Systematic expression analysis of WEE family kinases reveals the importance of PKMYT1 in breast carcinogenesis

Abstract Objectives Many cancer cells depend on G2 checkpoint mechanism regulated by WEE family kinases to maintain genomic integrity. The PKMYT1 gene, as a member of WEE family kinases, participates in G2 checkpoint surveillance and probably links with tumorigenesis, but its role in breast cancer remains largely unclear. Materials and Methods In this study, we used a set of bioinformatic tools to jointly analyse the expression of WEE family kinases and investigate the prognostic value of PKMYT1 in breast cancer. Results The results indicated that PKMYT1 is the only frequently overexpressed member of WEE family kinases in breast cancer. KM plotter data suggests that abnormally high expression of PKMYT1 predicts poor prognosis, especially for some subtypes, such as luminal A/B and triple‐negative (TNBC) types. Moreover, the up‐regulation of PKMYT1 was associated with HER2‐positive (HER2+), basal‐like (Basal‐like), TNBC statuses and increased classifications of Scarff, Bloom and Richardson (SBR). Co‐expression analysis showed PKMYT1 has a strong positive correlation with Polo‐like kinase 1 (PLK1), implying they may cooperate in regulating cancer cell proliferation by synchronizing rapid cell cycle with high quality of genome maintenance. Conclusions Collectively, this study demonstrates that overexpression of PKMYT1 is always found in breast cancer and predicts unfavourable prognosis, implicating it as an appealing therapeutic target for breast carcinoma.


| INTRODUC TI ON
Malignant tumours are the most threatening human diseases around the world. In 2018, there were about 18.1 million new cancer cases and 9.6 million cancer-related deaths. 1 Among them, breast cancer is the most commonly diagnosed cancer and the leading cause of cancer-related death among women. The incidence of this aggressive disease remains alarmingly high with more than one million newly diagnosed cases each year. [1][2][3] Understanding the molecular mechanisms of breast carcinogenesis is an important task for researchers to develop new methods for diagnosis and treatment of this malignancy. Despite years of research, the overall 5-year survival rate for patients with breast cancer remains low. 4,5 Therefore, there is still an urgent need for finding reliable biomarkers for early diagnosis, accurate prognosis and targeted therapy. 6 F I G U R E 1 PKMYT1 mRNA expression was elevated in human breast cancer. A, This graph generated by Oncomine indicates the numbers of datasets with statistically significant mRNA overexpression (red) or downexpression (blue) of PKMYT1, WEE1 and WEE1B (cancer tissues vs corresponding normal tissues). The threshold was defined with the following parameters: P-value of 1E-4, fold change of 2 and gene ranking of 10%. B, C, The GEPIA database verified that PKMYT1 gene expression was significantly upregulated in breast cancer tissues (BRCA) (n = 1085) compared with normal breast tissues (n = 291), *P < .05   During cell cycle, normal cells maintain the stability of the genome primarily through the DNA damage checkpoints, a surveillance mechanism that is frequently deregulated in cancers.
Because of the loss-of-function of tumour suppressor genes, such as mutations in p53 that leads to the inactivation of the G1 check- inhibitor renders apoptosis in TNBC cells, but its clinical application remains limited. 9,14,15 In other aspect, the role of PKMYT1 in breast cancer development remains unknown and awaits further investigations. In this work, we applied a wide range of integrated bioinformatics approach to assess the importance of PKMYT1 by analysing the expression, potential function and prognostic impact of PKMYT1 in human breast cancer.

| Data mining in Oncomine database
The Oncomine database (https ://www.oncom ine.org/resou rce/ login.html) is a publicly accessible, online cancer microarray database that helps facilitate research from genome-wide expression analysis. We used the Oncomine database to determine the transcription level of the PKMYT1 gene in breast cancer 16

| University of California Santa Cruz (UCSC) cancer genomics browser analysis
The UCSC Cancer Genomics Browser (http://xena.ucsc.edu/) 18,19 was used to verify the heat map of PKMYT1 expression, and the correlation between PKMYT1 and hub genes expression were analysed.

| Catalogue of somatic mutations in cancer (COSMIC) analysis for PKMYT1 mutations
The COSMIC database (http://cancer.sanger.ac.uk) is a high-resolution resource for studying the effects of somatic mutations in all forms of human tumours. We used this database to analyse mutations in PKMYT1 in breast cancer. 20,21 An overview of the distribution and substitutions on the coding strand in breast cancer was depicted in a pie chart.

| Gene correlation analysis in GEPIA
The online database Gene Expression Profiling Interactive Analysis Colour Mutation type Number of samples(%) were as follows: MCODE score > 5 points, degree cut-off = 2, node score cut-off = 0.2, Max depth = 100, and k-Score = 2.

| Functional and KEGG Pathway Enrichment Analysis
DAVID (http://david.abcc.ncifc rf.gov/) is a functional annotation tool that reveals the biological significance behind by entering a list of genes. 29,30 Based on the extracted co-expressed genes, GO analysis can be divided into three categories: biological processes (BP), cellular components (CC) and molecular functions (MF). 31 The KEGG pathway database is used to identify biological pathways for co-expressed gene enrichment. 32 Statistical significance was assessed using Fisher's exact test, and P-value < .05 was considered significant.

| Statistical analysis
All statistical analyses were performed by default as described by web resources. Briefly, Students' t test was conducted to compare mRNA expression in Oncomine database. Log-rank test was used for computing P-value in Kaplan-Meier (KM) plotter. GEPIA differential analysis was tested using one-way ANOVA by defining the disease state (Tumour or Normal) as variable. In DAVID annotation system, Fisher's exact test was adopted to measure the gene enrichment in annotation terms. In Breast Cancer Gene-Expression Miner v4.0, the linear dependence (correlation) between two variables was measured using Pearson's correlation coefficient. The correlation of gene expression in cBioPortal and UCSC databases was evaluated by Spearman's correlation. P < .05 was considered to be statistically significant (*, P < .05; **, P < .01; ***, P < .001).

| Up-regulation of PKMYT1 mRNA expression in human breast cancer
We analysed the expression profile of WEE family kinases using Oncomine database. The expression of PKMYT1, but not of WEE1 and WEE1B, was significantly elevated in several solid tumours, especially in breast cancer and colorectal cancer ( Figure 1A). The mining of GEPIA database further confirmed that PKMYT1 was the only member of WEE family kinases unregulated in breast cancer (BRCA) tissues in relative to normal tissues ( Figure 1B,C).

| PKMYT1 mutations are rare and high PKMYT1 expression predicts poor prognosis in breast cancer
We employed cBioPortal to evaluate the frequency of changes in PKMYT1 mutations in breast cancer. The frequency of mutation is very low, only 0.1% ( Figure 3A). The mutations of PKMYT1 in breast cancer were analysed using the COSMIC database. The pie chart describes the types of mutations, including nonsense mutations, missense mutations, and in-frame deletions, the largest proportion of which are missense mutations, up to 55.56% ( Figure 3B).   Nucleotide changes included C > T, C > G, G > C and T > C mutations, with the largest proportion being C > G and G > C ( Figure 3C).

| The associations of PKMYT1 expression profiles and clinical parameters in breast cancer patients
The expression profiles of PKMYT1 were examined across PAM50  Figure 4C). In addition, patients with Basal-like status showed significantly increased PKMYT1 expression (P < .0001) compared with patients with negative Basal-like status ( Table 2 and Figure 4D). Compared with non-TNBC group, PKMYT1 mRNA expression was significantly higher in TNBC patients (P < .0001) ( Table 2 and Figure 4E), but not in the case with Nodal Status (P = .8173) ( Table 2 and Figure 4F). In the Scarff, Bloom and Richardson (SBR) grade 34 status criteria, increased SBR levels were significantly associated with increased PKMYT1 transcript levels in relative to the SBR1 group (P < .0001) ( Figure 4G).
There was no significant relationship between ages (P = .3099) ( Figure 4H). With higher rate of Nottingham Prognostic Index (NPI) classification, the lower of the survival rate was associated ( Figure 4I).

| KEGG and GO enrichment analysis revealing functional association of PKMYT1 with cell proliferation
The Oncomine database (Stickeler Breast dataset) ( Figure 5A (Table 3). Collectively, these data suggest an essential role of PKMYT1 in regulating cell proliferation in breast cancer.

| PKMYT1 PPI network construction and analysis of 10 hub genes
Using the STRING database, the co-expressed 80 genes were constructed into a protein-protein network, and the most important module was obtained using Cytoscape (MCODE plug-in) ( Figure 6A). The top ten genes, including PLK1, NCAPH, TRIP13, KIF4A, SPAG5, CDCA5, FOXM1, ESPL1, PRC1 and CENPN, were identified as potential hub genes according to the degree score generated by CytoHubba plug-in (the cytoHubba plug-in, top 10 nodes ranked by DMNC) (Figure 6B), consistent with their enrichment in the top module analysed by MCODE (highlighted in yellow) ( Figure 6A). The biological process analysis of hub genes was further performed using BINGO plug-in. Particularly, peptide biosynthetic process, phytochelatin biosynthetic process, cellular biosynthetic process, peptide metabolic process, secondary metabolic process and phytochelatin metabolic process were largely altered, suggesting that they may participate in the protein anabolism required for cell division ( Figure 6C). Hierarchical clustering of the hub genes was performed using UCSC Cancer Genomics Browser

| Co-expression of PKMYT1 and PLK1
cBioportal regression analysis showed that PKMYT1 and PLK1 had high correlation coefficients (Spearman's correlation = 0.79; Pearson's correlation = 0.60) ( Figure 7A). This positive correlation between PKMYT1 and PLK1 transcript was substantiated by the analysis via both the bc-GenExMiner 4.0 database ( Figure 7B) and GEPIA ( Figure 7C). This was further confirmed using UCSC Xena with consistent correlative patterns in different subtypes ( Figure 7D). These data demonstrate that PKMYT1 has a strong association with PLK1, suggesting that they may be functional partners in breast carcinoma.

| High PLK1 expression predicts unfavourable prognosis in patients with breast cancer
To determine the genetic alteration of PLK1 in breast cancer, the expression profile of PLK1 was investigated using the Oncomine  Figure 8A). Subsequently, the prognostic value of PLK1 in breast cancer was studied by Kaplan-Meier plotter database, and it was confirmed that high expression of PLK1 mRNA was significantly associated with the decrease of RFS, OS, DMFS and PPS in breast cancer ( Figure 8B).

F I G U R E 8
The expression of PLK1 is upregulated in breast cancer and associated with poor prognosis. A, Invasive breast carcinoma, invasive ductal breast carcinoma, mixed lobular and ductal breast carcinoma, invasive lobular breast carcinoma, intraductal cribriform breast adenocarcinoma, and invasive ductal and lobular carcinoma were included in the box plots derived from the Oncomine database. B, Survival analyses of PLK1 in breast cancer using KM plotter. OS, overall survival; RFS, relapse-free survival; DMFS, distant metastasis-free survival. PPS, post-progression survival

| D ISCUSS I ON
Breast cancer is one of the most common malignancies in the mid- Polo-like kinase (PLK1), a key regulatory kinase involved in mitosis and cell cycle progression, 50,51 plays an important role in tumour cell anabolism by activating the pentose phosphate pathway. 52 The positive correlation of PLK1 and PKMYT1 in cancer cells may indicate a particular G2 checkpoint mechanism which synchronizes the rapid cell proliferation in accordance with maintenance of genomic stability. Mechanistically, PKMYT1 is highly expressed in cancer cells, and G2/M check is performed to ensure genomic stability. Simultaneously, the duration for G2/M checkpoint should be precisely controlled by PLK1 regulatory pathway for rapid cell proliferation. Co-targeting these two collaborative kinases might be an efficient way to treat breast carcinoma.
In summary, we have confirmed the up-regulation of PKMYT1 and its partner, PLK1, in breast cancer and validated their importance as prognostic factors. We propose that PKMYT1 could be a promising molecular target for the diagnosis and treatment of breast cancer.

ACK N OWLED G EM ENTS
We thank Dr Chun-Yan Lim from University of California, Berkeley,

CO N FLI C T O F I NTE R E S T
All authors declare that they have no conflict of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
Source data of this study were derived from the public repositories, as indicated in the section of "Materials and Methods" of the manuscript. And all data that support the findings of this study are available from the corresponding author upon reasonable request.