Identification and assessment of PLK1/2/3/4 in lung adenocarcinoma and lung squamous cell carcinoma: Evidence from methylation profile

Abstract Lung cancer is a very aggressive cancer characterized with molecular heterogeneities in different subtypes, including lung adenocarcinoma and lung squamous cell carcinoma. However, few related molecular signatures have been established for the treatment of lung cancer subtypes. Polo‐like kinase (PLK) family is a crucial regulator during cell division. Aberrant genetic and epigenetic alteration of PLK members plays a controversial role among different cancers. In this study, we performed an analysis of transcriptional and protein expression to identify overexpressed PLK1/4 and under‐expressed PLK2/3 in lung cancer subtypes. We then analysed biological function of PLKs and related genes. Besides, we estimated a correlation of PLKs with patient's genders and TP53 mutation in lung cancer. Higher PLK1/4 expression was significantly associated with male patient and TP53 mutant status, separately. Moreover, we carried out a methylation profile analysis including methylation level, DNA methyltransferases correlation and survival analysis of global methylation. Global methylation survival analysis showed that prognostic value of PLK1/2/4 methylation remained the same significant trend between two lung cancer subtypes, whereas prognostic value of PLK3 methylation lacked consistency. Taken together, these results provided instructive insights into a comprehensive evaluation for advanced therapeutic strategy based on epigenetic evidences.


| INTRODUC TI ON
Lung cancer is one of the most refractory malignancies with high incident and mortality. Although the curative effect of lung cancer therapy has improved recent years, its prognosis remains generally poor. The non-small-cell lung cancer (NSCLC) comprises >80% of all lung cancer cases and therefore is the focus of lung cancer studies worldwide. 1 The major pathologic subtypes of NSCLC include lung adenocarcinoma with 40%-50% proportion and lung squamous cell carcinoma with 25%-30%. 2 With great advances in understanding the molecular pathogenesis, lung cancer therapeutic strategy has been switching from a stage-directed therapeutic pattern to a systematic therapy with histological type and molecular feature. 3 Initiation and progression of different cancer subtypes is considered a distinct multistep process. An understanding of how genomics alteration and epigenetic modification contribute to tumorigenicity in distinct lung canner subtypes may optimize the lung cancer management.
Precise treatment for lung cancer is dependent on an accurate classification and detailed mechanism of cancer subtypes, otherwise the treatment for lung cancer-not otherwise specified (NOS)-is full of intractable problems. For example, pemetrexed is currently the drug for adenocarcinoma, and bevacizumab is not suitable for squamous cell carcinoma. 4 Given each subtype of cancer with specific histopathological property, subtype-specific strategy is predicted to improve a beneficial clinical outcome.
The individual therapy for advanced lung cancer treatment required a better understanding of mechanism not only at histopathological level but also at molecular level, which was because that genomic and epigenetic alteration could drive tumorigenesis of lung cancer subtypes. Prior study has reported that crizotinib is used to target ALK gene rearrangement in NSCLC. 5 Gefitinib and erlotinib are approved for EGFR mutant status in lung cancer cell. 6 However, little is known about the clinical settings when distinct lung cancer subtypes cover different molecular alteration.
Polo-like kinase (PLK) family has 5 members, which are implemented in cell cycle regulation and promote subsequent biological process, such as cell division. 7 It is suggested that PLKs play an important role in cancer cell proliferation. In this study, we identified mRNA and protein expression of PLK family members based on public data platforms, including Oncomine, TCGA, and Human Protein Atlas (HPA) databases. Then, we performed the biological function including GO and KEGG pathway enrichment, as well as a correlation with patient's gender and TP53 mutant status, separately.

| Transcriptional expression of PLK1/2/3/4/5 in lung cancer identified in Oncomine database
Oncomine database (http://ualcan.path.uab.edu/index.html) is a bioinformatics initiative to develop a collection for a cancer transcriptome data. 8 In this study, we analysed the individual gene expression level of PLK1/2/3/4/5 in lung cancer by Oncomine. The mRNA levels of PLK1/2/3/4/5 were compared between lung cancer and normal patient data sets. The threshold in our study was 1.5-fold change, Pvalue = .0001 and top 10% gene rank.

| Transcriptional expression of PLK1/2/3/4 in both lung cancer subtypes validated by TCGA database
UALCAN (http://ualcan.path.uab.edu/index.html) is an interactive web-portal for providing the in-depth analyses of TCGA gene expression data. 9 TCGA consortium has enabled clinical researchers and cancer scientists to investigate the molecular characteristics and expression profile of gene across various cancers. The relative expression of PLK1/2/3/4 was analysed between cancer and normal samples in lung cancer subtypes including lung adenocarcinoma and lung squamous cell carcinoma. A P-value less than .05 is statistically significant.

| The protein expression patterns of PLK1/2/3/4 based on immunohistochemistry
The HPA (https://www.proteinatlas.org) is a web tool which provides protein localization and expression in human tissues. 10 The result of immunohistochemistry was generated by unique antibody application. In the current study, the immunohistochemistry images were collected to show an overview of PLK1/2/3/4 between cancer and normal tissue in lung adenocarcinoma and lung squamous cell carcinoma, respectively. In our collection, antibodies for PLK1/2/3/4 in lung cancer were HPA053229, CAB009624, HPA060318 and HPA035026.

| Construction of related gene network for PLK1/2/3/4 in relation to biological function and pathway
STRING version 11.0 (https://string-db.org/) integrates publicly available sources to generate connectivity network of proteinprotein association. 11 The proteins of interest were uploaded to achieve a global network in order to analyse physical and functional interaction. Here, the network of PLK1/2/3/4-related genes was performed via STRING with confidence score = 0.400.
GeneMANIA (http://www.genemania.org) is a prediction server for biological function of input genes from a wealth of genomics and proteomics data. 12 Based on a large set of functional association data, input gene was conducted to show functional association network.
We submitted PLK1/2/3/4 to perform an interactive network among  The P-value less than .05 was considered as statistically significant.

| Methylation analysis of PLK1/2/3/4 genes in both lung cancer subtypes
Given the different expressions of PLK1/2/3/4 genes between lung cancer and normal tissue, we carried out a methylation analysis in UALCAN web to evaluate epigenetic regulation of PLK1/2/3/4 expression by promoter methylation. A P-value less than .05 showed a statistical significance.
The cBio Cancer Genomics Portal (cBioPortal; http://cbioportal.org) is a web resource to exhibit a multidimensional investigation for cancer genomics data. 15 Based on TCGA data, DNA methylation value was released and added to the portal. The mRNA expressions of DNMT1/3A/3B were correlated with the mRNA expression of PLK1/2/3/4 in lung adenocarcinoma and lung squamous cell carcinoma.
Spearman and Pearson correlation were used to measure the association between DNMT1/3A/3B and PLK1/2/3/4 in both lung cancer subtypes.
The P-value less than .05 was considered as statistically significant.
MethSurv (https://biit.cs.ut.ee/methsurv/) is a web tool to perform multivariable survival analysis based on CpG methylation patterns. 16 MethSurv utilizes TCGA methylation data to provide an overview analysis of methylation differences. In the present study, survival probability of patients with lung cancer was assessed using likelihood-ratio (LR) test to evaluate global DNA methylation survival of PLK1/2/3/4 in lung adenocarcinoma and lung squamous cell carcinoma.

F I G U R E 1
The flow chart of this study based on public databases. Firstly, we identified the differentially expressed polo-like kinase (PLK)s at transcriptional and protein level. Then, we constructed the related gene network for PLKs for GO and pathway enrichment. Furthermore, we analysed the clinicopathology of PLKs, the promoter methylation level of PLKs, the correlation analysis of PLKs and DNA methyltransferases (DNMTs), and the survival analysis based on global methylation of PLKs

| Transcriptional expression of PLK1/2/3/4/5 in patients with lung cancer
In order to assess expression pattern of PLK family members in lung cancer, we compared the transcriptional levels of PLK1/2/3/4/5 with those in normal samples in Oncomine database (Table 1 and Figure S1). The   Compared to normal samples, PLK4 was found to be higher in lung cancer sample. 19 While compared to normal sample, PLK5 showed no differential expression either in lung adenocarcinoma or in lung squamous cell carcinoma. As a result, our findings indicated that PLK family members had distinct expression characteristics in lung cancer.

| Identification of PLK1/2/3/4 expression in two subtypes of lung cancer
Given that we found the differential expression of PLK1

| Protein expression patterns of PLK1/2/3/4 in lung cancer subtypes
After examining the transcriptional expression of PLK1/2/3/4 in lung cancer, we continued to compare the protein expression patterns of PLK1/2/3/4 in both lung adenocarcinoma and lung squamous cell carcinoma. As shown in immunohistochemistry image from HPA platform, PLK1 protein showed a differentiation in neither lung adenocarcinoma nor lung squamous cell carcinoma ( Figure 3A,B). The staining of PLK2 protein was found high in both subtypes of lung cancer with a low level in normal tissue ( Figure 3C,D). PLK3 was higher expressed in pneumocytes of lung tissue than that in either lung adenocarcinoma or lung squamous cell carcinoma ( Figure 3E,F).
In normal lung tissue, PLK4 protein was not detected but at low level in lung adenocarcinoma and medium level in lung squamous cell carcinoma ( Figure 3G,H). In summary, these immunohistochemistry images basically supported that PLK1/2/3/4 proteins were in accordance with their transcriptional expression in lung cancer.

| Biological function and KEGG pathway of PLK1/2/3/4 and their related genes
Since PLK1/2/3/4 exhibited differential expression in lung cancer subtypes, we carried out to a biological analysis including protein-protein interaction, GO and KEGG pathway enrichment.
The PPI network of PLK1/2/3/4 with their 21 related genes was obtained by STRING followed by a visualization of GeneMANIA ( Figure 4A,B). The primary interaction comprised physical interactions, co-expression, prediction and co-localization.
GO function analysis was enriched including biological process (BP), cellular component (CC) and molecular function (MF). In biological process, the majority of 25 genes were associated with cell division, mitotic nuclear division and anaphase-promoting complexdependent catabolic process ( Figure 4C). In cellular component, it was obviously found that cytosol, nucleoplasm and centrosome were the most component involved with PLK1/2/3/4 as well as their related genes ( Figure 4D). In molecular function, the proteins of these molecules mainly engaged in the binding function including protein binding and ATP binding ( Figure 4E). According to the KEGG pathway analysis, cell cycle and oocyte meiosis were the most two pathways involved by these 25 molecules ( Figure 4F). Thus, we concluded that the function of PLK1/2/3/4 and their related genes were primarily related to cell division and cell cycle.

| The correlation of PLK1/2/3/4 with patient's gender and TP53 mutant status in lung cancer subtypes
Due to the oocyte meiosis pathway involved by PLK1/2/3/4 and their related genes, we further explored a correlation of PLK1/2/3/4 with TP53 mutation. In both lung cancer subtypes, significantly higher expressions of PLK1/4 were found in TP53 mutant lung cancer than that in TP53-wide subtypes ( Figure 5A,D,E,H). PLK2 expression was not related with TP53 mutation in either of lung cancer subtypes ( Figure 5B,F). PLK3 presented a lower expression in TP53 mutant lung squamous cell carcinoma. Taken together, our results showed that PLK1/3/4 expressions were remarkably associated with TP53 mutant status in lung cancer subtypes.
Next, we aimed to distinguish the differential expression of PLK1/2/3/4 between genders of patients with lung cancer subtypes. PLK1 expression was higher in male than that in female patients ( Figure S2A,E), so was PLK4 expression between genders ( Figure S2D,H), whereas PLK2 presented no differential expression between male and female of lung cancer patients ( Figure S2B,F).
PLK3 was higher expressed in female than male only in lung adenocarcinoma ( Figure S2C,G). These results indicated that PLK1/3/4 played a differential role between genders of patients with lung cancer subtypes.

| The methylation analysis of PLK1/2/3/4 in lung cancer
To explore an epigenetic mechanism driving the differential expression of PLKs in lung cancer, we investigated the promoter methylation level of PLK1/2/3/4 in lung adenocarcinoma and lung squamous cell carcinoma. In lung adenocarcinoma, PLK1/2/3 promoters had a lower level in primary tumour than those in normal tissue ( Figure 6A-C), and PLK4 promoter showed no significant alteration between normal and tumour primary ( Figure 6D). In lung squamous cell carcinoma, PLK1/3/4 promoters presented a higher methylation in primary tumour, whereas PLK2 promoter methylation was at lower level in primary tumour than that in normal sample.
After an observation of methylation in PLK promoters, we performed a correlation between PLK1/2/3/4 and DNMT1/3A/3B. DNMT1/3A/3B were positively related with PLK1/4 in lung cancer subtypes ( Figure S3), except the correlation between DNMT3A and PLK1 without significance in lung squamous cell carcinoma ( Figure S3E). In lung adenocarcinoma, DNMT3A/3B had a negative correlation with PLK3 ( Figure S3B,C). In lung squamous cell carcinoma, DNMT1 expression was negatively related with PLK2 level ( Figure S3D). Our findings revealed the correlation between mRNA levels of PLKs and DNMTs, providing a new insight into epigenetic regulation of PLK1/2/3/4 expression in lung cancer subtypes. report about NSCLC showed that PLK1/cMet axis mediated apoptosis of NSCLC cell. 25 Another prior study presented an antitumour activity of PLK1 inhibitor in NSCLC. 26 In the current study, To gain more insights into the methylation profile of PLKs, we performed a multivariate analysis based on in silico data. However, the main weakness of this study was the paucity of concrete data from in vivo and in vitro experiments to validate our findings, such as the prognostic value of global methylation pattern of PLKs. Another limitation is that large sample size is required in future studies to explore clinical features of PLKs, such as stages or grades. The detailed mechanism is worth to be investigated for the integrated prognostic role in lung cancer subtypes.

| Prognostic value of global methylation of PLK1/2/3/4 in patients with lung cancer subtypes
In summary, we herein identified the transcriptional and protein

ACK N OWLED G EM ENTS
The authors acknowledge the HelixLife for the introduction of public databases.

CO N FLI C T O F I NTE R E S T
The authors declare that they have no competing interests. Mi Li: Conceptualization (equal); Data curation (equal). Shilin Xia:

AUTH O R CO NTR I B UTI O N S
Conceptualization (equal); Data curation (equal).

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study were generated at TCGA, CPTAC and HPA. Derived data supporting the findings of this study are available from the corresponding author on reasonable request.