLC mRNA expression samples shown in the VENN plots ( Figure 1B). The mutual DEGs were uploaded to DAVID respectively, to identify the further insight of those genes. The detailed results of GO and KEGG pathway analysis were showed in Table 1 and Figure   1C. The GO and KEGG analysis results revealed that the mutual up-regulated DEGs were mainly associated with cell division, nucleoplasm, protein binding and Cell cycle.
Meanwhile, for the mutual down-regulated DEGs, the GO and KEGG analysis results primarily enriched in immune response, plasma membrane, scavenger receptor activity and Complement and coagulation cascades.
Module screening from the PPI network All DEGs among these samples were analyzed with PPI network, and the hub genes were screened with degrees > 90 based on the STRING database. Top 20 genes were identified as hub genes listed in Table 2. And heat maps of these hub genes expressions were showed in Figure 3A. Among those genes, the node degree of TOP2A was the highest one, which degree was 112. Moreover, after MCODE analysis, the top 3 significant modules were obtained, showed in Figure 2. The functional annotation and enrichment of these modules were also performed, showed in Table 3. Enriched function analysis revealed that genes in module 1 were primarily related to protein binding, cell division and Cell cycle. In module 2-3, genes were mainly enriched in inflammatory response, Chemokine signaling pathway and cellular defense response.

Validation of common hub genes by qRT-PCR
To validate the expression of KIF11, NDC80, PBK and TOP2A in human normal lung cells (BEAS-2B, MRC5), human small cell lung cancer cells (H446) and human lung adenocarcinoma cells (A549), qRT-PCR was performed. The results, showed in Figure 4A, revealed among those cell lines that genes KIF11, NDC80, PBK and TOP2A consistently expressed higher in cancer cells than normal cell lines (P < 0.05).

Monastrol reduces proliferation of SCLC cells
To evaluate the sensitivity of lung cancer cells to Monastrol, the survival cells after treatment were calculated by MTT assay. As showed in Figure 4B, following the augment of drug concentrations, the cellular viabilty (ratio to control) in cell lines H446 and A549 decreased significantly. However, BEAS-2B and MRC5 still had a high cellular viability even subjected to the highest doze. Monastrol was relatively tolerated for normal lung cells lines. To determine the anti-cancer effects of Monastrol in SCLC cells, we performed colony-forming assay. The results showed us that less and smaller clonogenicities in Petri dishes with Monastrol than that in control group (Figure 4C). The percentage of clone formation in control was significantly higher than in drug groups (0.25μmol/L, 1μmol/L).

Monastrol inhibits migration of SCLC cells
The widths of scratched areas were measured after the scratch, after 12h, and after 24h, to research the migration of SCLC cells. In Figure 4D, the width of scratched was significantly smaller after 24h in control group. However, there was only a slight decrease in Monastrol group. In addition, after 24h, the wounds in control group were also smaller than in drug group significantly.

Discussion
SCLC often diagnosed in an advanced stage, which is a highly aggressive malignancy associated with early metastasis, rapid progression, and poor survival. First line chemotherapy provides good response rates in advanced disease, but SCLC often relapses, progression free and overall survival are limited. Few related genes and molecular pathways have been studied by previous researchers, there is an urgent need for comprehensive analysis regarding to SCLC [8,9] . And clinical diagnosis, treatment and prognosis would be significantly improved if the appropriate targets are identified. Our study used bioinformatics analysis techniques combined with complicated and refined algorithm to identify key genes and pathways, which could provide potential targets for SCLC diagnosis and treatment. And screening anti-SCLC drugs based on targets to improve clinical treatment.
GSE40275 and GSE43346 datasets were downloaded, and then we screened normal tissues and SCLC tissues samples from them, also identified the DEGs between normal tissues and These results further explain the tumor formation mechanism of SCLC, and also provide novel ideas for further study of SCLC in the future.
PPI network of DEGs was constructed using STRING database and Cytocape. Three hub genes (KIF11, NDC80 and PBK) were identified as driver genes closely related to SCLC development and they had never been reported related to SCLC in previous studies. KIF11, kinesin family member 11, located in Chr10q23.33, encodes a motor protein which is known to be involved in spindle assembly, control mitotic spindle structure and chromosome behaviour during mitosis [10,11] . The function of this protein includes chromosome positioning, centrosome separation and establishing a bipolar spindle during cell mitosis. Kinesin motor domains couple cycles of ATP hydrolysis to cycles of microtubule binding and conformational changes that result in directional force and movement on microtubules [12] . Meanwhile, KIF11 dephosphorylation can led to mitotic exit, and the expression level of this gene is low in normal lung tissue [13,14] . In SCLC cells, overexpression of this gene may lead to a rise in phosphorylated Erk1/2 levels, promote bipolar spindle assembly and chromosome segregation [15][16][17] . And this gene also plays a role in DNA repair, gives assistance for some protein's transport from the trans-Golgi network to the cell surface and contributes to mitotic spindle checkpoint activation and Tat-mediated apoptosis in CD4-positive T-lymphocytes, which may boosts the progress of small cell lung cancer development. Given these circumstances, KIF11 might be a potential microtubule-related target for proliferating SCLC cells [18][19][20][21] . It prompted us to hypothesize that Monastrol, a potent and cell-permeable inhibitor targeted KIF11 with an IC 50 value of 14 μM, which causes aberrant interactions with the microtubule, and reversals at the ATP hydrolysis step, might have anti-SCLC therapeutic effects [22][23][24] . And it had also been verified by the following assays in this studies. The drug in protein binding structure 1X88 was downloaded in PDB dataset (Protein data bank, http://www.rcsb.org/), and shown in Figure3B.
NDC80, kinetochore complex component, located in Chr18p11.32, encoded protein consists of an N-terminal microtubule binding domain and a C-terminal coiled-coiled domain that interacts with other components of NDC80 kinetochore complex, which directly modulates microtubule dynamics [25] . This protein functions to organize and stabilize microtubule-kinetochore interactions and is required for proper chromosome segregation. Aurora A kinase phosphorylates NDC80 to regulate metaphase kinetochoremicrotubule dynamics [26] . And Aurora B-NDC80-Mps1 signaling axis is governing accurate chromosome segregation in mitosis [27] . Overproduction of Ndc80 in cancer cells unfavourably absorb protein interactors through the internal loop domain and lead to a change in the equilibrium of microtubule-associated proteins [28] . NDC80's interaction with either growing or shrinking microtubule ends such as differentially regulates mammalian kinetochore coupling to polymerizing and depolymerizing microtubules can be tuned by the phosphorylation state of its tail [29,30] . N-terminus-modified NDC80 can suppress tumour growth by interfering with kinetochore-microtubule dynamics [31] . In our study, it showed that this gene, which was the core of the interaction with multiple genes, had a pivotal position in SCLC tissues. The expression of this gene was abnormal regulated, in view of NDC80 was a driver gene of SCLC, we hypothesized its abnormal activation would promote the initiation of SCLC, it might play an important role during the course of SCLC development and it might be a biomarker in the early diagnosis of SCLC. Accordingly, NDC80 could be considered as an important therapeutic target for SCLC. PBK, PDZ binding kinase, this gene locates at Chr8p21.1 and encodes a serine/threonine protein kinase related to the dual specific mitogen-activated protein kinase kinase (MAPKK) family. This kinase can increase the rate of mitosis and expands malignant T cells [32] . FoxM1-regulated PBK exerts oncogenic activities via the activation of beta-Catenin pathway [33] . CDK1-mediated mitotic phosphorylation of PBK is involved in cytokinesis and inhibits its oncogenic activity [34] . PBK promotes lung cancer resistance to EGFR tyrosine kinase inhibitors by phosphorylating and activating c-Jun [35] . PBK mediates promyelocyte proliferation via Nrf2-regulated cell cycle progression and apoptosis [36] .
TOPK/PBK promotes cell migration via modulation of the PI3K/PTEN/AKT pathway and is associated with poor prognosis in lung cancer [37] . Increased levels of PBK may contribute to tumor cell development and progression through suppression of p53 function and consequent reductions in the cell-cycle regulatory proteins such as p21 [38] . PBK augments tumor cell growth following transient appearance in different types of progenitor cells in vivo as reported [39] . Overexpression of this gene has been implicated in tumorigenesis.
Overexpression of PBK contributes to tumor development and poor outcome of SCLC [40] .
PBK might play a pivotal role in SCLC invasion and metastasis [41] . Our research confirms that the expression of this gene was abnormal regulated and it is a core driver gene of SCLC.
The effects of Monastrol were evaluated in present study with MTT assay, colony-forming assay, and scratch assay in vitro. In MTT assay, the cellular viability (ratio to control) in H446 and A549 cell lines were revealed dose-depended decreased when treated with Monastrol. In colony-forming assay, the numbers and size of clonogenicities in Monastrol group were significantly less than control group, which was consistent with the results that Monastrol can reduce the proliferation of SCLC cells. In scratch assay, the wound widths in control group decreased sharply along with time, and were smaller than that in Monastrol group after 48h significantly. That implied that Monastrol strongly inhibited migration of SCLC cells. All these results suggest that Monastrol have the potential effect for treating SCLC, which is worthy of further study.

Conclusions
Our study selected out the DEGs and key pathways in the NFPA tissue. The DEGs identified in this study provided comprehensive insight into the molecular mechanism of the SCLC pathogenesis. Three hub genes: KIF11, NDC80 and PBK were aberrantly expressed and could be potentially applied as diagnostic biomarkers, therapeutic targets and prognostic biomarkers. Meanwhile, Monastrol, a potent inhibitor regarding to KIF11, suppressed proliferation and migration of SCLC cells, was a promising drug in treatment of SCLC patients.

Compliance with ethical standards
Conflict of interest The authors declare that they have no conficts of interest (including fnancial and non-fnancial interests).
Ethical approval All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and its later amendments or comparable ethical standards.